Maleidride biosynthesis – construction of dimeric anhydrides – more than just heads or tails

Katherine Williams *^a, Agnieszka J. Szwalbe ^b, Kate M. J. de Mattos-Shipley ^a, Andy M. Bailey ^a, Russell J. Cox ^c and Christine L. Willis ^d
aSchool of Biological Sciences, Life Sciences Building, University of Bristol, 24 Tyndall Ave, Bristol BS8 1TQ, UK. E-mail: katherine.williams@bristol.ac.uk
^bCelon Pharma, Warsaw, Poland
^cInstitute for Organic Chemistry and BMWZ, Leibniz University of Hannover, Schneiderberg 38, 30167, Hannover, Germany
^dSchool of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 1TS, UK

First published on 21st September 2022

---

---

Katherine Williams

Katherine Williams received her PhD from the University of Bristol in 2010, before moving to the Bristol Polyketide Group, investigating the biosynthesis of fungal metabolites with interesting bioactive properties. Research posts at the Leibniz Universität Hannover, Germany with Prof. Russell Cox, and at Cardiff University with Prof. Ruedi Allemann followed. Subsequently, Katherine moved back to the University of Bristol, working with Dr Andy Bailey and Prof. Chris Willis, on a project developing a high-throughput heterologous production platform for fungal natural product antibiotic discovery.

Agnieszka J. Szwalbe

Agnieszka Szwalbe obtained her bachelor's degrees in both Chemistry (2012) and Biotechnology (2012) at the University of Warsaw, Poland. She then studied biosynthesis of maleidrides during PhD studies under the supervision of Prof. Russell Cox and Prof. Tom Simpson at Bristol University, UK, and graduated in 2016. She is currently pursuing a career as an analytical chemist in medicinal chemistry division at Celon Pharma (Poland).

Kate M. J. de Mattos-Shipley

Kate de Mattos-Shipley obtained her PhD from Bristol University in 2011, for her investigations into the biosynthesis of a mushroom-derived antibiotic called pleuromutilin. She spent the next 10 years studying a wide range of fungal natural products, with a focus on genome mining, bioinformatic analyses and pathway elucidation through genetic manipulations and heterologous production. In 2021 she left the bench behind after accepting a position as a biotechnology editor for Nature Communications.

Andy M. Bailey

Andy Bailey is a senior lecturer at the University of Bristol, UK. His research interests are based around different aspects of fungal biology and their analysis using molecular genetic approaches. This includes genome mining to explore fungal secondary metabolism, fungi as pathogens of plants and invertebrates and other fungi, plus establishing methods for genetic analysis of basidiomycetes.

Russell J. Cox

Russell Cox was born in 1967 in the New Forest in the UK where he grew up. He studied chemistry at the University of Durham, and then worked with Prof. David O'Hagan at the same institution for his PhD, studying the biosynthesis of fungal metabolites. Post-doctoral periods with Professor John Vederas FRS in Edmonton Alberta, and Professors David Hopwood FRS and Tom Simpson FRS at Norwich and Bristol in the UK were followed by his appointment as a lecturer in the School of Chemistry at the University of Bristol where he rose to become full Professor of Organic and Biological Chemistry. He moved to become Professor of Microbiological Chemistry at the Leibniz Universität Hannover in Germany in 2013. He has served as an editorial board member of Natural Product Reports until 2012, and has been past chair of the Directing Biosynthesis series of scientific conferences. He is currently chair of the editorial board of RSC Advances, and Head of the Institute for Organic Chemistry at the Leibniz Universität Hannover.

Christine L. Willis

Chris Willis is Professor of Organic Chemistry and Head of Organic and Biological Chemistry at the University of Bristol. Her research focuses on natural product biosynthesis including the application of total synthesis, isotopic labelling, pathway engineering and mechanistic studies to produce biocatalysts and new bioactive molecules. She was the recipient of the Natural Product Chemistry Award of the Royal Society of Chemistry in 2020.

---

1. Introduction

Maleidrides are a group of biosynthetically related polyketide-based natural products that have been isolated from diverse filamentous fungi.^1,2 They contain at least one maleic anhydride moiety fused to a central carbocyclic core. There are three groups of maleidrides classified by the number of carbons in the central ring structure, the nonadrides (nine carbons), octadrides (eight carbons) and heptadrides (seven carbons) (Fig. 1A).¹ Other maleic anhydride based metabolites are known,² for example the cordyandhydrides³ and the tropolones.⁴ However, maleidrides are specifically formed by the coupling of two monomer units (1–3, Fig. 1B) to form a central carbocycle, with differing regiochemical dimerisation modes leading to significant structural diversity (Fig. 1C).^5–7 Dimerisation is proposed to occur in a head-to-head, head-to-tail, or head-to-side manner leading to the observed maleidride core structures (Fig. 1). The initial position of the pendant alkyl chains varies dependent on the mode of dimerisation, with head-to-head coupling leading to neighbouring side chains and head-to-tail to side chains on opposite sides of the central carbocycle (see Fig. 1A, Sections 5.1 and 5.2 for further details). Further tailoring modifications and rearrangements increase the structural complexity of maleidride natural products and can influence their bioactive properties.

The numbering systems used for the maleidrides varies greatly in the literature and shows no consistency. Hence in 2020 we proposed a more systematic method based on the size of the ring (1–9, 1–8, 1–7 as appropriate) beginning at the carbon alpha to the maleic anhydride ring, which gives the lowest numbers to the side chains. The maleic anhydride carbons are numbered with a prime, appropriate to the ring numbering, hence 3′ 4′ and 8′ 9′ for byssochlamic acid, and 1′′, 2′′, etc. for the first side chain, numbering from the ring junction, and 1′′′, 2′′′, etc. for the second chain.⁸ We have used this numbering system throughout.

This review aims to bring together studies on the chemical, genetic, and enzymatic aspects of maleidride biosynthesis. We will explore the literature regarding the biosynthesis of the monomer, evidence for dimerisation, and maleidride tailoring, by reviewing feeding studies, biomimetic syntheses, bioinformatics, gene deletions, heterologous expression and in vitro enzyme assays.

2. Maleidride structures and their bioactivities

2.1. Nonadrides

In 1931 Wijkman and co-workers isolated the first maleidrides from culture extracts of Penicillium glaucum, glauconic and glaucanic acids 4 and 5, (Fig. 2).⁹ Soon after, an isomer of glaucanic acid 5, (+)-byssochlamic acid 6 was isolated from Paecilomyces fulvus, a common contaminant of pasteurised goods.¹⁰ In the 1960s full structural elucidation of these compounds was achieved through both chemical degradation studies and X-ray crystallography.^11–15 In 1965, Barton and Sutherland named this family of related compounds (4–6) the ‘nonadrides’ in reference to the C₉-monomers thought to be involved in their construction,⁵ however this name has later become associated with the number of carbons in the central carbocyclic core of the maleidrides.

None of the initially discovered nonadrides have shown any significant bioactivities.^16,17 Many years later, (−)-byssochlamic acid 7, along with (−)-hydroxybyssochlamic acid 8 were extracted from a fungus that was isolated from a mangrove swamp.^18,19 (−)-Byssochlamic acid 7 was shown to have medium cytotoxic activity against HEp-2 and HepG2 cells, whereas (−)-hydroxybyssochlamic acid 8 showed weak activity.¹⁹ A reduced derivative of (+)-byssochlamic acid, dihydrobyssochlamic acid 9 was isolated from P. fulvus in 2015 (Fig. 2).¹

The rubratoxins A and B, 10 and 11 were first isolated from Penicillium rubrum in 1962,²⁰ and identified as the likely causative agents of fatal hepatotoxic poisoning events that occurred from contaminated foodstuffs. By 1970 their structures had been elucidated using a combination of degradation studies and X-ray crystallography, with the only difference between A and B being the reduction of one maleic anhydride moiety to a γ-hydroxybutenolide in rubratoxin A 10 (Fig. 3).^21–24 These compounds are strikingly more complex than the nonadrides (4–9) that had been previously characterised, and also the first nonadrides which appear to be formed not from two C₉-monomers, but instead by coupling of C₁₃-units. Despite their complexity, it is apparent that the mode of dimerisation is head-to-tail coupling, as occurs in byssochlamic acid 6 biosynthesis, as their pendant alkyl chains are positioned on opposite sides of the central carbocycle (Fig. 1 and 3). A desaturated derivative of rubratoxin B 11, rubratoxin C 12 was later isolated from a Penicillium sp.²⁵ Rubratoxin A 10 is a potent and highly specific inhibitor of protein phosphatase 2A, (PP2A), a target for anticancer drug development. Notably, it has approximately 100-fold stronger inhibition of PP2A than rubratoxin B 11.²⁶ The γ-hydroxybutenolide motif has been shown to be an important pharmacophore in other compounds.^27–29 Rubratoxin B 11 exhibits antitumour activity, likely linked to blocks in the progression of the cell cycle.³⁰ Rubratoxin C 12 shows weak activity against human cancer cell lines.²⁵ Ceramidastin 13, an analogue of the rubratoxins, has been isolated, also from a Penicillium sp.³¹ Inoue et al.³¹ state that the ¹H and ¹³C chemical shifts and coupling constants of ceramidastin 13 were very similar to those reported for rubratoxin B 11, suggesting the same stereochemistry between the two compounds, as shown in Fig. 3. Ceramidastin 13 was shown to be a novel inhibitor of bacterial ceramidase,³¹ an enzyme which is believed to contribute to skin infections of patients with atopic dermatitis.³² In 2019, a rubratoxin producing fungus, Talaromyces purpurogenus³³ was shown to produce five other nonadride compounds (14–18), one of which is an analogue of rubratoxin B 11 with one of the maleic anhydride moieties hydrolysed to a diacid (rubratoxin acid A 14).³⁴ Maleic anhydride ring-open forms of nonadrides may be artefacts of extraction protocols, and are known to interconvert with the ring-closed forms.^1,35,36 Hence it is difficult to determine whether 14 is a true natural product, although the authors note that 14 appears stable in their hands.³⁴ Compounds 15, 16, 17 and 18 also all contain one ring-open diacid and appear to be intermediates/shunts from the rubratoxin pathway.³⁷ All five compounds (14–18) were tested for their in vitro anti-inflammatory activities, with rubratoxin acid A 14 showing significant inhibitory activity against nitric oxide production (thought to play a crucial role in inflammatory responses)³⁸ from liposaccharide (LPS)-induced RAW264.7 cells.³⁴

