Detection of Unprecedented CYP74 Enzyme in Mammal: Hydroperoxide Lyase CYP74C44 of the Bat Sturnira hondurensis

The genome of the neotropical fruit bat Sturnira hondurensis was recently sequenced, revealing an unexpected gene encoding a plant-like protein, CYP74C44, which shares ca. 90% sequence identity with the putative CYP74C of Populus trichocarpa. The preparation and properties of the recombinant CYP74C44 are described in the present work. The CYP74C44 enzyme was found to be active against the 13- and 9-hydroperoxides of linoleic and α-linolenic acids (13-HPOD, 13-HPOT, 9-HPOD, and 9-HPOT, respectively), as well as the 15-hydroperoxide of eicosapentaenoic acid (15-HPEPE). All substrates studied were specifically transformed into chain cleavage products that are typical for hydroperoxide lyases (HPLs). The HPL chain cleavage reaction was validated by the identification of NaBH4-reduced products (Me/TMS) of 15-HPEPE and 13- and 9-hydroperoxides as (all-Z)-14-hydroxy-5,8,11-tetradecatrienoic, (9Z)-12-hydroxy-9-dodecenoic, and 9-hydroxynonanoic acids (Me/TMS), respectively. Thus, CYP74C44 possessed the HPL activity that is typical for the CYP74C subfamily proteins.


Introduction
Cytochromes of the P450 superfamily play numerous roles in the endogenous metabolism of aerobic organisms as well as the control of detoxification of xenobiotics [1,2]. The diversity of P450s is especially high in plants. For example, the rice species Oryza sativa (japonica cultivar-group) possesses 356 P450 genes and 99 pseudogenes [3]. The majority of P450s are monooxygenases, oxidizing some hydrophobic substrates and utilizing atmospheric dioxygen as a second substrate [1,2]. All monooxygenases depend on the specific electron transfer chains required for oxygen activation. In contrast to monooxygenases, there are less common non-classical P450s that are not dependent on electron donors and do not need atmospheric oxygen [4,5]. Instead, these P450s control the conversions of peroxides of hydrophobic compounds. For instance, these are the prostacyclin (PGI2) synthase (CYP8A1) and thromboxane (TXA2) synthase (CYP5A1) of mammals that convert the prostaglandin endoperoxides [6]. Plants possess non-classical P450s of a large and diverse CYP74 family, members of which control the conversions of fatty acid hydroperoxides to bioactive oxylipins [7,8]. The CYP74s are the distant congeners of mammalian PGI2 synthase (CYP8A1). The diversity of CYP74 proteins was previously expanded from family to clan after the detection of related genes in proteobacteria, brown algae, and some non-mammalian metazoans, including lancelets [9]. No CYP74 genes have been detected in mammals yet.
The recent sequencing of the bat Sturnira hondurensis Goodwin, 1940, genome [10] uncovered the unexpected CYP74 gene, possessing almost 90% identity with the putative CYP74C of the poplar Populus trichocarpa. This gene was assigned the name CYP74C44 (this assignment was generously done by Professor David R. Nelson). This work reports 2 of 12 the preparation of the recombinant CYP74C44 of S. hondurensis and its identification as a hydroperoxide lyase (HPL).

Bioinformatics Analyses
CYP74C44 of the neotropical fruit bat S. hondurensis possesses the greatest kinship with plant proteins of the CYP74C subfamily. The BLAST analyses revealed a high extent of CYP74C44 identity to the putative CYP74C protein XP_002305404.3 of the poplar Populus trichocarpa (Figure 1). It also shares about 65.5% identity with the putative CYP74C protein XP_019262969.1 of Nicotiana attenuata and CYP74C4 of Solanum lycopersicum L. In contrast, CYP74C44 shares only low identity with other P450s of S. hondurensis. For example, it has about 16.5% identity with the CYP7B1 (isoform X1, XP_036898685.1) of S. hondurensis.

