Carboxylic acid reductase: Structure and mechanism

Journal Pre-proof Carboxylic acid reductase: Structure and mechanism Deepankar Gahloth, Godwin A. Aleku, David Leys

PII:

S0168-1656(19)30890-9

DOI:

https://doi.org/10.1016/j.jbiotec.2019.10.010

Reference:

BIOTEC 8527

To appear in:

Journal of Biotechnology

Received Date:

21 June 2019

Revised Date:

10 October 2019

Accepted Date:

11 October 2019

Please cite this article as: Gahloth D, Aleku GA, Leys D, Carboxylic acid reductase: Structure and mechanism, Journal of Biotechnology (2019), doi: https://doi.org/10.1016/j.jbiotec.2019.10.010

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

Carboxylic acid reductase: structure and mechanism Deepankar Gahloth, Godwin A. Aleku, David Leys* School of Chemistry, MIB, Princess street 131, M1 7DN, Manchester, UK

*[email protected]



ro of

Highlights

Carboxylic acid reductase (CAR) enzymes catalyse the ATP- and NADPH-dependent reduction of wide range of acids to aldehydes

Recent structural and solution studies reveal the dynamic

-p





re

mechanism of carboxylic acid reduction.

This provides a blueprint for future structure based improvement of

lP

carboxylic acid reduction activity/substrate scope to develop CAR as robust biocatalyst.

ur na

Abstract:

Carboxylic acid reductase (CAR) enzymes are large multi-domain proteins that catalyse the ATP- and NADPH-dependent reduction of wide range of

Jo

acids to the corresponding aldehydes. This particular reaction is of considerable biotechnological interest. Recent advances in the structural and solution studies of isolated domain, di-domain and full-length CAR enzymes revealed valuable insights into the mechanism of carboxylic acid reduction activity. This review features the phylogenetic, sequence and structural insight into the CAR and implications of these observations in

1

order to improve carboxylic acid reduction activity to develop CAR as robust biocatalyst.

Introduction Biocatalysis has been extensively used in the chemical industries for the manufacture of products ranging from specialty to commodity chemicals (Bornscheuer et al., 2012; Ciriminna and Pagliaro, 2013). In the pharmaceutical

ro of

sector where chirality plays critical roles in biological actions of medicinal

products, the high specificity and stereoselectivity of enzyme-catalysed reactions are particularly attractive; hence biocatalysis is emerging as preferred synthetic

-p

routes for stereoselective synthesis of chiral pharmaceuticals (Bornscheuer et al., 2012; Ciriminna and Pagliaro, 2013; Hollmann et al., 2011; Straathof, 2014).

re

Aldehydes have applications in fine chemicals, pharmaceutical and flavor

lP

fragrance industries, and are important intermediates for the preparation of several high-value compounds including alcohols and amines. Owing to the

ur na

high reactivity of aldehydes and poor chemo-selectivity of abiotic catalysts, the chemical preparation of aldehydes is a relatively difficult process, requiring appropriate protection and deprotection steps with hazardous reagents yielding harmful waste (Gross and Zenk, 1969; Kato et al., 1991; Kunjapur and Prather,

Jo

2015; Li and Rosazza, 1997; Napora-Wijata et al., 2014). Carboxylic acid reductase (CAR) enzymes catalyse the ATP- and NADPH-dependent reduction of carboxylic acids to the corresponding aldehydes under ambient conditions and thus offer tremendous potential for future applications in organic synthesis. The first fully characterised CAR biocatalyst was isolated from Nocardia iowensis

2

(CARNi)(Venkitasubramanian et al., 2007); and provided the valuable template for structural and mechanistic studies of these enzymes family (Li and Rosazza, 1997; Napora-Wijata et al., 2014; Venkitasubramanian et al., 2007). Subsequently, several other homologues have been characterised, and the enzyme family has been shown to reduce a broad range of carboxylic acids to aldehydes under physiological conditions (Akhtar et al., 2013; Duan et al., 2015; Finnigan, 2016; Schwendenwein et al., 2016; Winkler and Winkler, 2016). The

ro of

strict chemo-selectivity of CARs, reducing carboxylic acids solely to the aldehyde

products, has inspired their incorporation into multi-enzyme cascades production of biofuel (Akhtar et al., 2013; Kallio et al., 2014). Recently, CARs