In 1972 scytalidin 19 was isolated from a Scytalidium species and characterised, however the relative and absolute configurations were not determined.³⁹ Later analysis of various Scytalidium species revealed that deoxyscytalidin 20 is also produced by scytalidin 19 producers.⁴⁰ Nonadrides 19 and 20 possess the same ring structure as byssochlamic acid 6, but with longer alkyl chains, providing further confirmation that the maleidrides are not limited to compounds formed from the dimerisation of C₉-units. Scytalidin 19 shows antifungal activity with low phytotoxicity, and was first identified due to its fungitoxic effects towards Poria carbonica, a wood-rotting fungus.³⁹ Recent work has confirmed the absolute and relative configurations of both scytalidin 19 and deoxyscytalidin 20.⁸ In 1989 a ring hydroxylated analogue of scytalidin 19 named castaneiolide 21 was isolated from Macrophoma castaneicola, which causes ‘black root rot disease’ in chestnut trees. Assays using the purified castaneiolide 21 showed that it induced wilting in chestnut leaves.⁴¹ More recent studies have confirmed the structure of castaneiolide 21 (Fig. 4).⁸

The structure of heveadride 22, isolated from Bipolaris heveae, was solved in 1973 by MacMillan and co-workers through degradation studies.⁴² Interestingly this nonadride shows a different substitution around the 9-membered ring compared with the byssochlamic acids, scytalidins and rubratoxins and has neighbouring side-chains on the same side of the molecule, reminiscent of glauconic and glaucanic acids 4 and 5, arising from a head-to-head dimerisation. In 1987 a longer chain analogue of 22, homoheveadride 23 was isolated from the lichen symbiont Cladonia polycarpoides.⁴³

Dihydroepiheveadride 24, a γ-hydroxybutenolide analogue of heveadride 22, as well as epiheveadride 25, were later isolated from an unidentified fungus, with 24 providing significant antifungal activity.⁴⁴ Heveadride 22 and epiheveadride 25 also produced a fungitoxic effect, albeit significantly weaker than dihydroepiheveadride 24.⁴⁴ In 2010 Wicklowia aquatica was shown to be a prolific producer of heveadride analogues, producing epiheveadride 25, dihydroepiheveadride 24, deoxoepiheveadride 26, tetrahydroepiheveadride 27, dideoxoepiheveadride 28, and deoxodihydroepiheveadride 29 (Fig. 4).⁴⁵ Of these, 27–29 did not appear to show antifungal activity.⁴⁵ Another heveadride analogue, curvulariahawadride 30 has recently been isolated from a Curvularia sp. and was shown to have nitric oxide production inhibitory activity (Fig. 4).⁴⁶

In contrast to all the nonadrides discussed above, cornexistin 31 and its derivatives contain only one maleic anhydride moiety (Fig. 5). Cornexistin 31 was isolated and characterised in 1992 by the Sankyo pharmaceutical company.⁴⁷ It is produced by the thermotolerant fungus Paecilomyces divaricatus, which is closely related to the byssochlamic acid 6 producer, P. fulvus.⁴⁸ Cornexistin 31 has significant broad-spectrum phytotoxic activity and is of especial interest due to its low toxicity to the crop plant maize (Zea mays).⁴⁷ It also appears to have a unique mode of action, possibly involving inhibition of the plant aspartate amino transferase.³⁵ A derivative of cornexistin 31, hydroxycornexistin 32, was later isolated from P. divaricatus, which has significantly stronger activity against broadleaf weeds.⁴⁹ Intermediates 33, 34 and 35 from the cornexistin biosynthetic pathway were later isolated from a P. divaricatus strain engineered to produce fewer competing metabolites, thus allowing for greater flux towards the cornexistin pathway.⁵⁰

In 1997 the phomoidrides A 36 and B 37 were isolated from cultures of a fungus (ATCC 74256), later identified as belonging to the pleosporales order.^51–53 Trace amounts of an epimer, phomoidride D 38 were also isolated.^51,52,54 The phomoidrides A 36 and B 37 have been shown in vitro to inhibit squalene synthase and Ras farnesyl transferase and therefore are attractive lead structures for the development of both cholesterol lowering and anticancer drugs.⁵¹ A further isomer named phomoidride C 39 was isolated in 2001.⁵⁵ Recently, three further phomoidrides have been isolated from ATCC 74256, phomoidrides E 40, G 41 and F 42 (Fig. 5).⁵³ The phomoidrides are nonadrides assembled on a complex central core with functionalised side chains at C-2 and C-3. It is apparent however that they are formed from a head-to-head dimerisation in a manner somewhat similar to the glauconic and glaucanic acids 4 and 5. They are unique amongst the maleidrides discovered thus far in that the carboxylic acid of one of the monomers appears to be retained in the mature structure. This is corroborated by feeding studies which demonstrate that the C-10 carboxylic acid is derived from succinate.⁵⁶

Very recently, six further nonadrides, the talarodrides A–F 43–48 were isolated from an Antarctic sponge derived fungus, Talaromyces sp. HDN1820200 (Fig. 6).⁵⁷ These unusual maleidrides also appear to be formed in a similar manner to glauconic and glaucanic acids 4 and 5, and share the bridgehead olefin present in most phomoidrides e.g.37. Talarodrides A 43 and B 44 show specific antibacterial activity against Proteus mirabilis and Vibrio parahemolyticus.⁵⁷ The methoxy groups present in talarodrides B 44 and C 45 are potentially artefacts due to the use of methanol during isolation.⁵⁷

The structures of the nonadrides have attracted significant attention from the scientific community not only because of their fascinating biosynthesis but also their structures have proved a challenge to the skills of synthetic chemists. Stork completed the first total synthesis of racemic byssochlamic acid in 1972 (ref. 58) and was later followed by White's “photoaddition–cyclodimerisation” strategy for the efficient assembly of the functionalised 9-membered ring.⁵⁹ The first enantioselective synthesis was reported by White and co-workers in 2000 following a similar approach used in the synthesis of the racemate.⁶⁰ The molecular complexity of the phomoidrides has demanded the development of selective strategies and several elegant total syntheses have been achieved.^61–64 Cornexistin 31 and related compounds have been of particular recent interest due to their potential value as herbicides.^47,49 Clark and Taylor^65–67 have explored synthetic routes towards cornexistin 31 and in 2020 the first total synthesis of (+)-cornexistin was reported by Magauer and co-workers.^68,69 Starting from malic acid, key steps included a Hiyama–Kishi coupling, stereoselective aldol reaction and intramolecular alkylation to deliver >150 mg of cornexistin 31. This approach could be readily adapted for the preparation of analogues.

2.2. Octadrides

Zopfiellin 49 was the first octadride to be reported, and was isolated from Zopfiella curvata in 1994, by Nissan Chemical Corp.⁷⁰ It shows promising antifungal activity against many plant pathogenic fungi, as well as various fungi that cause human diseases.⁷⁰ Zopfiellin 49 readily interconverts between the ring-closed dianhydride form and the ring-open tetracarboxylate 50, which is favoured at low pH (Fig. 7).³⁶ The dianhydride form does not appear to have significant fungicidal activity.^36,71 The activity of zopfiellin 49/50 is ameliorated by addition of oxaloacetate to fungal cultures, suggesting that the mode of action is associated with oxaloacetate metabolism.³⁶ Zopfiellin 49 was recently isolated from a close relative of Z. curvata, Diffractella curvata, and using a combination of NMR spectroscopy and the X-ray structure of a crystalline derivative, the absolute and relative configurations of zopfiellin 49 were confirmed.⁸

Another antifungal octadride, viburspiran 51, was isolated from Cryptosporiopsis sp. in 2011.⁷² Viburspiran 51 contains an ethylene bridge between C-3 and C-8. A similar metabolite, botryoanhydride 52, was recently isolated from an uncharacterised fungus which has an n-propyl group attached to C-1, instead of the n-pentyl group present in viburspiran (Fig. 7).⁷³

2.3. Heptadrides

The first natural heptadrides, agnestadrides A 53 and B 54, were isolated from the byssochlamic acid 6 producer, P. fulvus in 2015 (Fig. 7).¹ Baldwin and co-workers had previously characterised a compound with a heptadride structure during their biomimetic investigations into nonadride monomer dimerisation.⁷⁴ A head-to-side mode of dimerisation can explain the formation of the seven-membered central carbocycle (see Fig. 1 and Section 4.2 for more detail).^1,75

3. Origin of the monomers

Soon after the first structure elucidation of the maleidrides, Sutherland and co-workers^5,11 proposed that their biosynthesis may proceed via the coupling of two monomeric units. They were prescient in their hypotheses, proposing that monomer units could be derived from a citric acid intermediate, and that an anionic type coupling mechanism in either head-to-head or head-to-tail coupling could account for the structural differences between glauconic and glaucanic acids 4 and 5, and byssochlamic acid 6.⁵

To investigate the biosynthetic construction of the putative monomers, Sutherland and co-workers⁷⁶ performed a series of feeding experiments with ¹⁴C-labelled putative biosynthetic precursors combined with degradation studies. As the degradation of glauconic acid 4 into characteristic fragments had been previously established,^5,9 Sutherland and co-workers⁷⁶ selected 4 for these studies as it would undergo controlled decomposition to two known products: glauconin 55 and diethylacrolein 56, and then further degraded to CO₂ and the radioactivity measured (Scheme 1). The identified carbons could then be referenced to the putative monomer unit, 57.

