PLoS ONE plos plosone PLOS ONE 1932-6203 Public Library of Science San Francisco, CA USA 10.1371/journal.pone.0270863 PONE-D-22-07594 Research Article Biology and life sciencesMolecular biologyMacromolecular structure analysisRNA structure Biology and life sciencesBiochemistryNucleic acidsRNARNA structure Biology and life sciencesParasitologyParasite groupsApicomplexaPlasmodium Biology and life sciencesMolecular biologyMolecular biology techniquesSequencing techniquesRNA sequencing Research and analysis methodsMolecular biology techniquesSequencing techniquesRNA sequencing Biology and life sciencesBiochemistryNucleic acidsRNANon-coding RNARibosomal RNA Biology and life sciencesBiochemistryRibosomesRibosomal RNA Biology and life sciencesCell biologyCellular structures and organellesRibosomesRibosomal RNA Biology and life sciencesBiochemistryNucleic acidsRNANon-coding RNATransfer RNA Biology and life sciencesBiochemistryNucleotides Biology and life sciencesComputational biologyGenome analysisTranscriptome analysis Biology and life sciencesGeneticsGenomicsGenome analysisTranscriptome analysis Biology and life sciencesGeneticsGene expressionProtein translation The in vivo RNA structurome of the malaria parasite Plasmodium falciparum, a protozoan with an A/U-rich transcriptome RNA structurome of the malaria parasite https://orcid.org/0000-0001-8790-9986 Dumetz Franck Data curation Formal analysis Investigation Methodology Visualization Writing – original draft 1 ¤ Enright Anton J. Data curation Formal analysis Visualization 1 Zhao Jieyu Investigation Writing – original draft 2 Kwok Chun Kit Conceptualization Funding acquisition Methodology Supervision Writing – review & editing 2 3 * https://orcid.org/0000-0001-7583-2176 Merrick Catherine J. Conceptualization Funding acquisition Project administration Supervision Writing – review & editing 1 * Department of Pathology, University of Cambridge, Cambridge, United Kingdom Department of Chemistry and State Key Laboratory of Marine Pollution, City University of Hong Kong, Kowloon Tong, Hong Kong SAR, China Shenzhen Research Institute of City University of Hong Kong, Shenzhen, China Frischknecht Friedrich Editor Heidelberg University: Ruprecht Karls Universitat Heidelberg, GERMANY

The authors have declared that no competing interests exist.

Current address: Institute for Genome Sciences, University of Maryland School of Medicine, Baltimore, Maryland, United States of America

 * E-mail: ckkwok42@cityu.edu.hk (CKK); cjm48@cam.ac.uk (CJM) 1 9 2022 2022 17 9 e0270863 14 3 2022 19 6 2022 2022 Dumetz et al This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Plasmodium falciparum, a protozoan parasite and causative agent of human malaria, has one of the most A/T-biased genomes sequenced to date. This may give the genome and the transcriptome unusual structural features. Recent progress in sequencing techniques has made it possible to study the secondary structures of RNA molecules at the transcriptomic level. Thus, in this study we produced the in vivo RNA structurome of a protozoan parasite with a highly A/U-biased transcriptome. We showed that it is possible to probe the secondary structures of P. falciparum RNA molecules in vivo using two different chemical probes, and obtained structures for more than half of all transcripts in the transcriptome. These showed greater stability (lower free energy) than the same structures modelled in silico, and structural features appeared to influence translation efficiency and RNA decay. Finally, we compared the P. falciparum RNA structurome with the predicted RNA structurome of an A/U-balanced species, P. knowlesi, finding a bias towards lower overall transcript stability and more hairpins and multi-stem loops in P. falciparum.

This unusual protozoan RNA structurome will provide a basis for similar studies in other protozoans and also in other unusual genomes.

 http://dx.doi.org/10.13039/501100000265 Medical Research Council MR/K000535/1 and MR/L008823/1 https://orcid.org/0000-0001-7583-2176 Merrick Catherine J. Royal Society of Great Britain and Ireland Kan Tong Po fellowship Kwok Chun Kit Shenzhen Basic Research Project JCYJ20180507181642811 Kwok Chun Kit Research Grants Council of the Hong Kong SAR China Projects CityU 11101519, CityU 11100218, N_CityU110/17, CityU 21302317 Kwok Chun Kit http://dx.doi.org/10.13039/501100001692 Croucher Foundation 9509003, 9500030 Kwok Chun Kit http://dx.doi.org/10.13039/100007567 City University of Hong Kong 6000711, 7005503, 9680261, 9667222 Kwok Chun Kit To CJM: UK Medical Research Council (grants MR/K000535/1 and MR/L008823/1) To CKK: Royal Society Kan Tong Po fellowship, Shenzhen Basic Research Project (JCYJ20180507181642811), Research Grants Council of the Hong Kong SAR China Projects (CityU 11101519, CityU 11100218, N_CityU110/17, CityU 21302317), Croucher Foundation (Project No. 9509003), State Key Laboratory of Marine Pollution Director Discretionary Fund, City University of Hong Kong (projects 7005503, 9680261, 9667222). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data Availability Raw sequence data (FASTQ) available from the European Nucleotide Archive (ENA) (accession number PRJEB44384). All other relevant data are within the paper and its Supporting information files. Introduction

RNA molecules, composed of single strands of four different ribonucleotides, do not adopt a single canonical structure like the double helix of DNA; instead they can fold into complex secondary structures made of loops, hairpins and bulges [1]. They serve a variety of roles in the cell [2, 3], including structural roles carried out by transfer and ribosomal RNAs (tRNAs, rRNAs) and protein coding roles for messenger RNAs (mRNAs). These roles are facilitated by the ability of RNA molecules to adopt, and transition between, highly specific secondary structures [4, 5]. Different secondary structures can impact on the location of the RNA molecules, their metabolism and stability [6], their interaction with RNA binding proteins [7], their function and their regulation of protein expression. Two outstanding examples of this are the function-related shape of tRNA [8] and the translation expression controlling element, the Riboswitch [9].

One of the common strategies to investigate RNA structures in vivo is Selective 2′-Hydroxyl Acylation analyzed by Primer Extension (SHAPE) [10]. SHAPE uses chemicals that modify the 2’ hydroxyl group of flexible nucleotides, in which the 2’ O-adduct will inhibit reverse transcription during cDNA synthesis [11]. The pattern of stalling is then detected on a sequencing gel, which determines the identity of all unpaired/flexible bases in the RNA, and hence allows its secondary structure to be inferred. The technicalities of SHAPE only permit the analysis of one transcript at the time, requiring a high amount of starting RNA and potentially more than one sequencing gel to analyse larger transcripts. All these technical limitations rendered impossible the transcriptome-wide study of RNA structures, which is referred to as a structurome [12]. However, the advent of Next Generation Sequencing (NGS) and bio-informatics revolutionised the field of RNA structure and within the past decade, Structure-seq [13], and many other related methods [14–16], were developed.

To date, only a few in vivo structuromes are available. Many viruses have been studied, probably due to their small number of RNA molecules, such as HIV [17, 18], DENV and ZIKV [19, 20], influenza [21] and more recently SARS-CoV-2 [22–27]. Regarding eukaryotes, structuromes were established for model organisms like the plants Arabidopsis thaliana [13] and Oryza sativa (rice) [28], the yeast Saccharomyces cerevisiae [16], the bacteria Escherichia coli [29] and amongst mammals, mouse embryonic stem cells [29, 30] and human cell lines [6, 16, 29, 31]. These have revealed fundamental conserved features across the landscape of RNA structures.

In this paper we have established the in vivo structurome of a protozoan, Plasmodium falciparum. P. falciparum is one of the etiological agents of malaria, the most widely lethal human parasitic disease in the world [32]. Little is known about RNA structures in P. falciparum, yet the parasites offer many opportunities to understand the roles of RNA structures in eukaryotic cells. Firstly, the natural life cycle of the parasite oscillates from a warm-blooded host to an insect vector, meaning that the same organism is exposed to very different temperatures and metabolic environments. This allows the study in natural conditions of the effect of the environment and temperature upon the RNA structurome. Indeed, it has been shown that heat stress can affect the structuromes of plants [33] and bacteria [34, 35], which can likewise experience temperature extremes in nature. Furthermore, one intriguing piece of evidence concerning rRNAs suggests that temperature-responsive RNA structures are also important in Plasmodium life cycles. Plasmodium does not encode rRNAs in conventional tandem arrays, but rather in several isolated units which are transcribed independently in either mosquito stages (S-type) or mammalian stages (A-type) [36, 37]. rRNA transcription apparently switches in response to differential temperature and nutrient availability in the two hosts [38, 39]. The two types of rRNA were predicted, based on their sequences, to have different structures [40] and they do perform differently when complemented into S. cerevisiae [41], so there is clear potential for them to influence translation in insects versus humans. Thus, Plasmodium may have evolved variant RNA structures as a key element in its life cycle transitions.

Secondly, while all the previously cited organisms have a more balanced ratio of A/T and G/C in their genomes, P. falciparum is 80.6% A/T-rich [42]: one of the most biased genome compositions ever to be sequenced. This offers a critical point of view when studying genome evolution, and it may well lead to unusual RNA structures.

This RNA structurome of P. falciparum sets out intends to investigate the relationship between RNA shapes and genome composition, as well as transcription efficiency, in the asexual erythrocytic cycle of this important human parasite.

 Material and methods Parasite culture

Plasmodium falciparum 3D7, obtained from the Malaria Research and Reference Reagent Resource Center (MR4), was cultured in O+ human erythrocytes at 4% haematocrit in RPMI 1640 supplemented with 0.25% albumax (Invitrogen), 5% heat-inactivated human serum and 0.25% sodium bicarbonate in gassed chambers at 1% O2, 3% CO2, and 96% N2 [43]. Parasite count was assessed on thin blood smears stained with Hemacolor (Merck).

 RNA structure probing and poly(A) enrichment

An asynchronous culture was grown up to 1.2x1010 parasites as previously described [43]. Erythrocytes were pelleted and then incubated for 30 minutes either with 10 mM DMS (dimethyl sulfate, Sigma) in ethanol or 20 mM NAI (2-methylnicotinic acid imidazolide) in anhydrous DMSO, at 37°C, agitated at 200 rpm, in the same gas mixture described above. The NAI was prepared as previously described [44]. The reaction was stopped with 40 mM dithiothreitol (DTT). Parasites were extracted by erythrocyte lysis using 1v of 0.2% saponin in PBS and then washed 3 times in ice-cold PBS. RNA was extracted using Qiagen RNeasy plus mini kit (Qiagen) following manufacturer’s instructions and then submitted to DNase treatment for 15 minutes (Qiagen). The total RNA recovered was split into samples of 5 μg each and enriched in poly(A) transcripts using NEBNext Poly(A) mRNA magnetic isolation module (NEB) according to manufacturer’s instructions.