Bioinformatics Analyses
CYP74C44 of the neotropical fruit bat S. hondurensis possesses the greatest kinship with plant proteins of the CYP74C subfamily. The BLAST analyses revealed a high extent of CYP74C44 identity to the putative CYP74C protein XP_002305404.3 of the poplar Populus trichocarpa (Figure 1). It also shares about 65.5% identity with the putative CYP74C protein XP_019262969.1 of Nicotiana attenuata and CYP74C4 of Solanum lycopersicum L. In contrast, CYP74C44 shares only low identity with other P450s of S. hondurensis. For example, it has about 16.5% identity with the CYP7B1 (isoform X1, XP_036898685.1) of S. hondurensis.
The CYP74C44 gene is located at the genomic locus LOC118984682. A gene at the nearby LOC118984681 locus encodes a vacuolar-processing enzyme-like sequence possessing a high identity with proteins of different Populus species. On the other hand, the adjacent LOC118984680 and LOC118984683 genes encode the elongation factor 1-alpha 1-like protein and adhesion G protein-coupled receptor E2, respectively. The first gene is common for all animal species, while the second one is typical for different bat species. The construction of a phylogenetic tree of selected S. hondurensis P450s ( Figure 2) revealed that CYP74C44 is built into a separate branch ( Figure 2, left side of the tree), including also the putative prostacyclin PGI2 synthase CYP8A1 and several phylogenetically related proteins, namely CYP8B1, CYP7A1, CYP7B1, and CYP39A1. Another enzyme of prostaglandin endoperoxide metabolism, the putative thromboxane A2 synthase, is built into a separate branch (Figure 2, upper side of the tree) together with the CYP4 and CYP11 proteins. The protein BLAST analyses using CYP74C44 as a query vs. the S. hondurensis sequences (the partial multiple alignment of I-helix regions, SRS-4, is presented in Figure 3) showed poor homology. The I-helix groove motif of CYP74C44 possesses some alterations typical for CYP74 enzymes. Firstly, the D/E residue is conserved in monooxygenases such as CYP1A1, CYP2C21, and CYP4V2 ( Figure 3, I-helix groove motif, position 4). In contrast, the Asp (N) residue at this position is conserved in all CYP74s, including CYP74C44. The putative PGI2 synthase (CYP8A1) has a Gly (G) residue at this position and an N residue at the next position (5). The CYP74C44 gene is located at the genomic locus LOC118984682. A gene at the nearby LOC118984681 locus encodes a vacuolar-processing enzyme-like sequence possessing a high identity with proteins of different Populus species. On the other hand, the adjacent LOC118984680 and LOC118984683 genes encode the elongation factor 1-alpha 1-like protein and adhesion G protein-coupled receptor E2, respectively. The first gene is common for all animal species, while the second one is typical for different bat species.
The construction of a phylogenetic tree of selected S. hondurensis P450s ( Figure 2) revealed that CYP74C44 is built into a separate branch ( Figure 2, left side of the tree), including also the putative prostacyclin PGI2 synthase CYP8A1 and several phylogenetically related proteins, namely CYP8B1, CYP7A1, CYP7B1, and CYP39A1. Another enzyme of prostaglandin endoperoxide metabolism, the putative thromboxane A2 synthase, is built into a separate branch ( Figure 2, upper side of the tree) together with the CYP4 and CYP11 proteins. The protein BLAST analyses using CYP74C44 as a query vs. the S. hondurensis sequences (the partial multiple alignment of I-helix regions, SRS-4, is presented in Figure 3) showed poor homology. The I-helix groove motif of CYP74C44 possesses some alterations typical for CYP74 enzymes. Firstly, the D/E residue is conserved in monooxygenases such as CYP1A1, CYP2C21, and CYP4V2 ( Figure 3, I-helix groove motif, position 4). In contrast, the Asp (N) residue at this position is conserved in all CYP74s, including CYP74C44. The putative PGI2 synthase (CYP8A1) has a Gly (G) residue at this position and an N residue at the next position (5).
To test its biochemical behavior, the recombinant CYP74C44 was incubated with fatty acid hydroperoxides. The results are described below (Section 2.3).      To test its biochemical behavior, the recombinant CYP74C44 was incubated with fatty acid hydroperoxides. The results are described below (Section 2.3).

Kinetics and Substrate Specificity of the Recombinant CYP74C44
The CYP74C44 coding sequence consisted of 1435 nucleotides and encoded a 478 amino acid polypeptide. This sequence was custom synthesized and cloned into the vector pET-23a (Novagen, Madison, WI, USA) to yield the target recombinant protein with a His-tag at the C-terminus. The His-tagged recombinant protein was obtained in BL21-CodonPlus-RIL host strain cells (Stratagene, San Diego, CA, USA) and purified by metal affinity chromatography. The enzymatic activity was controlled using ultraviolet spectroscopy by the decrease in fatty acid hydroperoxide absorbance at 234 nm. The pH optimum of the recombinant CYP74C44 was 7.0 ( Figure 4).