In an initial experiment, a P. purpurogenum culture was fed separately [1-¹⁴C]- and [2-¹⁴C]-acetate 58, subsequently, labelled glauconic acid 4 was isolated (with 9.4% and 13.2% incorporation radiolabel respectively) and the site of isotopic labelling determined by degradation studies as shown in Schemes 1 and 2.⁷⁶ From these experiments it was deduced that the C₉-precursor 57 was assembled from two different components coupled to generate the double bond of the maleic anhydride (Scheme 2). The observed labelling pattern was consistent with the longer C₆-chain of the monomer unit being the product of a typical polyketide/fatty acid synthase (PKS/FAS), derived from a head-to-tail condensation of an acetate and two malonate units (Scheme 2). Two adjacent carbons from the C₃-chain showed similar incorporation of radioactivity from [2-¹⁴C]-acetate implying that these carbons have become equivalent in a precursor. To account for this, Sutherland and co-workers⁷⁶ proposed that labelled acetate also enters the citric acid cycle (Scheme 2), where it subsequently labels the truly symmetrical intermediate, succinate 59. Succinate 59 is then converted to oxaloacetate 60, where the [2-¹⁴C]-acetate 58 activity is distributed equally between the methylene and carbonyl groups.^76,77

The above experiments⁷⁶ were supported by feeding [2,3-¹⁴C₂]-succinate 59, which was observed to be efficiently incorporated into the C₃-chain. The authors concluded that oxaloacetate 60 is the likely direct precursor of the C₃ chain.^76,77

A complementary experiment was undertaken by Cox and Holker with [2,3-¹³C₂]-succinate 59 fed to P. purpurogenum⁷⁸ confirming that intact succinate 59 (or its derivative) was incorporated into the C₃-chain of the glauconic acid 4 precursor.⁷⁸ Further evidence for the biosynthetic origin of the monomers came from feeding studies using the rubratoxin producer P. rubrum. Analysis of the isolated rubratoxin B 11 revealed a labelling pattern in accordance with the longer chain (here C₁₀) being derived from a fatty acid and the shorter C₃ from the citric acid cycle.⁷⁹

The origin of the putative monomers that form phomoidride B 37 has also been investigated.⁵⁶ The producing organism, unidentified fungus ATCC 74256, was fed a series of carbon-13 labelled precursors, and phomoidride B 37 isolated and analysed by ¹³C NMR. The deduced labelling pattern shown in Scheme 3 was in full accordance with the longer C₁₂-chain being derived from a polyketide/fatty acid synthase.

In more recent investigations by Willis and co-workers⁸ on the biosynthesis of the nonadrides scytalidin 19 and deoxyscytalidin 20, [1,2-¹³C₂]-acetate 58 was fed to cultures of S. album and analysis of the ¹³C-NMR data of both metabolites was in accord with the polyketide and oxaloacetate origin of the natural products (Fig. 8).

4. Evidence for dimerisation during maleidride biosynthesis

As discussed in Section 3, in 1965 Barton and Sutherland⁵ with immense prescience had proposed that the biosyntheses of glauconic and glaucanic acids 4 and 5, and byssochlamic acid 6 may originate from similar building blocks (monomers) but coupled in different ways to generate the various carbon skeletons. The head-to-head anionic coupling mechanism proposed for the biosynthesis of glauconic and glaucanic acids 4 and 5, requires two identical 57 monomers (Scheme 4). The head-to-tail coupling required for byssochlamic acid 6 biosynthesis would require one monomer 57 and the exo-diene analogue 62 (Scheme 4).

The exo-diene 62 (herein named waquafranone B) had been reported to have been isolated from W. aquatica, a producer of a variety of heveadride analogues (e.g.25).⁴⁵ However, recent biomimetic dimerisation studies by Willis and co-workers⁸⁰ revised the structure of waquafranone B to be diacid 63 (Fig. 9). This is in accord with biomimetic studies by Sutherland and co-workers⁸¹ who demonstrated that exo-diene 62 is unstable.

The instability of the exo-diene 62 does not preclude its veracity as a true intermediate in maleidride biosynthesis, as unstable intermediates may be chaperoned by enzymes in vivo. The equilibrium represented between 57 and 62 in Scheme 4 is a regiochemical rationalisation depicted to describe a potential enzyme catalysed mechanism that remains to be proven.

In 2000 Sulikowski, Agnelli and Corbett were the first to propose that the maleidride monomer might contain a carboxylic acid, likely due to their specific interest in the phomoidrides, where one carboxylic acid is retained in the mature natural product.⁸² They proposed that the reactive anionic monomer is derived from decarboxylation of monomer 1.

Isolation of the carboxylated analogue of the anhydride 57, monomer 64, from the byssochlamic acid 6 producer P. fulvus, and the previous feeding studies by Sulikowski and co-workers,⁸² led Simpson and co-workers¹ to speculate that carboxylated monomer 64 coupled with exo-diene 62 may be the true intermediates for byssochlamic acid 6 biosynthesis, as well as for the newly discovered heptadrides 53 and 54 also isolated from P. fulvus (Scheme 5). The authors noted that in their hands carboxylated anhydride 64 was unstable, and completely decomposed to 57 in under 48 hours.¹

Key evidence for the involvement of a dimerisation step during maleidride biosynthesis has come from four sources: (i) feeding experiments performed in vivo; (ii) in vitro chemical investigations of the substrates, reaction conditions and their products; (iii) from combined chemical and genetic studies in maleidride producers; and (iv) from cell free biocatalysis with the proposed dimerisation enzymes.

4.1. In vivo studies

The first direct evidence for in vivo incorporation of maleic anhydride-based monomers into the structure of a nonadride metabolite was reported for glauconic acid 4 (Scheme 6).⁸³ The study by Moppett and Sutherland⁸³ involved separately feeding two isotopically labelled substrates, tritiated 65 and carbon-14 labelled 57, into liquid cultures of the glauconic acid 4 producer, P. purpurogenum. Feeding compound 65 afforded [1,4-³H₂]-glauconic acid 4 which was confirmed by degradation studies leading to an equal label distribution between glauconin 55 (C-1) and diethylacrolein 56 (C-4) (degradative studies are shown in Scheme 1). A 1:1 ratio of activities established that dimerisation had taken place, however the incorporation was very low (0.25%).

Incubating growing cultures of P. purpurogenum with the ¹⁴C-labelled 57 resulted in the isolation of glauconic acid 4 with 51.5% incorporation of carbon-14, with 97.5% of the total activity localised at C-7 and C-10 (Scheme 6).⁸³ In both experiments, the radiolabels were found at positions expected for the product of head-to-head dimerisation of the fed monomer units, and the higher level of incorporation of 57 suggested that the unsaturated anhydride is the correct monomer unit.⁸³

Sulikowski and co-workers sought a biomimetic approach towards the total synthesis of phomoidrides A 36 and B 37,⁵⁶ and this led the group to pursue biosynthetic studies in the unidentified fungus ATCC 74256 using precursors incorporating stable isotopic labels. Although phomoidrides A 36 and B 37 and glauconic acid 4 differ in the length of the pendant side-chains, the same symmetrical pattern can be discerned and consequently phomoidrides A 36 and B 37 were proposed to be formed through coupling of analogous C₁₆-precursor units.⁵²

Sulikowski and co-workers⁸⁴ prepared synthetic analogues of the predicted precursors incorporating deuterium (Scheme 7). The first synthetic substrate was thiol ester 66, as N-acetylcysteamine (SNAC) has been shown to be a valuable CoA substitute in biosynthetic studies, as it can readily pass through cell membranes, unlike CoA adducts. These CoA mimics are often used where carrier protein-bound thioesters are required in enzyme biosynthetic machinery, for example when investigating polyketide biosynthesis.⁸⁵ Sulikowski and co-workers⁸⁴ fed [²H₂]-thiol ester 66 to a culture of ATCC 74256 and phomoidride B 37 was isolated with incorporation of 3 deuterium atoms as determined by ²H NMR and ESIMS analysis. This provided evidence for a homodimerisation process having occurred (Scheme 7). A similar experiment with [²H₂]-67, with a pendant methyl group rather than the thiol ester, did not show any incorporation into phomoidride B 37 (Scheme 7). This important experiment provided the first evidence that dimerisation requires decarboxylation, at least in the case of the phomoidrides.⁸⁴

4.2. Biomimetic studies

Several biomimetic synthetic studies aimed at reconstructing the maleidride dimerisation event under laboratory conditions provide interesting insights into the mechanism of the reaction. Upon completing feeding studies with anhydride 57, Huff, Moppett and Sutherland set out to test self-dimerising properties in vitro.^81,86 To this end, maleic anhydride 57 was treated with base in order to generate the required carbanion intermediate. The reaction afforded a crystalline solid in a very low yield (2% with NaH, improved to 4% by using Et₃N), which was not the expected glaucanic acid 5, but believed to be iso-glaucanic acid 68, a stereoisomer of the natural product formed in vivo (Scheme 8).⁸⁶ In parallel, an attempt was made to synthesise fulgenic anhydride 62, in order to test a hypothesis that this compound might be involved in the reaction leading specifically to the formation of byssochlamic acid 6.⁸¹ However, the base-catalysed in vitro dimerisation reaction of the fulgenic anhydride 62 again yielded iso-glaucanic acid 68 and not byssochlamic acid 6 (Scheme 8). This was rationalised to be due to the instability of anhydride 62, which under the reaction conditions was found to isomerise to 57.