 Visualisation of 5.8S rRNA structure using RT-stalling

RT stalling was performed using a primer targeting the 5.8S rRNA gene (PF3D7_0531800). Briefly, 1.5 μg of total RNA from DMS, NAI and DMSO-treated cultures in 5.5 μl nuclease-free water was mixed with 1 μl of 5 μM Pfa_5.8S primer labelled at the 5’ end with Cy5 [Cy5] ATTTTCTGTAGGAGTACCACT (Eurofins Genomics) in 3 μl of reverse transcription buffer (20 mM Tris, pH 7.5, 4 mM MgCl2, 1 mM DTT, 0.5 mM dNTPs, 150 mM LiCl in final). For the sequencing reactions (A/G/C/T), an extra 1 μl of 10 mM dideoxynucleoside triphosphate (Roche) was added to replace 1 μl of nuclease-free water. All samples were heated at 75°C for 3 min, 35°C for 5 min and then held at 50°C. Next, 0.5 μl of Superscript III (200U/μl) (Thermo Scientific) was added and reverse transcription was performed at 50°C for 15 min, followed by addition of 0.5 μl of 2M NaOH at 95°C for 10 min to degrade the RNA template. After reverse transcription, 10 μl of 2x formamide orange G dye (94% deionized formamide, 20 mM Tris pH 7.5, 20 mM EDTA, orange G dye) was added and heated at 95°C for 3 min before loading into a pre-heated (at 90W for 45 min) 8% denaturing polyacrylamide gel. The gel was run at constant power of 90 W for 90 min after loading 3.2 μl of each sample. Fujifilm FLA 9000 was used to scan the gel.

 Library preparation and sequencing details

Random fragmentation of polyA-enriched RNA was performed with 150 ng polyA-enriched RNA in fragmentation buffer (40 mM Tris-HCl pH 8.2, 100 mM LiCl, 30 mM MgCl2) at 95°C for 60s to generate average fragment size of ~250 nt, and then purified with RNA Clean & Concentrator (Zymo Research). 3’ dephosphorylation reactions included 7 μl fragmented sample, 1 μl of 10x PNK buffer (NEB), 1 μl of rSAP enzyme (NEB) and 1 μl of PNK enzyme (NEB), carried out at 37°C for 30 min. 3’ adapter ligation was performed by adding 3’ adapter (5’-/5rApp/NNNNNNAGATCGGAAGAGCACACGTCTG/3SpC3/-3’ with 1:5 molar ratio of RNA to 3’ adapter), PEG 8000 (17.5% final) and T4 RNA ligase 2 (NEB) in 1x T4 RNA ligase buffer at 25°C for 1 h. Excess adapter was digested by adding 1 μl RecJf (NEB) and 1 μl 5’deadenylase (NEB) at 30°C for 30 min and removed by RNA Clean & Concentrator (Zymo Research). Reverse transcription reactions including the ligated RNA above, reverse primer (5’-CAGACGTGTGCTCTTCCGATCT-3’ with 1:2.5 molar ratio of RNA to reverse primer), reverse transcription buffer (see RTS method above) and Superscript III (1U/μl final) (Thermo Scientific) were performed at 75°C for 3 min, 35°C for 5 min, then 50°C for 50 min. In this step, Superscript III should be added before the 50°C and after reverse transcription, NaOH (0.1 M final) was added at 95 °C for 10 min for RNA degradation, followed by mixing with 5 μl 1M Tris-HCl (pH 7.5) before RNA Clean & Concentrator (Zymo Research). The purified cDNA was then ligated with 5’ adapter (5’/5Phos/AGATCGGAAGAGCGTCGTGTAGCTCTTCCGATCTN10/3SpC3/-3’ with 1:20 molar ratio of RNA to 5’adapter) using Quick Ligation Kit (NEB) at 37°C overnight. Ligated reaction mixture was heated to 95°C for 10 min for inactivation and mixed with 1 volume of 2x formamide orange G dye before purifying with 10% denaturing urea-TBE acrylamide gel (Thermo Scientific) at 300V for 20 min. The size of 100–400 nt was sliced, crushed and soaked in 1x TEN 250 buffer (1x TE pH 7.4, 0.25 M NaCl) with incubation at 80°C for 30 min with 1300 rpm shaking, and then purified with RNA Clean & Concentrator (Zymo Research). The purified ssDNA was mixed with forward primer (5’-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT-3’, 0.5 μM final) and reverse primer with different indexes (5’-CAAGCAGAAGACGGCATACGAGAT-(6 nt index seq)-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT-3’, 0.5 μM final) for PCR amplification by using KAPA HiFi HotStart ReadyMix (KAPA Biosystems). The PCR program included 95°C for 3 min, 16 cycles of 3 steps (98°C: 20s, 68°C: 15s, 72°C: 40s), 72°C for 90s, then cooled to 4°C for size selection. PCR samples were resolved on a 1.8% TAE agarose gel at 120 V for 55 min. The size of 150–400 nt was cut and recovered by Zymo DNA agarose gel extraction kit (Zymo Research). DNA libraries were quantified, pooled and sequenced on the Illumina Hiseq System in 150 bp paired-end (PE) configuration. A more detail protocol was described in reference [45].

 Structurome analysis

Forward and reverse reads from each sample were cleaned of adapter sequence contamination using reaper (v15-065) from the kraken package [46]. Paired reads were then aligned to an indexed reference transcriptome (PlasmoDB-45_Pfalciparum3D7) using Bowtie2 (v2.4.2) in paired-end mode and converted to SAM and sorted BAM files [47]. Reverse transcription stops were assessed using StructureFold2 for each individual sample [48]. Coverage and overlap of RT stop data were then computed along with cross-replicate coverage and overlap and stop correlation was assessed among all samples via StructureFold2. These data were normalised and structural reactivity data generated for each transcript with enough supporting data across samples using StructureFold2. These reactivities were supplied to RNAStructure for experimental reactivity constrained folding [49]. Resulting structures were then assessed for structural features using Forgi from the ViennaRNA package [50]. Unconstrained folding of every transcript was also performed using RNAStructure with the same parameters and assessed similarly to generate a cohort of unconstrained purely in silico folded transcripts. Parameters for the number of reads and mapping efficiency for each library are shown (Table 1).

10.1371/journal.pone.0270863.t001 Sequencing parameters for each NGS library generated in this study.
Sample Number of Read Pairs Yield (Gigabases) Mapped Reads Mapped Reads (%)
Total-RNA-DMS-3 148,865,508 44.6597 108,912,838 73.1%
Total-RNA-DMSO-3 163,681,632 49.1045 115,215,035 70.3%
Total-RNA-NAI-3 192,424,134 57.7272 144,128,728 74.9%
Total-RNA-DMS-2 166,742,223 50.0227 117,106,285 70.2%
Total-RNA-DMSO-2 186,503,272 55.951 128,887,798 69.1%
Total-RNA-NAI-2 145,601,085 43.6803 108,451,225 74.5%
Functional dataset analysis

Microarray datasets from parasites exposed to hyperoxia were extracted from [51], and the chloroquine-exposed dataset was extracted from [52]. Lists of genes were updated to the current genome annotation. Ribosome profiling data were obtained from [53] and translational efficiency was calculated as ribosome density per messenger RNA. The RNA decay dataset was extracted from [54] at 10hpi for rings, 24hpi for early trophozoites, 30hpi for late trophozoites and 46hpi for schizonts. Structural parameters per transcript (numbers of nucleotides involved in stems, hairpins, multi-stem loop and bulges) were normalized to transcript length before comparing these structural parameters with the translational efficiency of each gene.

 Results Assessment of NAI and DMS as RNA structure probing chemicals for <italic>P</italic>. <italic>falciparum</italic> inside the erythrocyte host cell

We conducted in vivo structure-seq on Plasmodium falciparum: an organism with a >80% AT-rich genome, and one that lives for the majority of its lifecycle inside another host cell. It was necessary first to test and optimise the parameters for in vivo RNA probing in this unusual organism. We used two different probing agents with different nucleotide reactivity: NAI (2-methylnicotinic acid imidazolide) and DMS (dimethyl sulfate). NAI reacts with all four unpaired nucleotides while DMS selectively reacts with unpaired adenosine and cytosine. Neither of these chemicals had been used in P. falciparum before. In order to check if the chemicals could penetrate inside the parasite inside the host erythrocyte, we exposed an asynchronous culture of P. falciparum to NAI or DMS. After RNA extraction we performed a primer extension with a Cy5-conjugated primer targeting the 5.8S rRNA transcript. When a nucleotide has been modified by one of the chemicals, meaning that the nucleotide was unpaired in the original RNA, the reverse transcriptase will stall one nucleotide beforehand, stopping the elongation reaction (Fig 1A). This can be seen on a denaturing polyacrylamide sequencing gel (Fig 1B, S1 Raw images). Reading the sequence confirmed that it was identical to the PF3D7_0531800 gene sequence available on PlasmoDB [55]. In NAI and DMS probed parasites, base reactivity was seen with all four nucleotides after NAI-probing, and with only adenosine and cytosine after DMS-probing. (Modification of adenosine and cytosine was not, however, identical at all positions with DMS and NAI, because these chemical probes react with different moieties of the RNA nucleotides, and with different efficiencies.) This indicated that both chemicals could penetrate into P. falciparum, and our probing strategy could investigate the RNA structures in P. falciparum.

10.1371/journal.pone.0270863.g001 <italic>In vivo</italic> probing of RNA structures in <italic>P</italic>. <italic>falciparum</italic>.

(A) Schematic of RNA structure probing. An asynchronous parasite culture was split into two, one exposed to NAI (+NAI) and the other exposed to DMSO as control (-NAI). Reverse transcription (RT) was performed using a 5’-Cy5 conjugated primer (dark red star) specific for the 5.8S rRNA of P. falciparum (in green). During RT, nucleotides that have been modified by NAI in the +NAI samples (red dots) will stall the reverse transcription, interrupting the elongation process (black line). Finally, elongation cDNA products are resolved on a PAGE gel. (B) Sequencing gel showing the 5.8S rRNA gene of P. falciparum. Lanes 1 to 4 are sequencing lanes where one of the four nucleotides in the reaction mix has been switched with the dideoxynucleotide, provoking RT-termination. Lanes 5 and 6 represent respectively DMS and NAI modifications on 5.8S rRNA.

To prepare the structurome, we then extracted the total RNA from two independent asynchronous parasite cultures treated with NAI, DMS, or DMSO only, as in Fig 1, and used poly(T) magnetic beads to enrich poly(A) RNA. The enriched poly(A) fraction represented 2–3.4% of the total RNA, which is commensurate with the literature for eukaryotic cells [56]. The enriched fractions were then used to generate NGS libraries (Fig 2A). The sequencing outputs were processed using StructureFold2 and RNAStructure for experimental reactivity constrained folding [57] to determine the secondary structure of the RNA molecules. Each replicate for each chemical treatment was analysed separately. After preliminary analysis, because the distribution of structure differences between replicates was marginal and the correlation between replicates and pooled samples was good (S1 Fig), reads from both replicates were merged for subsequent analysis. To ensure structure calls of a high quality, mappings were filtered such that there were no more than 3 mismatches/indels from the reference and the first 5’ base was matching. Coverage filtering also required transcripts with complete coverage between samples (i.e. meeting the StructureFold2 default filtering threshold of an average coverage >1.0 per nucleotide throughout the transcript (S2 Fig). Structural information was more comprehensive (and agreement between replicates was also better), for the NAI-probed datasets because all four unpaired nucleotides were detected. Subsequent analysis was therefore focused on the NAI dataset as the primary in vivo structurome.