Kinetics and Substrate Specificity of the Recombinant CYP74C44
The CYP74C44 coding sequence consisted of 1435 nucleotides and encod amino acid polypeptide. This sequence was custom synthesized and cloned vector pET-23a (Novagen, Madison, WI, USA) to yield the target recombinan with a His-tag at the C-terminus. The His-tagged recombinant protein was ob BL21-CodonPlus-RIL host strain cells (Stratagene, San Diego, CA, USA) and pu metal affinity chromatography. The enzymatic activity was controlled using u spectroscopy by the decrease in fatty acid hydroperoxide absorbance at 234 nm optimum of the recombinant CYP74C44 was 7.0 ( Figure 4).  (Table 1), the affinity of CYP74C44 for 15(S)-HPEPE wa than that for α-linolenate and linoleate hydroperoxides. At the same time, the activity (kcat) of this enzyme was higher towards linoleate hydroperoxides. How CYP74C44 enzyme exhibited the highest catalytic efficiency for C20 hydroperox times higher than C18 hydroperoxides).

Substrate and Product Specificities of the Recombinant CYP74C44
The recombinant CYP74C44 was incubated with the 13-and 9-hydropero linoleic and α-linolenic acids (13-HPOD, 13-HPOT, 9-HPOD, and 9-HPOT, resp as well as the 15-hydroperoxide of eicosapentaenoic acid (15-HPEPE). The (Me/TMS, with or without preliminary NaBH4 reduction) were subjected to analyses. The GC-MS chromatograms of the NaBH4-reduced products (Me/T presented in Figure 5. The structural formulae of the products formed by the nant enzyme are also presented in Figure 5. The kinetic data revealed that the recombinant CYP74C44 efficiently used 9(S)-HPOD, 9(S)-HPOT, 13(S)-HPOD, 13(S)-HPOT, and 15(S)-HPEPE as substrates. According to the K m values (Table 1), the affinity of CYP74C44 for 15(S)-HPEPE was greater than that for α-linolenate and linoleate hydroperoxides. At the same time, the catalytic activity (k cat ) of this enzyme was higher towards linoleate hydroperoxides. However, the CYP74C44 enzyme exhibited the highest catalytic efficiency for C20 hydroperoxides (3-6 times higher than C18 hydroperoxides).
The The last spectrum allowed for ascribing a structure of 14-hydroxytetradecanoic acid (Me/TMS) to the hydrogenation product and a structure of 14-hydroxy-5,8,11-tetradecatrienoic acid (Me/TMS) to compound 5. Overall, the described data indicated that the (all-Z)-14-oxo-5,8,11-tetradecatrienoic acid (5a) was formed as the product of 15-HPEPE conversion. Based on the results obtained, the name ShHPL (S. hondurensis hydroperoxide lyase) has been assigned to CYP74C44 and the name ShHPL to the corresponding gene.
For a long time, hydroperoxide lyase (HPL) has been proposed to control the chain cleavage of fatty acid hydroperoxides [12]. However, more recent studies have revealed that the enzyme is in fact an isomerase that converts the fatty acid hydroperoxides into a shortlived hemiacetal, spontaneously decomposing into two aldehyde fragments [17,28,29]. In view of these findings, the synonymous name "hemiacetal synthase", corresponding to the true catalytic function, was proposed [29]. Aldehydes and aldoacids, the products of hemiacetal decomposition, play a defensive and regulatory role in plants [12,13]. HPL products such as (2E)-hexenal ("leaf aldehyde") may cause gene damage by their addition to deoxyguanosine residues of DNA [30] and exhibit general genotoxicity, cytotoxicity, and antimicrobial and fungicidal activity [12,13]. Interestingly, recent research [31,32] demonstrated the fungicidal effects of exogenous (2E)-hexenal ("leaf aldehyde") against the psychrophilic fungus Pseudogymnoascus destructans, which causes the white-nose syndrome disease in bats. The results of the present work showed the presence of an endogenous HPL, a key enzyme of aldehyde biosynthesis, in S. hondurensis. The antifungal resistance might justify the physiological requirements of the HPL genes in the bats. However, the mechanism of the proposed quite uncommon gene transfer from plant to mammal remains to be revealed. Further genomic sequencing of the bats is needed to shed more light on this phenomenon.