Interest in the dimerisation was reinvigorated almost 30 years later, inspired by the discovery of the phomoidrides^51,52 and driven by the pursuit of an efficient total synthesis route. The reports on in vitro dimerisation came in a series of papers from the groups of Baldwin^74,87 and Sulikowski,^82,88 who both set out to investigate the chemical mechanism driving the reaction.

Studies were reopened by Baldwin and co-workers,⁷⁴ who reinvestigated the in vitro dimerisation studies towards glaucanic acid 5.^81,86 Beside obvious differences in the lengths of the side-chains (and consequently in the structure of the dimerising monomer), there are key differences in the stereochemistry between iso-glaucanic acid 68 and the phomoidrides. Despite this the authors viewed this biomimetic dimerisation as a potential synthetic route towards the phomoidrides.⁷⁴ Thus, 2-[(E)-1′-pentyl]-methyl maleic anhydride 69 was synthesised and treated with base under a range of conditions. Although mostly polymeric products were formed, iso-glaucanic acid analogue 70 together with two other minor dimerisation products, the spiro compound 71, as well as the heptadride 72 were isolated in low yields (Scheme 9). A common structural feature of all three products is the linkage of the two anhydride moieties via a CH₂ bridge. Hence Baldwin and co-workers⁷⁴ proposed that a stepwise Michael addition is more likely than a concerted 6π + 4π cycloaddition. Furthermore they suggest that the anion in intermediate 73 is able to attack at different electrophilic centres, accounting for the formation of the different products.⁷⁴

Further optimisation of the reaction conditions was carried out, with the highest yield (8.5%) of 70 achieved using DMSO/Et₃N (0.66 eq.)/MgCl₂ (0.5 eq.). X-ray crystallography confirmed the relative stereochemistry of the side-chains in accord with Sutherland's assignment of the configuration of iso-glaucanic acid 68.⁸⁶

In 2000 Sulikowski, Agnelli and Corbett investigating the in vitro dimerisation of phomoidride precursors⁸² proposed that within an in vivo system at least one of the dimerising units is likely to be covalently linked to an enzyme so imposing conformation constraints. Furthermore, if the dimerisation process is stepwise rather than concerted, in vitro studies linking the two monomers prior to cyclisation may lead to cleaner reactions.

In an initial experiment, Sulikowski and co-workers⁸² covalently linked the two units as bis-esters with varying chain-lengths (compounds 74a–f, Scheme 10). Treating a mixture of the six substrates, 74a–f (Scheme 10a, n = 1–6) with DBU in anhydrous MeCN triggered dimerisation with only substrate 74b (n = 2), to produce 75a and 75b (different stereoisomers at the newly formed stereocentres C-13 and C-17). A single stereoisomer 76 was obtained in an analogous reaction with symmetric diol 77 (Scheme 10b). A mechanism involving a Michael addition was proposed and it was assumed that the observed compounds were the thermodynamic products of the reaction. To trap kinetic products, the reaction using substrates 74a–f was repeated in the presence of excess acetic anhydride (Scheme 10c). Three additional dehydrated products 78a–c were identified, which were derived from substrates 74c–e (Scheme 10c). The position desired for the biomimetic synthesis of phomoidrides requires formation of C-13, C-14 bond. To the authors' disappointment, in all the in vitro products, the ring-closing C–C bond was formed exclusively between C-13 of the enolate and C-17 of the Michael acceptor instead.⁸²

Sulikowski and co-workers⁸⁸ modified the substrate by using a tertiary amide linker, to produce substrate 79 (Scheme 11). Whilst products 80, 81, 82 and 83 were formed, no products with the desired phomoidride core were detected.^88,89

Baldwin and co-workers⁸⁷ also investigated the influence of a covalent tether on the stereo- and regioselectivity of cyclisation. Substrates (84a–d and 85) were exposed to a range of reaction conditions and DBU in THF:DMSO (1:4) led to cyclisation (Scheme 12). Only three out of the five prepared substrates, 84b, 84c and 85, gave products which could be isolated and characterised showing the structures to be 86, 87, 88 and 89 (Scheme 12). The authors proposed that these cyclic products were the result of exo-orientated double Michael additions.⁸⁷

A recent study by Willis and co-workers⁸⁰ into maleic anhydride and related diacid natural products used a biomimetic approach to investigate in vitro dimerisations of the proposed monomers required for scytalidin 19 biosynthesis. The authors noted that in all previous biomimetic studies, the focus has been on homodimerisation of analogues of 57, rather than heterodimerisation using 57 and the exo-diene 62, which is proposed to be involved in maleidride biosynthesis during various modes of dimerisation (see Schemes 4, 5 and 17 and Section 5). However exo-diene 62 was unstable even when kept at −78 °C and after 96 h was converted to a mixture of products including the corresponding maleic anhydride 57. Homodimerisation of the maleic anhydride tetraketide monomer 90 using Et₃N, MgCl₂ in DMSO (as used by Baldwin and co-workers⁷⁴) gave iso-glaucanic acid analogue 91 in 10% yield. However, efforts to heterodimerise 90 with either 92 or 93 (avoiding the unstable exo-diene), gave iso-glaucanic acid analogue 91 in similar yields, with 92 and 93 recovered from the reaction unchanged. Use of freshly prepared exo-diene 94 in a heterodimerisation reaction with maleic anhydride 90 led to a complex mixture of products, none of which could be characterised (Scheme 13).⁸⁰

5. Molecular reconstruction of maleidride biosynthesis

5.1. Core genes for monomer biosynthesis

The genetic and enzymatic basis of maleidride biosynthesis remained cryptic until 2015, when Oikawa and co-workers⁹⁰ investigated the biosynthetic pathway for the production of maleidride monomers. In fungi the genes required for the biosynthesis, regulation and transport of a specific natural product are generally co-located as a single biosynthetic gene cluster (BGC).^91,92 Therefore Oikawa and co-workers⁹⁰ initially sequenced the genome of the phomoidride (e.g.37) producer, the unidentified fungus, ATCC 74256, to identify a putative BGC for the production of the phomoidrides (e.g.37). As previous feeding studies had demonstrated,^56,76,78 the likely origin of the maleidride monomer is the condensation of the product of a FAS/PKS with oxaloacetate. Oikawa and co-workers⁹⁰ proposed that a putative maleidride BGC might contain either an FAS/PKS clustered with a gene encoding a citrate synthase-like (CS) enzyme (Scheme 14). They identified a BGC they named phi (Fig. 10) which consisted of a highly-reducing PKS (hrPKS), phiA, clustered with phiI, a gene encoding a CS-like enzyme,⁹³ as well as a gene encoding a 2-methylcitrate dehydratase-like enzyme (2MCD, phiJ),⁹⁴ which is a likely candidate for the dehydration reaction required to form the unsaturated monomer 1 (Scheme 14). At the time, no genes encoding hydrolytic enzymes for hydrolysis of ACP-bound polyketide chains were detected, although more recent analysis has determined that phiM encodes a hydrolase, which is a homologue of the esterase from the asperlin BGC (alnB – C8VJR6.1).^95,96

Phylogenetic analysis of citrate synthase-like and 2-methylcitrate dehydratase-like enzymes from the likely phomoidride BGC, along with other subsequently discovered maleidride homologues, has determined that these enzymes form a separate clade with those that are known and predicted to produce or accept alkylcitrate.⁹⁶ It is therefore accepted that these enzymes should be referred to as alkylcitrate synthases (ACSs) and alkylcitrate dehydratases (ACDHs).⁹⁶

Oikawa and co-workers⁹⁰ reconstructed phiA, I, J in the heterologous host Aspergillus oryzae (a suitable host for the production of fungal natural products).^97–99 This resulted in the production of a new metabolite which possessed the characteristic UV absorption (λ_max 312 nm) for a maleic anhydride conjugated with an olefin.⁹⁰ Due to low titres, no specific product of phiA, I, J, was isolated and so the attention of the authors turned to a homologous cluster, tst, which they had identified in the publicly available Talaromyces stipitatus genome. Although T. stipitatus itself has not been reported to produce maleidrides, many Talaromyces species are known to produce glauconic and glaucanic acids 4 and 5, as well as the more complex rubratoxins e.g.10 (although no Talaromyces species are reported to produce phomoidrides).¹⁰⁰ Expression of the phiA, I, J homologues, tstA, I, J in A. oryzae resulted in the production of a compound with similar LCMS characteristics to that which was produced by the heterologous expression of phiA, I, J. The structure was confirmed to be 67 by NMR and HRMS (Scheme 15). Compound 67 is the predicted monomer required for phomoidride biosynthesis, and is an analogue of the substrate 66 successfully utilised in the phomoidride feeding studies conducted by Sulikowski and co-workers (Scheme 7).⁸⁴

Further evidence for the relatedness of the phi and tst BGCs comes from phylogenetic analyses by Williams et al.⁹⁶ This work showed that maleidride PKSs appear to clade according to the expected or confirmed chain length of their polyketide product, with PhiA and TstA forming a separate ‘hexaketide’ producing clade, which suggests that the T. stipitatus cluster may encode phomoidride biosynthesis or a related analogue formed from hexaketide based monomers.⁹⁶

Oikawa and co-workers⁹⁰ also expressed the tstI, J genes in Escherichia coli, followed by purification and enzyme assays utilising 2-decenoyl-CoA 95 and oxaloacetate 60 as substrates. This assay produced compound 96, which is carboxylated, with the polyketide derived moiety one acetate unit shorter than the compound isolated from A. oryzae (Scheme 15). Details of any further substrates tested were not available, therefore it is difficult to determine if 2-decenoyl-CoA 95 is the true substrate for TstI (the alkylcitrate synthase), or whether TstI may have some substrate flexibility regarding chain length.