10.1371/journal.pone.0270863.g002 RNA structurome of <italic>P</italic>. <italic>falciparum</italic> using Structure-seq.

(A) Schematic showing the Structure-seq pipeline. An asynchronous parasite culture was split into two, one exposed to NAI (+NAI) and the other exposed to DMSO as control (-NAI). Polyadenylated-enriched RNA fractions were then used to prepare the sequencing libraries in parallel. Fragmentation was performed to generate an average length (~250 nt) of RNA for sequencing followed by 3’ dephosphorylation, which replaces the 2’3’ cyclic phosphate group with a 3’-OH group for 3’ adapter ligation. Then a 3’ adapter (black line) was ligated to the 3’-OH group of dephosphorylated RNA. Next, RT of RNA molecules (orange line) was carried out using a designed reverse primer (dark blue line) at the 5’ end. During RT, nucleotides that had been modified by NAI in the +NAI samples (red dots) would stall the reverse transcription, interrupting the elongation process (pink line). Later, a 5’ adaptor (light blue line) was added by ssDNA ligation at the 3’ end of the newly synthesised cDNA strand (pink line). Finally, forward primers (brown+light blue line) and reverse primers (dark blue+red+green line) with different indexes were added by PCR for NGS and bioinformatic analysis. Reads were aligned to the PF3D7 reference genome and subsequently analysed with StructureFold2 to determine the structural reactivity sequence (transcribed as a dot/bracket sequence denoting paired/unpaired nucleotides), and with RNAStructure to predict the folding. (B) Representation of the diversity of the dataset regarding RNA families. (C) Structure-seq derived structures of glutamic acid (PF3D7_0411600), methionine (PF3D7_1339100), asparagine (PF3D7_0714700), alanine (PF3D7_0620800) and proline (PF3D7_1339200) tRNAs, all of which are similar to canonical tRNA structures, and of serine (PF3D7_041000) tRNA, which is non-canonical.

 Structure of the NAI-probed dataset

After filtering for low-coverage genes, the NAI structurome covered 51.23% of the total transcriptome of P. falciparum (2699 transcripts detected out of 5268 total). We first looked at the representation of the various families of RNA molecules. Within our dataset 95.5% of the transcripts were protein coding transcripts, 1.27% corresponded to mitochondrial transcripts (1.16% mitochondrial rRNA, 0.11% mitochondrial mRNA) and 0.94% were uncharacterised non-protein coding transcripts. Other types of non-coding RNA were also detected, 1.31% of snoRNA, 0.19% of snRNA and 0.34% of tRNA (Fig 2B). Within the protein coding transcripts, nine were encoded in the apicoplast, a subcellular organelle that is unique to apicomplexans and represents a relic chloroplast (S1 Table).

As well as the different families of transcripts, we detected 114 transcripts with splice variants (S2 Table). However, since we used a short-read technology to sequence the library, it was impossible to associate all the reads within a variant transcript to one variant. A second limitation of short-read sequencing was that it resulted in very limited coverage of untranslated regions, which in P. falciparum tend to be extremely A/U-biased (often over 90%), and also highly repetitive with many low-complexity regions. This precluded any comprehensive assessment of RNA structure in the UTRs versus CDSs of mRNAs. Nevertheless, this overall analysis indicates that our in vivo chemical probing, RNA extraction and polyA selection were robust because all the different RNA families in P. falciparum were detected, with a strong bias towards mRNAs, and if a long-read sequencing technique had been used, even the detection of transcript variants would likely have been possible.

Some RNA families have their function dictated by their structure. This is especially the case of tRNA. Within our dataset 9 tRNAs were identified. Using RNAStructure we were able to call their structures and 5 were similar to canonical eukaryotic tRNA molecules. Those tRNAs were for glutamic acid (PF3D7_0411600, PF3D7_0527700), methionine (PF3D7_1339100), asparagine (PF3D7_0714700), alanine (PF3D7_0620800) and proline (PF3D7_1339200) (Fig 2C). In the other 3 tRNAs, for valine (PF3D7_0312600), serine (PF3D7_0410100) and leucine (PF3D7_0620900), the lowest-free-energy structures predicted by RNAStructure were not canonical (S3 Fig). For some but not all of these, alternative folding with less favourable energy produced canonical structures, albeit less compatible with the data defining free/flexible bases (e.g. S3A Fig). It is theoretically possible that dynamic structural variants of some tRNAs exist in vivo, although it is unclear why some tRNAs would differ from others in the dataset. Importantly, however, tRNA structure is also dependent on nucleotide base modification such as m2A, m2G, pseudouridine, etc. [46]. Those modifications have not been assessed in this study, but it is possible that a) they would influence the apparently-noncanonical tRNA structures and b) they might cause extra RT-stalling and thus appear erroneously as ‘probed’ bases, impeding the determination of the true structures of hyper-modified RNAs such as tRNAs.

In general, opportunities to ‘verify’ the structurome were limited because of the lack of Plasmodium RNA structures obtained independently, by orthogonal experimental methods. The canonical structural model of a blood-stage P. falciparum rRNA is available from the Comparative RNA Web (CRW) site and this was used to map base reactivities obtained in vivo (S4 Fig). The result demonstrates visually the A/C reactivity of DMS versus the all-base reactivity of NAI, and makes it clear that many bases modelled as unpaired are nevertheless not highly reactive in vivo. Other factors such as protein binding and base modification probably affect their reactivity, emphasising the over-simplification imposed by in silico modelling and the extra level of information obtained by in vivo structural probing.

 Divergence between <italic>in vivo</italic> RNA structures and structures modelled <italic>in silico</italic>

Using the RNAStructure algorithm, RNA structures can be predicted from primary sequence alone, or they can be predicted with the addition of base-pairing information gained experimentally. We generated both types of structure, transcriptome-wide, taking only the lowest-free-energy structure in all cases where the algorithm generated several alternatives. This was simply because, in the absence of other constraints, the lowest-free-energy structure is the most likely to form.

Firstly, we sought to discover how much these two types of structures would differ in P. falciparum, i.e. how much new information was gained by performing the in vivo structurome? We calculated the ratio of bases that differed (i.e. that were paired versus unpaired) between each in-silico-predicted and probing-informed structure, and termed this ratio the ‘divergence’. Most of the non-protein coding transcripts had relatively low divergence, meaning that they were minimally informed by probing (Table 2). This was expected for highly conserved, stereotypical structures like rRNAs and tRNAs—although the divergence for mitochondrial rRNAs, which are encoded in a highly fragmentary fashion in the P. falciparum mitochondrial genome, was notably greater than for nuclear-encoded rRNAs. By contrast, the divergence of protein coding transcripts was much higher: a median of 0.7067 for the nuclear-encoded mRNAs and 0.8117 for mitochondrial mRNAs (Table 2).

10.1371/journal.pone.0270863.t002 Distribution of ‘divergence’ normalised by transcript length between base pairing determined by Structure-seq and <italic>in silico</italic>.
Protein coding mito_mRNA Uncharacterized non protein coding snoRNA tRNA rRNA snRNA mito_ rRNA
# transcript 2549 3 25 35 9 8 5 36
Median 0.71 0.81 0.53 0.44 0.17 0.22 0.46 0.54
Minimum 0.04 0.75 0.07 0.08 0.06 0.02 0.19 0.00
25% Percentile 0.60 0.75 0.34 0.26 0.10 0.06 0.19 0.35
75% Percentile 0.79 0.89 0.69 0.63 0.60 0.35 0.56 0.74
Maximum 1.03 0.89 0.91 0.92 0.73 0.47 0.66 1.05

Data are categorised by family of RNA.

Since structure and function in non-protein coding transcripts, like tRNAs, are deeply intertwined, we then sought any relationship between folding complexity and biological function. Using the ‘divergence’ as a metric, we split the protein coding transcripts into 3 tiers: minimal value to 25th percentile, 25th to 75th percentile and 75th percentile to the highest value (Table 2). The list of transcripts was then submitted for gene ontology enrichment and using Revigo [58] we obtained broad families of GO terms and networks of terms belonging to a similar biological pathway. In tier 1, with the lowest ‘divergence’, we observed an enrichment of 21 terms, with the strongest enrichment in terms related to ribosomal functions (see GO term 7, ‘translation’, in Fig 3A, S3 Table). This was expected, since ribosomal RNA structures are highly conserved and are likely to be predicted accurately in silico with little alteration in P. falciparum. Twenty-one terms were found in tier 2. Two terms were again strongly enriched: ‘translation’ and ‘cellular amide metabolism’, and these fell into a loosely-associated collection of similar terms around ‘protein’ and ‘protein regulation’. There was a second cluster restricted to two terms around RNA splicing, and a third cluster around mitochondrian/organellar organisation (Fig 3B, S3 Table). In tier 3, with the lowest level of identity between in vivo structures and in silico predictions, only seven terms were enriched and there was only one cluster, composed of four terms strongly enriched around different kinds of amino acid metabolism (Fig 3C, S3 Table).

10.1371/journal.pone.0270863.g003 GO terms enrichment scatterplot for protein coding transcripts, tier 1 to 3.

All plots represent the GO enrichment provided by REVIGO for (A) tier 1, (B) tier 2 and (C) tier 3. Detailed values are available in S3 Table. Each GO term is represented by a numbered circle on the plot, with placement indicating its similarity to nearby GO terms. The colour scale represents the adjusted p-value of the GO term and the size of each circle represents the log10 of the number of annotations per GO term.

Secondly, we assessed the value of performing Structure-seq instead of just in silico prediction for modelling the shape of RNA molecules. We used ViennaFold on the P. falciparum transcriptome to fold the structures with or without reactivity data. We first aligned the dot/bracket sequences obtained from in silico folding with the Structure-seq-determined sequence of the U2 snRNA (PF3D7_1137000). It showed a difference in the pairing of 74 nucleotides out of 198, meaning a 37.4% discrepancy in the structure determined in silico compared to in vivo (Fig 4A). Structurally, the difference between the two transcripts is major and only 3 loops are common to both structures (Fig 4B). More interestingly, an RNA-appropriate measurement of Minimum Free Energy, MFEden [59], revealed that the Structure-seq U2 snRNA structure had a lower MFEden (-24.76 versus 15.50). Extending this to the whole dataset, we observed a similar effect in each RNA category (Table 3). Thus, structures called from Structure-seq are more likely to form and are more stable than the ones predicted only using an in silico prediction.

10.1371/journal.pone.0270863.g004 Difference in U2 snRNA structure determined by Structure-seq versus <italic>in silico</italic> prediction.