-p

have been combined with amine transaminases, imine reductases (France et al.,

2016) and reductive aminases ( Aleku et al. 2017, Ramsden et al., 2019) for the

re

synthesis of cyclic amines and N-alkylated secondary amines.

lP

The huge biocatalytic potential of CARs and the desire to improve their catalytic properties to enhance their industrial application have inspired significant

ur na

interest in the mechanistic aspects of this interesting enzyme class. Here, we review recent structural and mechanistic studies of CAR-catalysed reaction and we highlighted how these insights can guide the tailoring of these enzymes for applications.

Jo

CARs are large multi-domain enzymes and form a distant relative of the NRPS family (Drake et al., 2016; Gulick, 2016; Reimer et al., 2016; Strieker et al., 2010). CAR consists of a substrate activating adenylation domain (A-domain), which is more closely related to the acyl-CoA synthetase members of the ANL superfamily of adenylating enzymes. The C-terminus of the CAR adenylation domain is fused to a phosphopantetheine carrier protein domain (PCP-domain) (Figure 1). To

3

obtain active CAR enzyme, a phosphopantetheine group must be attached covalently to a conserved serine in the PCP domain mediated by a suitable phosphopantetheine transferase enzyme. Finally, the C-terminal region of the PCP domain further extends into the terminal reductase domain (R-domain) that yields the aldehyde product. Related NRPS reductase domains catalyse the progressive four-electron reduction of the PCP bound acyl group to the corresponding alcohol (Chhabra et al., 2012). In contrast, CARs catalyse a strict

ro of

two-electron reduction, producing the corresponding aldehyde. Thus, in CARs,

the irreversible reduction of a carboxylic acid occurs via a multistep process involving (i) Adenylation: ATP-dependent activation of the acid mediated by the

-p

A-domain to form an acyl adenylate intermediate (ii) Thiolation: the transfer of the acyl group onto the PCP phosphopantethiene linker and finally (iii)

re

Reduction: reduction of the acyl-thioester at the R-domain yielding the aldehyde

lP

product (Figure 1).

Recently, the first crystal structures and associated solution studies of CARs from

ur na

Segniliparus rugosus (CARsr, #WP_007468889), Nocardia iowensis (CARni, #Q6RKB1.1) and Mycobacterium marium (CARmm, #WP_012393886.1) reveal substantial domain dynamics are required in order to complete the catalytic cycle (Gahloth et al., 2017). Furthermore, they provide an insight into the

Jo

mechanism ensuring specific 2-electron reduction occurs as opposed to the 4electron reduction observed in distantly related reducing systems.

Phylogenetic and sequence analysis of CAR family Winkler and co-workers described a detailed phylogenetic and bio-informatics analysis of CAR family (E.C.1.2.1.30) (Stolterfoht et al., 2017). CARs sequences

4

are found in both bacteria (Actinobacteria, Firmicutes, and Proteobacteria) and fungi (Ascomycota and Basidiomycota). The CAR family (E.C.1.2.1.30) can be classified into four major clades: Type 1 (Corynebacteriales, Streptomycetales, Terrabacteria), Type 2 (Bacillales, Terrabacteria), Type 3 (Pseudomonadales, gProteobacteria), and Type 4 (a mixed group, mainly Enterobacteriales). A minor clade Type 5 (Burkholderiales) was included later (Khusnutdinova et al., 2017). Bioinformatics sequence analysis of CARs identifies highly conserved signature

ro of

sequences in each of the catalytic domains (Stolterfoht et al., 2017). However, most of characterized CARs are largely from the Type 1 class (Finnigan et al., 2017). CAR sequences vary in the sequence similarity 50-90%, suggestive of a

-p

wide substrate specificity of CAR enzymes (Figure 2).

