Alternative splicing of coq-2 controls the levels of rhodoquinone in animals

  1. June H Tan
  2. Margot Lautens
  3. Laura Romanelli-Cedrez
  4. Jianbin Wang
  5. Michael R Schertzberg
  6. Samantha R Reinl
  7. Richard E Davis
  8. Jennifer N Shepherd  Is a corresponding author
  9. Andrew G Fraser  Is a corresponding author
  10. Gustavo Salinas  Is a corresponding author
  1. The Donnelly Centre, University of Toronto, Canada
  2. Laboratorio de Biología de Gusanos. Unidad Mixta, Departamento de Biociencias, Facultad de Química, Universidad de la República - Institut Pasteur de Montevideo, Uruguay
  3. Department of Biochemistry and Molecular Genetics, RNA Bioscience Initiative, University of Colorado School of Medicine, United States
  4. Department of Biochemistry and Cellular and Molecular Biology, University of Tennessee, United States
  5. Department of Chemistry and Biochemistry, Gonzaga University, United States

Abstract

Parasitic helminths use two benzoquinones as electron carriers in the electron transport chain. In normoxia, they use ubiquinone (UQ), but in anaerobic conditions inside the host, they require rhodoquinone (RQ) and greatly increase RQ levels. We previously showed the switch from UQ to RQ synthesis is driven by a change of substrates by the polyprenyltransferase COQ-2 (Del Borrello et al., 2019; Roberts Buceta et al., 2019); however, the mechanism of substrate selection is not known. Here, we show helminths synthesize two coq-2 splice forms, coq-2a and coq-2e, and the coq-2e-specific exon is only found in species that synthesize RQ. We show that in Caenorhabditis elegans COQ-2e is required for efficient RQ synthesis and survival in cyanide. Importantly, parasites switch from COQ-2a to COQ-2e as they transit into anaerobic environments. We conclude helminths switch from UQ to RQ synthesis principally via changes in the alternative splicing of coq-2.

Introduction

Parasitic helminths are major human pathogens. Soil-transmitted helminths (STHs) such as Ascaris, hookworms, and whipworms infect well over a billion people, and 7 out of the 18 neglected diseases categorized by WHO are caused by helminths (CDC, 2019). Despite the huge impact on global health of these infections, there are few classes of available anthelmintics and resistance is increasing in humans and is widespread in some species that infect animals — for example, Haemonchus contortus, a common parasite in small ruminants, has developed multidrug resistance (Jackson and Coop, 2000; Kotze and Prichard, 2016). Thus, there is a serious need to develop new classes of anthelmintics that target the parasites while leaving their animal hosts unaffected. One potential target for anthelmintics is their unusual anaerobic metabolism which differs from that of their hosts. When STHs are in the free-living stages of their life cycles, they use the same aerobic respiration as their hosts and use ubiquinone (referred to here as UQ and as Q in other papers) as an electron carrier in their mitochondrial electron transport chain (ETC). However, when they infect their hosts, they encounter highly anaerobic environments. This is particularly the case for species that inhabit the host gut, for example Ascaris can live for many months in this anaerobic environment (Dold and Holland, 2011). To survive, they use an alternate form of anaerobic metabolism that relies on the electron carrier rhodoquinone (RQ). Since hosts do not synthesize or use RQ, RQ-dependent metabolism could be a key pharmacological target as it is required by the parasite but absent in mammalian hosts.

RQ is an electron carrier that functions in the mitochondrial ETC of STHs (Van Hellemond et al., 1995). RQ is a prenylated aminobenzoquinone that is similar to UQ (Figure 1), but the slight difference in structure gives RQ a lower standard redox potential than UQ (−63 mV and 110 mV, respectively) (Erabi et al., 1976; Unden and Bongaerts, 1997). This difference in redox potential means that RQ, but not UQ, can play a unique role in anaerobic metabolism. In aerobic metabolism, UQ can accept electrons from a diverse set of molecules via quinone-coupled dehydrogenases, including succinate dehydrogenase and electron-transferring-flavoprotein (ETF) dehydrogenase. In RQ-dependent anaerobic metabolism, RQ does the reverse — it carries electrons to these same dehydrogenase enzymes and drives them in reverse to act as reductases that transfer electrons onto a diverse set of terminal acceptors (Tielens and Van Hellemond, 1998; van Hellemond et al., 2003). UQ thus allows electrons to enter the ETC via dehydrogenases; RQ can drive the reactions that let electrons leave the ETC and this difference in the direction of electron flow is driven by the difference in redox potential (Figure 1A). This ability of RQ to provide electrons to an alternative set of terminal electron acceptors allows helminths to continue to use a form of mitochondrial ETC to generate ATP without oxygen. In this RQ-dependent anaerobic metabolism, electrons enter the ETC from NADH through Complex I and onto RQ, and this is coupled to proton pumping to generate the proton motive force required for ATP synthesis by F0F1-ATPase (van Hellemond et al., 2003). The electrons are carried by RQ to the quinone-coupled dehydrogenases which are driven in reverse as reductases and the electrons thus exit onto a diverse set of terminal acceptors, allowing NAD+ to be regenerated and the redox balance to be maintained. This RQ-dependent metabolism does not occur in the hosts and is thus an excellent target for anthelmintics (Kita et al., 2003).

Rhodoquinone and ubiquinone biosynthesis and function in electron transport chains.

(A) In aerobic metabolism, ubiquinone (UQ) shuttles electrons in the ETC from Complex I (CI; yellow box) and quinone-coupled dehydrogenases (QDHs), such as Complex II. These electrons are ultimately transferred to oxygen. In anaerobic metabolism, rhodoquinone (RQ) reverses electron flow in QDHs and facilitates an early exit of electrons from the ETC onto anaerobic electron acceptors (Ae-A), such as fumarate. (B) The RQ biosynthetic pathway in C. elegans requires L-tryptophan, a precursor in the kynurenine pathway. L-Trptophan is transformed into 3-hydroxyanthranilic acid (3HA) in four steps. It is proposed that 3HA is a substrate for COQ-2, producing 3-hydroxy-5-nonaprenylanthranilic acid (NHA), where n=9. The transformation of NHA to RQ requires several shared proteins from the UQ biosynthetic pathway. (C) Eukaryotes can use either p-aminobenzoic acid (pABA) or 4-hydroxybenzoic acid (4HB) as precursors to UQ. Prenylation is facilitated by Coq2 to form 3-hexaprenyl-4-hydroxybenzoic acid (HHB) or 3-hexaprenyl-4-aminobenzoic acid (HAB), where n = varies between species. Further functionalization of these intermediates occurs through a Coq synthome (Coq3–Coq9 and Coq11) to yield UQ.

In the animal kingdom, RQ is present in several facultative anaerobic lineages that face environmental anoxia or hypoxia as part of their life cycle. Among animals, RQ has only been described in nematodes, platyhelminths, mollusks, and annelids (Van Hellemond et al., 1995). The key steps of RQ biosynthesis in animals were recently elucidated (Roberts Buceta et al., 2019; Del Borrello et al., 2019). In contrast to bacteria, where RQ derives from UQ (Bernert et al., 2019; Brajcich et al., 2010), RQ biosynthesis in animals requires precursors derived from tryptophan (Figure 1B). Recent studies performed on C. elegans have demonstrated that animals that lack a functional kynureninase pathway (e.g. strains carrying mutations in kynu-1, the sole kynureninase) are unable to synthesize RQ (Del Borrello et al., 2019; Roberts Buceta et al., 2019). It is presumed that 3-hydroxyanthranilic acid (3HA, also sometimes referred to as 3HAA) from the kynurenine pathway is prenylated in a reaction catalyzed by COQ-2. The prenylated benzoquinone ring can then be modified by methylases and hydroxylases to form RQ. This proposed pathway is analogous to the biosynthesis of UQ from 4-hydroxybenzoic acid (4HB) or para-aminobenzoic acid (pABA) (Figure 1C). The key insight from these previous studies is that the critical choice between UQ and RQ synthesis is the choice of substrate by COQ-2 — if 4HB is prenylated by COQ-2, UQ will ultimately be produced, but if 3HA is used, the final product will be RQ. In most parasitic helminths, there is a major shift in quinone composition as they move from aerobic to anaerobic environments, for example, RQ is less than 10% of H. contortus total quinone when the parasite is in an aerobic environment but is over 80% of total quinone in the anaerobic environment of the sheep gut and similar shifts occur in other parasites (Lümmen et al., 2014; Sakai et al., 2012; Van Hellemond et al., 1995). Somehow, COQ-2 must, therefore, switch from using 4HB to using 3HA as a substrate, but the mechanism for this substrate switch is completely unknown. Understanding this mechanism is important — if we could interfere pharmacologically with the switch to RQ synthesis, it could lead to a new class of anthelmintics.

In this study, we reveal that two variants of COQ-2, derived from alternative splicing of mutually exclusive exons, are the key for the choice to synthesize RQ or UQ. We find that one of the mutually exclusive exons is only found in the genomes of animals that synthesize RQ and that its inclusion remodels the core of the COQ-2 enzyme. We show that the removal of this exon abolishes almost all RQ biosynthesis in C. elegans. Finally, we find that inclusion of this RQ-specific exon expression is increased in the stages of the parasite life cycle where they encounter hypoxic conditions and increase RQ production, while the alternative exon is increased in normoxic life stages. We thus conclude that alternative splicing of COQ-2 is the key mechanism that causes the switch from UQ to RQ synthesis in the parasite life cycle.

Results

The C. elegans COQ-2 polyprenyl transferase required for quinone biosynthesis has two major alternative splice forms

Our research groups previously showed that if COQ-2 uses 4HB as a substrate, it would lead to the synthesis of UQ; however, if COQ-2 uses 3HA, it would ultimately yield RQ. As parasites move from aerobic environments to the anaerobic niches in their host, they change their quinone composition from high UQ to high RQ. For this to occur, COQ-2 must switch its substrate from 4HB to 3HA, but the mechanism for this switch is unknown.