(A) U2 snRNA (PF3D7_1137000) reactivity sequence alignment, comparing in vitro probing using NAI and in silico prediction using the same constraint parameters. The first line shows the nucleotide sequence, the second line shows the in silico predicted dot/bracket sequence and the third line, the Structure-seq dot/bracket sequence extracted from reactivity measurement. A dot shows a nucleotide not paired with any other nucleotide. An open bracket shows a nucleotide paired with another and starting a structural motif, while a closed bracket shows a nucleotide paired with another but finishing a structural motif. Red-highlighted dot or bracket indicates a change of pairing between the NAI and in silico predicted structural sequence. A green-highlighted bracket indicates a change in motif orientation between the NAI and in silico predicted structural sequence. Boxes with roman numerals indicate the conserved structures between the two shapes. (B) U2 snRNA structure predicted in silico from RNA sequence alone, on the left, and with the addition of reactivity data on the right. The red and green colouring have the same meaning as in (A).

 10.1371/journal.pone.0270863.t003 MFEden average for every type of RNA molecule.
protein coding mito_mRNA uncharacterised non protein coding snoRNA tRNA rRNA snRNA mito_rRNA
Structure-seq -20.5 -27.0 -29.8 -29.6 -63 -31.1 -37.9 -25
in silico 14.1 10.3 10.7 10.2 -8.8 3.2 10.7 2.7

All MFEden values per transcript are in S4 Table.

 Effect of A/U bias in the structurome of <italic>P</italic>. <italic>falciparum</italic> by comparison with an A/U-balanced <italic>Plasmodium</italic> species

The Plasmodium genus offers the possibility to compare very different genome compositions. The genome of P. falciparum is ~80% A/T rich while P. knowlesi, a macaque Plasmodium species which also infects humans, is 62.5% A/T rich [60]. In order to assess whether the base composition of the genome, and hence the transcriptome, could affect the structurome, we compared the in silico folding of P. falciparum and P. knowlesi structuromes. Firstly, we assessed the stability of the two predicted structuromes by comparing the average free energy (MFEden [59]) of all folded structures: P. falciparum had a significantly higher MFEden than P. knowlesi (respectively 14.03 and 7.1, t-test p-value <0.0001). Secondly, we looked at the occurrence of various structures by comparing the length-normalised number of nucleotides participating in a structure (e.g. a hairpin or bulge) in each P. falciparum transcript with their P. knowlesi orthologs. The structurome of P. falciparum had significantly more nucleotides involved in hairpins and multi-stem loops than the structurome of P. knowlesi (t-test, p-value <0.0001), whereas P. knowlesi had significantly more nucleotides involved in stems and bulges (t-test, p-value <0.0001) (S5 Table).

 RNA structures affect translation efficiency in a stage specific manner

Fig 2 shows that various families of RNAs display very conserved structures that are related to their particular functions. This is clearly true for structural RNAs (tRNAs, rRNAs, etc.), but it may also be true for mRNAs, whose main function is to be translated into proteins. The stability of the transcript and the efficiency of its translation could both be affected by mRNA structure. Therefore, if mRNA structures can vary with varying cellular conditions, transcript and protein abundance may both vary accordingly.

We first explored the concept that transcript abundance could vary with cellular conditions due to changes in mRNA structure. For this, we used published microarray and transcriptomic datasets from P. falciparum under physiological stresses like hyperoxia [51], or drug exposure like chloroquine treatment [52]. These datasets reveal which transcripts are upregulated and downregulated in each condition. We compared the most differentially expressed genes in each condition with the structurome dataset, to see if they were particularly structured or unstructured. Neither the overall degree of ‘divergence’ nor the representation of diverse structures (the proportion of hairpins, loops, etc), correlated with differential expression under stress (S6 Table).

Secondly, we looked at the effect of RNA structures on translation efficiency (TE). To establish the translation efficiency for each transcript in our dataset, we extracted from previously published work [53] the average level of ribosome attachment per transcript, determined by Ribo-seq. We obtained TE for the transcriptome of rings, early and late trophozoites and schizonts. Applying a linear regression model, we compared TE with transcript structuredness (‘NAI ratio’, i.e. the ratio of paired to unpaired bases) and also the proportion of different motifs such as stems, hairpins, multi-loop stems and bulges (Fig 5). We observed no correlation between the transcript-length-normalised ‘structuredness’ and TE in any of the four stages (Fig 5A). Therefore, there was no overall relationship between the proportion of paired bases in a transcript and the efficiency of its translation. However, breaking down the different possible structures (Fig 5B–5E) we observed a positive correlation of TE with the presence of stems and a negative correlation between TE and hairpins, multi-loop stems and bulges. The correlations were only seen in the more mature parasite stages, late trophozoites and schizonts.

10.1371/journal.pone.0270863.g005 Impact of various RNA motifs on transcription efficiency across the erythrocytic cycle of <italic>P</italic>. <italic>falciparum</italic>.

The columns represent the different erythrocytic stages of P. falciparum and the rows represent the correlation between transcription efficiency and (A) the transcript structuredness normalised by the transcript length, (B) the number of stems, (C) the number of hairpins, (D) the number of multi-stem loops and (E) the number of bulges. Statistically significant positive correlations have a p-value framed in green and statistically significant negative correlations have a p-value framed in red.

 RNA decay is mediated by the same structures across the cell cycle

The final critical part of an RNA molecule’s life inside the cell is its rate of decay. Following a similar reasoning as for translation efficiency, we extracted RNA decay values from data published by Painter et al. [54] and tested for any correlation between RNA decay and RNA secondary structures in the different cell cycle stages (rings, early and late trophozoites and schizonts). For this, we applied a linear regression model to compare RNA decay (the higher the value, the less stable the molecule is) to the number of different structures per transcript. We observed no correlation between RNA decay and RNA secondary structures involving high amounts of base pairing: stems and hairpins (Fig 6A and 6B). However, when nucleotides were more exposed, as in bulges and multi-stem loops (Fig 6C and 6D) we observed correlations between RNA decay and the presence of those structures in all stages (except for multi-stem loops in ring-stage) (Fig 6C).

10.1371/journal.pone.0270863.g006 Correlation between RNA secondary structures and RNA decay across the erythocytic cycle of <italic>P</italic>. <italic>falciparum</italic>.

The columns represent the different erythrocytic stages of P. falciparum and the rows represent the correlation between RNA decay and (A) the number of stems, (B) the number of hairpins, (C) the number of multi-stem loops and (D) the number of bulges. Statistically significant negative correlations have a p-value framed in red.

 Discussion

In this study, we produced the SHAPE-seq-determined structurome of a parasitic protozoan, P. falciparum. This had never previously been performed for any protozoan parasite, although a second study, appearing after the submission of this manuscript, has also now described an in vivo P. falciparum structurome [61]. This study covered only 23–35% of the transcriptome (in trophozoites and rings respectively), compared to >51% here, and had approximately 35% fewer reads per sample with markedly lower mappability of reads. Our study therefore remains the most comprehensive structurome in P. falciparum to date. A third study, equally recently published, mapped duplex RNA regions by RNA-seq after degradation of ssRNA [62]; however this does not produce data directly comparable to an NAI-probed in vivo structurome.

Firstly, we demonstrated that it is possible to use probing chemicals on the parasite within the host erythrocyte, meaning that the chemicals can get through the host cell membrane, the parasitophorous vacuole and the parasite membrane. The fact that we could also detect mitochondrial and apicoplast transcripts shows that the chemicals also passed through the inner organellar membranes. Thus, we obtained in vivo structures for about half the transcripts in the transcriptome, including structural and noncoding RNAs as well as mRNAs. Most of the structural RNAs had stereotypical structures, although a subset of tRNAs did not model stereotypically—an interesting observation that merits further investigation. Amongst mRNAs, the in vivo structures tended to have lower free energy, making them more likely to form, than the equivalent structures predicted in silico, thus demonstrating the value of the experimental approach. Highly conserved RNAs such as ribosomal RNAs tended to be predicted well, whereas classes of genes whose structures were poorly predicted in silico included genes involved in metabolism, biosynthesis of complex molecules, and transport. These genes may have become particularly divergent in P. falciparum due to its parasitic intracellular lifestyle.

We also took advantage of the striking disparities in genome composition in the Plasmodium genus to see if the extreme base composition of the P. falciparum transcriptome might have an impact on the different structures formed, by comparing the predicted structurome of P. knowlesi with that of P. falciparum. We observed that in P. falciparum transcripts were more prone to form structures like hairpins and multi-stem loops, whereas a more balanced genome like P. knowlesi gave rise to orthologous transcripts with more stems. Overall, P. falciparum transcripts were predicted to fold with higher free energy than their orthologs in P. knowlesi, suggesting that an A/U-biased transcriptome is less stable. However, this might be mitigated by in vivo factors, since we also found that overall the P. falciparum in vivo structurome had lower free energy than was predicted in silico. It would be interesting in future to perform an in vivo structurome of P. knowlesi and see whether in vivo transcripts tend towards greater stability in this species as well.

To examine how the RNA structurome might influence parasite biology, we looked at the impacts of the different structures on translation efficiency and RNA decay. We showed that RNA structures had no impact on translation efficiency in rings and early trophozoites, two stages of the erythrocytic cycle that are translationally less active [53, 63]. On the contrary, late trophozoites and schizonts are more translationally active and negative correlations between translation efficiency and complex structures like hairpins, multi-loop stem and bulges were observed. Recently, we studied a particularly stable RNA structure in P. falciparum, the RNA G-quadruplex, and demonstrated that it likewise had a negative impact on translation [64]. We also observed a positive correlation between translation efficiency and a simpler RNA structure, stems. This is consistent with the idea that stem structures can facilitate ribosome attachment, whereas other complex structures can inhibit it [65].

The correlations described above were not detected in either of the parallel P. falciparum ‘structurome’ studies. A study comparing duplex RNA regions with translation efficiency failed to observe any clear correlations [62], probably because it did not distinguish the subtleties of different types of RNA secondary structure: this highlights the power of the structurome approach taken here. Meanwhile, the comparable study by Wang and co-workers [61] reported that the least-structured transcripts had the highest abundance, but since abundance was derived only from a snapshot RNA-seq measurement, it was not possible to attribute this specifically to transcript half-life, decay rate, etc. as individual factors that might be affected by RNA structure. Our study examined two such factors individually—translation efficiency and also RNA decay rate—and therefore generated different insights.

There was a negative correlation between structures and RNA decay, so we can hypothesise that RNA structures influence protein expression on at least two levels: the rate of translation and the degree of transcript stability. Similar effects were previously demonstrated in human cell lines, and furthermore RNA base modifications were also found to be important, impacting protein expression by changing the functional half-life of transcripts [66]. This latter area has been little explored in P. falciparum as yet, but a recent study of base modification in P. falciparum likewise implicated RNA base modification in protein expression [67].

