Plasmodium falciparum protein Pfs16 is a target for transmission-blocking antimalarial drug development

Phenotypic cell-based screens are critical to the discovery of new antimalarial lead compounds. However, identification and validation of cellular targets of lead compounds is required following discovery in a phenotypic screen. We recently discovered a Plasmodium transmission-blocking N-((4-hydroxychroman-4-yl)methyl)-sulfonamide (N-4HCS) compound, DDD01035881, in a phenotypic screen. DDD01035881 and its potent derivatives have been shown to block Plasmodium male gamete formation (microgametogenesis) with nanomolar activity. Here, we synthesised a photoactivatable N-4HCS derivative, probe 2, to identify the N-4HCS cellular target. Using probe 2 in photo-affinity labelling coupled with mass spectrometry, we identified the 16 kDa Plasmodium falciparum parasitophorous vacuole membrane protein Pfs16 as the likely cellular target of the N-4HCS series. Further validating Pfs16 as the cellular target of the N-4HCS series, the Cellular Thermal Shift Assay (CETSA) confirmed DDD01035881 stabilised Pfs16 in lysate from activated mature gametocytes. Additionally, photo-affinity labelling combined with in-gel fluorescence and immunoblot analysis confirmed the N-4HCS series interacted with Pfs16. High-resolution, widefield fluorescence and electron microscopy of N-4HCS-inhibited parasites was found to result in a cell morphology entirely consistent with targeted gene disruption of Pfs16. Taken together, these data strongly implicate Pfs16 as the target of DDD01035881 and establish the N-4HCS scaffold family as a powerful starting point from which future transmission-blocking antimalarials can be developed.

To advance the N-4HCS scaffold here we sought to identify the mode of action and/or cellular 1 0 1 target of DDD01035881. Using photo affinity labelling, label-free Cellular Thermal Shift Assay 1 0 2 (CETSA) and cellular analysis of treated parasites, we present strong interdisciplinary evidence  The N-4HCS scaffold was recently identified in a high-throughput screen for transmission blocking 1 1 0 antimalarials 14 , with subsequent development to improve its activity by medicinal chemistry 15 . We 1 1 1 sought to determine a potential mode of action and cellular target for the N-4HCS scaffold, starting 1 1 2 out from the original hit DDD01035881, given its potency for inhibiting Plasmodium falciparum 1 1 3 parasite microgametogenesis, the process of male gamete formation from blood-circulating mature To facilitate identification of potential cellular target(s) via photoaffinity labelling we first derivatised 1 1 7 DDD01035881 to incorporate both a photo-activatable group and clickable alkyne moiety onto the the tolerance of the N-4HCS scaffold to large changes in structure was found to be limited ( Table   1 2 0 S1), consistent with our previous medicinal chemistry study 15 . Ultimately, a strategy to incorporate 1 2 1 an alkyne handle and aryl azide moiety separately on the molecule was developed, leading to the 1 2 2 synthesis of probe 2. Parent molecule 1 was synthesised to resemble probe 2 by retaining similar 1 2 3 structural changes to the N-4HCS scaffold without the photoactivatable and clickable moieties, 1 2 4 thus mimicking biological activity of probe 2 as a control or active competitor. Critically, parent 1 2 5 molecule 1 and probe 2 retained micromolar to nanomolar IC 50 s in the in vitro male gamete 1 2 6 formation assay, respectively ( Table 1).

2 7
Preliminary testing of probe 2 was performed to test cross-linking and ligation to azido-1 2 8 TAMRA/biotin capture reagent (AzTB) 16 . Cell lysate derived from activated mature gametocytes 1 2 9 was irradiated in the presence of increasing concentrations of probe 2 or DMSO to test a probe 2 1 3 0 concentration-dependent crosslinking of proteins. Lysate containing crosslinked proteins was then 1 3 1 ligated to AzTB capture reagent in a copper catalysed azide-alkyne cycloaddition (CuAAC). presence of probe 2 and was additionally shown to be AzTB dependent, thus confirming Target identification was subsequently carried out using a 9plex tandem mass tag (TMT) 1 3 7 methodology. Live P. falciparum stage V gametocyte cultures were treated with either DMSO, 1 3 8 probe 2 (10 µM) or a combination of probe 2 and parent molecule 1 to act as a competitor (10 µM 1 3 9 each), acquiring triplicate samples of each condition. Live treated gametocytes were irradiated at 1 4 0 254nm to crosslink probe to protein targets, then purified and lysed before ligating to AzTB in a 1 4 1 CuAAC reaction. AzTB-ligated proteins were subsequently enriched with Neutravidin agarose 1 4 2 beads before preparing peptides for TMT labelling and quantification. Peptide samples were then  Of 129 protein hits identified, the specific probe-protein interaction profile was determined by 1 4 7 omitting any hits identified in DMSO-treated samples, excluded as non-specific binding shown in Figure 1A. To further confirm the specificity of the interaction between probe and protein 1 5 1 hits, samples treated with probe 2 in the presence of parent molecule 1, serving as a competitor, 1 5 2 were compared to samples treated with probe 2 alone. Compared in this way, out-competed 1 5 3 probe-specific interactions revealed 9 protein hits, depicted as negatively enriched proteins in  Table 2 and see also Extended SI and Figure S1B-C).