We identified two distinct splice forms of C. elegans coq-2: coq-2a and coq-2e. These are annotated in the genome and confirmed by RNA-sequencing by nanopore-based direct RNA sequencing, and by targeted validation studies (Kuroyanagi et al., 2014; Ramani et al., 2011; Roach et al., 2020). These two isoforms differ by the mutually exclusive splicing of two internal exons (6a and 6e), both of 134 nucleotides (see Figure 2A). We note that mutually exclusive splicing of cassette exons is very rare in the C. elegans genome and fewer than 100 such splicing events have been identified (Kuroyanagi et al., 2014; Ramani et al., 2011). Both coq-2a and coq-2e splice forms are abundant in C. elegans across all stages of development, where ~30–50% of coq-2 is the coq-2e isoform (Gerstein et al., 2010; Grün et al., 2014; Ramani et al., 2011; Roach et al., 2020).

C. elegans coq-2 gene model.

(A) The coq-2 gene contains two mutually exclusive exons, 6e (blue box) and 6a (red box), that are alternatively spliced (blue and red lines, respectively) generating two COQ-2 isoforms. Light gray boxes represent coding sequences of exons 1–5 and 7–8, black lines represent introns, and dark gray boxes denote 5´and 3´ untranslated regions of exons 1 and 8. (B) Alternative splicing of COQ-2 changes the enzyme core. The sequences of C. elegans coq-2a and coq-2e were threaded onto the crystal structure of the apo-form of the Aeropyrum pernix COQ-2 homolog (PDB: 4OD5) in Chimera using Modeller. The region switched by mutually exclusive alternative splicing is shown magenta color. (C) The alternative exons found in all RQ-synthesizing species have two residues that are invariant (L204 and S243 show in cyan; C. elegans numbering) that are near the binding site of the two substrates. Substrates are the polyprenyl tail (dark green; geranyl S-thiolodiphosphate in the crystal structure), and the aromatic ring (light green; p-hydroxybenzoic acid [4HB] in the crystal structure). Note that COQ-2 is rotated from panel B to panel C for clarity.

To examine how the alternative splicing of COQ-2 might affect its function, we threaded the predicted COQ-2a and COQ-2e protein sequences onto the solved crystal structure of the archaean A. pernix COQ-2 ortholog (PDB:4OD5). We found that the splicing change causes a switch in two α-helices at the core of the COQ-2 structure (Figure 2B). This is a region of the protein that is believed to form a hydrophobic tunnel along which aromatic substrates must pass to the active site for the key polyprenylation reaction (Desbats et al., 2016), suggesting that the change in splicing could affect COQ-2 substrate selection and thus could explain a shift from UQ to RQ synthesis. We, therefore, examined whether similar COQ-2 alternative splicing is seen in parasitic helminths and how the different splice forms compare to COQ-2 sequences in parasite hosts which do not synthesize RQ.

Parasitic helminths have a distinct splice form of coq-2 that is not present in any of the parasitic hosts

C. elegans has two major isoforms of coq-2 which resulted from mutually exclusive alternative splicing of two internal exons and affects the core of the enzyme. If this alternative splicing affects the choice of COQ-2 substrate and thus the switch from UQ to RQ synthesis, we reasoned that parasitic helminths that synthesize RQ should have a similar gene structure and that the hosts that do not synthesize RQ should not. We used both gene predictions and available RNA-seq data to examine the coq-2 gene structure and splicing in parasitic helminths and their hosts (human, sheep, cow, and rat, respectively; Figure 3). Since many of the parasite gene structures were not correctly annotated, all the relevant exon junctions in Figure 3 were manually annotated and confirmed with RNA-seq data (see Materials and methods). We note that while the alternative splicing of coq-2 in C. elegans we observe using RNA-seq has been validated by nanopore-based direct transcript sequencing, microarray analysis, and RT-PCR, the data on the other helminths only derives from RNA-seq and genome analysis. Although the same robust RNA-seq analysis pipeline has been used, it would be ideal to validate these changes in the future. Remarkably, we find a similar gene structure with the same mutually exclusive exons in all parasites known to synthesize RQ (Figure 3). Furthermore, annelids and mollusks are the only other phyla known to synthesize RQ and their coq-2 orthologs also show similar mutually exclusive alternative splicing of homologous exons and we only find evidence for alternative splicing between a coq-2a form and a coq-2e like form in animals that synthesize RQ. Crucially, no mammalian hosts show any evidence for this kind of alternative splicing of their coq-2 orthologs, nor do they have an e-like exon either in their gene predictions or genome sequences or in any available RNA-seq or direct transcriptome sequence data (human and mouse are shown as representatives in Figure 3). Note that the RNA-seq datasets are much deeper and more extensive in humans and mouse than in helminths, mollusks, or annelids, so this lack of evidence for such alternative splicing is unlikely to be due to a failure to detect such events due to low coverage transcriptome data. We conclude that the e-exon and the mutually exclusive alternative splicing of coq-2 is unique to animals that synthesize RQ and is not found in any other species. This suggests that this coq-2 alternative splicing could indeed be linked to the ability to synthesize RQ.

Gene models for coq-2 orthologs in various species.

Parasitic helminths, as well as annelids and mollusks, have two internal exons that are spliced in a mutually exclusive manner. By contrast, humans and other hosts only have one exon that is homologous to exon 6a of C. elegans coq-2. The a-form exon (red) shares greater similarity to the exon present in species that do not synthesize RQ, while the e-form (blue) is present only in RQ-producing species. The genes used for each species are listed in Supplementary file 3. The gene structures shown are based on genome annotations but in many cases include manual reannotations — in all such cases, the manual annotations are confirmed with RNA-seq data.

We aligned the two mutually exclusive exons across helminth species and compared them to the similar regions of their host COQ-2 sequences, and of other eukaryotes that cannot synthesize RQ (S. cerevisiae and S. pombe) (Figure 4). We find that all the coq-2a-specific exons are similar to the pan-eukaryotic COQ-2 sequence, whereas all the coq-2e-specific exons have a distinct sequence to this. We examined the alignments of the a- and e-specific exons and identified two residues that are strictly conserved in pan-eukaryotic COQ-2 sequences, Phe204 and Ala243 (C. elegans numbering), that are switched to a Leu and a Ser residue in all COQ2-e-specific exons that we examined (Figure 4). These residues sit very close to the substrates in the active site of the enzyme (Figure 2C) and we note that mutation of the equivalent Ala243 residue dramatically affects the ability of human COQ-2 to synthesize UQ (Desbats et al., 2016). Altogether these results suggest that animals that synthesize both UQ and RQ synthesize two forms of COQ-2 — one looks similar to that in all other eukaryotic species, whereas the other has a single exon that appears to be specific for species that synthesize RQ. To test whether these two COQ-2 isoforms have distinct roles in UQ and RQ synthesis, we turned to C. elegans.

Conserved changes between a- and e-form exons across RQ-producing species.

Amino acid sequences of COQ-2 orthologs were aligned using Clustal Omega (Madeira et al., 2019). The sequences of exons homologous to exon 6a/e in C. elegans, as well as the flanking five amino acid sequences, were used to generate the alignment. Sequences of the mutually exclusive exons are shaded in red (a-form) or blue (e-form). Two residue changes between the a- and e- forms are highlighted (Phe to Leu, Ala to Ser) and are invariant across diverse species that synthesize RQ. The COQ-2 orthologs and exons used for each species are listed in Supplementary file 3.

Efficient RQ synthesis requires the coq-2e isoform

We found that helminths synthesize two major isoforms of coq-2, whereas their hosts only synthesize a single isoform. To test the requirement for each of the two major helminth isoforms of coq-2 for RQ synthesis, we used CRISPR engineering to generate C. elegans mutant strains that either lack coq-2 exon 6a (coq-2(syb1715)) or coq-2 exon 6e (coq-2(syb1721)) — we refer to these as coq-2∆6a and coq-2∆6e, respectively, from here on (see Figure 5A and Supplementary file 1 for details of engineering). We find that the coq-2∆6e strain synthesizes essentially no detectable RQ but has higher levels of UQ, whereas coq-2∆6a has greatly reduced UQ levels but higher RQ levels (Figure 5B, Supplementary file 2). We conclude that the coq-2e isoform, which includes the helminth/annelid/mollusc-specific exon 6e, is required for efficient RQ synthesis. While these engineered changes could potentially affect RQ levels via some indirect mechanisms such as alterations in the levels of 4HB and 3HA, it is unclear what mechanism would drive those metabolic rewirings. We thus suggest that the more parsimonious explanation is the one we propose here that COQ-2a and COQ-2e have different substrate preferences due to the large change to the core of the enzyme.

Figure 5 with 1 supplement see all
The C. elegans coq-2 edited strains and effects of exon 6a and 6e deletions on quinone biosynthesis.

(A) Mutant strains were generated in C. elegans by deletion of exon 6a (coq-2∆6a) or exon 6e (coq-2∆6e). (B) Deletion of exon 6a from the coq-2 gene significantly increased the level of RQ9 (p=0.013) and significantly decreased UQ9 (p<0.001) compared to the N2 control. By contrast, the deletion of exon 6e decreased RQ9 to a negligible level (p<0.001) and slightly increased the level of UQ9 (p=0.130) compared to N2. Statistically significant increases and decreases with respect to N2 levels are denoted with ★ and ◊, respectively; error bars reflect standard deviation where N = 4. (C) Deletion of coq-2 exon 6e affects the ability of worms to survive extended KCN treatment. Wild-type (N2) and coq-2 mutant L1 worms were exposed to 200 µM KCN for 15 hr. KCN was then diluted 6-fold and worm movement was measured over 3 hr to track recovery from KCN exposure (see Materials and methods). Worms without exon 6e could not survive extended treatment with KCN while deletion of exon 6a had little effect on KCN survival. Cyanide titration is shown in Figure 5—figure supplement 1. Curves show the mean of four biological replicates and error bars are standard errors of the mean.

To further examine whether COQ-2e is required for RQ synthesis and thus for RQ-dependent metabolism, we tested whether the coq-2∆6a and coq-2∆6e strains could survive long-term exposure to potassium cyanide (KCN) (Figure 5C and Figure 5—figure supplement 1). We previously showed that when C. elegans is exposed to KCN, it switches to RQ-dependent metabolism and that while wild-type worms can survive a 15 hr exposure to KCN, C. elegans strains that do not synthesize RQ cannot survive. We found that while the coq-2∆6a strain (that can synthesize RQ) survives 15 hr of KCN exposure, as well as wild-type animals, the coq-2∆6e strain that synthesizes no RQ does not survive, confirming the functional relevance of the coq-2e isoform as being critical for RQ synthesis.