This study focuses on RNA structure in the erythrocytic life cycle of Plasmodium parasites. Two more parts of the full life cycle remain to be studied in future: the liver stage and the sexual cycle which happens in the host vector, a mosquito. The liver stages are particularly interesting because two human-infective Plasmodium species, P. vivax and P. ovale, can make hypnozoites—a dormant form that can stay inside the liver of infected people for months and then reactivate to cause disease [68]. Currently the biology of hypnozoites remains complicated to investigate. However, translational repression may be key to long-term dormancy, so the RNA structurome of hypnozoites could be intriguing. A more accessible stage, the gametocyte, also shows dormancy (albeit a dormancy that lasts days rather than months), and female gametocytes are known to translationally repress hundreds of transcripts [69]. Once in the mosquito midgut, the translational repression is lifted, and we can hypothesise that the change of temperature between a warm-blooded host and a non-thermoregulatory insect would impact RNA structures, playing a role in lifting the repression.

Within the mosquito, the sexual life cycle encompasses four different life stages for which no RNA structures are known. This step of the life cycle is crucially important in completion of the transmission chain and also in generating genetic diversity [70]. Recent progress in single cell RNA sequencing give us the opportunity to look at the transcriptome of single parasites at all life cycle stages, including those in the host vector [71]. In parallel, novel RNA structure determination techniques, like SHAPE-MaP-seq associated with the latest algorithms like DRACO, would make the determination of single cell RNA structures possible [72]. Other technological breakthroughs like SMRT-seq (Single Molecule, Real Time) associated with SHAPE-MaP would allow access to transcriptional variant structures [73], which were not possible to analyse in this study, and this could shed new light on post-transcriptional regulation in Plasmodium. Finally, the key roles of RNA-binding proteins in determining RNA structures in vivo have yet to be examined. Broadly, Wang and co-workers found that de-proteinised RNAs that had been refolded in vitro were less structured, and/or protein-bound, than the same RNAs in vivo, which is not unexpected [61]. However, more specific future experiments could potentially knock out single RNA binding proteins and then compare the resultant RNA structures in vivo. This would shed more detailed light on the determinants of the P. falciparum RNA structurome.

Overall, this work paves the way to decipher another level of complexity in the molecular biology of Plasmodium parasites.

 Supporting information Polyacrylamide gels showing the reverse-transcribed 5.8S rRNA gene of <italic>P</italic>. <italic>falciparum</italic>.

(A) results after probing with DMS (2 concentrations), NAI, vehicle control, i.e. DMSO, or no treatment (UT). DMS and NAI modifications are detected as reverse-transcriptase stops, with very similar patterns at both DMS concentrations. The vehicle control confirms that there is negligible ‘background’, i.e. almost no reverse-transcriptase stops, yielding a near-empty lane similar to the untreated control. (B) full-width gel that appears cropped (as highlighted in red) in Fig 1B, visualised with Fujifilm FLA 9000.

(PDF)

 DMS and NAI samples: Correlation and pooling data.

(A) For each of the different treatments (both individual replicates and pooled), the distribution of ‘RNAdistance’ is plotted, i.e. the computed distances between the purely computationally-determined structures and those guided by reactivity data. The y-axis represents the frequency of RNAdistance metrics, which are represented on the x-axis. Replicates of each treatment appear similar, as quantified in (B). (RNAdistance calculated as per: R. Lorenz, S.H. Bernhart, C. Hoener zu Siederdissen, H. Tafer, C. Flamm, P.F. Stadler and I.L. Hofacker (2011), "ViennaRNA Package 2.0", Algorithms for Molecular Biology: 6:26). B) Pearson correlation heat map showing how the RNAdistance-derived distributions in panel A correlate across samples. Two distinct groupings are clearly formed for NAI and DMS.

(TIFF)

 Plots showing the coverage of each transcript (averaged per nucleotide) in each replicate of each treatment condition.

More than 50% of transcripts in every condition met the coverage threshold of >1.0 per nucleotide.

(TIF)

 <italic>RNAStructure</italic> folding of three non-canonical-shaped tRNAs.

(A) valine tRNA (PF3D7_0312600), (B) serine tRNA (PF3D7_0410100) and (C) leucine (PF3D7_0620900). The structures in (A) and (C) could also be folded into a less-energetically favourable canonical shape: to exemplify this, NAI-reactive bases in (A) are marked on both structures (red, orange, green from highest to lowest relative reactivity). The structure in B could not be folded into a canonical shape, even at 50% less favourable energy, when using the constraints obtained by NAI probing.

(TIFF)

 <italic>P</italic>. <italic>falciparum</italic> 18S rRNA structure.

The canonical structure of the P. falciparum 18S rRNA (from https://crw-site.chemistry.gatech.edu/) was compared with base reactivity data for the assembled sequence of the PF3D7_0725600 gene (encoding blood-stage-expressed 18S rRNA). Maximum base reactivities for both the NAI and DMS datasets were mapped: NAI reactivities in the blue channel and DMS reactivities in the red channel, hence dually-reactive bases appear pink. Reactivities were scaled to colour intensity.

(TIFF)

 Complete and detailed list of transcripts found during the Structure-seq analysis.

The file contains all the genomic information, product description, gene type, transcript length, GC and AT content, reactivity for each replicated of DMS and NAI experiment plus the combined one, and the gene ontology information.

(XLSX)

 Table of transcript variants.

Same information available as in S1 Table.

(XLSX)

 REVIGO output for clustering of GO terms.

Each tab represents the data per tier of divergence.

(XLSX)

 Table of free energy and transcript structures from Structure-seq and <italic>in silico</italic> analysis.

(XLSX)

 List of the <italic>P</italic>. <italic>falciparum</italic> transcripts and their <italic>P</italic>. <italic>knowlesi</italic> orthologs with <italic>in silico</italic> folding information normalised by transcript length.

(XLSX)

 Table of raw data from 2 previous studies on perturbation of the transcriptome via environmental conditions.

Sheet 1 and 2, respectively named 4h_hyperoxia and 8h_hyperoxia, are the raw microarray data from Torrentino-Madamet et al. [51]. Sheet 3, named CQ_dataset, is the list of differentially expressed genes after long chloroquine treatment from Untaroiu et al. [52]. Original data have a column header in red, all column headers in black were added for the purpose of this study. Rows 1 and 2 show the average value for each category for up- and down- regulated genes.

(XLSX)

We would like to thank Dr. Betty Chung from the Cambridge Pathology Department for her help with the transcription efficiency data processing.

