The Plasmodium falciparum cytoplasmic translation apparatus: a promising therapeutic target not yet exploited by clinically approved anti-malarials

Background The continued spectre of resistance to existing anti-malarials necessitates the pursuit of novel targets and mechanisms of action for drug development. One class of promising targets consists of the 80S ribosome and its associated components comprising the parasite translational apparatus. Development of translation-targeting therapeutics requires a greater understanding of protein synthesis and its regulation in the malaria parasite. Research in this area has been limited by the lack of appropriate experimental methods, particularly a direct measure of parasite translation. Methods An in vitro method directly measuring translation in whole-cell extracts from the malaria parasite Plasmodium falciparum, the PfIVT assay, and a historically-utilized indirect measure of translation, S35-radiolabel incorporation, were compared utilizing a large panel of known translation inhibitors as well as anti-malarial drugs. Results Here, an extensive pharmacologic assessment of the PfIVT assay is presented, using a wide range of known inhibitors demonstrating its utility for studying activity of both ribosomal and non-ribosomal elements directly involved in translation. Further, the superiority of this assay over a historically utilized indirect measure of translation, S35-radiolabel incorporation, is demonstrated. Additionally, the PfIVT assay is utilized to investigate a panel of clinically approved anti-malarial drugs, many with unknown or unclear mechanisms of action, and show that none inhibit translation, reaffirming Plasmodium translation to be a viable alternative drug target. Within this set, mefloquine is unambiguously found to lack translation inhibition activity, despite having been recently mischaracterized as a ribosomal inhibitor. Conclusions This work exploits a direct and reproducible assay for measuring P. falciparum translation, demonstrating its value in the continued study of protein synthesis in malaria and its inhibition as a drug target. Electronic supplementary material The online version of this article (10.1186/s12936-018-2616-7) contains supplementary material, which is available to authorized users.


Introduction
6 116 generating ambiguous or misleading results, it has remained a commonly used assay for studying parasite 117 translation for lack of a better alternative [9]. 118 Commercially available compounds that directly interact with the eukaryotic ribosome to inhibit 119 translation initiation and/or elongation via a variety of mechanisms and binding sites, as well as several 120 inhibitors of translation known to act upon non-ribosomal components of the translational machinery were 121 tested (Tables 1 & 2). The eukaryote-specific inhibitors bruceantin and verrucarin A inhibit translation 122 initiation through binding of mutually exclusive sites [12][13][14][15]. Suramin, also a specific inhibitor of the 123 eukaryotic ribosome, inhibits both initiation and elongation through binding of multiple sites on the 40S, 60S 124 and 80S ribosomes [16]. The eukaryote-specific elongation inhibitors tested are also distinct in their activities: 125 cycloheximide and lactimidomycin overlap in their binding of the ribosome A-site, but differences in size and 126 side-chains yield unique effects; anisomycin also overlaps cycloheximide's binding site, but the two drugs 127 bind the ribosome in distinct rotational conformations at different steps of elongation; homoharringtonine 128 binds the A-site, but specifically inhibits re-initiating ribosomes; and nagilactone C inhibits both eEF-1α- After determining the EC50 of each drug for the P. falciparum W2 strain in a 72-hour parasite growth 141 assay, the drugs were characterized in both the S35 incorporation and PfIVT assays (Table 3) (Figs 2 and 3).
142 Drugs were tested in the S35 and PfIVT assays at 0.1-, 1-, 10-, and 100-fold their determined growth assay 143 EC50 in W2 parasites, except in cases where the highest concentration was constrained by solubility or 144 available stock solution. The translation initiation inhibitors bruceantin and verrucarin A were both potent 8 145 (nanomolar) inhibitors of S35 incorporation and PfIVT (Fig 2). All translation elongation inhibitors 146 (anisomycin, cycloheximide, homoharringtonine, lactimidomycin, and nagilactone C) also strongly inhibited 147 both S35 incorporation and PfIVT (Fig 2). Cycloheximide was additionally tested at 1000-fold its EC50, as it 148 did not inhibit S35 incorporation at the lower concentrations tested, but did at this higher concentration, in line 149 with inhibitory concentrations in recent reports, which also show that significantly higher concentrations of 150 cycloheximide are required for complete, measurable inhibition of translation than for rapid and total parasite 151 killing in vivo (S1 Fig) [8,9,35]. Suramin, which has been shown to inhibit both translation initiation and 152 elongation, robustly inhibited PfIVT, but not S35 incorporation, likely due to poor cell permeability and the 153 short timeframe of the S35 assay (2 hour drug pre-incubation followed by 2 hour radiolabel incorporation) 154 (Fig 2). The tRNA mimetic puromycin, which induces premature termination of nascent polypeptides, 155 inhibited both S35 incorporation and PfIVT with similar efficacy (Fig 2). Elucidating an even greater range of 156 utility for the PfIVT assay, we found it to be capable of identifying inhibitors of non-ribosomal components of 157 translation. The glutamyl-prolyl-tRNA synthetase inhibitor halofuginone inhibits both S35 incorporation and 158 the PfIVT assay (Fig 2). In sum, these data demonstrate the ability of the PfIVT assay to interrogate both 159 direct ribosomal activity, as well as extra-ribosomal components of the translational machinery.
160 Importantly, all of the eukaryotic ribosome-specific inhibitors, which therefore should inhibit only P.
166 falciparum cytoplasmic and not apicoplast or mitochondrial ribosomes, displayed inhibition in the PfIVT 167 assay, with several achieving complete or near complete blocking of translation (suramin, anisomycin, 168 lactimidomycin, nagilactone C) (Fig 2). In addition, the prokaryotic ribosome-specific inhibitor thiostrepton 169 did not inhibit the PfIVT assay at any concentration tested, despite inhibiting the S35 assay at concentrations 170 above its determined EC50 (Fig 3). Thiostrepton is known to have multiple targets apart from ribosomes in 171 eukaryotes and has been shown to induce an ER stress response with a phenotype similar to thapsigargin, 172 which likely accounts for its activity in the S35 assay [36-38]. 232 However, when subjected to stringent quality control and careful optimization, the PfIVT assay reliably and 233 specifically identifies inhibitors of translation initiation and elongation, as well as inhibitors of non-ribosomal 234 proteins necessary for translation, such as tRNA synthetase.

235
The PfIVT assay is particularly valuable to the study of P. falciparum translation as a direct measure of 236 translation, as opposed to the indirect measures to which the field has historically been constrained, such as 237 incorporation of radiolabeled amino acids in vivo. Importantly, our data show the PfIVT assay is significantly 238 more specific, and in some cases more sensitive, than S35-radiolabel incorporation in identifying small 239 molecule inhibitors of translation. Indeed, the PfIVT assay specifically identified all eukaryotic translation 240 inhibitors tested, while S35-radiolabel incorporation was prone to false-positives. We found that none of the 241 clinically approved antimalarials tested are inhibitors of translation, emphasizing the potential for translation 242 as a useful therapeutic target, as there is unlikely to be pre-existing mechanism-specific resistance to any 243 identified candidates resulting from use of these drugs. It is notable that mefloquine, in contrast to other 244 previously reported translation inhibitors, did not exhibit any inhibitory activity. Mefloquine was likely 13 245 mischaracterized as an 80S ribosome inhibitor through a combination of non-specific inhibition of S35 246 incorporation, as well as artifacts arising from cryo-EM structures obtained under the non-physiologic 247 condition of 10mM magnesium -well above the ~4mM magnesium that we have found to be optimal for 248 translation (Fig 1)[9].

249
While the PfIVT assay exhibits clear benefits over existing methodologies for the study of P.