Regulation of the alternative splicing of coq-2 in helminths

Helminths synthesize two isoforms of COQ-2 — COQ-2a resembles the pan-eukaryotic consensus and cannot synthesize RQ, whereas COQ-2e includes an exon that is only found in species that synthesize RQ and COQ-2e is required for RQ synthesis. Changing the levels of coq-2a and coq-2e splice forms could thus regulate the switch from UQ synthesis in the aerobic environment outside the host to RQ synthesis in the host gut. We thus examined RNA-seq data to see whether parasites switch between these isoforms as they switch between these environments.

Ascaris suum has a relatively simple life cycle (schematic Figure 6A) and is the pig equivalent of Ascaris lumbricoides which infects ~900 M humans. Eggs are laid in the host and emerge via defecation and the L1-L3 larval stages develop within the egg outside the host. The L3 infective larval stage then enters the host via ingestion into the digestive tract. These then leave the digestive tract and synthesize their way to the lungs where they develop into L4 larvae and, finally, the L4 re-enter the digest tract and move to the small intestine where they develop into adulthood. The adults remain in this anaerobic environment for the remainder of their life. The free-living larval stages have relatively low RQ (~35% of total quinones), whereas adults have high RQ ( ~100% of total quinones) (Takamiya et al., 1993). We used RNA-seq data (Wang et al., 2011; Wang et al., 2012) to analyze coq-2 isoforms in free-living stages and in the adults to examine whether there was a switch from coq-2a to coq-2e as the parasites switch from low RQ aerobic-respiring embryos and larvae to high RQ anaerobic adults. We see a clear switch approximately 60% of coq-2 transcripts are coq-2e in the free-living, aerobic stages but >90% is the RQ-synthesizing coq-2e form in adults (Figure 6B). Increased coq-2e levels thus correlate with increased RQ levels.

Correlation of COQ-2 splicing with change from aerobic to anaerobic life stages.

(A) Schematic of the life cycle of A. suum. ‘OUT’ denotes aerobically respiring free-living stages; ‘IN’ indicates stages living inside the host intestine. (B) Graph indicates the percentage of all COQ-2 transcripts that include the RQ-specific exon (Percent e-form) in a number of life cycle stages, sexes, and tissues. Timing of embryogenesis shows the post-fertilization time in hours. (C) Schematic of the life cycle of Strongyloides stercoralis. ‘OUT’ denotes aerobically respiring free-living stages; ‘IN’ indicates stages living inside the host. Note that egg, L1, L2, and L3 can either develop inside or outside the host. (D) Graph indicates the percentage of all COQ-2 transcripts that include the RQ-specific exon (Percent e-form) in L3 larvae that either developed outside the host (OUT) or inside the host (IN). Data were derived from three individual replicates taken from published RNA-seq data (Stoltzfus et al., 2012).

The analysis in Ascaris is complicated by the life cycle: while we are comparing coq-2 splicing in two distinct environments, this is also necessarily a comparison between different life stages (we are comparing developing embryos or larvae with adults). It is possible that the changes in splicing we see are not environmentally-induced by the switch from normoxia to anaerobic conditions but are simply a developmentally programmed switch. To address this, we turned to Strongyloides stercoralis. The life cycle of S. stercoralis (schematic Figure 6C) is broadly similar to Ascaris — L1-L3 stages are free-living, the L3 infective stage infects hosts, and L4 larvae and adults develop and live in the host. However, they have an alternative life cycle where instead of L1-L3 developing outside the host, the eggs can hatch in the host and the entire life cycle takes place inside the host. This allows us to compare the same larval stage in two conditions — here we compare L3 animals that developed inside the host anaerobic environment with L3 animals that developed outside the host. The difference is clear about <20% of coq-2 transcripts are coq-2e in the free-living L3s but >60% is the RQ-synthesizing coq-2e form in L3s that developed inside the host (Figure 6D; Stoltzfus et al., 2012). We conclude that increased inclusion of the coq-2e-specific exon correlates with increased synthesis of RQ during the parasite life cycles.

In summary, our data show that animal species that synthesize RQ regulate the choice between synthesizing UQ and RQ by alternative splicing of the polyprenyltransferase COQ-2 gene. A switch between two mutually exclusive exons changes the core of the COQ-2 enzyme and switches it from primarily generating UQ precursors to primarily RQ precursors. We propose that this switch ultimately results in a shift in quinone content from high UQ in aerobic conditions to high RQ in anaerobic conditions. This alternative splicing event is only seen in species that synthesize RQ and the switch correlates with the change from aerobic to anaerobic metabolism in parasitic helminths.

Discussion

Organisms are continually challenged by changes in their environments and they must be able to respond for them to survive. Hypoxia is one such challenge and animals have evolved diverse strategies to alter their metabolism to cope with low oxygen levels. For example, humans rapidly switch to anaerobic glycolysis, generating lactate; goldfish on the other hand can adapt to hypoxia by fermenting carbohydrates to generate ethanol (Shoubridge and Hochachka, 1980). In this paper, we focus on how helminths can survive in anaerobic conditions by switching from ubiquinone (UQ)-dependent aerobic metabolism to rhodoquinone (RQ)-dependent anaerobic metabolism.

The ability to use RQ is highly restricted among animal species — only helminths, mollusks, and annelids are known to synthesize and use RQ. This is a key adaptation since it allows them to rewire their mitochondrial electron transport chain (ETC) to use a variety of terminal electron acceptors in the place of oxygen. They can, therefore, still use Complex I to pump protons and generate the proton motive force need to power the F0F1-ATPase in the absence of oxygen. This allows them to survive without oxygen for long periods — they are thus facultative anaerobes. C. elegans is a free-living nematode that does not face extended anaerobic conditions as an obligate part of its life cycle. However, Pseudomonas aeruginosa, a natural pathogen of C. elegans, uses cyanide to kill the worm (Gallagher and Manoil, 2001). Cyanide blocks the conventional ETC at Complex IV preventing the use of oxygen as a final electron acceptor. We speculate that in C. elegans, the alternative ETC may be an adaptive strategy to withstand transient cyanide stress from pathogens. For parasitic helminths like Ascaris, the need to respire anaerobically is critical for their life cycle since they must survive for long periods in the anaerobic environment of the human gut. Since the host does not synthesize RQ or use RQ-dependent metabolism if we could interfere with RQ synthesis, this would be an excellent way to target the parasite and leave the host untouched.

We previously showed that the key decision on whether to synthesize UQ to power aerobic metabolism or RQ to synthesize anaerobic metabolism is dictated by the choice of substrate of the polyprenyltransferase COQ-2 (Del Borrello et al., 2019; Roberts Buceta et al., 2019). COQ-2 must switch from using 4HB to synthesize UQ in aerobic conditions to 3HA to synthesize RQ in anaerobic conditions. Here we reveal the simple mechanism for that switch in substrate specificity in helminths: they use the mutually exclusive alternative splicing of two internal exons to remodel the core of COQ-2. Inclusion of exon 6a results in the COQ-2a enzyme that can synthesize UQ but not RQ; switching exon 6a for the alternative exon 6e yields COQ-2e which principally synthesizes RQ (Figure 7). All eukaryotes synthesize a homolog of COQ-2a — only the species known to synthesize RQ (helminths, annelids, and mollusks) have genomes encoding the RQ-specific exon 6e and this exon is introduced by alternative splicing in a similar mutually exclusive splicing event in all these phyla. These different lineages thus have the same solution to the problem of substrate switching in COQ-2 — to have two distinct forms of COQ-2 due to alternative splicing — and all do it with the same structural switch in COQ-2.

Discrimination between RQ and UQ biosynthesis in C.

elegans. There are two variants of exon 6 in the C. elegans coq-2 gene (6a and 6e) which undergo mutually exclusive alternative splicing leading to COQ-2a and COQ-2e isoforms, respectively. Synthesis of UQ originates from 4-hydroxybenzoic acid (4HB) and prenylation is catalyzed by COQ-2a (and marginally by COQ-2e) to form 4-hydroxy-3-nonaprenylbenzoic acid (NHB). By contrast, RQ is most likely synthesized from 3HA, and prenylation is facilitated by COQ-2e to form NHA. Several additional steps are required to convert NHB to UQ and NHA to RQ, respectively. Supplementary Documentation.

The alternative splicing of COQ-2 in all animal lineages that synthesize RQ draws focus to COQ-2 as a potential anthelmintic target. COQ-2e is required for RQ generation and has a distinct sequence to the pan-eukaryotic COQ-2a. The switch between COQ-2a to COQ-2e causes a change in the core of the COQ-2 active site and all species that synthesize RQ have a pair of conserved residues in COQ-2e. We suggest that small molecule inhibitors that selectively target COQ-2e and not the host COQ-2 enzyme could be potent anthelmintics and our research groups are initiating these screens at this point. We also note that it is rare to see such a clear and profound change in enzyme specificity due to a single splice event affecting the core of an enzyme catalytic site. There are other examples, most notably a switch in the substrate specificity of the cytochrome P450 CYP4F3 from LTB4 to arachidonic acid (Christmas et al., 1999; Christmas et al., 2001). This switch is tissue specific rather than environmentally induced, however, and there are few other examples to our knowledge. The regulation of COQ-2 specificity by alternative splicing is thus a beautiful and rare example of this type of enzyme regulation by alternative splicing. We note that while alternative splicing of COQ-2 appears to be the key regulated step in determining RQ or UQ biosynthesis, we see no such splicing regulation for other enzymes that are required for both RQ and UQ synthesis downstream of COQ-2 (Roberts Buceta et al., 2019), for example, there are no known splice variants of COQ-3 and COQ-5, which are quinone methylases downstream of COQ-2. This suggests that while COQ-2 can clearly discriminate between substrates that have/lack a 2-amino group, COQ-3 and COQ-5 would be more promiscuous than COQ-2 and act on both RQ and UQ precursors.