 References Mathews DH, Sabina J, Zuker M, Turner DH. Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol. 1999;288: 911–40. doi: 10.1006/jmbi.1999.2700 10329189 Sharp PA. The Centrality of RNA. Cell. 2009;136: 577–580. doi: 10.1016/j.cell.2009.02.007 19239877 Cech TR, Steitz JA. The noncoding RNA revolution-trashing old rules to forge new ones. Cell. 2014;157: 77–94. doi: 10.1016/j.cell.2014.03.008 24679528 Wan Y, Kertesz M, Spitale RC, Segal E, Chang HY. Understanding the transcriptome through RNA structure. Nat Rev Genet. 2011;12: 641–655. doi: 10.1038/nrg3049 21850044 Vandivier LE, Anderson SJ, Foley SW, Gregory BD. The Conservation and Function of RNA Secondary Structure in Plants. Annu Rev Plant Biol. 2016;67: 463–488. doi: 10.1146/annurev-arplant-043015-111754 26865341 Sun L, Fazal FM, Li P, Broughton JP, Lee B, Tang L, et al. RNA structure maps across mammalian cellular compartments. Nat Struct Mol Biol. 2019;26: 322–330. doi: 10.1038/s41594-019-0200-7 30886404 Casas-Vila N, Sayols S, Pérez-Martínez L, Scheibe M, Butter F. The RNA fold interactome of evolutionary conserved RNA structures in S. cerevisiae. Nat Commun. 2020;11: 1–12. doi: 10.1038/s41467-020-16555-4 32493961 Hasler D, Meister G. From tRNA to miRNA: RNA-folding contributes to correct entry into noncoding RNA pathways. FEBS Lett. 2016;590: 2354–63. doi: 10.1002/1873-3468.12294 27397696 Wachter A. Riboswitch-mediated control of gene expression in eukaryotes. RNA Biol. 2010;7. doi: 10.4161/rna.7.1.10489 20009507 Wilkinson KA, Merino EJ, Weeks KM. Selective 2’-hydroxyl acylation analyzed by primer extension (SHAPE): quantitative RNA structure analysis at single nucleotide resolution. Nat Protoc. 2006;1: 1610–6. doi: 10.1038/nprot.2006.249 17406453 Spitale RC, Crisalli P, Flynn RA, Torre EA, Kool ET, Chang HY. RNA SHAPE analysis in living cells. Nat Chem Biol. 2013;9: 18–20. doi: 10.1038/nchembio.1131 23178934 Westhof E, Romby P. The RNA structurome: high-throughput probing. Nat Methods. 2010;7: 965–967. doi: 10.1038/nmeth1210-965 21116245 Ding Y, Tang Y, Kwok CK, Zhang Y, Bevilacqua PC, Assmann SM. In vivo genome-wide profiling of RNA secondary structure reveals novel regulatory features. Nature. 2014;505: 696–700. doi: 10.1038/nature12756 24270811 Kwok CK, Tang Y, Assmann SM, Bevilacqua PC. The RNA structurome: transcriptome-wide structure probing with next-generation sequencing. Trends Biochem Sci. 2015;40: 221–32. doi: 10.1016/j.tibs.2015.02.005 25797096 Zhao J, Qian X, Yeung PY, Zhang QC, Kwok CK. Mapping In Vivo RNA Structures and Interactions. Trends Biochem Sci. 2019;44: 555–556. doi: 10.1016/j.tibs.2019.01.012 30853244 Rouskin S, Zubradt M, Washietl S, Kellis M, Weissman JS. Genome-wide probing of RNA structure reveals active unfolding of mRNA structures in vivo. Nature. 2014;505: 701–705. doi: 10.1038/nature12894 24336214 Watts JM, Dang KK, Gorelick RJ, Leonard CW, Bess JW, Swanstrom R, et al. Architecture and secondary structure of an entire HIV-1 RNA genome. Nature. 2009;460: 711–6. doi: 10.1038/nature08237 19661910 Tomezsko PJ, Corbin VDA, Gupta P, Swaminathan H, Glasgow M, Persad S, et al. Determination of RNA structural diversity and its role in HIV-1 RNA splicing. Nature. 2020;582: 438–442. doi: 10.1038/s41586-020-2253-5 32555469 Li P, Wei Y, Mei M, Tang L, Sun L, Huang W, et al. Integrative Analysis of Zika Virus Genome RNA Structure Reveals Critical Determinants of Viral Infectivity. Cell Host Microbe. 2018;24: 875–886.e5. doi: 10.1016/j.chom.2018.10.011 30472207 Huber RG, Lim XN, Ng WC, Sim AYL, Poh HX, Shen Y, et al. Structure mapping of dengue and Zika viruses reveals functional long-range interactions. Nat Commun. 2019;10. doi: 10.1038/s41467-019-09391-8 30926818 Simon LM, Morandi E, Luganini A, Gribaudo G, Martinez-Sobrido L, Turner DH, et al. In vivo analysis of influenza A mRNA secondary structures identifies critical regulatory motifs. Nucleic Acids Res. 2019;47: 7003–7017. doi: 10.1093/nar/gkz318 31053845 Manfredonia I, Nithin C, Ponce-Salvatierra A, Ghosh P, Wirecki TK, Marinus T, et al. Genome-wide mapping of SARS-CoV-2 RNA structures identifies therapeutically-relevant elements. Nucleic Acids Res. 2020; 1–17. doi: 10.1093/nar/gkaa1053 33166999 Sun L, Li P, Ju X, Rao J, Huang W, Ren L, et al. In vivo structural characterization of the SARS-CoV-2 RNA genome identifies host proteins vulnerable to repurposed drugs. Cell. 2021;184: 1865–1883.e20. doi: 10.1016/j.cell.2021.02.008 33636127 Ziv O, Price J, Shalamova L, Kamenova T, Goodfellow I, Weber F, et al. The Short- and Long-Range RNA-RNA Interactome of SARS-CoV-2. Mol Cell. 2020;80: 1067–1077.e5. doi: 10.1016/j.molcel.2020.11.004 33259809 Huston NC, Wan H, Strine MS, de Cesaris Araujo Tavares R, Wilen CB, Pyle AM. Comprehensive in vivo secondary structure of the SARS-CoV-2 genome reveals novel regulatory motifs and mechanisms. Mol Cell. 2021;81: 584–598.e5. doi: 10.1016/j.molcel.2020.12.041 33444546 Ling Yang S, DeFalco L, Anderson DE, Zhang Y, Aw AJ, Lim Y, et al. Comprehensive mapping of SARS-CoV-2 interactions in vivo reveals 1 functional virus-host interactions 2 3. bioRxiv. 2021; 2021.01.17.427000. doi: 10.1101/2021.01.17.427000 Lan Tammy C. T., Allan Matthew F., Malsick Lauren E., Khandwala Stuti, Nyeo Sherry S. Y., Sun Yu, Guo Junjie U., Bathe Mark, Anthony Griffiths SR, Lan TCT, Allan MF, Malsick LE, Khandwala S, Nyeo SSY, et al. Insights into the secondary structural ensembles of the full SARS-CoV-2 RNA genome in infected cells. bioRxiv. 2021; 2020.06.29.178343. doi: 10.1101/2020.06.29.178343 Deng H, Cheema J, Zhang H, Woolfenden H, Norris M, Liu Z, et al. Rice In Vivo RNA Structurome Reveals RNA Secondary Structure Conservation and Divergence in Plants. Mol Plant. 2018;11: 607–622. doi: 10.1016/j.molp.2018.01.008 29409859 Guo JU, Bartel DP. RNA G-quadruplexes are globally unfolded in eukaryotic cells and depleted in bacteria. Science (80-). 2016;353. doi: 10.1126/science.aaf5371 27708011 Incarnato D, Neri F, Anselmi F, Oliviero S. Genome-wide profiling of mouse RNA secondary structures reveals key features of the mammalian transcriptome. Genome Biol. 2014;15: 491. doi: 10.1186/s13059-014-0491-2 25323333 Wu X, Bartel DP. Widespread Influence of 3′-End Structures on Mammalian mRNA Processing and Stability. Cell. 2017;169: 905–917.e11. doi: 10.1016/j.cell.2017.04.036 28525757 WHO. World malaria report 2018. 2018. https://www.who.int/malaria/publications/world-malaria-report-2018/en/ Su Z, Tang Y, Ritchey LE, Tack DC, Zhu M, Bevilacqua PC, et al. Genome-wide RNA structurome reprogramming by acute heat shock globally regulates mRNA abundance. Proc Natl Acad Sci U S A. 2018;115: 12170–12175. doi: 10.1073/pnas.1807988115 30413617 Righetti F, Nuss AM, Twittenhoff C, Beele S, Urban K, Will S, et al. Temperature-responsive in vitro RNA structurome of Yersinia pseudotuberculosis. Proc Natl Acad Sci U S A. 2016;113: 7237–7242. doi: 10.1073/pnas.1523004113 27298343 Twittenhoff C, Brandenburg VB, Righetti F, Nuss AM, Mosig A, Dersch P, et al. Lead-seq: Transcriptome-wide structure probing in vivo using lead(II) ions. Nucleic Acids Res. 2020;48: E71–E71. doi: 10.1093/nar/gkaa404 32463449 Waters AP, Syin C, McCutchan TF. Developmental regulation of stage-specific ribosome populations in Plasmodium. Nature. 1989;342: 438–40. doi: 10.1038/342438a0 2586613 Gunderson JH, Sogin ML, Wollett G, Hollingdale M, de la Cruz VF, Waters AP, et al. Structurally distinct, stage-specific ribosomes occur in Plasmodium. Science. 1987;238: 933–7. doi: 10.1126/science.3672135 3672135 Fang J, McCutchan TF. Thermoregulation in a parasite’s life cycle. Nature. 2002;418: 742. doi: 10.1038/418742a 12181557 Fang J, Zhou H, Rathore D, Sullivan M, Su XZ, McCutchan TF. Ambient glucose concentration and gene expression in Plasmodium falciparum. Mol Biochem Parasitol. 2004;133: 125–129. doi: 10.1016/j.molbiopara.2003.09.004 14668019 John Rogers M, Gutell RR, Damberger SH, Li J, Mcconkey GA, Waters AP, et al. Structural features of the large subunit rRNA expressed in Plasmodium falciparum sporozoites that distinguish it from the asexually expressed large subunit rRNA. RNA. 1996;2: 134–145. Velichutina I V., Rogers MJ, McCutchan TF, Liebman SW. Chimeric rRNAs containing the GTPase centers of the developmentally regulated ribosomal rRNAs of Plasmodium falciparum are functionally distinct. RNA. 1998;4: 594–602. doi: 10.1017/s1355838298980049 9582100 Gardner MJ, Hall N, Fung E, White O, Berriman M, Hyman RW, et al. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature. 2002;419: 498–511. doi: 10.1038/nature01097 12368864 Trager W, Jensen JB. Human malaria parasites in continuous culture. Science. 1976;193: 673–5. doi: 10.1126/science.781840 781840 Kwok CK, Ding Y, Tang Y, Assmann SM, Bevilacqua PC. Determination of in vivo RNA structure in low-abundance transcripts. Nat Commun. 2013;4: 2971. doi: 10.1038/ncomms3971 24336128 Yeung PY, Zhao J, Chow EYC, Mou X, Hong HQ, Chen L, et al. Systematic evaluation and optimization of the experimental steps in RNA G-quadruplex structure sequencing. Sci Rep. 2019;9: 1–12. doi: 10.1038/s41598-019-44541-4 31147619 Davis MPA, van Dongen S, Abreu-Goodger C, Bartonicek N, Enright AJ. Kraken: A set of tools for quality control and analysis of high-throughput sequence data. Methods. 2013;63: 41–49. doi: 10.1016/j.ymeth.2013.06.027 23816787 Langmead B, Wilks C, Antonescu V, Charles R. Scaling read aligners to hundreds of threads on general-purpose processors. Hancock J, editor. Bioinformatics. 2019;35: 421–432. doi: 10.1093/bioinformatics/bty648 30020410 Tack DC, Tang Y, Ritchey LE, Assmann SM, Bevilacqua PC. StructureFold2: Bringing chemical probing data into the computational fold of RNA structural analysis. Methods. 2018;143: 12–15. doi: 10.1016/j.ymeth.2018.01.018 29410279 Reuter JS, Mathews DH. RNAstructure: software for RNA secondary structure prediction and analysis. BMC Bioinformatics. 2010;11: 129. doi: 10.1186/1471-2105-11-129 20230624 Lorenz R, Bernhart SH, Höner Zu Siederdissen C, Tafer H, Flamm C, Stadler PF, et al. ViennaRNA Package 2.0. Algorithms Mol Biol. 2011;6: 26. doi: 10.1186/1748-7188-6-26 22115189 Torrentino-Madamet M, Alméras L, Desplans J, Le Priol Y, Belghazi M, Pophillat M, et al. Global response of Plasmodium falciparum to hyperoxia: A combined transcriptomic and proteomic approach. Malar J. 2011;10: 1–15. doi: 10.1186/1475-2875-10-4 21223545 Untaroiu AM, Carey MA, Guler JL, Papin JA. Leveraging the effects of chloroquine on resistant malaria parasites for combination therapies. BMC Bioinformatics. 2019;20: 1–9. doi: 10.1186/s12859-019-2756-y 30987583 Caro F, Ahyong V, Betegon M, DeRisi JL. Genome-wide regulatory dynamics of translation in the Plasmodium falciparum asexual blood stages. Elife. 2014;3: 1–24. doi: 10.7554/eLife.04106 25493618 Painter HJ, Chung NC, Sebastian A, Albert I, Storey JD, Llinás M. Genome-wide real-time in vivo transcriptional dynamics during Plasmodium falciparum blood-stage development. Nat Commun. 2018;9: 2656. doi: 10.1038/s41467-018-04966-3 29985403 Aurrecoechea C, Brestelli J, Brunk BP, Dommer J, Fischer S, Gajria B, et al. PlasmoDB: a functional genomic database for malaria parasites. Nucleic Acids Res. 2009;37: D539–D543. doi: 10.1093/nar/gkn814 18957442 Westermann AJ, Gorski SA, Vogel J. Dual RNA-seq of pathogen and host. Nat Rev Microbiol. 2012;10: 618–630. doi: 10.1038/nrmicro2852 22890146 Ritchey LE, Su Z, Tang Y, Tack DC, Assmann SM, Bevilacqua PC. Structure-seq2: sensitive and accurate genome-wide profiling of RNA structure in vivo. Nucleic Acids Res. 2017;45: e135. doi: 10.1093/nar/gkx533 28637286 Supek F, Bošnjak M, Škunca N, Šmuc T. REVIGO Summarizes and Visualizes Long Lists of Gene Ontology Terms. Gibas C, editor. PLoS One. 2011;6: e21800. doi: 10.1371/journal.pone.0021800 21789182 Trotta E. On the normalization of the minimum free energy of RNAs by sequence length. PLoS One. 2014;9. doi: 10.1371/journal.pone.0113380 25405875 Pain A, Böhme U, Berry AE, Mungall K, Finn RD, Jackson AP, et al. The genome of the simian and human malaria parasite Plasmodium knowlesi. Nature. 2008;455: 799–803. doi: 10.1038/nature07306 18843368 Alvarez DR, Ospina A, Barwell T, Zheng B, Dey A, Li C, et al. The RNA structurome in the asexual blood stages of malaria pathogen plasmodium falciparum. RNA Biol. 2021;18: 2480–2497. doi: 10.1080/15476286.2021.1926747 33960872 Alvarez DR, Ospina A, Barwell T, Zheng B, Dey A, Li C, et al. The RNA structurome in the asexual blood stages of malaria pathogen plasmodium falciparum. RNA Biol. 2021;00: 1–18. doi: 10.1080/15476286.2021.1926747 33960872 Bunnik EM, Batugedara G, Saraf A, Prudhomme J, Florens L, Le Roch KG. The mRNA-bound proteome of the human malaria parasite Plasmodium falciparum. Genome Biol. 2016;17: 1–18. doi: 10.1186/s13059-016-1014-0 27381095 Dumetz F, Chow EY, Harris LM, Umar MI, Jensen A, Chung B, et al. G-quadruplex RNA motifs influence gene expression in the malaria parasite Plasmodium falciparum. bioRxiv. 2021; 2021.03.08.434398. doi: 10.1093/nar/gkab1095 34792144 Kramer MC, Gregory BD. Does RNA secondary structure drive translation or vice versa? Nat Struct Mol Biol. 2018;25: 641–643. doi: 10.1038/s41594-018-0100-2 30061597 Mauger DM, Cabral BJ, Presnyak V, Su SV., Reid DW, Goodman B, et al. mRNA structure regulates protein expression through changes in functional half-life. Proc Natl Acad Sci U S A. 2019;116: 24075–24083. doi: 10.1073/pnas.1908052116 31712433 Baumgarten S, Bryant JM, Sinha A, Reyser T, Preiser PR, Dedon PC, et al. Transcriptome-wide dynamics of extensive m6A mRNA methylation during Plasmodium falciparum blood-stage development. Nat Microbiol. 2019;4: 2246–2259. doi: 10.1038/s41564-019-0521-7 31384004 White NJ, Imwong M. Relapse. Advances in Parasitology. 2012. pp. 113–150. doi: 10.1016/B978-0-12-397900-1.00002-5 23199487 Guerreiro A, Deligianni E, Santos JM, Silva PAGC, Louis C, Pain A, et al. Genome-wide RIP-Chip analysis of translational repressor-bound mRNAs in the Plasmodium gametocyte. Genome Biol. 2014;15: 1–16. doi: 10.1186/s13059-014-0493-0 25418785 Arez AP, Pinto J, Pålsson K, Snounou G, Jaenson TGT, do Rosário VE. Transmission of mixed Plasmodium species and Plasmodium falciparum genotypes. Am J Trop Med Hyg. 2003;68: 161–8. Available: http://www.ncbi.nlm.nih.gov/pubmed/12641406 12641406 Bogale HN, Pascini T V., Kanatani S, Sá JM, Wellems TE, Sinnis P, et al. Transcriptional heterogeneity and tightly regulated changes in gene expression during Plasmodium berghei sporozoite development. Proc Natl Acad Sci. 2021;118: e2023438118. doi: 10.1073/pnas.2023438118 33653959 Morandi E, Manfredonia I, Simon LM, Anselmi F, van Hemert MJ, Oliviero S, et al. Genome-scale deconvolution of RNA structure ensembles. Nat Methods. 2021;18: 249–252. doi: 10.1038/s41592-021-01075-w 33619392 Mays AD, Schmidt M, Graham G, Tseng E, Baybayan P, Sebra R, et al. Single-molecule real-time (SMRT) full-length RNA-sequencing reveals novel and distinct mRNA isoforms in human bone marrow cell subpopulations. Genes (Basel). 2019;10: 1–17. doi: 10.3390/genes10040253 30934798 10.1371/journal.pone.0270863.r001 Decision Letter 0 Frischknecht Friedrich Academic Editor 2022 Friedrich Frischknecht This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Submission Version 0