carrier di-domain module

re

Strucutral insights into the Adenylation domain and Adenylation-peptidyl

lP

The adenylation domain acts as a gatekeeper for incoming substrates, and hence determines the CAR substrate specificity. The A-domain structures of CARsr and

ur na

CARni in complex with AMP and substrates have been reported (Gahloth et al., 2017). Adenylation domains can be divided into an N-terminal Acore and a Cterminal Asub domain, as observed in case of homologous structures such as bacterial benzoate-CoA ligase and human medium-chain acyl-coenzyme A

Jo

synthetase ACSM2A (Kochan et al., 2009; Thornburg et al., 2015). In these and other NPRS systems, the C-terminal domain (Asub) adopts distinct conformation according to the catalytic state (Gulick, 2009). Both isolated A-domain CAR structures adopt a conformation of the Asub region that corresponds to the adenylation state of the catalytic cycle. The interface established between Acore and Asub is more extensive in CAR than the other members of the ANL

5

superfamily. The substrate binding site in CAR adenylation domain is mostly lined by hydrophobic residues, with conserved His300 (CARni numbering) located close to both AMP phosphate and substrate ligands. Mutagenesis of His300 abolishes activity, demonstrating a key role in the catalytic process. Sequence alignment of the A-domain from all four type of CARs demonstrates that members of different type/class are only distantly related and in fact share rather low sequence homology (Figure 3). Given that residues around the active

ro of

site are not strictly conserved, this is likely to affect shape and size of active site

and hence the substrate specificity. For example, the active site shape and size

are different in CARsr and CARni due to substitution of Phe294 (CARsr) to

-p

Ser280 (CARni) and insertion of Ala425 in CARni. Homology models of adenylation domain of CARs from Type II: Aspergillus terreus (CARat)

re

#XP_001212808.1; Type III: Neurospora crassa (CARnc) #YP_955820.1; Type IV:

lP

Trametes versicolor (CARtv) #XP_008043822.1 were built using either CARni or CARsr adenylation structure as template. These models reveal that differences

ur na

were observed mostly due to the insertion/deletions in the amino acid sequence as suggested by multiple sequence alignment (Figure 3). Clearly, crystal structures of adenylation domain from distinct CAR classes are required to provide

further

insights

into

substrate

specificity

and

guide

future

Jo

evolution/application of these enzymes. The CARsr A-PCP di-domain module crystal structures yielded both the Adenylation and Thiolation state conformation of the Asub region, confirming that A-PCP module is a dynamic entity. In the adenylation state, the PCP domain is positioned distantly from the A domain, with the Asub-PCP linker region adopting an extended alpha helical conformation.

This positions the conserved

6

phosphopantetheine attachment site serine residue at ~52 Å away from the bound AMP phosphate. In contrast, in the thiolation state the relative orientation of both PCP and Asub domain is altered dramatically with both the Asub and PCP regions adopting a distinct conformation (Figure 4): the adenylation Asub domain rotates ~165° at the Lys528 hinge region. Furthermore, the relative orientation of the PCP domain and Asub domain also changes, due to an additional ~75° rotation at Ala651. The combined rotational motions at Lys528 and Ala651 leads

ro of

to the displacement of the PCP centre of gravity by ~50 Å (Figure 4), forming a

new interaction surface between Acore and the PCP domain placing PCP Ser-702

-p

at ~19 Å to AMP phosphate.

An Asp on-off activity switch in the CAR R-domain

re

Crystal structure of the isolated reductase domain from CARmm combined with

lP

the structure of the PCP-R didomain modules from CARsr revealed the underlying mechanism ensuring strict two-electron reduction. Two distinct

ur na

conformation of Asp984 (Asp998 in CARsr) in CARmm reductase were observed, corresponding to an inactive (off) and active (on) conformation. In the active state (on), the NADPH nicotinamide moiety is ordered and placed adjacent to conserved residues Thr921 and Tyr956, key residues that assists in thioester

Jo

reduction. In the on state, Asp984 is buried within the protein and pointing away from the nicotinamide. However, in inactive state (off), Asp984 adopts a different conformation leading to a disordered nicotinamide moiety. The motion of Asp984 between both conformations is concomitant with reorientation of the backbone of residues 983-985. The on-off equilibrium of Asp984 appears poised towards the off state in absence of substrate. In case of PCP-R didomain