Finally, the mechanism of alternative splicing to regulate the synthesis of UQ or RQ may explain some part of the evolution of RQ synthesis in animals and its phylogenetic distribution. One of the enduring mysteries of RQ synthesis is that it only occurs in very restricted animal phyla. RQ synthesis could either be the ancestral state with widespread loss across most animal species or else RQ synthesis could have arisen independently in helminths, annelids, and mollusks. It is unclear that our findings here definitively answer this but we do note that there is a precedent for a mutually exclusive splicing event that has arisen independently in helminths, annelids, and mollusks in the gene mrp-1 (Yue et al., 2017) which encodes a transporter that is required for the uptake of vitamin B12 into mitochondria. The mrp-1 gene undergoes alternative splicing in these species and just like coq-2, mrp-1 has mutually exclusive exons of exactly the same length. Importantly, in mrp-1, a mutually exclusive alternative splicing has arisen independently in each lineage in which it is seen. This example indicates it is at least plausible that alternative splicing of coq-2 might have originated independently in helminths, annelids, and mollusks, but we do not believe that our current data can distinguish between an ancestral animal origin of RQ biosynthesis followed by independent losses, or independent animal origin of RQ biosynthesis due to an ‘mrp-1-like’ independent evolution of alternative splicing of coq-2. Future studies should shed light on this issue. Whatever the evolutionary scenario, our results show that a simple innovation of a mutually exclusive alternative splicing event can have a profound effect rewiring animal metabolism and it will be interesting to establish whether alternative splicing of COQ-2 is sufficient to drive RQ biosynthesis, or whether additional innovations are needed.

Materials and methods

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional
information
Gene (Caenorhabditis elegans) coq-2WormBaseWBGene00000762
Strain, strain background
(Caenorhabditis elegans)
N2Caenorhabditis Genetics Center (CGC)N2Wild-type
Strain, strain background
(Caenorhabditis elegans)
coq-2 (syb1715)This paperPHX1715coq-2∆6a
Supplementary file 1
Strain, strain background
(Caenorhabditis elegans)
 coq-2 (syb1721)This paperPHX1721coq-2∆6e
Supplementary file 1
Strain, strain background (Escherichia coli)OP50Caenorhabditis Genetics Center (CGC)OP50
Sequence-based reagentsgRNAs used in this studyThis paperSupplementary file 1
Sequence-based reagentPCR primers used in this studyThis paperSupplementary file 1
Chemical compound, drugPotassium cyanide (KCN)Sigma-Aldrich60178–25GStock solution: 50 mM in PBS buffer
OtherHexane, HPLC gradeSigma-Aldrich650552–1L
OtherAcetonitrile, LC-MS gradeFisher ScientificA955-4
OtherUbiquinone-3 standardCampbell et al., 2019
OtherUbiquinone-9 standardSigma-Aldrich27597–1 MG
OtherRhodoquinone-9 standardRoberts Buceta et al., 2019Isolated from Ascaris suum
Software, algorithmWhippetSterne-Weiler et al., 2018RRID:SCR_018349https://github.com/timbitz/Whippet.jl
Software, algorithmHISAT2Kim et al., 2019RRID:SCR_015530https://daehwankimlab.github.io/hisat2/
Software, algorithmUCSF ChimeraResource for Biocomputing Visualization and Informatics Pettersen et al., 2004RRID:SCR_004097http://plato.cgl.ucsf.edu/chimera/
Software, algorithmMODELLERUniversity of California at San Francisco Webb and Sali, 2016RRID:SCR_008395https://salilab.org/modeller/
Software, algorithmImage analysis pipelineSpensley et al., 2018https://github.com/fraser-lab-UofT/acute_assay
Software, algorithmPythonPythonRRID:SCR_008394https://www.python.org/
Software, algorithmNIS-ElementsNikonRRID:SCR_014329https://www.microscope.healthcare.nikon.com/products/software/nis-elements
Software, algorithmImageMagickImageMagickRRID:SCR_014491https://imagemagick.org/
Software, algorithmBioFormatsOME - Open Microscopy EnvironmentRRID:SCR_000450https://docs.openmicroscopy.org/bio-formats/5.7.1/users/comlinetools/index.html

Sequence identification and analysis

Request a detailed protocol

To identify COQ-2 sequences from lineages known to synthesize RQ, we searched and analyzed from genomes and transcriptomes of platyhelminths (Schistosoma mansoni, Fasciola hepatica, and Schmidtea mediterranea), nematodes (Ascaris suum, Brugia malayi, Haemonchus contortus, Trichuris muris, Strongyloides stercoralis, Strongyloides ratti, Ancylostoma ceylanicum, and C. elegans), mollusks (Crassostrea virginica, and Biomphalaria glabrata), and annelids (Capitella teleta). Sequences were retrieved from https://parasite.wormbase.org (WBPS14), S. mediterranea database (http://smedgd.neuro.utah.edu), NCBI protein and nucleotide databases and UniProt (https://www.uniprot.org). Human, mice, and Saccharomyces cerevisiae COQ-2, eukaryotic lineages are known to be unable to synthesize RQ, were also identified for comparison. Searches were performed initially with BLASTP (protein databases) using human and C. elegans COQ-2 sequences as queries. Additionally, TBLASTN searches were performed using genomic sequences and cDNAs databases. This served to confirm the annotated protein sequences and to identify the non-annotated ones. Identified sequences were confirmed by best reciprocal hits in BLAST. Multiple sequence alignments were made with MUSCLE 3.8 (Chojnacki et al., 2017). Gaps were manually refined after alignment inspection.

RNA-seq analysis of mutually exclusive exons

Request a detailed protocol

To confirm if coq-2 exons are spliced in a mutually exclusive manner in other helminths, mollusks and annelids, we analyzed the existing RNA-seq data for evidence of alternative splicing (listed in Supplementary file 3). Whippet (v0.11) (Sterne-Weiler et al., 2018) was used to analyze RNA-seq data for quantification of AS events. To create a splicing index of exon-exon junctions in Whippet, genome annotations were taken from WormBase Parasite (WBPS14) and Ensembl Metazoa (Release 45). To identify novel exons and splice sites, reads were first aligned to the genome using HISAT2 (Kim et al., 2015). The BAM file generated was then used to supplement the existing genome annotations to create a splicing index of known and predicted exon-exon junctions in Whippet (using the --bam --bam-both-novel settings). Where required, TBLASTN data was also used to guide manual re-annotation of the coq-2 gene. Quantification of AS events was then performed by running whippet quant at default settings. This analysis was repeated for all species listed in Figure 5 using publicly available RNA-seq datasets (details on datasets used can be found in Supplementary file 3). A summary of coq-2 exons with reads that mapped to alternative exon-exon junctions are listed in Supplementary file 3. We also identified cases where both coq-2 exons were either included or skipped. However, since these are likely to be non-productive transcripts due to a pre-mature termination codon, we expressed exon usage as the proportion of events where either only the ‘a’ or the ‘e’ form is included.

Structural analysis

Request a detailed protocol

Multiple sequence alignment was performed using Clustal Omega (Madeira et al., 2019). The substrate-bound structure of a UbiA homolog from A. pernix (PDB: 4OD5) was displayed on Chimera (Pettersen et al., 2004) and the C. elegans sequence was threaded by homology using Modeller (Sali and Blundell, 1993; Webb and Sali, 2016).

Caenorhabditis elegans strains and culture conditions

Request a detailed protocol

The C. elegans wild-type Bristol strain (N2) was obtained from the Caenorhabditis Genetics Center (CGC, University of Minnesota, USA), which is supported by the National Institutes of Health-Office of Research Infrastructure Programs. The C. elegans mutant strains in coq-2 exon 6A (PHX1715, coq-2(syb1715)) and coq-2 exon 6E (PHX1721, coq-2(syb1721)) were generated by Suny Biotech Co., Ltd (Fuzhou City, China) using CRISPR/Cas9 system. The precise deletion of both mutant strains (134 bp) was verified by DNA sequencing the flanking region of exons 6a and 6e. The wild-type sequence, the deleted sequence in each strain, the sgRNAs, and primers used are listed in Supplementary file 1.

The general methods used for culturing and maintenance of C. elegans are described in Brenner, 1974. All chemical reagents were purchased from Sigma-Aldrich (St. Louis, MO). The E. coli OP50 strain, used as C. elegans food, was also received from CGC.

Lipid extraction and LC-MS quantitation

Request a detailed protocol

Lipid extractions of C. elegans N2 and mutant strains were performed on ~100 mg worm pellets after adding 1000 pmol UQ3 internal standard (Roberts Buceta et al., 2019). LC-MS samples were prepared from lipid extracts and diluted 1:100 (Bernert et al., 2019). Standards were extracted using the same method as for worm samples at the following concentrations: UQ3 (10 pmol/10 µL injection), RQ9 (0.75, 1.5, 3.0, 4.5, or 6.0 pmol/10 µL injection), and UQ9 (3.75, 7.5, 15.0, 22.5, or 30.0 pmol/10 µL injection). The UQ3 standard was synthesized at Gonzaga University (Campbell et al., 2019), the RQ9 standard was isolated by preparative chromatography from A. suum lipid extracts (Roberts Buceta et al., 2019) and the UQ9 standard was purchased (Sigma-Aldrich, St. Louis, MO). The general LC-MS conditions and parameters were previously reported (Bernert et al., 2019; Campbell et al., 2019). Samples were analyzed in quadruplicate and the pmol quinone was determined from the standard curve and corrected for recovery of internal standard. Samples were normalized by mg pellet mass.

Image-based KCN recovery assay

Request a detailed protocol

The KCN recovery assay was performed as previously described (Del Borrello et al., 2019; Spensley et al., 2018). Briefly, L1 worms were isolated by filtration through an 11 µm nylon mesh filter (Millipore: S5EJ008M04). Approximately, 100 L1 worms in M9 were dispensed to each well of a 96-well plate and an equal volume of potassium cyanide (KCN) (Sigma-Aldrich, St. Louis, MO) solution was then added to a final concentration of 200 µM KCN. Upon KCN addition, the plates were immediately sealed and incubated at room temperature for 15 hr on a rocking platform. After 15 hr, KCN was diluted 6-fold by addition of M9 buffer. Plates were immediately imaged on a Nikon Ti Eclipse microscope every 10 min for 3 hr. Fractional mobility scores (FMS) were then calculated using a custom image analysis pipeline (Spensley et al., 2018). For each strain, FMSfor the KCN-treated wells were normalized to the M9-only control wells at the first timepoint. Three technical replicates were carried out in each experiment and the final FMS taken from the mean of four biological replicates.

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files.