The SciFinder search revealed no mentions of (all-Z)-14-oxo-5,8,11-tetradecatrienoic acid as an HPL product in the literature. This compound has only been described as a product of chemical [33] or combined enzymatic-chemical [34] conversions of arachidonic acid. The loss of information on C 14 aldoacid production by HPLs is not surprising since 15-HPEPE and 15-HPETE are not the physiological substrates for plant HPLs. The occurrence of HPL in the bat creates prerequisites for biosynthetic pathways from eicosapentaenoic or arachidonic acids to (all-Z)-14-oxo-5,8,11-tetradecatrienoic acid (see the mechanistic scheme in Figure 7). Moreover, 15-HPEPE is the preferred substrate for CYP74C44. and AOS [21][22][23][24] members. Thus, the identification of recombinant CYP74C44 as an u specific HPL is not surprising. The majority of the CYP74C proteins studied thus far b long to Solanaceae (Asterids). No CYP74 enzymes have been detected in mammals b fore. Only non-mammalian metazoans such as lancelets [9,25], sea anemones [26,27], a stony corals [9] have been shown to possess the CYP74 clan enzymes. For a long time, hydroperoxide lyase (HPL) has been proposed to control the cha cleavage of fatty acid hydroperoxides [12]. However, more recent studies have reveal that the enzyme is in fact an isomerase that converts the fatty acid hydroperoxides int short-lived hemiacetal, spontaneously decomposing into two aldehyde fragme [17,28,29]. In view of these findings, the synonymous name "hemiacetal synthase", c responding to the true catalytic function, was proposed [29]. Aldehydes and aldoaci the products of hemiacetal decomposition, play a defensive and regulatory role in pla [12,13]. HPL products such as (2E)-hexenal ("leaf aldehyde") may cause gene damage their addition to deoxyguanosine residues of DNA [30] and exhibit general genotoxici cytotoxicity, and antimicrobial and fungicidal activity [12,13]. Interestingly, recent search [31,32] demonstrated the fungicidal effects of exogenous (2E)-hexenal ("leaf dehyde") against the psychrophilic fungus Pseudogymnoascus destructans, which cau the white-nose syndrome disease in bats. The results of the present work showed t presence of an endogenous HPL, a key enzyme of aldehyde biosynthesis, in S. hondur sis. The antifungal resistance might justify the physiological requirements of the H genes in the bats. However, the mechanism of the proposed quite uncommon ge transfer from plant to mammal remains to be revealed. Further genomic sequencing the bats is needed to shed more light on this phenomenon.
The SciFinder search revealed no mentions of (all-Z)-14-oxo-5,8,11-tetradecatrien acid as an HPL product in the literature. This compound has only been described a product of chemical [33] or combined enzymatic-chemical [34] conversions of arac donic acid. The loss of information on C14 aldoacid production by HPLs is not surprisi since 15-HPEPE and 15-HPETE are not the physiological substrates for plant HPLs. T occurrence of HPL in the bat creates prerequisites for biosynthetic pathways fro eicosapentaenoic or arachidonic acids to (all-Z)-14-oxo-5,8,11-tetradecatrienoic acid (s the mechanistic scheme in Figure 7). Moreover, 15-HPEPE is the preferred substrate CYP74C44.  The detection of a typical CYP74C gene in the bat genome appears surprising. The CYP74C44 gene is highly homologous to the CYP74s of poplar and other plant species but has only weak homology to other P450s of S. hondurensis. These facts raise the question of the origin of the CYP74C44 gene in the S. hondurensis genome. Since S. hondurensis is a fruitfeeding bat, one can propose that the CYP74C44 gene could emerge from fruits. However, the BLAST search revealed another CYP74 family gene (CYP74B) in the genome of the distinct bat species Rhinolophus ferrumequinum (information generously communicated by Professor David R. Nelson). This species is insect-feeding. Thus, the food could not serve as a direct source of a "plant gene" in this species. Moreover, the genes located at adjacent loci encode proteins that are typical for animals but not for plants. Further studies of bat genomes are needed to resolve the intriguing question of the origin of these genes.
Concluding remarks: (1) The neotropical fruit bat S. hondurensis has the plant-like protein CYP74C44, possessing ca. 90% sequence identity with the putative CYP74C of Populus trichocarpa. (2) The recombinant CYP74C44 was active towards the 13-and 9-hydroperoxides of linoleic and α-linolenic acids as well as the 15-hydroperoxide of eicosapentaenoic acid; the last one was the preferred substrate. (3) The recombinant CYP74C44 possessed hydroperoxide lyase (HPL) activity towards all tested fatty acid hydroperoxides and converted them into the chain cleavage products, aldehydes and aldoacids. (4) Presumably, CYP74C44 and its products may play a defensive role against fungi infecting the bats.