7

structures, the PCP-domain is docked onto the larger NADPH-binding domain, located directly above the ribose 2’-phosphate binding pocket. The PCP phosphopantetheine linker Ser702 is located 16 Å from the nicotinamide binding pocket (Figure 5). Comparison of apo- and holo- PCP-R didomain crystal structures established that binding of the phosphopanteine affects the conformational equilibrium of Asp984, which remains in the inactive state in the absence of the

ro of

phospopantheine linker. Hence, only the docking of the acyl-thioester substrate leads to reductase domain activation, ensuring two electron reduction occurs, as

the aldehyde product is unable to alter the equilibrium towards the active state.

-p

Indeed, it is suggested that the R-domain thioester substrate affinity is largely

dependent on the phosphopantetheine linker region, as opposed to substrate

re

derived acyl-moiety. This is supported by the fact that mutation of residues

lP

involves in phosphopentetheine binding by the R domain completely abolishes activity, while only benzoyl-CoA but not other thioester substrates can act as

ur na

alternative substrate. Furthermore, weak benzoic acid reduction activity using isolated A- and R-domains could be obtained only by addition of R-pantetheine, effectively bypassing the need of PCP domain and demonstrating the need for the

Jo

pantetheine group for activity.

A dynamic CAR-model suggests future routes to application A full-length model of CAR has been built based on the crystal structures and solution data in the adenylation-reduction conformation, which may suggest that the reduction of the substrate-phosphopantethine linkage might occur simultaneously with activation of the next substrate molecule in the adenylation

8

domain. A putative model for thiolation state has also been proposed by combining crystal structures and solution models, which suggests a dramatic reorientation of reductase with respect to adenylation domain for the transition from adenylation-reduction state to thiolation state. Comparison of two models revealed that despite the intricate mechanism of the individual catalytic domains, the communication and transfer of the activated acyl group appears to depend on a relatively simple system of beads on a string type arrangement with

ro of

no interaction between A and R domains (Figure 6). The lack of sophisticated communication between the CAR terminal domains suggests that CAR can be

engineered/modified by domain exchange. Chimeric CARs can provide a

-p

promising and previously unexplored application of these multi-domain enzymes, potentially extending the application of the enzymes beyond the

re

synthesis of aldehydes. Given that free (R)-pantetheine can be used as a

lP

substitute to rescue the activity of isolated CAR domains, this
 circumvents the
 initial need to
 optimise domain
 linkers in full-
 length chimeric
 enzymes.

ur na

Indeed, as a ‘chemical bypass’, this method could be used to probe whether an adenylation domain could be used in combination with alternative C-terminal domains in place of the R-domain, in order to create novel chimeras capable of

Jo

catalysing alternative reactions with carboxylic acids as substrates.

Engineering carboxylic acid reductases Hence, CARs represent a versatile set of biocatalysts with potential for industrial applications. Although incorporation into main stream industrial synthetic processes is yet to be realised, recent interests in the development of these enzymes by industry groups (Finnigan et al., 2017) indicates their potential for

9

industrial applications. One notable trait of CARs is their remarkably broad substrate scope, displaying activity against a wide range of structurally diverse carboxylic acid substrates.

In order to enhance their application however,

certain properties, for example, their catalytic efficiency and activity towards sterically demanding substrates will need to be improved via directed evolution or structure-guided (semi)-rational approaches. Efforts to evolve CAR enzymes by directed evolution has been impeded by the

ro of

lack of robust high-throughput screening and selection systems to monitor CAR activity in vivo. Moreover, the complexity and the size of CARs (typically >120

kDa) make the evolution of this enzyme class by random approach a non-trivial

-p

task; hence limited attempts have been made to evolve CARs using random enzyme engineering techniques. (Ressmann et al., 2019)

re

Recently, Winkler and co-workers developed a UV and fluorescence-based high-

lP

throughput assay for aldehyde quantification in the presence of microbial cells by exploiting the reaction of amino benzamidoxime (ABAO) with aldehydes to

ur na

generate fluorescing dihydroquinazoline derivatives (Figure 7a) (Ressmann et al., 2019). This method allows the detection and quantification of aldehyde formation in the presence of resting E. coli cells. To address the problem of over reduction of the carboxylate to the corresponding alcohol product resulting from

