1.1 Functions of the cellular prion protein (PrP^C)

Transmissible spongiform encaphalopathies (TSEs) are fatal neurodegenerative diseases that are characterized by the accumulation of a protease resistant form (PrP^SC) of a cellular protein (PrP^C), that is normally sensitive to digestion 〚1〛. There has been considerable speculation as to the function of the cellular prion protein (PrP^C), which is a widely expressed glycoprotein, anchored in the plasma membrane via a glycosylphosphatidylinosol link. PrP^C null mice appear to develop and reproduce normally and are no longer susceptible to TSEs, suggesting that normally PrP^C has no essential function that cannot be compensated 〚2〛. However evidence suggests that PrP^C performs a multiplicity of functions including the binding of Cu²⁺ which may be associated with superoxide dismutase activity 〚3〛; the capacity to bind cell surface receptors, in particular that for laminin 〚4〛; the transduction of extracellular signals impinging at the plasma membrane 〚5〛; and the ability to bind nucleic acids 〚6–8〛.

The long standing interest of our laboratory with respect to nucleic acid binding proteins, and in particular the nucleic acid chaperone functions of retroviral nucleocapsid proteins (NC), prompted us to examine a possible activity of PrP^C in the chaperoning of viral nucleic acids in comparison with the well established, specific chaperone activity of the HIV-1 nucleocapsid protein NCP7 〚9–10〛.

1.2 Nucleic acid chaperone functions of NCP7

The nucleic acid chaperone properties of NCP7 with respect to retroviral replication have been reviewed elsewhere 〚9–10〛 and are summarized in figure 1. In general, nucleic acid chaperone activity depends on the capacity of the chaperone protein to bind to a nucleic acid and destabilise its secondary structure. Protein–nucleic acid and protein–protein interactions destabilize nucleic acid structures of intermediate stability. This results in the progressive accumulation of the elements in their most stable structural conformation. In this way the activity of a nucleic acid chaperone can for example promote the annealing of complementary strands of nucleic acids under physiological conditions when in the absence of chaperones high temperatures and salt concentrations would be required to break down the single-stranded initial structure to allow hybridization. However because chaperone activity destabilizes structures of intermediate stability, only stable interactions between appropriate complementary partners will be facilitated.

2.1 PrP mimics NCP7 chaperone activity in tRNA^Lys annealing and initiation of ss-cDNA synthesis by RT

The nucleic acid chaperone activity of NCP7 is essential during several phases of the life cycle of HIV-1. After receptor binding, penetration of the viral core inside a cell and uncoating, the chaperone activity is required for the generation of the proviral DNA from the genomic RNA template. The initial step in a sequence of events is the specific production of the HIV strong stop cDNA (ss-cDNA) from the RNA template and requires the annealing of a tRNA^Lys primer to the complementary primer binding site (PBS) in the 5’ region of the genomic RNA and reverse transcriptase (RT) to catalyse elongation of the ss-cDNA. That a nucleic acid chaperone is also required for this process is illustrated by the failure of the reaction to proceed under physiological conditions in vitro in the absence of NCP7 (figure 2, lane 1 and 5) 〚10〛. However, in the presence of NCP7 annealing of the radioactively labelled tRNA^Lys occurs allowing production of the ss-cDNA by RT (figure 2, lanes 2–4). The capacity to chaperone the annealing of the tRNA to the PBS is a highly specific property of retroviral NC proteins that is not shared by well characterized cellular nucleic acid chaperones such as Rec A, p53 and others 〚10, 11〛. For this reason we were very surprised to find that PrP^C of human origin could substitute for NCP7 in this reaction (figure 2, lanes 6–8).

2.2 PrP mimics NCP7 during strand transfers required to generate LTRs of HIV-1 proviral DNA

In order to generate the proviral DNA flanked by the LTRs the ss-cDNA must be transfered to the 3’ end of the genomic RNA to act as primer for complete minus strand DNA (cDNA(-)) synthesis. For this process to occur RT with RNAse H activity, the 5’ and 3’ RNA sequences each containing the R sequences present in these regions of the genomic RNA and the nucleic acid chaperone activity of a retroviral NC are needed 〚9〛. In vitro this can be modelled using a labelled oligonucleotide complementary to the primer binding site as a primer together with the other elements and NC. In the absence of NC no specific reaction products are visible (figure 3, lane 1). However in the presence of NCP7 a dose-dependent synthesis of the specific ss-cDNA and HIV-1 cDNA(-) reaction products occurs (figure 3, lanes 2-4). HuPrP can replace NCP7 at similar concentrations in this assay of strand transfer (figure 3, lanes 6–8) indicating that huPrP has a chaperoning activity similar to NCP7. In vitro experiments also established that huPrP can replace NCP7 during the second strand transfer (data not shown) 〚12〛.

2.3 PrP mimics NCP7 with respect to the inhibition of self-primed reverse transcription

Besides chaperoning the specific interaction between complementary nucleic acid sequences (primer binding, strand transfers), that must occur for the synthesis of proviral DNA from the genomic RNA template, NC proteins also function to inhibit non-specific replicative events that can take place. The repair of nicks in the genome or refolding of the RNA template can cause self-initiation of reverse transcription. Needless to say these non-specific phenomena will prevent the faithful production of proviral DNA. Retroviral NC proteins bind with low affinity over the whole length of the RNA template. The chaperone activity, which destabilizes RNA structures of intermediate stability, thus prevents self initiation 〚9〛. In the absence of a nucleic acid chaperone non-specific cDNA is synthesized as a result of self-priming of the RNA template in the presence of RT (figure 4, lanes 1 and 5). However, when NCP7 (figure 4, lanes 2-4) or huPrP^C (figure 4, lanes 6–8) is included in the reaction, self priming is inhibited in a dose dependent manner.