Bioinformatic Methods
The search for CYP74-related genes was carried out in the NCBI database. Primer construction was performed using the Vector NTI Advance 11.5 program (Invitrogen, Madison, WI, USA). The BLAST analyses of the CYP74s were performed using the protein NCBI BLAST tool. The multiple alignments of selected CYP74 amino acid sequences were made with Clustal Omega and MEGA7 software [37]. The phylogenetic analysis was carried out using the maximum likelihood method based on the Poisson correction model [38], and the bootstrap consensus tree was inferred from 1000 replicates. The iTOL tool (https://itol.embl.de/ (accessed on 1 February 2022)) was used to visualize the phylogenetic model output.

Expression and Purification of Recombinant Enzyme
The target sequence encoding the CYP74C44 sequence was synthesized by the Lumiprobe Company (Saint Petersburg, Russia). The plasmid containing the target se-quence was subsequently cut with NdeI and XhoI and subcloned into the same sites of the expression vector pET-23a (Novagen, Madison, WI, USA) to yield the target recombinant protein with a His-tag at the C-terminus. The resulting construction was transformed into the Escherichia coli host strain BL21-CodonPlus-RIL (Stratagene, San Diego, CA, USA). Using the pET-23a vector allowed the preparation of the target recombinant protein with a His-tag at the C-terminus. The resulting construction was sequenced to verify the presence of the CYP74C44 sequence. The recombinant gene was expressed in host cells as described before [39]. Purification of the His-tagged recombinant protein was performed using a Bio-Scale Mini Profinity IMAC (immobilized metal affinity chromatography) cartridge in the BioLogic LP chromatographic system (Bio-Rad, Hercules, CA, USA) (Supplementary Figure S1). The recombinant enzyme was eluted from the cartridges using 50 mM histidine. The homogeneity of the purified protein was confirmed by SDS-PAGE. The protein concentration was estimated as described before [40].

Kinetic Studies
The enzymatic activity of the purified recombinant CYP74C44 was determined by monitoring the 234 nm signal decrease with a PB 2201 B UV-VIS spectrophotometer (SOLAR, Minsk, Belarus) with substrate concentrations ranging from 5 to 150 µM. The measurements were carried out in 0.6 mL of Na phosphate buffer (pH 7.0) at 25 • C. A molar extinction coefficient of 25,000 M −1 cm −1 at 234 nm for fatty acid hydroperoxides was used. Kinetic parameters were calculated by fitting the datasets to a one-site saturation model for simple ligand binding using the SigmaPlot 11 software (Systat Software Inc., Palo Alto, CA, USA). Five independent experiments were performed for each specified variant.

Incubations of Recombinant Enzyme with Substrates
The recombinant enzyme (10 µg) was incubated with 100 µg of 9-HPOD, 9-HPOT, 13-HPOD, 13-HPOT, or 15-HPEPE for 15 min at 23 • C in 10 mL of Na phosphate buffer (100 mM, pH 7.0). The reaction mixture was acidified to pH 6.0, and the products were extracted with a hexane/ethyl acetate (1:1, by volume) mixture. The products were methylated with ethereal diazomethane and trimethylsilylated with a pyridine/hexamethyldisilazane/ trimethylchlorosilane (2:1:2, by volume) mixture at 23 • C for 15 min. The silylation reagents were evaporated in vacuo. The dry residue was dissolved in 100 µL hexane and subjected to GC-MS analyses. When specified, the products were reduced with NaBH 4 , then methylated and trimethylsilylated. In separate cases, the products (Me esters) were subjected to hydrogenation over PtO 2 , followed by trimethylsilylation. The product derivatives were analyzed as Me esters/TMS derivatives (Me/TMS) by GC-MS.

Methods of Spectral Analyses
The products were analyzed as Me/TMS derivatives by GC-MS as described before [17]. The GC-MS analyses were performed using a Shimadzu QP2020A mass spectrometer connected to a Shimadzu GC-2010 Plus gas chromatograph equipped with a Macherey-Nagel Optima-5-MS (5% phenyl, 95% methylpolysiloxane) fused capillary column (length, 30 m; ID, 0.25 mm; film thickness, 0.25 µm). Helium at a linear velocity of 30 cm/s was used as the carrier gas. Injections were made in the split mode using an initial column temperature of 120 • C and an injector temperature of 230 • C. Then, the column temperature was raised at 10 • C/min until 240 • C. Electron impact ionization (70 eV) was used. Funding: The authors are thankful for the financial support from the government assignment for the FRC Kazan Scientific Center of the Russian Academy of Sciences (bioinformatics and phylogenetic analyses). The preparation of the recombinant enzyme, the studies of its catalytic activities, and the structural and mechanistic studies were carried out with funding from the Russian Science Foundation (grant 20-14-00338).
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.