Following isolation of the carboxylated monomer 96 from the enzyme assays conducted by Oikawa and co-workers,⁹⁰ (Scheme 15) a mechanism was proposed for dimerisation of a carboxylated analogue of compound 67 (97) to produce the predicted phomoidride intermediate 98via an aldol like reaction (Scheme 16). Hu and co-workers⁵³ recently isolated further phomoidrides E 40, F 41, and G 42 which led them to propose that the key phomoidride intermediate 99 is more likely to be formed via a Claisen condensation (Scheme 16).

Oikawa and co-workers⁹⁰ have proposed a unified model for maleidride biosynthesis (Scheme 17). This model is based on the homo- and hetero-dimerisations of the carboxylated anhydride, ‘monomer A’ 1, the decarboxylated anhydride ‘monomer B’ 2 and the exo-diene anhydride ‘monomer C’ 3, and is driven by the formation of an enolate derived from A 1. The authors proposed that their model accounts for discrepancies in previous feeding experiments, as these appeared to be based on a single monomer.

In 2016 Cox and co-workers⁷⁵ reported the results of studies on maleidride biosynthesis via heterologous expression in the host A. oryzae. This study further characterised the pathway for byssochlamic acid 6 and agnestadrides A and B 53 and 54 following on from earlier predictions by Simpson and co-workers.¹ Genes homologous to those identified by Oikawa and co-workers⁹⁰ (encoding an hrPKS, an ACS, and an ACDH) were identified clustered within the P. fulvus genome (Fig. 11).

In addition, a gene (bfL1) encoding an enzyme with a putative hydrolytic function was identified, which is also homologous (35.48% identity) to the esterase from the asperlin BGC.⁹⁵ Expression of the P. fulvus hrPKS, ACS and ACDH in A. oryzae did not produce any novel compounds, whereas these genes, with the addition of bfL1, produced the carboxylated anhydride 64 and its decomposition product, 57. This is contradictory to the results obtained by Oikawa and co-workers⁹⁰ where the addition of a hydrolytic enzyme was not necessary for the production of monomers. Later work by Cox and co-workers⁵⁰ investigating the cornexistin 31 pathway via gene deletion experiments, also suggested that the homologous hydrolase (pvL1) in the cornexistin BGC (Fig. 12) is essential, as no maleidride related compounds accumulated in the hydrolase deletion strain.

Interestingly, all confirmed and putative maleidride BGCs contain a hydrolase homologue, suggesting that it is important for the biosynthesis of maleidride compounds.⁹⁶In vitro studies by Cox and co-workers¹⁰¹ showed that the P. fulvus hydrolase, BfL1, catalysed the hydrolysis of a series of a thiol esters, rather than being ACP-selective, therefore exactly how selectivity is controlled is unknown.¹⁰¹

Investigations into the ACS and ACDH enzymes through in vitro characterisation have also been reported.¹⁰¹ Assays using both unsaturated (a) and saturated (b) versions of the substrates 100, 101, 102 and 103, with oxaloacetic acid and purified BfL2 (ACS) showed that only the CoA thiol ester 103a/b could be turned over by BfL2 (Scheme 18) to produce 104a/b.

Comparison of 104a to synthetic standards revealed that the enzyme product is exclusively the anti diastereomer.¹⁰¹

The synthesis of citrate is catalysed in most organisms by a Si-citrate synthase, with known Re-citrate synthases phylogenetically unrelated to Si-citrate synthases.¹⁰² A structural model of BfL2 was built based on the primary metabolism citrate synthase from Acetobacter aceti,¹⁰³ which is phylogenetically related to other Si-citrate synthases. Furthermore, the crystal structure of the A. aceti citrate synthase is bound to oxaloacetate and an acetyl CoA mimic in positions that should result in an S stereocentre.¹⁰¹ The structural model of BfL2 showed that all of the residues involved in catalysis and binding oxaloacetate and acyl CoA are structurally highly conserved with the A. aceti citrate synthase.¹⁰¹ This led to the proposal that BfL2 also creates a 3S-stereocentre, and thus ultimately an 3S,4R configuration.¹⁰¹ Cox and co-workers¹⁰¹ also suggested that differences in the configuration at the 4-position of 104 must be controlled by the geometry of the enoyl CoA intermediate.¹⁰¹ Recent in silico analysis of maleidride BGCs by Williams et al.⁹⁶ has shown that many clusters contain an enoyl CoA isomerase, which may be involved in providing the appropriate substrate for the ACSs.

In vitro assays with purified ACDHs from the P. fulvus or P. divaricatus BGCs (BfL3/PvL2) demonstrated that only the anti diastereomer 104a can be dehydrated to produce the equilibrated products 105, the diacid, and 64, the anhydride (Scheme 19).¹⁰¹

5.2. Core genes for dimerisation

Comparison of the maleidride BGC from P. fulvus by Cox and co-workers⁷⁵ to putative maleidride BGCs identified from genome sequences available on NCBI, as well as the putative phomoidride e.g.37 BGC⁹⁰ revealed further genes in common. Each cluster encodes one or two proteins that have some similarity to ketosteroid isomerases (KSI-like) and one or two proteins that contain phosphatidylethanolamine-binding protein (PEBP) domains.⁷⁵ Expression of the monomer forming genes (PKS, hydrolase, ACS and ACDH) with both KSI-like genes in the host A. oryzae led to the production of both byssochlamic acid 6 and agnestadride A 53 demonstrating that within the context of the A. oryzae genome, there are sufficient catalytic activities to perform both head-to-tail and head-to-side dimerisations of maleidride monomers, and that the KSI-like enzymes catalyse that dimerisation. The presence of both KSI-like enzymes appeared to be required for the dimerisation to occur in vivo. Addition of the two genes containing PEBP domains led to an over 20-fold increase in dimerised products.⁷⁵

Further studies by Cox and co-workers¹⁰¹ showed that in contrast to the in vivo experiments, yeast cell-free extracts of either P. fulvus KSI-like enzyme are capable of catalysing dimerisation. Addition of the P. fulvus PEBP enzymes did not appear to appreciably increase yields of dimerised products, however the low-yielding nature of these experiments makes quantitative comparisons difficult.¹⁰¹ We have previously proposed that the KSI-like enzymes are renamed ‘maleidride dimerising cyclases’ (MDCs), as they alone are sufficient to perform the dimerisation reaction.⁹⁶ All known and putative MDCs contain an NTF2 domain (nuclear transport factor 2 – IPR032710), which categorises them within the NTF2-like superfamily.⁹⁶ This large group of proteins, which includes enzymes that have isomerase, cyclase, dehydratase and hydrolase activities, have low sequence identity but share a common structural fold that can be adapted to serve a range of functions.¹⁰⁴

Further gene deletions to the cornexistin 31 producer, P. divaricatus corroborated these results, and suggested at least a supplementary role for the PEBP enzymes.⁵⁰ Within the cornexistin BGC, only one MDC and one gene containing a PEBP domain are present (Fig. 12). Deletion of the MDC gene led to complete cessation of cornexistin 31 biosynthesis, with accumulation of the carboxylated anhydride monomer 64 and its spontaneous ring open form 105, which had not previously been detected from P. divaricatus extracts. Deletion of the gene containing the PEBP domain led to a decrease in the titre of cornexistin 31, and accumulation of 64, 105 and the decarboxylated monomer 57 (Fig. 13).⁵⁰

Further research investigating the biosynthesis of zopfiellin 49 by Oikawa and co-workers¹⁰⁵ identified a zopfiellin BGC (Fig. 14) from the genome of Z. curvata.

This work again demonstrated that the MDC and PEBP genes are involved in dimerisation of maleidride monomers; once introduced to an A. oryzae strain producing the zopfiellin monomer 106, two dimerised products were isolated, the nonadrides prezopfiellin 20 (which was identified as deoxyscytalidin 20 by Willis and co-workers⁸) and iso-prezopfiellin 107 (Scheme 20).¹⁰⁵ It is notable that the mode of dimerisation for these nonadrides is different, i.e.: head-to-tail to produce deoxyscytalidin 20 and head-to-head (mode B) for iso-prezopfiellin 107 (see Scheme 17 for dimerisation types). This is the second known system where different modes of dimerisation can occur within the same pathway, the first being the biosynthesis of the nonadride byssochlamic acid 6 (head-to-tail dimerisation) and the heptadrides, agnestadrides A and B 53 and 54 (head-to-side dimerisation).¹

No evolutionary relationship regarding mode of dimerisation appears to be displayed by the MDCs.⁹⁶ The lack of close homologues to the MDCs constrains our ability to predict a mechanism for these enzymes, with crystallisation, modelling and mutation studies likely required to further our understanding of these unique enzymes. Until then, exactly how the MDCs control dimerisation, including apparently simultaneously catalysing different modes of dimerisation, remains cryptic.