2.4 PrP mimics NCP7 with respect to the dimerization of HIV-1 genomic RNA

In viral particles the HIV-1 retroviral genome is present as a dimer coated with about 2 000 molecules of NCP7 with the replication primer tRNA^Lys annealed to the primer binding site. The dimerization of genomic RNA also functions to form the viral nucleocapsid and participates in the process of viral particle budding 〚9〛. The dimerization process can be investigated in vitro using 5ˈ leader RNA, labelled tRNA^Lys and nucleic acid chaperones. In the absence of an appropriate chaperone there is neither primer binding nor dimerization (figure 5A and B, lanes 1 and 5). However in the presence of NCP7 binding of the tRNA^Lys occurs and the dose-dependent formation of RNA dimers (and in vitro oligomers) can be observed (figure 5A and B, lanes 2–4). It is very suprising that huPrP^C shows an identical activity (figure 5A and B, lanes 6–8). It is well established that this is a specific characteristic of retroviral NC proteins and similar activity is not observed with other well characterized nucleic acid binding and chaperone proteins of either viral, bacterial or cellular origin e.g. RT, HIV-1 Vpr and Integrase, T4 gp32, RecA and p53 〚9–11〛.

The finding that huPrP^C behaves as a nucleic acid chaperone confirms and extends previous evidence that PrP^C is a nucleic acid binding protein 〚6–8〛. The nucleic acid binding activity can be demonstrated using genomic RNA of HIV-1 by gel shift assay 〚12〛. Interestingly PrP also interacts with viral DNA to form large nucleoprotein complexes similar to those formed by NC binding to DNA as revealed by dark field electron microscopy 〚12〛. Indeed it was these findings that prompted us to further investigate a possible function for PrP^C as a nucleic acid chaperone 〚12, 13〛. Our results, which show clearly that huPrP is a molecular mimic of NCP7, and retroviral NC in general, are of particularly great interest. HuPrP^C has all the specific chaperone functions of HIV-1 NCP7 being able to chaperone tRNA^Lys annealing to primer binding site, both the minus and plus strand transfers required for synthesis of proviral DNA flanked by LTR and dimerization of genomic RNA. Furthermore huPrP also possesses the capacity to inhibit self-initiation during elongation by RT to ensure the fidelity of the reverse transcription process. Many cellular or indeed viral proteins have been identified with nucleic acid chaperone properties, but apart from NC proteins, no other has these functions 〚9, 10〛. These nucleic acid chaperone properties are not unique to huPrP^C because similar results were obtained with PrP^C of ovine 〚12, 13〛 and murine origin. Studies of the nucleic acid chaperone domain of huPrP^C show that the nucleic acid chaperone properties map to the unstructured N-terminal region 〚12, 13〛. Interestingly there is no evident sequence or structural homology with identified retroviral NC proteins.

One question is: why should a cellular protein have these functions? The nucleic acid chaperone functions that PrP^C possesses are not apparently essential for any identified cellular function, and such chaperone activity might be expected to facilitate infection by retroviruses. On the contrary, a hypothesis under investigation in our laboratory is that PrP^C may function as a defence mechanism against retroviral infection. There is some preliminary evidence to support this. In the central nervous system, where cellular levels of expression of PrP^C are highest, astrocytes are the major neural cells infected by HIV-1. Part of the cellular response of astrocytes to HIV-1 infection is to increase expression of the PrP^C mRNA 〚14〛. Astrocytes infected with HIV-1 do not normally produce infective particles 〚14〛. We have compared the levels of expression of PrP^C in established stable MLV-based helper cell lines and in the parent cell lines from which they were derived. So far we have always found that the expression of PrP^C is either reduced or indeed not detectable in the helper cell lines although PrP^C is readily detectable in the parent cell line (E.D. unpublished). This would suggest that high levels of expression of PrP^C may not be compatible with efficient production of retrovirus (i.e. HIV, MLV).

Preliminary evidence suggests that PrP^C interacts with the GAG polyprotein of HIV-1 from which NCP7 is formed by proteolytic maturation. PrP might compete with NCP7 for binding sites on the HIV-1 genomic RNA and thus perturb discrete phases of the viral life cycle that we have yet to identify. GAG plays a critical function in the assembly of viral particles, perhaps PrP–GAG interactions may function to inhibit assembly. If PrP^SC were to remain bound to genomic RNA in the viral core, the question would arise as to whether viruses might behave as vectors for PrP^sc. If this were the case one might expect transmissible spongiform encaphalopathies to be more rapidly disseminated and produce lesions more quickly in animals co-infected with PrP^SC and retroviruses. Indeed this has been reported in the mouse-MLV system 〚15〛.

On the other hand interactions between PrP and viral nucleic acids might activate secondary messenger cascades that may influence processes required for efficient viral replication or perhaps induce apoptosis of the host cell. This would explain the absence, or the reduced expression of PrP^C in helper cells selected for efficient stable virus production. However, despite the fact that PrP^C has been implicated in signal transduction 〚5〛, the existence of such a mechanism remains to be established. Clearly the elucidation of such a phenomenon may be of great interest with respect to the development of new anti-viral strategies.