5 6
Among the 6 protein hits specific to probe 2 (Table 2), the protein most positively enriched when 1 5 7 comparing DMSO-treated and probe 2-treated samples was the gametocyte-specific 16 kDa  Plasmodium falciparum female gametocyte-specific protein Pfg377 (PlasmoDB ID, 1 6 0 PF3D7_1250100) 18 was also identified. As the N-4HCS compounds have been shown to solely 1 6 1 inhibit gametocyte viability but to exert no effect on other Plasmodium lifecycle stages 14 , Pfs16 and 1 6 2 DISCUSSION 4 0 2 As resistance inevitably threatens the long-term success of artemisinin and its derivatives in    targeting was localised to 42 AAs comprising the transmembrane domain (22 AA) and part of the 4 3 0 C-terminal tail (20 of the 31 AA). The protein membrane interaction was found to be stabilised with 4 3 1 11 C-terminal AA, which when removed reduced the level of retention but did not affect PVM 4 3 2 targeting. In contrast, amino acids between the N-terminal secretory signal sequence and 4 3 3 transmembrane domain were found to be crucial for capping involved in egress during region also shown to be conserved in other PVM proteins, is unlikely to be the target and therefore  gametogenesis to occur following a mosquito feed, the compound would need to last long enough 4 5 2 in circulation to be taken up by a feeding mosquito. This likely represents the key challenge for this drug uptake to the mosquito not the human host 46,47 . Such an approach may work well with the N-4 6 0 4HCS compound series.

6 1
A constant challenge with mode of action identification in Plasmodium sexual stages is that these 4 6 2 stages of the lifecycle do not replicate. This means that in vitro evolution and whole genome 4 6 3 analysis (IVIEWGA), a method in which parasites are continually exposed to antimalarial 4 6 4 compounds of interest to yield resistance before determining the genetic basis of resistance 48 , will 4 6 5 not work 48 . Thus, whilst this chemogenetic approach has been hugely valuable in mapping the 4 6 6 druggable genome it remains unattainable for studying targets of compounds which specifically of-concept study used CETSA coupled to mass spectrometry, with both a whole cell and cell lysate 4 7 0 approach, to validate targets of pyrimethamine and E64d. Having proven CETSA to be efficacious 4 7 1 and robust, the study went on to define to MoA of quinolone drugs, quinine and mefloquine 50 . and validation of a PVM cellular target from cellular phenotypic screen-derived hits. In summary, we have identified Pfs16, described as the earliest marker of sexual conversion, as a  .   (  +  )  -A  T  P  a  s  e  l  o  c  a  l  i  z  e  d  i  n  p  l  a  s  m  a  m  e  m  b  r  a  n  e  s  o  f  m  a  l  a  r  i  a  p  a  r  a  s  i  t  e  c  e  l  l  s  ,  5  4  5   P  l  a  s  m  o  d  i  u  m  f  a  l  c  i  p  a  r  u  m  ,  i  s  i  n  v  o  l  v  e  d  i  n  r  e  g  i  o  n  a  l  a  c  i  d  i  f  i  c  a  t  i  o  n  o  f  p  a  r  a  s  i  t  i  z  e  d  e  r  y  t  h  r  o  c  y  t  e  We thank Kathrin Witmer, Alisje Churchyard, Irene Garcia-Barbazan, Oriol Llorà-Batlle, Farah 6 2 9

REFERENCES AND NOTES
Dahalan, Eliana Real and David Grimson, for assisting with parasite culture and for sharing expert and Tate lab for assistance with this study. We thank George Ashdown (Baum lab) for assistance Howard and Henry Benns for assisting in the design of photoaffinity labelling experiments. We helpful discussions and experimental support. We also acknowledge Mathieu Brochet and 6 3 7 colleagues for providing assistance in the design of the flow cytometry-based assay for The authors declare no competing interests. The PlasmoDB database (https://plasmodb.org/) was used to analyse protein expression.   The ion target value for MS was set to 106 and for MS/MS to 2 x 105, and the intensity threshold  silico digests of the reference proteome, the following peptide bond cleavages were allowed: 8 2 3 arginine or lysine followed by any amino acid (a general setting referred to as Trypsin/P). Up to two 8 2 4 missed cleavages were allowed. The false discovery rate was set to 0.01 for peptides, proteins, and 20 ppm for precursor and product ions, respectively, minimum peptide length = 7, minimum 8 2 7 razor unique peptides = 2, minimum scores for unmodified and modified peptides = 0 and 40, 8 2 8 respectively). "Match between runs" option (time window 0.7 min) was allowed and "Unique and 8 2 9 razor peptides" mode was selected to allow identification and quantification of proteins in groups 8 3 0 (razor peptides are uniquely assigned to protein groups and not to individual proteins), and for TMT 8 3 1 quantification (MS2 mode) the minimal ratio count 2 was selected. were filtered to remove rows based on 'contaminants' and 'reverse' columns. The data was log 2  For in-gel fluorescence and western blot analysis (Figures 1C-F, S2, S3), live P. falciparum NF54 8 4 3 gametocytes were treated with either DMSO or probe 2 (2.5 µM, 5 µM or 10 µM) and irradiated and centrifuged (17,000 g, 4°C) in lysis buffer (1% Triton-X100/10mM Tris/150mM 8 4 7 NaCl/cOmplete™ ULTRA EDTA-free Protease Inhibitor Cocktail at pH 7.5 in H2O).

4 8
Protein concentration was determined using the Pierce™ BCA Protein Assay Kit following the manufacturer's instructions. Absorbance was measured using a NanoDrop™ 2000 spectrophotometer, using BSA as a protein standard. Lysed protein was made up to 0.5-1 mg/ml in PBS to a volume of 100 µl to perform the CuAAC reaction and protein precipitation, as described above. Samples of the clicked sample and crude lysate, with a total of 10 µg protein each, were set 8 5 3 aside to be analysed by in-gel fluorescence and western blot. immunoblots and gels which were analysed to obtain (D-F). Significance was determined by 1 0 2 0 performing an unpaired two-tailed t-test and is denoted as *** (p < 0.001), * (p < 0.05) and ns (p ≥ 1 0 2 1 0.05).  Protein hits identified as specific to the probe 2-protein interaction profile, including *Pfs16 and