Jo

E. coli endogenous carbonyl reductases, an engineered E. coli strain with compromised aldehyde reducing activity was used as the expression host. Applying this assay to screen libraries generated by random mutagenesis and targeting the adenyaltion and the reductase domains of CARni led to the identification of Q283 as hotspot for further evolution. (Schwendenwein et al., 2019). In particular, the screening evaluated the binding affinity of sterically

10

hindered 2-substituted benzoic acid substrates which are poor substrates of the wild-type enzyme. Gln283Arg variant was reported to exhibit an improved binding affinity for the substrate resulting in a 2-fold improvement in yield when compared with the wild-type enzyme. The identified Gln283 position was subjected to site saturation mutagenesis; screening of this single site saturation library uncover Gln282Pro mutant, which displayed 9‐fold improvement in the substrate binding affinity and a 4‐fold increase in biotransformation yield

ro of

(Schwendenwein et al., 2019).

Importantly, the development of HTA to quantify aldehyde formation in wholecells is exciting and should enable more focused saturation libraries of key

-p

residues identified from recent structural, mechanistic and site-directed mutagenesis studies (Gahloth et al., 2017; Qu et al., 2019; Stolterfoht et al.,

re

2018).

lP

Molecular clues and functional roles of amino acid residues revealed from recent structural and mechanistic studies of CARs should inspire interest in focused

ur na

(semi)rational evolution techniques. Indeed, Flitsch and co-workers have successfully exploited structural and mechanistic clues to repurpose CARs for amide bond formation (Wood et al., 2017). Ser689 which is an essential residue involved in the formation of the thioester from the acyl adenylate intermediate

Jo

generated by the enzyme’s adenylation domain was replaced by alanine. The Ser689Ala variant retains adenylation activity but has lost thiolation activity, hence allowing the accumulation of the acyl adenylate from the carboxylic acid substrate. Bio-transformation reactions supplying an amine nucleophile and employing Ser689Ala as the catalyst generated the corresponding amide from carboxylic acid. The amidation reaction is a consequence of nucleophilic

11

interaction between the formed adenylate and the supplied amine nucleophile. Furthermore, truncated CAR constructs were generated to exclude the terminal reductase domain. These were shown to exclusively catalyse amide formation , affording conversion of up to 69% (Figure 7b). The selectivity of these rationally designed CAR variants will be particularly applicable as amide synthetases in synthetic cascades especially those involving other NAD(P)H-dependent

factors are generated in situ from cellular metabolism.

Conclusions

ro of

enzymes and in whole-cell biotransformations where the nicotinamide co-

-p

Significant progress has been achieved in the past decade with regards to the

mechanistic aspects of CAR- catalysed reaction. Structural and mechanistic

re

studies published previously (Gahloth et al., 2017 ) and highlighted in this

lP

review provide valuable insights into the dynamic mechanism of A-PCP domain and induced fit activation of reductase domain. Furthermore, important protein-

ur na

protein interaction surfaces were determined as in case of A-PCP and PCP-R domain. These interactions reveal how the different parts of these enzymes communicate during the catalytic cycle. Altogether, these insights should guide future direction and efforts aimed at evolving CARs as suitable and robust

Jo

biocatalysts for industrial applications.

Declaration of interests X The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

12

Jo

ur na

lP

re

-p

Acknowledgements D.L. is a Royal Society Wolfson Merit Award holder.