29 Apr 2022

PONE-D-22-07594The in vivo RNA structurome of the malaria parasite Plasmodium falciparum, a protozoan with an A/T-rich transcriptomePLOS ONE

Dear Dr. Merrick,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Jun 09 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Friedrich Frischknecht

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at 

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match. 

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

3. Thank you for stating the following in the Funding Section of your manuscript: 

"UK Medical Research Council (grants MR/K000535/1 and MR/L008823/1) to CJM, Royal Society Kan Tong Po fellowship, Shenzhen Basic Research Project (JCYJ20180507181642811), Research Grants Council of the Hong Kong SAR, China Projects CityU 11101519, CityU 11100218, N_CityU110/17, CityU 21302317, Croucher Foundation Project No. 9509003, 9500030, State Key Laboratory of Marine Pollution Director Discretionary Fund, City University of Hong Kong projects 6000711, 7005503, 9680261, 9667222 to CKK. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication."

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Funding section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. 

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows: 

"The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. PLOS ONE now requires that authors provide the original uncropped and unadjusted images underlying all blot or gel results reported in a submission’s figures or Supporting Information files. This policy and the journal’s other requirements for blot/gel reporting and figure preparation are described in detail at https://journals.plos.org/plosone/s/figures#loc-blot-and-gel-reporting-requirements and https://journals.plos.org/plosone/s/figures#loc-preparing-figures-from-image-files. When you submit your revised manuscript, please ensure that your figures adhere fully to these guidelines and provide the original underlying images for all blot or gel data reported in your submission. See the following link for instructions on providing the original image data: https://journals.plos.org/plosone/s/figures#loc-original-images-for-blots-and-gels. 

  

In your cover letter, please note whether your blot/gel image data are in Supporting Information or posted at a public data repository, provide the repository URL if relevant, and provide specific details as to which raw blot/gel images, if any, are not available. Email us at plosone@plos.org if you have any questions.

5. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

6. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

Dear Catherine,

thanks for sending your nice paper to Plos One. It has now been seen by two experts in the field, who, as you will see, liked the work but have some suggestions to improve the paper. I would encourage you to revise the manuscript according to their comments.

Sincerely

Freddy

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

********** 

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

********** 

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

********** 

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

********** 

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this manuscript Dumetz et al. probe RNA secondary structures during the asexual replication of Plasmodium falciparum within human red blood cells. They use the well-established method Strucutre-seq that uses the chemical NAI to probe unpaired nucleotides and subsequent RT fall-off to interrogate RNA secondary structures on a transcriptome wide scale. Comparisons of their in vivo derived structurs to known, well characterized structures (e.g. tRNA) show the robustness of their method and I particularly like the ‘divergence’ score to categorize different RNA classes. Finally, the authors compare their RNA structure to translational efficiencies and RNA decay rates and find that some structures within RNA transcripts can influence those. Overall, the study is well designed, the methodology is clear, their findings well-presented and the study will overall lay the foundation to study dynamics of the RNA structurome across the P. falciparum life-cycle in the future.

Please find my comments to specific points below:

line 344ff: The structure-seq relies on RT fall-off. This also happens very frequently at sites of RNA modifications, especially in hypermodified molecules such as tRNA. For the tRNA where the authors do not find a ‘canonical’ structure, the RT fall-off because of modifications might make the probing of the actual structure impossible. This information should be added to the Discussion.

What about different structures in different regions of mRNA transcripts, i.e. 5’UTR-CDS-3’UTR. UTRs in P. falciparum are substantially more A/U rich than the CDS. Please provide a paragraph/table/figure showing where different structures are located and if there is a potential bias.

line 361: The authors do mention the previously published P. falciparum structuromes. How similar are the structures especially of mRNAs identified in the different studies?

line 325: It is not fully clear if all presented structures are derived from the NAI samples. If so, why did the authors chose to present only those in the main text although all samples where sequenced and present in the supplement? How different are the structures of the NAI and DMS samples?

line 502: Different transcript regions have substantially different functions within a mRNA molecule. 5’UTRs are essential for translation and can even initiate protein synthesis cap-idenpendent with secondary structures, i.e. IRES. On the otherhand 3’UTR are crucial for RNA decay. So how does the comparison with translational efficiencies and mRNA decay change when focusing only on structures on certain regions, i.e. 5’UTR and translation and 3’UTR and decay?

Are there certain conserved structures in particular regions present in multiple different transcripts?

Please upload all raw and processed sequencing data to the SRA and GEO databases.

Minor

Please ensure to accurately state “A/U” when talking about the transcriptome (e.g. in the title)

Reviewer #2: Merrick and co-workers used a combination of chemical modification and RNA seq to inform on the secondary structure of mRNAs of the malaria-causing parasite Plasmodium falciparum. The authors were able to provide data on more than half of the parasite’s transcriptome and compare a AT-rich to a AT-balanced genome. The current study complements and expands insights from two recent studies on the same subject.

The presented data appear solid and the conclusions are scientifically sound. But some minor changes (for example Fig. 3) would help the reader to better access the significance of this work. Below are some specific minor comments, which should help to further strengthen this manuscript.

Line 214: Technically the parasite does not live “entirely” inside a host cell, many stages are extracellularly. Please revise.

Sentence starting in line 250 may be clearer if stating that nucleotides were unpaired before labelling reaction.

The labels of Suppl. Fig. 3 are very small.

For Fig. 2C, would it be possible to have the experimentally-determined structures superimposed on the “canonical” folding? This would strengthen the point. While Fig. 2C and Suppl. Fig. 4 clearly show the strength of the approach in detecting the expected canonical structures of some tRNAs, showing some of the non-canonical structures in the main figure would be appreciated.

Revise formatting of the reference in line 358.

End of the sentence in line 403: . instead of ,

Fig. 3, please provide a scale for the different shades of red/fold enrichment. Which P-value cut-off was used to identify significantly enriched GO terms? Also, the figure would be easier to understand, if panels A, B, and C had labels indicating the divergence tiers. It is unclear how the numbers of enriched GO terms in the text correspond to the number of GO terms shown in the figure, please explain.

The Fig. 4A would benefit from a legend describing what the green/red shading and the boxes indicate. Please also explain here what the dot/bracket sequences mean.

Is the term “significant” (line 435) used to indicate statistical significance? If so, please provide information on the tests. If not, please rephrase.

Fig. 5 and 6 caption: It is not the “life cycle” but the development inside erythrocytes.

While the other two studies investigating the structure of P. falciparum mRNAs are mentioned, this manuscript would benefit from an extended discussion of the results from the other studies.

********** 

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Sebastian Baumgarten

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

 10.1371/journal.pone.0270863.r002 Author response to Decision Letter 0 Submission Version 1

15 Jun 2022

Dear Prof Frischknecht

Thank you for the opportunity to revise this manuscript. Below is a list of revisions (red text) in response to the reviewers’ comments: we hope that the paper will now prove acceptable for publication.

Sincerely,

Catherine Merrick

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

All checked

2. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

This has been added in the online submission system

3. Thank you for stating the following in the Funding Section of your manuscript:

"UK Medical Research Council (grants MR/K000535/1 and MR/L008823/1) to CJM, Royal Society Kan Tong Po fellowship, Shenzhen Basic Research Project (JCYJ20180507181642811), Research Grants Council of the Hong Kong SAR, China Projects CityU 11101519, CityU 11100218, N_CityU110/17, CityU 21302317, Croucher Foundation Project No. 9509003, 9500030, State Key Laboratory of Marine Pollution Director Discretionary Fund, City University of Hong Kong projects 6000711, 7005503, 9680261, 9667222 to CKK. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication."

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Funding section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

"The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript." Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

The cover letter now contains the correct funding statement, and it has been removed from the actual manuscript.

