Protein Kinase R, a protein found in our bodies, is a broad antiviral which is critical to our resistance to many different viruses. It serves as a kind of sentinel; it recognizes and binds to double-stranded RNA (dsRNA) in our cells (Sadler et al. 2007). Our cells never produce this form of RNA on their own - only viruses do - so presence of dsRNA is a warning sign that our cells are under attack. When PKR binds to a piece of dsRNA, it gains the ability to modify another protein within our cells, eIF2α, which is a protein allows our cellular machinery to “translate” messenger RNA into protein (Donnelly et al. 2013). When PKR modifies eIF2α in cells hijacked by a virus, it prevents eIF2α from initiating this translation process (Sadler et al. 2007). Without translation, viruses can no longer replicate within these cells and become immensely less worrisome to human beings. Thus, PKR has been the target of viruses for as long as it has existed, and a sort of “arms race” has persisted; viruses adapt mutations to interfere with PKR’s function, while PKR adapts mutations to evade the various viruses which bombard it. Kathryn S. Carpentier et al. have discovered that in fact, a single amino acid (only a few atoms large) within PKR that controls its susceptibility to multiple antagonistic proteins introduced by different viruses.
To understand how a tiny region only a few atoms large within a protein could determine our ability to resist certain viruses, we must first examine the interactions between various iterations of cytomegalovirus (CMV), a virus which can cause lifelong life-threatening disease in patients with compromised immune systems, and PKR. Carpentier et al. combined a virus called Vaccinia virus (VacV) with the PKR-antagonist gene, known as TRS1, from various primate cytomegaloviruses as well as from human cytomegalovirus (HCMV); TRS1 is known to prevent PKR from interacting with eIF2α, allowing human cytomegalovirus to replicate within human cells without being ousted. They used VacV to infect human cells called HeLa cells which contained the PKR gene and the PKR protein. They noted that HCMV was able to significantly decrease the function of PKR in HeLa cells, as the virus replicated successfully at a high percentage.[fig3] Then, they took advantage of the interactions of different primate species cytomegaloviruses with PKR in order to map their specificity. They pinpointed a single amino acid - the position 489 phenylalanine found within PKR’s αG helix - which, when mutated to a serine, significantly increased its resistance to the TRS1 antagonist protein coded by human cytomegalovirus.
Upon discovering that mutating just one amino acid (the mutation is abbreviated F489S) would significantly increase PKR’s resistance to TRS1, Carpentier et al. ran replication tests which showed that HCMV replicated significantly less in the cells which contained the mutated PKR. It became clear that not only does this mutation resist a damaging viral protein, it actually prevents HCMV from replicating successfully in human cells. They also studied the modification activity of eIF2α (which PKR interacts with) and noted “robust” activity within the mutated PKR cells, giving proof that the mutant PKR was functioning properly, unhindered by the virus.
Carpentier et al. knew from previous studies that the viral antagonist protein TRS1 binds directly to PKR in order to inhibit its function. Curious to understand how this PKR mutation actually affected its interaction with its antagonist TRS1. They suspected that the mutation inhibited binding, and performed a study in which they used a molecule to “tag” TRS1 proteins. They were able to identify the TRS1 proteins which successfully bound to PKR using antibodies which targeted the tag molecule and the PKR. The experiment confirmed their suspicion, revealing that TRS1-PKR binding was heavily hindered by the presence of the phenylalanine-to-serine mutation.
Phenylalanine and serine are only two of the twenty amino acids which are the building blocks of protein, so naturally the researchers had to ask - what of the other 18? They were interested to see how tolerant the 489 position of PKR was to different amino acid mutations (how well it could still function when they were applied), and what better way to find out, they figured, than to try them all? They induced every possible amino acid mutation at PKR position 498 and studied each mutant’s interaction with eIF2α in absence of any virus. In each case, PKR successfully modified eIF2α. This meant that the 489 position, while crucial to regulating TRS1 binding, was not significant to its eIF2α modification; PKR can mutate this position to evade viruses in any manner without the concern of losing its function as collateral! The next step was to identify which of these 20 mutations (besides the serine, which was already known) would also inhibit TRS1 binding. In fact, all mutations were preventative except for phenylalanine, tryptophan, methionine, and tyrosine. It is important to note that each of these amino acids has a hydrophobic side chain, which means they might associate with a hydrophobic region of TRS1.
Figure 7a: Cells expressing wild type PKR and the F489S mutant gene were infected with HCMV-TRS1, VacV K3L, and VacV K3L H47R (the functional mutant which has a negative effect on PKR activity). Replication assay results showed that the F489S mutation to the PKR gene gave PKR significant resistance to TRS1 and K3L H47R, allowing it to function properly and limit viral replication.
Position 489 is found on the αG helix of PKR which inserts into a hydrophobic region of eIF2α during activation. Poxviruses encode a protein called K3L to mimic the structure eIF2α and bind competitively to PKR, thus inhibiting its ability to modify eIF2α. Perhaps the most fascinating aspect of the study was that Carpentier et al. discovered that the same phenylalanine-to-serine mutation at position 489 in PKR also provided it with resistance to K3L, which is unrelated to TRS1 except that they evolved convergently. Although wild type K3L had little effect on the function of PKR, a functional mutant K3L H47R was combined with VacV and introduced via VacV infection of cells containing F489S PKR and wild type PKR. The mutated PKR cells exhibited a significantly lower percent infection than the wild type PKR cells did, showing that this single amino acid mutation was able to curtail infection of multiple viruses by avoiding harmful interactions with two unrelated viral proteins. Such a result highlights not only the incredible ability of tiny mutations to have powerful antiviral effects, but also how different viruses have converged evolutionarily to target weak points in cellular antiviral defense (e.g. PKR’s αG helix).
Now that such a weak point in our antiviral defense has been examined, how do we proceed? Understanding how viruses have evolved convergently to target such weak points could potentially give scientists the opportunity to develop broad antiviral drugs which beneficially mutate and bolster weak points in our antiviral defense.
Carpentier KS, Esparo NM, Child SJ, Geballe AP (2016) A Single Amino Acid Dictates Protein Kinase R Susceptibility to Unrelated Viral Antagonists. PLoS Pathog 12(10): e1005966. doi:10.1371/journal.ppat.1005966 http://journals.plos.org/plospathogens/article/file?id=10.1371/journal.ppat.1005966&type=printable
Sadler AJ, Williams BR. Structure and function of the protein kinase R. Curr Top Microbiol Immunol. 2007;316:253–292. pmid:17969452
2. From paper, figure 7a