The putative accessory role of the PEBP containing enzymes has been hypothesised to involve the chaperoning of unstable intermediates such as 1 and/or the known anionic binding ability of PEBP containing enzymes.^75,106

5.3. Comparison of maleidride BGCs

To date there are six BGCs which have been linked to specific maleidrides through experimental approaches: the byssochlamic acid 6/agnestadrides e.g.53 BGC,⁷⁵ the rubratoxins e.g.10 BGC,³⁷ the cornexistin 31 BGC,⁵⁰ two zopfiellin 49 BGCs,^8,105 and the scytalidin 19 BGC.⁸ Two maleidride BGCs have been identified from confirmed maleidride producing strains – linked to phomoidrides e.g.37 (ref. 90) and epiheveadride 25 biosynthesis (Fig. 15).⁹⁶ A further fourteen putative maleidride BGCs have been identified from publicly available genomes.⁹⁶ Bioinformatic comparison of these maleidride BGCs supported the conserved core set of genes required for basic maleidride biosynthesis in all clusters – those encoding monomer biosynthesis – the hrPKS, the hydrolase, the alkylcitrate synthase and the alkylcitrate dehydratase, and those involved in dimerisation – the maleidride dimerising cyclases and the PEBP-like. In all cases, the clusters contain one or two MDC genes. Most clusters have one or two genes that contain a PEBP domain.⁹⁶ The hypothesised ancillary nature of the PEBP enzymes does not preclude those clusters without genes that contain a PEBP domain from encoding maleidride biosynthesis.⁹⁶

There are further sets of genes in common between the maleidride BGCs, some of which are common to many fungal natural product BGCs, the cytochrome P450s, α-ketoglutarate-dependent dioxygenases (αKGDDs), regulators and transporters, and some of which are more specific to maleidride BGCs, for example the isochorismatase-like, and a group of genes with sequence homology to each other, but with no characterised homologues (conserved maleidride proteins) (Fig. 15). Many of the genes which encode for catalytic enzymes are likely to be involved in post-dimerisation tailoring (see Section 5.4.2), however, the function of many others currently remains obscure.⁹⁶

5.4. Genes responsible for maleidride structural diversification

6. Overview of maleidride compounds

The structures of all maleidride compounds discussed in this review have been classified in Fig. 18 and 19 according to their mode of dimerisation, to demonstrate the structural relationships between these compounds. Furthermore, their known or predicted monomer chain length, producing species, and any known bioactivities have been collated in Table 1.

---

7. Conclusions

Since the first maleidride isolation in the 1930s,⁹ exactly how these compounds are formed have posed a challenge to our biosynthetic understanding, with increasing insight leading to the potential to synthesise and manipulate their structures in a rational manner. The core ring of 7-, 8- or 9-carbons is unusual in nature, and this class of compound has received growing interest as more representatives have been isolated, particularly given that the majority have important biological activities.²

Recent genetic and biochemical studies^{50,75,90,101,105} have added support to the original feeding studies^56,76–78 showing that the monomer for the maleidrides is derived from an oxaloacetate cross-linked via its β carbon to the β carbon of a polyketide. The core set of enzymes responsible for formation of the monomer have been characterised: a highly reducing-PKS, a hydrolase, an alkylcitrate synthase and an alkylcitrate dehydratase.^75,90,96,101 Moving beyond the monomer, the core enzyme required for dimerisation, and therefore ultimately controlling the structure of the mature maleidride, is the maleidride dimerising cyclase.^{50,75,96,101,105} This coupling reaction appears to be aided by the PEBP-like enzymes, although their exact role is currently obscure.^{50,75,96,101,105} The precise detail of how cyclisation is controlled remains cryptic, at present it is not possible to predict whether a biosynthetic gene cluster will deliver dimers showing head-to-head, head-to-tail or head-to-side modes of cyclisation, highlighting that there is still much to be discovered in this type of pathway.

In terms of the octadrides, we now have a far better understanding of how the octadride zopfiellin 49 is formed via a ring-contraction, with the oxidative elimination of a ring-carbon by an α-ketoglutarate dependent dioxygenase, converting the nonadride precursor to the octadride.^8,105 It is yet to be determined whether the ZopK/ZopL9 enzyme responsible for this step of zopfiellin 49 biosynthesis can be modified to ring-contract other nonadrides. Furthermore, with only limited yields recovered from both in vitro and in vivo reactions, a question remains as to whether additional, as yet unidentified, enzymes are required to elevate the yield of this type of reaction.^8,105

Various modes of post-cyclisation tailoring have been highlighted and, given the ongoing discovery of new maleidride BGCs from sequence data hinting at unidentified members of this class,⁹⁶ we expect the range of modifications available to continue to increase. The maleidrides are a challenging, but rewarding class of fungal natural product and the increasing knowledge about their biosynthesis raises interesting possibilities for combining synthetic biology approaches with semi-synthetic chemistry to deliver a wide range of maleidrides for future pharmacological assessment.

8. Author contributions

KW drafted the majority of the manuscript, with help from AJS. KMJdMS, AMB, RJC and CLW edited the manuscript with AMB, RJC and CLW contributing short sections.

9. Conflicts of interest

RJC is an Editor for the Special Issue “Engineering Fungal Biosynthetic Pathways”.

10. Acknowledgement

We would like to thank the MRC for funding KW and KMJdMS (MR/N029909/1), and the BBSRC and Syngenta for funding AJS (BB/J006289/1). We would like to thank Dr Claudio Greco for useful discussions on this work.