The following previously published data sets were used
    1. Gao F
    2. Liu X
    3. Wu XP
    4. Wang XL
    5. Gong D
    6. Lu H
    7. Xia Y
    8. Song Y
    9. Wang J
    10. Du J
    11. Liu S
    12. Han X
    13. Tang Y
    14. Yang H
    15. Jin Q
    16. Zhang X
    17. Liu M
    (2012) EBI European Nucleotide Archive
    ID PRJNA170655. Differential methylation in discrete developmental stages of the parasitic nematode Trichinella spiralis.
    1. Wang J
    2. Mitreva M
    3. Berriman M
    4. Thorne A
    5. Magrini V
    6. Koutsovoulos G
    7. Kumar S
    8. Blaxter ML
    9. Davis RE
    (2012) EBI European Nucleotide Archive
    ID SRP013573. Silencing of Germline-Expressed Genes by DNA Elimination in Somatic Cells.
    1. Choi YJ
    2. Ghedin E
    3. Berriman M
    4. McQuillan J
    5. Holroyd N
    6. Mayhew GF
    7. Christensen BM
    8. Michalski ML
    (2011) EBI European Nucleotide Archive
    ID ERP000948. Transcription profiling of of the human filarial nematode Brugia malayi across multiple life-cycle stages.
    1. The Wellcome Trust Sanger Institute
    (2013) EBI European Nucleotide Archive
    ID ERP002187. RNA-seq of Strongyloides ratti to investigate molecular basis of parasitism.
    1. Stoltzfus JD
    2. Minot S
    3. Berriman M
    4. Nolan TJ
    5. Lok JB
    (2012) EBI European Nucleotide Archive
    ID ERP001556. RNAseq of S. stercoralis PV001 strain developmental stages.
    1. Baskaran P
    2. Jaleta TG
    3. Streit A
    4. Rödelsperger C
    (2017) EBI European Nucleotide Archive
    ID PRJEB14543. Duplications and positive selection drive the evolution of parasitism associated gene families in the nematode Strongyloides papillosus.
    1. The Genome Institute
    (2016) EBI European Nucleotide Archive
    ID PRJNA72583. Ancylostoma ceylanicum Genome Sequencing.
    1. Eccles D
    2. Chandler J
    3. Camberis M
    4. Henrissat B
    5. Koren S
    6. Le Gros G
    7. Ewbank JJ
    (2018) EBI European Nucleotide Archive
    ID PRJEB20824. Nippostrongylus brasiliensis Genome & Transcriptome sequencing and assembly.
    1. Protasio AV
    2. Dunne DW
    3. Berriman M
    (2013) EBI European Nucleotide Archive
    ID ERP000427. RNA-Seq of Schistosoma mansoni (flatworms) larva and adult individuals at different life-stages.
    1. Kenny NJ
    2. Truchado-García M
    3. Grande C
    (2016) EBI European Nucleotide Archive
    ID PRJNA306682. Biomphalaria glabrata Transcriptome or Gene expression.
    1. Burns R
    2. Pechenik J
    (2016) EBI European Nucleotide Archive
    ID PRJNA379706. Capitella teleta RNA sequencing of competent and pre-competent larvae.

References

    1. Brenner S
    (1974)
    The genetics of Caenorhabditis elegans
    Genetics 77:71–94.
  1. Report
    1. Lümmen P
    2. Huang L
    3. Lisse D
    4. Berry EA
    (2014)
    Mitochondrial Respiratory Complex II: Molecular Target of Novel Active Ingredients for Pest Management
    San Francisco: 248th ACS National Meeting & Exposition.

Decision letter

  1. Gisela Storz
    Senior and Reviewing Editor; National Institute of Child Health and Human Development, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Thank you for submitting your article "Alternative splicing of COQ-2 determines the choice between ubiquinone and rhodoquinone biosynthesis in animals" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Gisela Storz as Reviewing and Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data.

Our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife, either of which would be linked to the original paper.

Summary:

Parasitic helminths are important human pathogens. It is important to find candidate therapies that specifically target these pathogens. Many parasitic helminths use rhodoquinone (RQ) instead of ubiquinone (UQ) when they live inside a host, in an anaerobic environment. Since the host does not make or use RQ, it could be an important pharmacological target. The authors previously showed that switch from UQ to RQ synthesis in parasitic helminths is driven by a change of substrate specificity of the polyprenyltransferase COQ-2 (Del Borrello et al., 2019; Roberts Buceta et al., 2019) when their living environment is switched from normoxia to anaerobic conditions. In this paper, they answer a simple but important question: what is the mechanism for substrate switching for COQ-2? The authors found that coq-2 has two different splicing isoforms, coq-2a and coq-2e, and that COQ-2a is specific for UQ synthesis while COQ-2e is specific for RQ synthesis. They use CRISPR/Cas9 to specifically delete each isoform in C. elegans and then measure UQ and RQ levels. The results clearly demonstrate the specificity of COQ-2a and COQ-2e for UQ and RQ, respectively. They further show the functional significance of this as animals that cannot synthesize RQ are hypersensitive to cyanide, which blocks the electron transport chain, and is a proxy for anaerobic conditions.

Revisions for this paper:

The proposed model is attractive and seems very promising and overall the reviewers were enthusiastic about the study. However, the paper is light on results. After discussion, the reviewers recommend that you consider the following experiments:

1) Directly validate splice forms' abundance on the RNA levels. The RNA-seq data meta-analysis is reasonable, but it would be good to see this validated.

2) Measure the level of 4HB and 3HA. If only the splice form selection drives the switch in quinone type then 4HB and 3HA levels should not change drastically.

3) Titrate KCN in Figure 5C experiment and also feed the helminths with 3HA. Data show that even the UQ-specific form can make a small amount of RQ. Is it enough to survive lower KCN treatment and can it be boosted by 3HA feeding?

4) Can the e-exon, which switches COQ2 to make more RQ, be found in any other organism? It was not clear in the text and methods if the sequence was "BLASTed" against all organisms to see if the e-exon exists in any non-Nematodes.

5) Analyze the in silico model and do binding simulations to explain why F204L and A243S changes would so strongly dictate precursor selection. In the human system, COQ2 can use many modified headgroups so why two mutations make it so specific?

Editorial points:

The text should also be revised to address the following:

– Throughout: The text could be improved to correctly describe the results of the experiments. For example, in the final conclusion authors say that "A switch between two mutually exclusive exons changes the core of the COQ-2 enzyme and switches it from generating UQ precursors to RQ precursors" however, Figure 5C shows that RQ-specific form makes both RQ and UQ, and also UQ-specific form can make a small amount of RQ. Therefore, it is somewhat misleading to suggest that individual splice forms can make only UQ or RQ.

– Introduction paragraph three: Maybe just start sentence with "RQ biosynthesis in animals requires precursors derived from tryptophan" (the evidence in Stairs is not convincing, it could be bacteria in the medium), the situation in microbes does not bear upon these finding. Bernet et al. is convincing, but concerns the R. rubrum gene.

– Please explain the conversion of the amino group in PABA to the hydroxyl group in HHB in Figure 2C.

– Introduction paragraph four: summarize the findings and restate the Abstract (mollusc and mollusk spelling occur). The meaning of "independent evolution" in is unclear.

– Figure 1 legend: ETF is not shown in the figure. For malate dismutation to operate (which it does, the endproduct stoichiometry is correct), fumarate reductase (CII) is the RQ oxidant (Müller et al., MMBR 2012).

– Subsection “Regulation of the alternative splicing of coq-2 in helminths” paragraph four: the data show that animal species

– Discussion paragraph three: all animal lineages

– Discussion: The authors favour independent origin of this mechanism and the exons. One needs to spell out what that means. That would mean that sequence homologous exons arose independently in independent lineages, which would mean that e arose from a each time, which is possible, and that e always arose in front of a in each of the 14 lineages surveyed, a 1/16000 proposition if all of the events were independent. Of course, if e arose from a in the common ancestor of these lineages followed by many differential losses in aerobic lineages, then we would have normal Darwinian evolution and no need for LGT or independent origins (googling differential gene loss in eukaryotes or evolution by gene loss returns a lot of hits, the evidence for the widespread occurrence of loss in eukaryotes is uncontroversial). Also possible. Of course, it is also possible that the alternative RQ specific exon here arose only once in one of these animals has been passed around via LGT among eukaryotic lineages (Leger et al., 2018) and specifically inserted into the COQ2 gene of the anaeobes in the conserved position (always in front of the UQ exon) in adaptation to anaerobic niches.

– What if we consider the possibility that a single origin and loss, not independent origins (by whatever mechanism) are the cause of the homologous exons (and exon order). What use would RQ be to the common ancestor of molluscs annelisds and nematodes? It would be essential.

– Animal aroses and diversified during a phase of Earth history in which oxygen was still low. Geologists have been saying for 20 years that the Proterozoic was anoxic (see Figure 1 in Lyons et al., 2014 or Figure 2 in Catling and Zahnle, 2020). Palaeontologists have been saying that animals arose and diversified in the Ediacaran (that is, Precambrian; <540 MY ago), nematode-mollusc divergence (do Reis et al., 2015) or the nematode annelid divergence (Parfrey et al., 2011). When I find 540 MY ago on the timescales in Lyons et al. and Catling and Zahnle, oxygen is low. Moreover, during the one billion years of eukaryote evolution before that, oxygen was low too. Some of us have been saying for 20 years (Zimorski et al., 2019; Gould et al., 2019) that survival in anaerobic environments was a normal thing that eukaryotes did for over a billlion years of evolution before life on land above the soli line. In that interpretation, the expectation would be that mechanisms of aerobic-anaerobic switching exist that are conserved across annelids, molluscs and nematodes (including free living and parasitic forms). My goodness, a Nobel Prize in Physiology and Medicine went for HIF last year, which senses low oxygen in animals. HIF is conserved across all animals. Did the HIF-dependent oxygen sensing cascade arise independently in all animals? Not likely. Was the HIF pathway laterally transferred? Not likely. I would think that the authors might wish to at least entertain the possibility that the anaerobic physiology that they see in these in animals (and eukaryotes) is conserved from the low oxygen and anaerobic past of animal and eukaryotic lineages. Is it possible that HIF sends the signal that regulates the RQ response? Yes, but I know of no evidence to suggest that, and there are HIF-independent possibilites. They mention Hochachka's work on vertebrates and the goldfish ethanol fermentation example. The vertebrate mechanisms (anaerobiosis in humans will work for about a minute) are not UQ dependent. This is the author's paper and they should write what they want. But if they say independent origins, that interpretation carries a lot of corollaries, none of which are spelled out in the paper, but are spelled out here. Of course, maybe 20 years of geochemical data on late oyxgen, the known age of fossil animals, the conservation of HIF and the conservation of pathways for eukaryote anaerobe survival are all wrong.