ro of

☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

13

References

Jo

ur na

lP

re

-p

ro of

Akhtar, M.K., N.J. Turner, and P.R. Jones. 2013. Carboxylic acid reductase is a versatile enzyme for the conversion of fatty acids into fuels and chemical commodities. P Natl Acad Sci USA. 110:87-92. Bornscheuer, U.T., G.W. Huisman, R.J. Kazlauskas, S. Lutz, J.C. Moore, and K. Robins. 2012. Engineering the third wave of biocatalysis. Nature. 485:185-194. Chhabra, A., A.S. Haque, R.K. Pal, A. Goyal, R. Rai, S. Joshi, S. Panjikar, S. Pasha, R. Sankaranarayanan, and R.S. Gokhale. 2012. Nonprocessive [2+2]e(-) offloading reductase domains from mycobacterial nonribosomal peptide synthetases. P Natl Acad Sci USA. 109:5681-5686. Ciriminna, R., and M. Pagliaro. 2013. Green Chemistry in the Fine Chemicals and Pharmaceutical Industries. Org Process Res Dev. 17:1479-1484. Drake, E.J., B.R. Miller, C. Shi, J.T. Tarrasch, J.A. Sundlov, C.L. Allen, G. Skiniotis, C.C. Aldrich, and A.M. Gulick. 2016. Structures of two distinct conformations of holo-non-ribosomal peptide synthetases. Nature. 529:235-U289. Duan, Y.T., P.Y. Yao, X. Chen, X.T. Liu, R. Zhang, J.H. Feng, Q.Q. Wu, and D.M. Zhu. 2015. Exploring the synthetic applicability of a new carboxylic acid reductase from Segniliparus rotundus DSM 44985. J Mol Catal B-Enzym. 115:1-7. Finnigan, W. 2016. Carboxylic acid reductases and their use as well defined enzyme building blocks for the construction of in vitro cascade reactions. New Biotechnol. 33:S15-S16. Finnigan, W., A. Thomas, H. Cromar, B. Gough, R. Snajdrova, J.P. Adams, J.A. Littlechild, and N.J. Harmer. 2017. Characterization of Carboxylic Acid Reductases as Enzymes in the Toolbox for Synthetic Chemistry. Chemcatchem. 9:1005-1017. France, S.P., S. Hussain, A.M. Hill, L.J. Hepworth, R.M. Howard, K.R. Mulholland, S.L. Flitsch, and N.J. Turner. 2016. One-Pot Cascade Synthesis of Monoand Disubstituted Piperidines and Pyrrolidines using Carboxylic Acid Reductase (CAR), omega-Transaminase (omega-TA), and Imine Reductase (IRED) Biocatalysts. Acs Catal. 6:3753-3759. Gahloth, D., M.S. Dunstan, D. Quaglia, E. Klumbys, M.P. Lockhart-Cairns, A.M. Hill, S.R. Derrington, N.S. Scrutton, N.J. Turner, and D. Leys. 2017. Structures of carboxylic acid reductase reveal domain dynamics underlying catalysis. Nat Chem Biol. 13:975-+. Gross, G.G., and M.H. Zenk. 1969. Reduction of Aromatic Acids to Aldehydes and Alcohols in Cell-Free Systems .1. Purification and Properties of Aryl Aldehyde - Nadp-Oxidoreductase from Neurospora Crassa. Eur J Biochem. 8:413-+. Gulick, A.M. 2009. Conformational Dynamics in the Acyl-CoA Synthetases, Adenylation Domains of Non-ribosomal Peptide Synthetases, and Firefly Luciferase. Acs Chem Biol. 4:811-827. Gulick, A.M. 2016. Structural insight into the necessary conformational changes of modular nonribosomal peptide synthetases. Curr Opin Chem Biol. 35:89-96.