11. Notes and references

1. A. J. Szwalbe, K. Williams, D. E. O'Flynn, A. M. Bailey, N. P. Mulholland, J. L. Vincent, C. L. Willis, R. J. Cox and T. J. Simpson, Chem. Commun., 2015, 51, 17088–17091 RSC.
2. X. L. Chen, Y. G. Zheng and Y. C. Shen, Chem. Rev., 2007, 107, 1777–1830 CrossRef CAS.
3. M. Isaka, M. Tanticharoen and Y. Thebtaranonth, Tetrahedron Lett., 2000, 41, 1657–1660 CrossRef CAS.
4. A. al Fahad, A. Abood, T. J. Simpson and R. J. Cox, Angew. Chem., Int. Ed., 2014, 53, 7519–7523 CrossRef CAS.
5. D. H. R. Barton and J. K. Sutherland, J. Chem. Soc., 1965, 1769–1772 RSC.
6. R. Schor and R. J. Cox, Nat. Prod. Rep., 2018, 35, 230–256 RSC.
7. J. Liu, A. Liu and Y. Hu, Nat. Prod. Rep., 2021, 38, 1469–1505 RSC.
8. K. M. J. de Mattos-Shipley, C. E. Spencer, C. Greco, D. M. Heard, D. E. O'Flynn, T. T. Dao, Z. Song, N. P. Mulholland, J. L. Vincent, T. J. Simpson, R. J. Cox, A. M. Bailey and C. L. Willis, Chem. Sci., 2020, 11, 11570–11578 RSC.
9. N. Wijkman, Justus Liebigs Ann. Chem., 1931, 485, 61–73 CrossRef CAS.
10. H. Raistrick and G. Smith, Biochem. J., 1933, 27, 1814–1819 CrossRef CAS.
11. J. E. Baldwin, D. H. Barton, J. L. Bloomer, L. M. Jackman, L. Rodriguez-Hahn and J. K. Sutherland, Experientia, 1962, 18, 345–352 CrossRef CAS.
12. T. A. Hamor, I. C. Paul, J. M. Robertson and G. A. Sim, Experientia, 1962, 18, 352–354 CrossRef CAS PubMed.
13. G. Ferguson, J. M. Robertson and G. A. Sim, Proceedings of the Chemical Society of London, 1962, 385 CAS.
14. D. H. R. Barton, L. M. Jackman, L. Rodriguez-Hahn and J. K. Sutherland, J. Chem. Soc., 1965, 1772–1778 RSC.
15. D. H. R. Barton, L. D. S. Godinho and J. K. Sutherland, J. Chem. Soc., 1965, 1779–1786 RSC.
16. T. Dethoup, D. Kumla and A. Kijjoa, J. Biopestic., 2015, 8, 107–115 CAS.
17. D. H. R. Barton, Reason and Imagination: Reflections on Research in Organic Chemistry, World Scientific Publishing Co Pte Ltd, 1996 Search PubMed.
18. C. Y. Li, R. Y. Yang, Y. C. Lin and S. N. Zhou, Chem. Nat. Compd., 2006, 42, 290–293 CrossRef CAS.
19. C. Y. Li, R. Y. Yang, Y. C. Lin, Z. G. She and S. N. Zhou, J. Asian Nat. Prod. Res., 2007, 9, 285–291 CrossRef CAS.
20. B. J. Wilson and C. H. Wilson, J. Bacteriol., 1962, 84, 283–290 CrossRef CAS PubMed.
21. R. J. Townsend, M. O. Moss and H. M. Peck, J. Pharm. Pharmacol., 1966, 18, 471–473 CrossRef CAS.
22. M. O. Moss, F. V. Robinson, A. B. Wood, H. M. Paisley and J. Feeney, Nature, 1968, 220, 767–770 CrossRef CAS.
23. M. O. Moss, A. B. Wood and F. V. Robinson, Tetrahedron Lett., 1969, 367–370 CrossRef CAS.
24. G. Buchi, K. M. Sander, J. D. White, J. Z. Gougoutas and S. Singh, J. Am. Chem. Soc., 1970, 92, 6638–6641 CrossRef CAS.
25. R. D. Chen, Z. Yan, J. H. Zou, N. Wang and J. G. Dai, Chin. Chem. Lett., 2014, 25, 1308–1310 CrossRef CAS.
26. S. Wada, I. Usami, Y. Umezawa, H. Inoue, S. Ohba, T. Someno, M. Kawada and D. Ikeda, Cancer Sci., 2010, 101, 743–750 CrossRef CAS PubMed.
27. J. L. Blanchard, D. M. Epstein, M. D. Boisclair, J. Rudolph and K. Pal, Bioorg. Med. Chem. Lett., 1999, 9, 2537–2538 CrossRef CAS PubMed.
28. A. R. Ortiz, M. T. Pisabarro and F. Gago, J. Med. Chem., 1993, 36, 1866–1879 CrossRef CAS.
29. M. D. Wu, M. J. Cheng, B. C. Wang, Y. J. Yech, J. T. Lai, Y. H. Kuo, G. F. Yuan and I. S. Chen, J. Nat. Prod., 2008, 71, 1258–1261 CrossRef CAS.
30. T. Wang, Y. Zhang, Y. Wang and Y. H. Pei, Toxicol. in Vitro, 2007, 21, 646–650 CrossRef CAS.
31. H. Inoue, T. Someno, T. Kato, H. Kumagai, M. Kawada and D. Ikeda, J. Antibiot., 2009, 62, 63–67 CrossRef CAS.
32. Y. Ohnishi, N. Okino, M. Ito and S. Imayama, Clin. Diagn. Lab. Immunol., 1999, 6, 101–104 CrossRef CAS PubMed.
33. S. Natori, S. Sakaki, H. Kurata, S. I. Udagawa, M. Ichinoe, M. Saito, M. Umeda and K. Ohtsubo, Appl. Microbiol., 1970, 19, 613–617 CrossRef CAS.
34. J. Y. Zhao, X. J. Wang, Z. Liu, F. X. Meng, S. F. Sun, F. Ye and Y. B. Liu, J. Nat. Prod., 2019, 82, 2953–2962 CrossRef CAS.
35. T. Amagasa, R. N. Paul, J. J. Heitholt and S. O. Duke, Pestic. Biochem. Physiol., 1994, 49, 37–52 CrossRef CAS.
36. M. Futagawa, D. E. Wedge and F. E. Dayan, Pestic. Biochem. Physiol., 2002, 73, 87–93 CrossRef CAS.
37. J. Bai, D. J. Yan, T. Zhang, Y. Z. Guo, Y. B. Liu, Y. Zou, M. C. Tang, B. Y. Liu, Q. Wu, S. S. Yu, Y. Tang and Y. C. Hu, Angew. Chem., Int. Ed., 2017, 56, 4782–4786 CrossRef CAS PubMed.
38. W. Sosroseno, I. Barid, E. Herminajeng and H. Susilowati, Oral Microbiol. Immunol., 2002, 17, 72–78 CrossRef CAS.
39. G. M. Strunz, M. Kakushima and M. A. Stillwell, J. Chem. Soc., Perkin Trans. 1, 1972, 18, 2280–2283 RSC.
40. W. A. Ayer, P. P. Lu, H. Orszanska and L. Sigler, J. Nat. Prod., 1993, 56, 1835–1838 CrossRef CAS.
41. K. Arai, S. Shimizu, H. Miyajima and Y. Yamamoto, Chem. Pharm. Bull., 1989, 37, 2870–2872 CrossRef CAS.
42. R. I. Crane, P. Hedden, J. MacMillan and W. B. Turner, J. Chem. Soc., Perkin Trans. 1, 1973, 194–200 RSC.
43. A. W. Archer and W. C. Taylor, Phytochemistry, 1987, 26, 2117–2119 CrossRef CAS.
44. T. Hosoe, K. Fukushima, T. Itabashi, K. Nozawa, K. Takizawa, K. Okada, G. M. D. Takaki and K. Kawai, J. Antibiot., 2004, 57, 573–578 CrossRef CAS PubMed.
45. T. Hosoe, J. B. Gloer, D. T. Wicklow, H. A. Raja and C. A. Shearer, Heterocycles, 2010, 81, 2123–2130 CrossRef CAS.
46. V. Suthiphasilp, A. Raksat, T. Maneerat, S. Hadsadee, S. Jungsuttiwong, S. G. Pyne, P. Chomnunti, W. Jaidee, R. Charoensup and S. Laphookhieo, J. Fungi, 2021, 7, 408 CrossRef CAS PubMed.
47. M. Nakajima, K. Itoi, Y. Takamatsu, S. Sato, Y. Furukawa, K. Furuya, T. Honma, J. Kadotani, M. Kozasa and T. Haneishi, J. Antibiot., 1991, 44, 1065–1072 CrossRef CAS.
48. R. A. Samson, J. Houbraken, J. Varga and J. C. Frisvad, Persoonia, 2009, 22, 14–27 CrossRef CAS PubMed.
49. S. C. Fields, L. Mireles-Lo and B. C. Gerwick, J. Nat. Prod., 1996, 59, 698–700 CrossRef CAS.
50. K. Williams, A. J. Szwalbe, C. Dickson, T. R. Desson, N. P. Mulholland, J. L. Vincent, J. M. Clough, A. M. Bailey, C. P. Butts, C. L. Willis, T. J. Simpson and R. J. Cox, Chem. Commun., 2017, 53, 7965–7968 RSC.
51. T. T. Dabrah, H. J. Harwood, L. H. Huang, N. D. Jankovich, T. Kaneko, J. C. Li, S. Lindsey, P. M. Moshier, T. A. Subashi, M. Therrien and P. C. Watts, J. Antibiot., 1997, 50, 1–7 CrossRef CAS PubMed.
52. T. T. Dabrah, T. Kaneko, W. Massefski and E. B. Whipple, J. Am. Chem. Soc., 1997, 119, 1594–1598 CrossRef CAS.
53. L. H. Zhang, Y. N. Wang, L. Zhang, B. Y. Liu, C. Zhang, D. J. Yan, J. Bai and Y. C. Hu, Org. Chem. Front., 2021, 8, 5926–5933 RSC.
54. D. F. Meng, Q. Tan and S. J. Danishefsky, Angew. Chem., Int. Ed., 1999, 38, 3197–3201 CrossRef CAS.
55. P. Spencer, F. Agnelli and G. A. Sulikowski, Org. Lett., 2001, 3, 1443–1445 CrossRef CAS.
56. P. Spencer, F. Agnelli, H. J. Williams, N. P. Keller and G. A. Sulikowski, J. Am. Chem. Soc., 2000, 122, 420–421 CrossRef CAS.
57. Y. Zhao, C. Sun, L. Huang, X. Zhang, G. Zhang, Q. Che, D. Li and T. Zhu, J. Nat. Prod., 2021, 84, 3011–3019 CrossRef CAS.
58. G. Stork, J. M. Tabak and J. F. Blount, J. Am. Chem. Soc., 1972, 94, 4735–4737 CrossRef CAS.
59. J. D. White, M. P. Dillon and R. J. Butlin, J. Am. Chem. Soc., 1992, 114, 9673–9674 CrossRef CAS.
60. J. D. White, J. Kim and N. E. Drapela, J. Am. Chem. Soc., 2000, 122, 8665–8671 CrossRef CAS.
61. C. Chen, M. E. Layton, S. M. Sheehan and M. D. Shair, J. Am. Chem. Soc., 2000, 122, 7424–7425 CrossRef CAS.
62. K. C. Nicolaou, J. K. Jung, W. H. Yoon, Y. He, Y. L. Zhong and P. S. Baran, Angew. Chem., Int. Ed., 2000, 39, 1829–1832 CrossRef CAS PubMed.
63. N. Waizumi, T. Itoh and T. Fukuyama, J. Am. Chem. Soc., 2000, 122, 7825–7826 CrossRef CAS.
64. J. C. Leung, A. A. Bedermann, J. T. Njardarson, D. A. Spiegel, G. K. Murphy, N. Hama, B. M. Twenter, P. Dong, T. Shirahata, I. M. McDonald, M. Inoue, N. Taniguchi, T. C. McMahon, C. M. Schneider, N. Tao, B. M. Stoltz and J. L. Wood, Angew. Chem., Int. Ed., 2018, 57, 1991–1994 CrossRef CAS.