– Why does C. elegans make both RQ and UQ? A paragraph in the Discussion would be helpful, even if it is speculative.

– Do the other enzymes involved in RQ or UQ biosynthesis change in expression under anaerobic conditions?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Alternative splicing of COQ-2 determines the choice between ubiquinone and rhodoquinone biosynthesis in animals" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by Gisela Storz as the Reviewing and Senior Editor. The reviewers have opted to remain anonymous.

Summary:

Parasitic helminths are important human pathogens. It is important to find candidate therapies that specifically target these pathogens. Many parasitic helminths use rhodoquinone (RQ) instead of ubiquinone (UQ) when they live inside a host, in an anaerobic environment. Since the host does not make or use RQ, it could be an important pharmacological target. The authors previously showed that switch from UQ to RQ synthesis in parasitic helminths is driven by a change of substrate specificity of the polyprenyltransferase COQ-2 (Del Borrello et al., 2019; Roberts Buceta et al., 2019) when their living environment is switched from normoxia to anaerobic conditions. In this paper, they answer a simple but important question: what is the mechanism for substrate switching for COQ-2? The authors found that coq-2 has two different splicing isoforms, coq-2a and coq-2e, and that COQ-2a is specific for UQ synthesis while COQ-2e is specific for RQ synthesis. They use CRISPR/Cas9 to specifically delete each isoform in C. elegans and then measure UQ and RQ levels. The results clearly demonstrate the specificity of COQ-2a and COQ-2e for UQ and RQ, respectively. They further show the functional significance of this as animals that cannot synthesize RQ are hypersensitive to cyanide, which blocks the electron transport chain, and is a proxy for anaerobic conditions.

Revisions:

Reviewer #2:

The authors have put worth a reasonable revision that provided very little new data, but nonetheless somewhat bolsters their compelling model. Given that the authors haven't, and perhaps can't, provide certain validations of their data, I think the title and some language should be toned down. In the end, both splice forms can make both quinones, but at a different ratio. Also, both splice forms are often co-expressed. The levels of 4HB and 3HA headgroups were not measured so how they impact levels of the quinones is unknown. It will require a significant amount of biochemical data to fully understand the exact differences between the COQ-2a and COQ-2e splice forms. However, it is quite unlikely that the general model put forth is incorrect.

Reviewer #3:

I continue to think this is a nice study that should be published in eLife. The authors did a nice job in their rebuttal and I am satisfied overall. I agree that RNA-seq is totally fine, given the depth and precision of the method. However, I find parts of the Discussion a bit unsatisfying. For instance, what is the function of alternative splicing of mrp-1? There is no reference for the statement that this transporter transports B12 into the mitochondria. Also, it is not clear that vitamin B12 is required for RQ synthesis in C. elegans, even though it is stated. Finally, the relationship with branched chain fatty acid synthesis is unclear to me. However, clarifying this is up to the authors and shouldn't prevent acceptance of the work.

https://doi.org/10.7554/eLife.56376.sa1

Author response

Revisions for this paper:

The proposed model is attractive and seems very promising and overall the reviewers were enthusiastic about the study. However, the paper is light on results. After discussion, the reviewers recommend that you consider the following experiments:

1) Directly validate splice forms' abundance on the RNA levels. The RNA-seq data meta-analysis is reasonable, but it would be good to see this validated.

We agree with the reviewers that it is crucial that central methods are carefully validated. Our primary way to analyse splicing throughout this paper is with RNA-seq. We absolutely agree that such an approach must be validated carefully and we have done rigorous validation of the RNA-seq splicing analysis pipeline that we employ here, specifically in C. elegans, the model helminth. In our publication https://genome.cshlp.org/content/21/2/342.abstract, we used this same RNA-seq pipeline along with a completely independent method using microarrays to analyse splicing in C. elegans and extensively validated the data using RT-PCR. We found that the 2 independent platforms correlated with an R2=0.85 and furthermore, that we could confirm 92% of identified splice forms using RT-PCR. The RNA-seq pipeline is thus robust and accurate. In addition to that validation in our group, the C. elegans transcriptome has been also been analysed using direct sequencing on a nanopore platform (which we refer to in the text), another independent platform — the RNAseq data match very closely to the nanopore data for coq-2. Finally, another independent group specifically validated the coq-2 mutually exclusive splicing event using RT-PCR ( https://pubmed.ncbi.nlm.nih.gov/25254147/ ). We therefore feel that our RNA-seq-based pipeline for analysing alternative splicing has been validated extremely carefully — nanopore, microarray, RT-PCR all confirm its accuracy and coverage in general and specifically in the case of coq-2, both nanopore and RT-PCR confirm our analysis.

This validation has all been done in C. elegans. We recognize that we did not validate the RNA-seq splicing analysis in the other species being discussed. But we are using exactly the same pipeline as has been extensively validated in C. elegans, a nematode, and are applying it to other nematodes. The data presented for Ascaris and S. stercoralis show extremely high agreement between independent biological repeats and these genomes are highly compact so splice analysis is easy and robust. We don’t fully understand why the reviewers think that our highly validated pipeline would be robust in one nematode but not in another — I’m unclear what the underlying hypothesis for this would be? It does feel that at some point a highly validated technology should be trusted unless there’s a clear reason for scepticism and I’m not sure what that would be in this case. At any rate, we are happy to include the following sentences in the paper:

“We note that while the alternative splicing of coq-2 in C. elegans we observe using RNA-seq has been validated by direct sequencing, microarray analysis, and RT-PCR, the data on the other helminths only derives from RNA-seq and genome analysis. While it uses the same robust RNA-seq analysis pipeline, it would be ideal to validate these changes in the future.”

2) Measure the level of 4HB and 3HA. If only the splice form selection drives the switch in quinone type then 4HB and 3HA levels should not change drastically.

We do indeed intend to carry out metabolomics analysis on the different C. elegans strains. We propose to add the following text to address this point.

“While these engineered changes could potentially affect RQ levels via some indirect mechanism such as alterations in the levels of 4HB and 3HA, it is unclear what mechanism would drive those metabolic rewirings. We thus suggest that the more parsimonious explanation is the one we propose here: that COQ-2a and COQ-2e have different substrate preference due to the large change to the core of the enzyme and that this is what drives the shift from UQ to RQ synthesis.”

3) Titrate KCN in Figure 5 C experiment and also feed the helminths with 3HA. Data show that even the UQ-specific form can make a small amount of RQ. Is it enough to survive lower KCN treatment and can it be boosted by 3HA feeding?

We have titrated the KCN and show these data now in Figure 5—figure supplement 1, we thank the reviewers for asking for this and should have included this initially. Doses of KCN below ~150 µM fail to drive worms into RQ-dependent metabolism so we focus on doses above 150 µM KCN and we can clearly see that the coq-2e strain is not distinguishable from the kynu-1 mutant strain that makes no detectable RQ (our previous paper https://elifesciences.org/articles/48165). We therefore conclude that the coq-∆2e strain behaves like a RQ-deficient strain in our assays so while it may make a very small amount of RQ, this is insufficient for survival. We have not tried to feed worms with large amounts of 3HA to try to drive the coq-∆2e strain to make increased levels of RQ for two reasons. First, it didn’t seem to us to affect the conclusions — that the splicing of coq-2 affects the balance between UQ and RQ synthesis. Would the finding that adding a massive excess of 3HA could drive some small change in RQ synth in the coq-∆2e strain affect that central conclusion? In our eyes it would not. The second is that 3HA connects with other pathways in both the bacterial food and the worms and so we felt it was not a clean experiment — it would be hard to ascribe changes specifically to coq-2 and changes in bacterial diet and metabolic state can have major outcomes on RQ-dependent metabolism and worm phenotypes. On this line, we tried something along these lines first by supplementing bacteria with additional Trp — this caused the worms fed on these supplemented bacteria to arrest and die. This was due to some metabolic change in the bacterial food since if we did the same with metabolically inert heat-killed bacteria, we do not see any obvious toxicity of Trp. However, worms grow substantially worse on the heat killed bacteria, adding yet another confounder. We therefore hope the reviewers are happy that we have not done this specific experiment and that they are satisfied with the figure supplement.

4) Can the e-exon, which switches COQ2 to make more RQ, be found in any other organism? It was not clear in the text and methods if the sequence was "BLASTed" against all organisms to see if the e-exon exists in any non-Nematodes.

We apologize if this was unclear in the paper — this is absolutely central to the findings. The e-exon is ONLY present in the species that make RQ — helminths, molluscs and annelids. None of the other animals we examined encode such an exon in their genome and we originally stated “This gene structure is only seen in species that make RQ — we find no evidence in any available data for similar alternative splicing in any mammalian hosts (human and mouse are shown as representatives in Figure 3) or in other lineages that lack RQ, such as yeasts.” We agree that this could be made stronger and so now write that “we only find evidence for alternative splicing between a coq-2a form and a coq-2e like form in animals that make RQ. Crucially, no mammalian hosts show any evidence for this kind of alternative splicing of their coq-2 orthologues, nor do they have an e-like exon either in their gene predictions or genome sequences or in any available RNA-seq or direct transcriptome sequence data (human and mouse are shown as representatives in Figure 3). Note that the RNA-seq datasets are much deeper and more extensive in humans and mouse than in helminths, molluscs or annelids, so this lack of evidence for such alternative splicing is unlikely to be due to a failure to detect such events due to low coverage transcriptome data. We conclude that the e-exon and the mutually exclusive alternative splicing of coq-2 is unique to animals that make RQ and is not found in any other species.”

5) Analyze the in silico model and do binding simulations to explain why F204L and A243S changes would so strongly dictate precursor selection. In the human system, COQ2 can use many modified headgroups so why two mutations make it so specific?

This is a fantastic question and we are beginning to undertake binding studies on COQ-2 and several other enzymes. However, we note that the structure we used to model the helminth COQ-2 structures is very evolutionarily distant — there are only 2 available COQ-2 structures and both are from Archaea. This means that while the structure is likely to be substantially correct, the precise spatial location of each side chain is not guaranteed with enough precision to do any reasonable modeling of binding (at least according to expert colleagues!). Thus while the location of the key helices that change from COQ-2a to COQ-2e is broadly correct, the precise positioning of the 2 side chains highlighted in the paper cannot be known with sufficient precision. It’s frustrating, and we would love to know more but that’s the best we can do at this stage and hope the reviewers understand the limitations of the available structures.