14

Jo

ur na

lP

re

-p

ro of

Hollmann, F., I.W.C.E. Arends, and D. Holtmann. 2011. Enzymatic reductions for the chemist. Green Chem. 13:2285-2314. Kallio, P., A. Pasztor, K. Thiel, M.K. Akhtar, and P.R. Jones. 2014. An engineered pathway for the biosynthesis of renewable propane. Nat Commun. 5. Kato, N., E.H. Joung, H.C. Yang, M. Masuda, M. Shimao, and H. Yanase. 1991. Purification and Characterization of Aromatic Acid Reductase from Nocardia-Asteroides Jcm-3016. Agr Biol Chem Tokyo. 55:757-762. Khusnutdinova, A.N., R. Flick, A. Popovic, G. Brown, A. Tchigvintsev, B. Nocek, K. Correia, J.C. Joo, R. Mahadevan, and A.F. Yakunin. 2017. Exploring Bacterial Carboxylate Reductases for the Reduction of Bifunctional Carboxylic Acids. Biotechnol J. 12. Kochan, G., E.S. Pilka, F. von Delft, U. Oppermann, and W.W. Yue. 2009. Structural Snapshots for the Conformation-dependent Catalysis by Human Mediumchain Acyl-coenzyme A Synthetase ACSM2A. J Mol Biol. 388:997-1008. Kunjapur, A.M., and K.L.J. Prather. 2015. Microbial Engineering for Aldehyde Synthesis. Appl Environ Microb. 81:1892-1901. Li, T., and J.P.N. Rosazza. 1997. Purification, characterization, and properties of an aryl aldehyde oxidoreductase from Nocardia sp. strain NRRL 5646. J Bacteriol. 179:3482-3487. Napora-Wijata, K., G.A. Strohmeier, and M. Winkler. 2014. Biocatalytic reduction of carboxylic acids. Biotechnol J. 9:822-843. Qu, G., M. Fu, L. Zhao, B. Liu, P. Liu, W. Fan, J.A. Ma, and Z. Sun. 2019. Computational Insights into the Catalytic Mechanism of Bacterial Carboxylic Acid Reductase. J. Chem. Inf. Model. Reimer, J.M., M.N. Aloise, P.M. Harrison, and T.M. Schmeing. 2016. Synthetic cycle of the initiation module of a formylating nonribosomal peptide synthetase. Nature. 529:239-U305. Ressmann, A.K., D. Schwendenwein, S. Leonhartsberger, M.D. Mihovilovic, U.T. Bornscheuer, M. Winkler, and F. Rudroff. 2019. Substrate-Independent High-Throughput Assay for the Quantification of Aldehydes. . Adv Synth Catal. Schwendenwein, D., G. Fiume, H. Weber, F. Rudroff, and M. Winkler. 2016. Selective Enzymatic Transformation to Aldehydes in vivo by Fungal Carboxylate Reductase from Neurospora crassa. Adv Synth Catal. 358:3414-3421. Schwendenwein, D., A.K. Ressmann, M. Doerr, M. Höhne, U.T. Bornscheuer, M.D. Mihovilovic, F. Rudroff, and M. Winkler. 2019. Random MutagenesisDriven Improvement of Carboxylate Reductase Activity using an Amino Benzamidoxime-Mediated High-Throughput Assay. Adv Synth Catal. Stolterfoht, H., D. Schwendenwein, C.W. Sensen, F. Rudroff, and M. Winkler. 2017. Four distinct types of EC 1.2.1.30 enzymes can catalyze the reduction of carboxylic acids to aldehydes Holly. J Biotechnol. 257:222-232. Stolterfoht, H., G. Steinkellner, D. Schwendenwein, T. Pavkov-Keller, K. Gruber, and M. Winkler. 2018. Identification of Key Residues for Enzymatic Carboxylate Reduction. Front Microbiol. 9. Straathof, A.J.J. 2014. Transformation of Biomass into Commodity Chemicals Using Enzymes or Cells. Chem Rev. 114:1871-1908. Strieker, M., A. Tanovic, and M.A. Marahiel. 2010. Nonribosomal peptide synthetases: structures and dynamics. Curr Opin Struc Biol. 20:234-240.

15

Jo

ur na

lP

re

-p

ro of

Thornburg, C.K., S. Wortas-Strom, M. Nosrati, J.H. Geiger, and K.D. Walker. 2015. Kinetically and Crystallographically Guided Mutations of a Benzoate CoA Ligase (BadA) Elucidate Mechanism and Expand Substrate Permissivity. Biochemistry-Us. 54:6230-6242. Venkitasubramanian, P., L. Daniels, and J.P.N. Rosazza. 2007. Reduction of carboxylic acids by Nocardia aldehyde oxidoreductase requires a phosphopantetheinylated enzyme. J Biol Chem. 282:478-485. Winkler, M., and C.K. Winkler. 2016. Trametes versicolor carboxylate reductase uncovered. Monatsh Chem. 147:575-578. Wood, A.J.L., N.J. Weise, J.D. Frampton, M.S. Dunstan, M.A. Hollas, S.R. Derrington, R.C. Lloyd, D. Quaglia, F. Parmeggiani, D. Leys, N.J. Turner, and S.L. Flitsch. 2017. Adenylation Activity of Carboxylic Acid Reductases Enables the Synthesis of Amides. Angew Chem Int Edit. 56:14498-14501.