65. J. S. Clark, F. Marlin, B. Nay and C. Wilson, Org. Lett., 2003, 5, 89–92 CrossRef CAS.
66. J. S. Clark, J. M. Northall, F. Marlin, B. Nay, C. Wilson, A. J. Blake and M. J. Waring, Org. Biomol. Chem., 2008, 6, 4012–4025 RSC.
67. J. C. Tung, W. S. Chen, B. C. Noll, R. E. Taylor, S. C. Fields, W. H. Dent and F. R. Green, Synthesis, 2007, 2388–2396 CAS.
68. C. Steinborn, R. E. Wildermuth, D. M. Barber and T. Magauer, Angew. Chem., Int. Ed., 2020, 59, 17282–17285 CrossRef CAS.
69. R. E. Wildermuth, C. Steinborn, D. M. Barber, K. S. Muhlfenzl, M. Kendlbacher, P. Mayer, K. Wurst and T. Magauer, Chem.–Eur. J., 2021, 27, 12181–12189 CrossRef CAS.
70. T. Watanabe, T. Yasumoto, M. Murata, M. Tagawa, H. Narushima, T. Furusato, M. Kuwahara, M. Hanaue and T. Seki, Europe Pat., EP0582267A1, Nissan Chemical Corp., United States, 1994 Search PubMed.
71. M. Futagawa, A. M. Rimando, M. R. Tellez and D. E. Wedge, J. Agric. Food Chem., 2002, 50, 7007–7012 CrossRef CAS.
72. M. Saleem, H. Hussain, I. Ahmed, S. Draeger, B. Schulz, K. Meier, M. Steinert, G. Pescitelli, T. Kurtan, U. Floerke and K. Krohn, Eur. J. Org. Chem., 2011, 808–812 CrossRef CAS.
73. M. Isaka, S. Palasarn, S. Sommai, P. Laksanacharoen and K. Srichomthong, Nat. Prod. Res., 2018, 32, 1506–1511 CrossRef CAS.
74. J. E. Baldwin, A. Beyeler, R. J. Cox, C. Keats, G. J. Pritchard, R. M. Adlington and D. J. Watkin, Tetrahedron, 1999, 55, 7363–7374 CrossRef CAS.
75. K. Williams, A. J. Szwalbe, N. P. Mulholland, J. L. Vincent, A. M. Bailey, C. L. Willis, T. J. Simpson and R. J. Cox, Angew. Chem., Int. Ed., 2016, 55, 6783–6787 Search PubMed.
76. J. L. Bloomer, C. E. Moppett and J. K. Sutherland, Chem. Commun., 1965, 24, 619–621 RSC.
77. J. L. Bloomer, C. E. Moppett and J. K. Sutherland, J. Chem. Soc. C, 1968, 588–591 RSC.
78. R. E. Cox and J. S. E. Holker, J. Chem. Soc., Chem. Commun., 1976, 583–584 RSC.
79. S. Nieminen, T. G. Payne, P. Senn and C. Tamm, Helv. Chim. Acta, 1981, 64, 2162–2174 CrossRef CAS.
80. D. M. Heard, E. R. Tayler, R. J. Cox, T. J. Simpson and C. L. Willis, Tetrahedron, 2020, 76, 130717 CrossRef CAS.
81. R. K. Huff, C. E. Moppett and J. K. Sutherland, J. Chem. Soc., Perkin Trans. 1, 1972, 2584–2590 RSC.
82. G. A. Sulikowski, F. Agnelli and R. M. Corbett, J. Org. Chem., 2000, 65, 337–342 CrossRef CAS PubMed.
83. C. E. Moppett and J. K. Sutherland, Chem. Commun., 1966, 21, 772–773 RSC.
84. G. A. Sulikowski, F. Agnelli, P. Spencer, J. M. Koomen and D. H. Russell, Org. Lett., 2002, 4, 1447–1450 CrossRef CAS PubMed.
85. J. Franke and C. Hertweck, Cell Chem. Biol., 2016, 23, 1179–1192 CrossRef CAS PubMed.
86. R. K. Huff, C. E. Moppett and J. K. Sutherland, Chem. Commun., 1968, 1192–1193 RSC.
87. J. E. Baldwin, R. M. Adlington, F. Roussi, P. G. Bulger, R. Marquez and A. V. W. Mayweg, Tetrahedron, 2001, 57, 7409–7416 CrossRef CAS.
88. G. A. Sulikowski, W. D. Liu, F. Agnelli, R. M. Corbett, Z. S. Luo and S. J. Hershberger, Org. Lett., 2002, 4, 1451–1454 CrossRef CAS PubMed.
89. P. G. Bulger, S. K. Bagal and R. Marquez, Nat. Prod. Rep., 2008, 25, 254–297 RSC.
90. R. Fujii, Y. Matsu, A. Minami, S. Nagamine, I. Takeuchi, K. Gomi and H. Oikawa, Org. Lett., 2015, 17, 5658–5661 CrossRef CAS.
91. N. P. Keller and T. M. Hohn, Fungal Genet. Biol., 1997, 21, 17–29 CrossRef CAS.
92. M. H. Medema, R. Kottmann, P. Yilmaz, M. Cummings, J. B. Biggins, K. Blin, I. de Bruijn, Y. H. Chooi, J. Claesen, R. C. Coates, P. Cruz-Morales, S. Duddela, S. Duesterhus, D. J. Edwards, D. P. Fewer, N. Garg, C. Geiger, J. P. Gomez-Escribano, A. Greule, M. Hadjithomas, A. S. Haines, E. J. N. Helfrich, M. L. Hillwig, K. Ishida, A. C. Jones, C. S. Jones, K. Jungmann, C. Kegler, H. U. Kim, P. Koetter, D. Krug, J. Masschelein, A. V. Melnik, S. M. Mantovani, E. A. Monroe, M. Moore, N. Moss, H.-W. Nuetzmann, G. Pan, A. Pati, D. Petras, F. J. Reen, F. Rosconi, Z. Rui, Z. Tian, N. J. Tobias, Y. Tsunematsu, P. Wiemann, E. Wyckoff, X. Yan, G. Yim, F. Yu, Y. Xie, B. Aigle, A. K. Apel, C. J. Balibar, E. P. Balskus, F. Barona-Gomez, A. Bechthold, H. B. Bode, R. Borriss, S. F. Brady, A. A. Brakhage, P. Caffrey, Y.-Q. Cheng, J. Clardy, R. J. Cox, R. De Mot, S. Donadio, M. S. Donia, W. A. van der Donk, P. C. Dorrestein, S. Doyle, A. J. M. Driessen, M. Ehling-Schulz, K.-D. Entian, M. A. Fischbach, L. Gerwick, W. H. Gerwick, H. Gross, B. Gust, C. Hertweck, M. Hofte, S. E. Jensen, J. Ju, L. Katz, L. Kaysser, J. L. Klassen, N. P. Keller, J. Kormanec, O. P. Kuipers, T. Kuzuyama, N. C. Kyrpides, H.-J. Kwon, S. Lautru, R. Lavigne, C. Y. Lee, B. Linquan, X. Liu, W. Liu, A. Luzhetskyy, T. Mahmud, Y. Mast, C. Mendez, M. Metsa-Ketela, J. Micklefield, D. A. Mitchell, B. S. Moore, L. M. Moreira, R. Mueller, B. A. Neilan, M. Nett, J. Nielsen, F. O'Gara, H. Oikawa, A. Osbourn, M. S. Osburne, B. Ostash, S. M. Payne, J.-L. Pernodet, M. Petricek, J. Piel, O. Ploux, J. M. Raaijmakers, J. A. Salas, E. K. Schmitt, B. Scott, R. F. Seipke, B. Shen, D. H. Sherman, K. Sivonen, M. J. Smanski, M. Sosio, E. Stegmann, R. D. Suessmuth, K. Tahlan, C. M. Thomas, Y. Tang, A. W. Truman, M. Viaud, J. D. Walton, C. T. Walsh, T. Weber, G. P. van Wezel, B. Wilkinson, J. M. Willey, W. Wohlleben, G. D. Wright, N. Ziemert, C. Zhang, S. B. Zotchev, R. Breitling, E. Takano and F. O. Gloeckner, Nat. Chem. Biol., 2015, 11, 625–631 CrossRef CAS PubMed.
93. G. Wiegand and S. J. Remington, Annu. Rev. Biophys. Biophys. Chem., 1986, 15, 97–117 CrossRef CAS PubMed.
94. A. R. Horswill and J. C. Escalante-Semerena, Biochemistry, 2001, 40, 4703–4713 CrossRef CAS.
95. M. F. Grau, R. Entwistle, Y. M. Chiang, M. Ahuja, C. E. Oakley, T. Akashi, C. C. C. Wang, R. B. Todd and B. R. Oaldey, ACS Chem. Biol., 2018, 13, 3193–3205 CrossRef CAS PubMed.
96. K. Williams, K. M. J. de Mattos-Shipley, C. L. Willis and A. M. Bailey, Fungal Biol. Biotechnol., 2022, 9, 2 CrossRef CAS.
97. R. Fujii, A. Minami, T. Tsukagoshi, N. Sato, T. Sahara, S. Ohgiya, K. Gomi and H. Oikawa, Biosci., Biotechnol., Biochem., 2011, 75, 1813–1817 CrossRef CAS.
98. K. Tagami, C. W. Liu, A. Minami, M. Noike, T. Isaka, S. Fueki, Y. Shichijo, H. Toshima, K. Gomi, T. Dairi and H. Oikawa, J. Am. Chem. Soc., 2013, 135, 1260–1263 CrossRef CAS PubMed.
99. C. W. Liu, K. Tagami, A. Minami, T. Matsumoto, J. C. Frisvad, H. Suzuki, J. Ishikawa, K. Gomi and H. Oikawa, Angew. Chem., Int. Ed., 2015, 54, 5748–5752 CrossRef CAS PubMed.
100. N. Yilmaz, C. M. Visagie, J. Houbraken, J. C. Frisvad and R. A. Samson, Stud. Mycol., 2014, 78, 175–341 CrossRef CAS PubMed.
101. S. Yin, S. Friedrich, V. Hrupins and R. J. Cox, RSC Adv., 2021, 11, 14922–14931 RSC.
102. F. L. Li, C. H. Hagemeier, H. Seedorf, G. Gottschalk and R. K. Thauer, J. Bacteriol., 2007, 189, 4299–4304 CrossRef CAS.
103. J. A. Francois, C. M. Starks, S. Sivanuntakorn, H. Jiang, A. E. Ransome, J. W. Nam, C. Z. Constantine and T. J. Kappock, Biochemistry, 2006, 45, 13487–13499 CrossRef CAS PubMed.
104. R. Y. Eberhardt, Y. Y. Chang, A. Bateman, A. G. Murzin, H. L. Axelrod, W. C. Hwang and L. Aravind, BMC Bioinf., 2013, 14, 327 CrossRef PubMed.
105. T. Shiina, T. Ozaki, Y. Matsu, S. Nagamine, C. W. Liu, M. Hashimoto, A. Minami and H. Oikawa, Org. Lett., 2020, 22, 1997–2001 CrossRef CAS.
106. L. Serre, B. Vallée, N. Bureaud, F. Schoentgen and C. Zelwer, Structure, 1998, 6, 1255–1265 CrossRef CAS PubMed.
107. C. L. M. Gilchrist and Y.-H. Chooi, Bioinformatics, 2021, 37, 2473–2475 CrossRef CAS.
108. L. F. Wu, S. Meng and G. L. Tang, Biochim. Biophys. Acta, Proteins Proteomics, 2016, 1864, 453–470 CrossRef CAS.
109. I. J. Clifton, M. A. McDonough, D. Ehrismann, N. J. Kershaw, N. Granatino and C. J. Schofield, J. Inorg. Biochem., 2006, 100, 644–669 CrossRef CAS.
110. B. Tudzynski, M. Mihlan, M. C. Rojas, P. Linnemannstons, P. Gaskin and P. Hedden, J. Biol. Chem., 2003, 278, 28635–28643 CrossRef CAS PubMed.

This journal is © The Royal Society of Chemistry 2023