Editorial points:

The text should also be revised to address the following:

– Throughout: The text could be improved to correctly describe the results of the experiments. For example, in the final conclusion authors say that "A switch between two mutually exclusive exons changes the core of the COQ-2 enzyme and switches it from generating UQ precursors to RQ precursors" however, Figure 5C shows that RQ-specific form makes both RQ and UQ, and also UQ-specific form can make a small amount of RQ. Therefore, it is somewhat misleading to suggest that individual splice forms can make only UQ or RQ.

We understand there was a lack of clarity in some of our explanations. We updated the summary paragraph at the end of the Results section to include the sentence:

“A switch between two mutually exclusive exons changes the core of the COQ-2 enzyme and switches it from primarily generating UQ precursors to primarily RQ precursors. We propose that this switch ultimately results in a shift in quinone content from high UQ in aerobic conditions to high RQ in anaerobic conditions.”

Prior to this paragraph we added another conclusion sentence:

“We conclude that increased inclusion of the coq-2e-specific exon correlates with increases synthesis of RQ during the parasite life cycles.”

We also rewrote the last paragraph in the Introduction to further clarify these concerns:

In this study, we reveal that two variants of COQ-2, derived from alternative splicing of mutually exclusive exons, are the key for the choice to make RQ or UQ. We find that one of the mutually exclusive exons is only found in the genomes of animals that make RQ and that its inclusion remodels the core of the COQ-2 enzyme. We show that the removal this exon abolishes RQ biosynthesis in C. elegans. Finally, we find that inclusion of this RQ-specific exon expression is increased in the stages of the life cycle of parasites where they encounter hypoxic conditions and where they increase RQ production, while the alternative exon is increased in normoxic life stages. We thus conclude that alternative splicing of COQ-2 is the key mechanism that causes the switch from UQ to RQ synthesis in the parasite life-cycle.”

– Introduction paragraph three: Maybe just start sentence with "RQ biosynthesis in animals requires precursors derived from tryptophan" (the evidence in Stairs is not convincing, it could be bacteria in the medium), the situation in microbes does not bear upon these finding. Bernet et al. is convincing, but concerns the R. rubrum gene.

This sentence was simplified. The protist example and the Stairs paper were removed from the sentence. There is a striking contrast between RQ biosynthesis in bacteria and animals and this will be the topic of an upcoming review that two of our co-authors are writing for BBA Bioenergetics. “In contrast to bacteria where RQ derives from UQ (Bernert et al., 2019; Brajcich et al., 2010), RQ biosynthesis in animals requires precursors derived from tryptophan (Figure 1B).”

– Please explain the conversion of the amino group in PABA to the hydroxyl group in HHB in Figure 2C.

The figure was changed to show prenylation of pABA or 4HB gives HHB or HAB, respectively. This was clarified in the figure caption:

“Prenylation is facilitated by Coq2 to form 3-hexaprenyl-4-hydroxybenzoic acid (HHB) or 3-hexaprenyl-4-aminobenzoic acid (HAB), where n = 6. Further functionalization of these intermediates occurs through a CoQ synthome (Coq3-Coq9 and Coq11) to yield UQ.”

– Introduction paragraph four: summarize the findings and restate the Abstract (mollusc and mollusk spelling occur). The meaning of "independent evolution" in is unclear.

Spelling was changed to mollusc throughout.

The statement on “Independent evolution” was removed from the Introduction and is addressed in the final paragraph of the Discussion.

– Figure 1 legend: ETF is not shown in the figure. For malate dismutation to operate (which it does, the endproduct stoichiometry is correct), fumarate reductase (CII) is the RQ oxidant (Müller et al., MMBR 2012).

ETF was removed from the figure caption. We would like to maintain Quinone-coupled dehydrogenases (QDH) in the scheme, since QDHs are broader than just complex II (it also includes quinone-dependent dehydro-orotate dehydrogenase and flavoprotein electron-transfer dehydrogenase). We think that something general is more appropriate. In any case, we included Complex II and fumarate as examples in the caption. For clarity, Panel A was rewritten as follows:

“In aerobic metabolism, ubiquinone (UQ) shuttles electrons in the ETC from Complex I and quinone-coupled dehydrogenases (QDHs), such as Complex II. These electrons are ultimately transferred to oxygen. In anaerobic metabolism, rhodoquinone (RQ) reverses electron flow in QDHs and facilitates an early exit of electrons from the ETC on to anaerobic electron acceptors (Ae-A), such as fumarate.”

– Subsection “Regulation of the alternative splicing of coq-2 in helminths” paragraph four: the data show that animal species

This was changed.

– Discussion paragraph three: all animal lineages

This was changed.

– Discussion: The authors favour independent origin of this mechanism and the exons. One needs to spell out what that means. That would mean that sequence homologous exons arose independently in independent lineages, which would mean that e arose from a each time, which is possible, and that e always arose in front of a in each of the 14 lineages surveyed, a 1/16000 proposition if all of the events were independent. Of course, if e arose from a in the common ancestor of these lineages followed by many differential losses in aerobic lineages, then we would have normal Darwinian evolution and no need for LGT or independent origins (googling differential gene loss in eukaryotes or evolution by gene loss returns a lot of hits, the evidence for the widespread occurrence of loss in eukaryotes is uncontroversial). Also possible. Of course, it is also possible that the alternative RQ specific exon here arose only once in one of these animals has been passed around via LGT among eukaryotic lineages (Leger et al., 2018) and specifically inserted into the COQ2 gene of the anaeobes in the conserved position (always in front of the UQ exon) in adaptation to anaerobic niches.

– What if we consider the possibility that a single origin and loss, not independent origins (by whatever mechanism) are the cause of the homologous exons (and exon order). What use would RQ be to the common ancestor of molluscs annelisds and nematodes? It would be essential.

We thank the editor for these thoughtful comments. We concur with the editor that we were overenthusiastic in supporting the independent evolution of coq-2 alternative splicing in different lineages. We changed our somewhat biased suggestions to a more balanced view, in which we discuss alternative evolutionary scenarios. We conclude that our current data can neither support an ancestral animal origin of RQ biosynthesis followed by independent losses, or independent animal origin of RQ biosynthesis. We added the following statement to the Discussion:

“It is unclear that our findings here definitively answer this but we do note that there is a precedent for a mutually exclusive splicing event that has arisen independently in helminths, annelids, and molluscs in the gene mrp-1 (Yue et al., 2017).”

We extensively modified the last paragraphs of the Discussion, which now reads:

This example indicates it is at least plausible that alternative splicing of coq-2 might have originated independently in helminths, annelids, and molluscs, but we do not believe that our current data can distinguish between an ancestral animal origin of RQ biosynthesis followed by independent losses, or independent animal origin of RQ biosynthesis due to an “mrp-1-like” independent evolution of alternative splicing of coq-2. Future studies should shed on light on this issue. Whatever the evolutionary scenario, our results show that a simple innovation of a mutually-exclusive alternative splicing event can have a profound effect rewiring animal metabolism and it will be interesting to establish whether alternative splicing of COQ-2 is sufficient to drive RQ biosynthesis, or whether additional innovations are needed.”

– Animal aroses and diversified during a phase of Earth history in which oxygen was still low. Geologists have been saying for 20 years that the Proterozoic was anoxic (see Figure 1 in Lyons et al., 2014 or Figure 2 in Catling and Zahnle, 2020). Palaeontologists have been saying that animals arose and diversified in the Ediacaran (that is, Precambrian; <540 MY ago), nematode-mollusc divergence (do Reis et al., 2015) or the nematode annelid divergence (Parfrey et al., 2011). When I find 540 MY ago on the timescales in Lyons et al. and Catling and Zahnle, oxygen is low. Moreover, during the one billion years of eukaryote evolution before that, oxygen was low too. Some of us have been saying for 20 years (Zimorski et al., 2019; Gould et al., 2019) that survival in anaerobic environments was a normal thing that eukaryotes did for over a billlion years of evolution before life on land above the soli line. In that interpretation, the expectation would be that mechanisms of aerobic-anaerobic switching exist that are conserved across annelids, molluscs and nematodes (including free living and parasitic forms). My goodness, a Nobel Prize in Physiology and Medicine went for HIF last year, which senses low oxygen in animals. HIF is conserved across all animals. Did the HIF-dependent oxygen sensing cascade arise independently in all animals? Not likely. Was the HIF pathway laterally transferred? Not likely. I would think that the authors might wish to at least entertain the possibility that the anaerobic physiology that they see in these in animals (and eukaryotes) is conserved from the low oxygen and anaerobic past of animal and eukaryotic lineages. Is it possible that HIF sends the signal that regulates the RQ response? Yes, but I know of no evidence to suggest that, and there are HIF-independent possibilites. They mention Hochachka's work on vertebrates and the goldfish ethanol fermentation example. The vertebrate mechanisms (anaerobiosis in humans will work for about a minute) are not UQ dependent. This is the author's paper and they should write what they want. But if they say independent origins, that interpretation carries a lot of corollaries, none of which are spelled out in the paper, but are spelled out here. Of course, maybe 20 years of geochemical data on late oyxgen, the known age of fossil animals, the conservation of HIF and the conservation of pathways for eukaryote anaerobe survival are all wrong.

The issues raised by the editor are fascinating. However, we have not included them in the discussion, since they remain highly speculative.

– Why does C. elegans make both RQ and UQ? A paragraph in the Discussion would be helpful, even if it is speculative.

We thank the editor for this comment. This is now discussed. We include the following text in the Discussion:

“C. elegans is a free-living nematode that does not face extended anaerobic conditions as an obligate part of its life-cycle. However, Pseudomonas aeruginosa, a natural pathogen of C. elegans, uses cyanide to kill the worm (Gallagher and Manoil, 2001). Cyanide blocks the conventional ETC at complex IV preventing the use of oxygen as a final electron acceptor. We speculate that in C. elegans, the alternative ETC may be an adaptive strategy to withstand a transient cyanide stress from pathogens. For parasitic helminths like Ascaris, the need to respire anaerobically is critical for their life cycle since they must survive for long periods in the anaerobic environment of the human gut.”

– Do the other enzymes involved in RQ or UQ biosynthesis change in expression under anaerobic conditions?