16

Figure legends Figure 1. Schematic overview of CAR. (A) CAR has three domains: an N-terminal Adenylation domain and a C-terminal reductase domain, fused together by middle phosphopantetheine carrier protein domain. (B) Substrate enters into the active site of A domain where the intermediate acyl-AMP formed at the expense of ATP. The acyl group is transferred to the phosphopantetheine arm attached to the PCP domain. The acyl-thioester intermediate is able to dock into the active site of the reductase domain by reorientation of the PCP domain. This activates the reductase domain, resulting in formation of the aldehyde product.

ro of

The isolated CAR domains (A and R domain) are active, and their activity can be

coupled to mimic full-length CAR (albeit inefficiently) by addition of Rpantetheine (shown in grey).

-p

Figure 2. The CAR enzymes(EC.1.2.1.30) have been classified into five

phylogenetic class (Type 1-5). A phylogenetic tree showing the five different

re

classes of CAR enzymes. Crystal structure and homology models of the

lP

corresponding adenylation domain are shown.

Figure 3: Structure based multiple sequence alignment of the CAR adenylation domain belonging to Type 1, 2, 3 and 4. Type 1: Seginilaparous rugosus (CARsr),

ur na

Seginiliparous rotundus (CARsro), Nocardia iowensis (CARni) and Mycobacterium smegmatis (CARms), Type 2: Aspergillus terreus (CARat), Type 3: Neurospora crassa (CARnc) and Type 4: Tramets vesicolor (CARtv). Conserved residues are highlighted in red. His 300 (CARni) is conserved across the entire CAR family. Secondary structure element of CARsr adenylation domain is shown at top of the

Jo

sequence alignment.

Figure 4: Structures of the CAR isolated A domain, and the A-PCP didomain, in adenylation and thiolation states. PCP domain is shown in magenta, Acore in cyan and Asub domain in yellow. The phosphopantetheine binding site on PCP domain (Ser702) is shown in blue.

17

Figure 5. Superposition of unmodified and phosphopantetheine modified PCP-R domain structures shows off and on conformations of the reductase domain. PCP domain is shown in magenta while the reductase domain is shown in cyan and in green cartoon corresponding to the inactive off and active on states. The inset shows the key Asp984(CARsr) in the on and off conformation. Figure 6. A dynamic model for CAR catalysis. Based on the various isolated domain and didomain crystal structures, it is proposed large scale domain reorientation is essential for catalysis by CAR. Furthermore, models suggest little

ro of

communication occurs directly between the A and R domains, and that the adenylation and reduction reaction can occur simultaneously.

Figure 7. Enzyme engineering approaches applied to carboxylic acid reductases (CAR) a). High-throughput assay developed to quantify CAR-catalysed aldehyde

-p

formation in the presence of microbial cells. Amination of the formed aldehyde using amino benzamidoxime (ABAO) as the nucleophile and spontaneous

re

cyclization generates dihydroquinazoline derivative which can quantified by absorbance/fluorescence based methods. b). Rationally-designed CAR variants

lP

(Single point mutant CARNi S689A and truncated CARMmΔ729‐1175 which lacks

Jo

ur na

the reduction domain) were repurposed as amidation catalysts.

18

B

ro of

-p

re

lP

ur na

Jo Figure 1

A

19

ro of

-p

re

lP

ur na

Jo Figure 2

20

ro of

-p

re

lP

ur na

Jo Figure 3

21

Figure 4

Adenylation State

Thiolation State

Jo

ur na

lP

re

-p

ro of

Adenylation State

22

Figure 5

Ppant arm Asp984 On

Jo

ur na

lP

re

-p

ro of

Asp984 Off

23

ro of

-p

re

lP

ur na

Jo Figure 6

24

ro of

-p

re

lP

ur na

Jo Figure 7

25