4. PLOS ONE now requires that authors provide the original uncropped and unadjusted images underlying all blot or gel results reported in a submission’s figures or Supporting Information files. This policy and the journal’s other requirements for blot/gel reporting and figure preparation are described in detail at https://journals.plos.org/plosone/s/figures#loc-blot-and-gel-reporting-requirements and https://journals.plos.org/plosone/s/figures#loc-preparing-figures-from-image-files. When you submit your revised manuscript, please ensure that your figures adhere fully to these guidelines and provide the original underlying images for all blot or gel data reported in your submission. See the following link for instructions on providing the original image data: https://journals.plos.org/plosone/s/figures#loc-original-images-for-blots-and-gels.

In your cover letter, please note whether your blot/gel image data are in Supporting Information or posted at a public data repository, provide the repository URL if relevant, and provide specific details as to which raw blot/gel images, if any, are not available. Email us at plosone@plos.org if you have any questions.

There are only 2 gels shown in this manuscript: both were already full-length (uncropped at top and bottom) and the one in supp Fig.1 was already full-width; the one in main Fig.1B has now been additionally provided as a full-width image in supp Fig .1. This is also noted in the cover letter.

5. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

The ’data not shown’ have been added as a supplementary table (6) in the revised manuscript.

6. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Done – reference formatting checked & one preprint updated to the published version.

Additional Editor Comments:

Dear Catherine,

thanks for sending your nice paper to Plos One. It has now been seen by two experts in the field, who, as you will see, liked the work but have some suggestions to improve the paper. I would encourage you to revise the manuscript according to their comments.

Sincerely

Freddy

Reviewer #1

In this manuscript Dumetz et al. probe RNA secondary structures during the asexual replication of Plasmodium falciparum within human red blood cells. They use the well-established method Strucutre-seq that uses the chemical NAI to probe unpaired nucleotides and subsequent RT fall-off to interrogate RNA secondary structures on a transcriptome wide scale. Comparisons of their in vivo derived structurs to known, well characterized structures (e.g. tRNA) show the robustness of their method and I particularly like the ‘divergence’ score to categorize different RNA classes. Finally, the authors compare their RNA structure to translational efficiencies and RNA decay rates and find that some structures within RNA transcripts can influence those. Overall, the study is well designed, the methodology is clear, their findings well-presented and the study will overall lay the foundation to study dynamics of the RNA structurome across the P. falciparum life-cycle in the future.

Please find my comments to specific points below:

line 344ff: The structure-seq relies on RT fall-off. This also happens very frequently at sites of RNA modifications, especially in hypermodified molecules such as tRNA. For the tRNA where the authors do not find a ‘canonical’ structure, the RT fall-off because of modifications might make the probing of the actual structure impossible. This information should be added to the Discussion.

This valid point has been added (at line ~400)

What about different structures in different regions of mRNA transcripts, i.e. 5’UTR-CDS-3’UTR. UTRs in P. falciparum are substantially more A/U rich than the CDS. Please provide a paragraph/table/figure showing where different structures are located and if there is a potential bias.

This has now been discussed briefly at line ~370: ‘A second limitation of short-read sequencing was that it resulted in very limited coverage of untranslated regions, which in P. falciparum tend to be extremely A/U-biased (often over 90%), and also highly repetitive with many low-complexity regions. This precluded any comprehensive assessment of RNA structure in the UTRs versus CDSs of mRNAs.’

line 361: The authors do mention the previously published P. falciparum structuromes. How similar are the structures especially of mRNAs identified in the different studies?

Unfortunately, the previously published studies do not provide their reactivity data gene-by-gene - e.g. in dot-bracket format, which would facilitate the comparison of their paired/unpaired nucleotides with ours. (This is a young, developing field and there are not yet any community standards for publishing this type of data). Without this, it would be necessary to take their raw sequencing-read data and re-fold the entire transcriptome. Besides this being a very large task, the resultant comparison could anyway be somewhat flawed if they did not use exactly the same algorithms and parameters as us. Systematic comparisons between structuromes determined by different groups via different methods will likely be an important endeavour as this field develops; however, this is beyond the scope of our current paper.

line 325: It is not fully clear if all presented structures are derived from the NAI samples. If so, why did the authors chose to present only those in the main text although all samples where sequenced and present in the supplement? How different are the structures of the NAI and DMS samples?

This is already addressed (now at line 321): ‘Structural information was more comprehensive (and agreement between replicates was also better), for the NAI-probed datasets because all four unpaired nucleotides were detected. Subsequent analysis was therefore focused on the NAI dataset as the primary in vivo structurome’. We initially tried 2 different probing methods (DMS as well as NAI) because it was unclear whether both chemicals would efficiently access the Plasmodium RNA: this proved to be the case, and the information from the NAI datasets was therefore more complete. Because there is no well-validated way (to our knowledge) to blend data from DMS and NAI together, we chose to focus on the NAI datasets alone.

line 502: Different transcript regions have substantially different functions within a mRNA molecule. 5’UTRs are essential for translation and can even initiate protein synthesis cap-idenpendent with secondary structures, i.e. IRES. On the other hand 3’UTR are crucial for RNA decay. So how does the comparison with translational efficiencies and mRNA decay change when focusing only on structures on certain regions, i.e. 5’UTR and translation and 3’UTR and decay?

As per the point above: our coverage of UTRs, via basic Illumina sequencing, was too patchy to be confident in comparing this with CDSs. In future, some recently published improvements in RNAseq methodologies specifically to preserve A/U-rich Plasmodium sequences (e.g. ‘DAFT-seq’, Chappell et al 2020), could perhaps be used to ameliorate this problem.

Are there certain conserved structures in particular regions present in multiple different transcripts?

We have struggled with this question, which is rather general: i.e. how should we meaningfully define ‘particular regions’ of multiple transcripts? The simplest way to define regions is 5’UTR/CDS/3’UTR – but as noted above, our coverage of UTRs is much lower than of CDSs, precluding a meaningful analysis. We then considered seeking common RNA structures in the codes for common protein domains, but there are many hundreds of protein domains, versus only 4 broad categories of RNA structure – hairpin, stem, bulge and multi-stem-loop, so we did not find this to be meaningful either.

Please upload all raw and processed sequencing data to the SRA and GEO databases.

This is done, apologies for its omission from this manuscript (in a previous submission, it was provided in the metadata rather than the manuscript itself, but it has now been added to Materials & Methods after the section on sequencing). ‘Raw data (FASTQ) from sequencing for each sample is available from the European Nucleotide Archive (ENA) under accession PRJEB44384’.

Please ensure to accurately state “A/U” when talking about the transcriptome (e.g. in the title)

This has been checked/corrected throughout

Reviewer #2

Merrick and co-workers used a combination of chemical modification and RNA seq to inform on the secondary structure of mRNAs of the malaria-causing parasite Plasmodium falciparum. The authors were able to provide data on more than half of the parasite’s transcriptome and compare a AT-rich to a AT-balanced genome. The current study complements and expands insights from two recent studies on the same subject.

The presented data appear solid and the conclusions are scientifically sound. But some minor changes (for example Fig. 3) would help the reader to better access the significance of this work. Below are some specific minor comments, which should help to further strengthen this manuscript.

Line 214: Technically the parasite does not live “entirely” inside a host cell, many stages are extracellularly. Please revise.

This has been corrected

Sentence starting in line 250 may be clearer if stating that nucleotides were unpaired before labelling reaction.

This has been corrected

The labels of Suppl. Fig. 3 are very small.

This has been corrected

For Fig. 2C, would it be possible to have the experimentally-determined structures superimposed on the “canonical” folding? This would strengthen the point. While Fig. 2C and Suppl. Fig. 4 clearly show the strength of the approach in detecting the expected canonical structures of some tRNAs, showing some of the non-canonical structures in the main figure would be appreciated.

In figure 2C, it is the experimentally-determined structures that are shown. In the original figure, we showed only experimentally-determined structures that were canonical (non-canonical versions were in the supp figure). As requested, we have now shown a non-canonical example as well, for direct side-by-side comparison within the main figure.

Revise formatting of the reference in line 358.

All references are checked/updated in the revision.

End of the sentence in line 403: . instead of ,

This has been corrected.

Fig. 3, please provide a scale for the different shades of red/fold enrichment. Which P-value cut-off was used to identify significantly enriched GO terms? Also, the figure would be easier to understand, if panels A, B, and C had labels indicating the divergence tiers. It is unclear how the numbers of enriched GO terms in the text correspond to the number of GO terms shown in the figure, please explain.

We appreciate the reviewer’s pointing out that Figure 3 (a representation of GO terms analysis) was not fully quantitative. We have therefore re-done the analysis, changing the representation to a scatter-plot format, which still represents the relatedness between similar GO terms, but is also quantitative both in terms of enrichment (see rainbow scale) and number of GO terms (see size of circles). We have also labelled the 3 plots with Tiers 1-3.

The Fig. 4A would benefit from a legend describing what the green/red shading and the boxes indicate. Please also explain here what the dot/bracket sequences mean.

This information is all in the legend, but we appreciate the reviewer’s pointing out that the figure is complex to look at without visual legend as well. A visual legend has been added to a revised figure 4.

Is the term “significant” (line 435) used to indicate statistical significance? If so, please provide information on the tests. If not, please rephrase.

This has been corrected to ‘major’

Fig. 5 and 6 caption: It is not the “life cycle” but the development inside erythrocytes.

This has been corrected

While the other two studies investigating the structure of P. falciparum mRNAs are mentioned, this manuscript would benefit from an extended discussion of the results from the other studies.

We do already have within our discussion several paragraphs of comparative insight concerning the two other existing studies:

a) lines 585-594 comparing the overall coverage of our structurome vz that of Qi et al.,

b) lines 636-647 comparing our findings about RNA structure vz translational-efficiency,

c) lines 685-690 concerning the effect of RNA binding proteins, as suggested by Qi et al.

It is difficult to provide much more insightful discussion, particularly of Alvarez et al., since these authors used an entirely different methodology.

Submitted filename: PlosOne rebuttal.docx

 10.1371/journal.pone.0270863.r003 Decision Letter 1 Frischknecht Friedrich Academic Editor 2022 Friedrich Frischknecht This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Submission Version 1

20 Jun 2022

The in vivo RNA structurome of the malaria parasite Plasmodium falciparum, a protozoan with an A/U-rich transcriptome

PONE-D-22-07594R1

Dear Dr. Merrick,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Friedrich Frischknecht

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

thanks for providing a constructive revision, in my view the paper can now be accepted. I wonder only if in the table 2 all numbers could be limited to two digits after the . ?

Reviewers' comments:

 10.1371/journal.pone.0270863.r004 Acceptance letter Frischknecht Friedrich Academic Editor 2022 Friedrich Frischknecht This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

16 Aug 2022

PONE-D-22-07594R1

The in vivo RNA structurome of the malaria parasite Plasmodium falciparum, a protozoan with an A/U-rich transcriptome 

Dear Dr. Merrick:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Dr. Friedrich Frischknecht

Academic Editor

PLOS ONE