This is a very interesting question that we would like to explore. However, we do not have data to provide for this manuscript.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Revisions:

Reviewer #2:

The authors have put worth a reasonable revision that provided very little new data, but nonetheless somewhat bolsters their compelling model. Given that the authors haven't, and perhaps can't, provide certain validations of their data, I think the title and some language should be toned down. In the end, both splice forms can make both quinones, but at a different ratio. Also, both splice forms are often co-expressed. The levels of 4HB and 3HA headgroups were not measured so how they impact levels of the quinones is unknown. It will require a significant amount of biochemical data to fully understand the exact differences between the COQ-2a and COQ-2e splice forms. However, it is quite unlikely that the general model put forth is incorrect.

Reviewer #3:

I continue to think this is a nice study that should be published in eLife. The authors did a nice job in their rebuttal and I am satisfied overall. I agree that RNA-seq is totally fine, given the depth and precision of the method. However, I find parts of the Discussion a bit unsatisfying. For instance, what is the function of alternative splicing of mrp-1? There is no reference for the statement that this transporter transports B12 into the mitochondria. Also, it is not clear that vitamin B12 is required for RQ synthesis in C. elegans, even though it is stated. Finally, the relationship with branched chain fatty acid synthesis is unclear to me. However, clarifying this is up to the authors and shouldn't prevent acceptance of the work.

We have tried to address the additional comments raised by the two reviewers regarding our previous revised manuscript. Specifically, we have changed the title to something that is more precise and is formally correct and have toned down some of the language as requested. For example instead of stating that the COQ-2e isoform is required for RQ synthesis, we now say that it is required for efficient RQ synthesis since the other isoform does indeed make an incredibly low level of RQ. We have also removed a small section in the Discussion to address the question raised by the other reviewer about mrp-1 and the B12 involvement in RQ-dependent metabolism. It was not central for the paper or conclusions in any way and we actually prefer it as it reads now, so we thank this reviewer for that suggestion.

References

dos Reis et al. (2015) Uncertainty in the timing of origin of animals and the limits of precision in molecular timescales. Curr Biol 25, 2939-2950,

Parfrey LW, Lahr DJ, Knoll AH, Katz LA. 2011. Estimating the timing of early eukaryotic diversification with multigene molecular clocks. Proc Natl Acad Sci USA 108: 13624-13629.

Catling and Zahnle, Sci. Adv. 2020;6: eaax1420

Lyons TW et al. 2014. The rise of oxygen in Earth's early ocean and atmosphere. Nature 506: 307-315.

Leger et al. 2018. Demystifying eukaryote lateral gene transfer. BioEssays, 40, 1700242

Gould SB et al. Adaptation to life on land and high oxygen via transition from ferredoxin- to NADH-dependent redox balance. Proc. Roy. Soc. Lond. B. 286: 20191491 (2019)

Zimorski V et al. Energy metabolism in anaerobic eukaryotes and Earth's late oxygenation. Free Radicals Biol. Med. 140:279-294 (2019).

https://doi.org/10.7554/eLife.56376.sa2

Article and author information

Author details

  1. June H Tan

    The Donnelly Centre, University of Toronto, Toronto, Canada
    Contribution
    Data curation, Formal analysis, Investigation, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6597-3952
  2. Margot Lautens

    The Donnelly Centre, University of Toronto, Toronto, Canada
    Contribution
    Data curation, Formal analysis, Investigation, Visualization
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8503-9603
  3. Laura Romanelli-Cedrez

    Laboratorio de Biología de Gusanos. Unidad Mixta, Departamento de Biociencias, Facultad de Química, Universidad de la República - Institut Pasteur de Montevideo, Montevideo, Uruguay
    Contribution
    Data curation, Formal analysis, Investigation
    Competing interests
    No competing interests declared
  4. Jianbin Wang

    1. Department of Biochemistry and Molecular Genetics, RNA Bioscience Initiative, University of Colorado School of Medicine, Aurora, United States
    2. Department of Biochemistry and Cellular and Molecular Biology, University of Tennessee, Knoxville, United States
    Contribution
    Data curation, Formal analysis, Investigation
    Competing interests
    No competing interests declared
  5. Michael R Schertzberg

    The Donnelly Centre, University of Toronto, Toronto, Canada
    Contribution
    Conceptualization, Data curation, Investigation
    Competing interests
    No competing interests declared
  6. Samantha R Reinl

    Department of Chemistry and Biochemistry, Gonzaga University, Spokane, United States
    Contribution
    Formal analysis, Investigation
    Competing interests
    No competing interests declared
  7. Richard E Davis

    Department of Biochemistry and Molecular Genetics, RNA Bioscience Initiative, University of Colorado School of Medicine, Aurora, United States
    Contribution
    Supervision, Funding acquisition, Investigation, Project administration
    Competing interests
    No competing interests declared
  8. Jennifer N Shepherd

    Department of Chemistry and Biochemistry, Gonzaga University, Spokane, United States
    Contribution
    Conceptualization, Supervision, Funding acquisition, Investigation, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    shepherd@gonzaga.edu
    Competing interests
    No competing interests declared
  9. Andrew G Fraser

    The Donnelly Centre, University of Toronto, Toronto, Canada
    Contribution
    Conceptualization, Supervision, Funding acquisition, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    andyfraser.utoronto@gmail.com
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9939-6014
  10. Gustavo Salinas

    Laboratorio de Biología de Gusanos. Unidad Mixta, Departamento de Biociencias, Facultad de Química, Universidad de la República - Institut Pasteur de Montevideo, Montevideo, Uruguay
    Contribution
    Conceptualization, Supervision, Funding acquisition, Investigation, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    gsalin@fq.edu.uy
    Competing interests
    No competing interests declared

Funding

Canadian Institutes of Health Research (501584)

  • Andrew G Fraser

Canadian Institutes of Health Research (5003009)

  • Andrew G Fraser

Agencia Nacional de Investigación e Innovación (FCE_2014_1_104366)

  • Gustavo Salinas

Agencia Nacional de Investigación e Innovación (FCE_1_2019_1_155779)

  • Gustavo Salinas

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

C. elegans strain N2 and E. coli OP50 were provided by the Caenorhabditis Genetics Center, which is funded by NIH Office of Research Infrastructure Program: P40 OD010440. We thank Exequiel Barrera, John Calarco, and Amy Caudy for helpful discussions. We would also like to thank Amy Buck, Niki Gounaris, Jim Collins, Rick Maizels, and Murray Selkirk for organizing the Hydra meeting on Parasitic Helminths where this collaboration started.

Senior and Reviewing Editor

  1. Gisela Storz, National Institute of Child Health and Human Development, United States

Version history

  1. Received: March 5, 2020
  2. Accepted: August 2, 2020
  3. Accepted Manuscript published: August 3, 2020 (version 1)
  4. Version of Record published: August 18, 2020 (version 2)

Copyright

© 2020, Tan et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,450
    Page views
  • 182
    Downloads
  • 8
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. June H Tan
  2. Margot Lautens
  3. Laura Romanelli-Cedrez
  4. Jianbin Wang
  5. Michael R Schertzberg
  6. Samantha R Reinl
  7. Richard E Davis
  8. Jennifer N Shepherd
  9. Andrew G Fraser
  10. Gustavo Salinas
(2020)
Alternative splicing of coq-2 controls the levels of rhodoquinone in animals
eLife 9:e56376.
https://doi.org/10.7554/eLife.56376

Further reading

    1. Epidemiology and Global Health
    2. Medicine
    Jeffrey Thompson, Yidi Wang ... Ulrich H von Andrian
    Research Article Updated

    Background:

    Although there are several efficacious vaccines against COVID-19, vaccination rates in many regions around the world remain insufficient to prevent continued high disease burden and emergence of viral variants. Repurposing of existing therapeutics that prevent or mitigate severe COVID-19 could help to address these challenges. The objective of this study was to determine whether prior use of bisphosphonates is associated with reduced incidence and/or severity of COVID-19.

    Methods:

    A retrospective cohort study utilizing payer-complete health insurance claims data from 8,239,790 patients with continuous medical and prescription insurance January 1, 2019 to June 30, 2020 was performed. The primary exposure of interest was use of any bisphosphonate from January 1, 2019 to February 29, 2020. Bisphosphonate users were identified as patients having at least one bisphosphonate claim during this period, who were then 1:1 propensity score-matched to bisphosphonate non-users by age, gender, insurance type, primary-care-provider visit in 2019, and comorbidity burden. Main outcomes of interest included: (a) any testing for SARS-CoV-2 infection; (b) COVID-19 diagnosis; and (c) hospitalization with a COVID-19 diagnosis between March 1, 2020 and June 30, 2020. Multiple sensitivity analyses were also performed to assess core study outcomes amongst more restrictive matches between BP users/non-users, as well as assessing the relationship between BP-use and other respiratory infections (pneumonia, acute bronchitis) both during the same study period as well as before the COVID outbreak.

    Results:

    A total of 7,906,603 patients for whom continuous medical and prescription insurance information was available were selected. A total of 450,366 bisphosphonate users were identified and 1:1 propensity score-matched to bisphosphonate non-users. Bisphosphonate users had lower odds ratios (OR) of testing for SARS-CoV-2 infection (OR = 0.22; 95%CI:0.21–0.23; p<0.001), COVID-19 diagnosis (OR = 0.23; 95%CI:0.22–0.24; p<0.001), and COVID-19-related hospitalization (OR = 0.26; 95%CI:0.24–0.29; p<0.001). Sensitivity analyses yielded results consistent with the primary analysis. Bisphosphonate-use was also associated with decreased odds of acute bronchitis (OR = 0.23; 95%CI:0.22–0.23; p<0.001) or pneumonia (OR = 0.32; 95%CI:0.31–0.34; p<0.001) in 2019, suggesting that bisphosphonates may protect against respiratory infections by a variety of pathogens, including but not limited to SARS-CoV-2.

    Conclusions:

    Prior bisphosphonate-use was associated with dramatically reduced odds of SARS-CoV-2 testing, COVID-19 diagnosis, and COVID-19-related hospitalizations. Prospective clinical trials will be required to establish a causal role for bisphosphonate-use in COVID-19-related outcomes.

    Funding:

    This study was supported by NIH grants, AR068383 and AI155865, a grant from MassCPR (to UHvA) and a CRI Irvington postdoctoral fellowship, CRI2453 (to PH).

    1. Epidemiology and Global Health
    Tianyi Huang
    Insight

    A large observational study has found that irregular sleep-wake patterns are associated with a higher risk of overall mortality, and also mortality from cancers and cardiovascular disease.