Research Article: How Does Domain Replacement Affect Fibril Formation of the Rabbit/Human Prion Proteins

Date Published: November 17, 2014

Publisher: Public Library of Science

Author(s): Xu Yan, Jun-Jie Huang, Zheng Zhou, Jie Chen, Yi Liang, Human Rezaei.

http://doi.org/10.1371/journal.pone.0113238

Abstract

It is known that in vivo human prion protein (PrP) have the tendency to form fibril deposits and are associated with infectious fatal prion diseases, while the rabbit PrP does not readily form fibrils and is unlikely to cause prion diseases. Although we have previously demonstrated that amyloid fibrils formed by the rabbit PrP and the human PrP have different secondary structures and macromolecular crowding has different effects on fibril formation of the rabbit/human PrPs, we do not know which domains of PrPs cause such differences. In this study, we have constructed two PrP chimeras, rabbit chimera and human chimera, and investigated how domain replacement affects fibril formation of the rabbit/human PrPs.

As revealed by thioflavin T binding assays and Sarkosyl-soluble SDS-PAGE, the presence of a strong crowding agent dramatically promotes fibril formation of both chimeras. As evidenced by circular dichroism, Fourier transform infrared spectroscopy, and proteinase K digestion assays, amyloid fibrils formed by human chimera have secondary structures and proteinase K-resistant features similar to those formed by the human PrP. However, amyloid fibrils formed by rabbit chimera have proteinase K-resistant features and secondary structures in crowded physiological environments different from those formed by the rabbit PrP, and secondary structures in dilute solutions similar to the rabbit PrP. The results from transmission electron microscopy show that macromolecular crowding caused human chimera but not rabbit chimera to form short fibrils and non-fibrillar particles.

We demonstrate for the first time that the domains beyond PrP-H2H3 (β-strand 1, α-helix 1, and β-strand 2) have a remarkable effect on fibrillization of the rabbit PrP but almost no effect on the human PrP. Our findings can help to explain why amyloid fibrils formed by the rabbit PrP and the human PrP have different secondary structures and why macromolecular crowding has different effects on fibrillization of PrPs from different species.

Partial Text

Transmissible spongiform encephalopathies, also known as prion diseases, are infectious fatal neurodegenerative diseases that affect the nervous system in humans and animals [1]. The key procedure of prion diseases is believed to be the conversion of the normal protease-sensitive cellular prion protein (PrPC) in such mammals into the aberrant protease-resistant pathogenic prion protein (PrPSc) [1]–[4]. Rabbits are among the few animal species that have resistance to prions from other animal species [5]. However, the three-dimensional structure of the recombinant protein rabbit PrPC is composed of an unstructured flexible N-terminal region and a C-terminal globular domain which comprises two short anti-parallel β-strands and three α-helices similar to the structures of other mammalian PrPC[6]–[8]. There are few difference among the C-terminal domains, and some research believes that the unique primary structure of rabbit PrPC inhibits formation of its abnormal isoform [6], [9].

Prion protein, a unique infectious amyloid disease-associated protein, causes many lethal human and animal prion diseases [1]. Rabbits are only sensitive to artificial rabbit prion [17] and resistant to prions from other animal species [5]. It has been demonstrated that the unique rabbit PrP sequence, β2-α2 helix-cap, and the residues surrounding the glycosylphosphatidylinositol anchor attachment site could contribute to its resistance to prion diseases [9], [39]–[41]. Furthermore, the presence of either serine (rabbit) or asparagine (hamster) residues in positions 170 and 174 of PrP not only affect the secondary structure of the β2-α2 loop but also the propensity with which the PrP misfolds into β-state-rich octamers [42]. A recent molecular dynamics study has indicated that the sites I214 and S173 of the rabbit PrP can influence the stability of rabbit PrP native state [43]. We have demonstrated previously that macromolecular crowding remarkably inhibits fibril formation of the rabbit PrP but significantly accelerates fibril formation of the human/bovine PrPs [27], [28]. In this study, we want to know which domains of PrPs cause such differences. To align the rabbit PrP23-228 with the human PrP23-231, we found that there are 88% identities between their sequences. Hydrogen/deuterium exchange and solid-state NMR results have demonstrated that PrP fibrils contain in-register parallel β-sheets and that the structurally ordered fibril core includes the C-terminal segment, approximately residues 175–225, which includes the H2H3 of monomeric PrP [44], [45]. In this study the effects of replacement certain domains in rabbit and human PrPs on fibril formation under natural crowding conditions were investigated by a series of biochemical experiments using PrP chimeras containing the H2H3 domain. We demonstrated that two rabbit PrP chimeras we designed (rabbit chimera and chimera R) did form amyloid fibrils with different structural features in absence and presence of crowding agents. By contrast, the rabbit PrP forms amyloid fibrils with same structural features in absence and presence of crowding agents [27]. We also demonstrated that two human PrP chimeras we designed (human chimera and chimera H) did form amyloid fibrils with same structural features in absence and presence of crowding agents, which is in agreement with those observed in human PrP fibrils [25], [27]. Furthermore, we found that PK resistance of the fibrils from the rabbit chimera is different from that observed in rabbit PrP fibrils [27] but PK resistance of the fibrils from the human chimera is similar to that observed in human PrP fibrils [25], [27]. We then investigated amyloid formation of rabbit PrP-H2H3 and human PrP-H2H3. To our surprise, macromolecular crowding accelerated the nucleation step of fibril formation of both rabbit PrP-H2H3 and human PrP-H2H3, and both rabbit PrP-H2H3 and human PrP-H2H3 formed amyloid fibrils with β-sheet-rich conformation from the native state which has predominant α-helix conformation, which are similar to those of the human PrP [27], [28]. Our data indicate that the structure of amyloid fibrils formed by rabbit PrP-H2H3 is different from that formed by the rabbit PrP [27] while the structure of amyloid fibrils formed by human PrP-H2H3 is similar to that formed by human PrP [25], [27]. All the results above suggest that the amino acids beyond PrP-H2H3 (herein PrP-B1H1B2 domain) have a remarkable effect on fibrillization of the rabbit PrP but almost no effect on the human PrP. In other words, not only the H2H3 domain could play an important role, but also the B1H1B2 domain could take part in fibrillization of the rabbit PrP. Our conclusion that the H2H3 domain plays an important role in PrP fibrillization is in agreement with previously published work [29]–[31], [44], [45]. Rabbit/human chimera formed amyloid fibrils with the same structural features as those of chimera R/H, and the difference between Rabbit/human chimera and chimera R/H is the N-terminal flexibly disordered tail, indicating that an N-terminal flexibly disordered tail of PrP is not important for fibrillization of the rabbit/human PrPs.

 

Source:

http://doi.org/10.1371/journal.pone.0113238