Two Distinct Enzymes Have Both Phytoene Desaturase and 3,4-Desaturase Activities Involved in Carotenoid Biosynthesis by the Extremely Halophilic Archaeon Haloarcula japonica

Abstract

The extremely halophilic archaeon Haloarcula japonica accumulates the C₅₀ carotenoid, bacterioruberin (BR). To reveal the BR biosynthetic pathway, unidentified phytoene desaturase candidates were functionally characterized in the present study. Two genes encoding the potential phytoene desaturases, c0507 and d1086, were found from the Ha. japonica genome sequence by a homology search using the Basic Local Align Search Tool. Disruption mutants of c0507 and d1086 and their complemented strains transformed with expression plasmids for c0507 and d1086 were subsequently constructed. High-performance liquid chromatography (HPLC) ana­lyses of carotenoids produced by these strains revealed that C0507 and D1086 were both bifunctional enzymes with the same activities as both phytoene desaturase (CrtI) and 3,4-desaturase (CrtD). C0507 and D1086 complemented each other during BR biosynthesis in Ha. japonica. This is the first study to identify two distinct enzymes with both CrtI and CrtD activities in an extremely halophilic archaeon.

Carotenoids, an important class of natural isoprenoid-derived pigments, are synthesized by all photosynthetic organisms and some non-photosynthetic bacteria, archaea, and fungi. They act as light-harvesting pigments in photosynthesis (Holt et al., 2005) and also as antioxidants (Miller et al., 1996; Naguib, 2000) and light-protecting pigments in halophilic archaea (Dundas and Larsen, 1963; Shahmohammadi et al., 1998).

All carotenoids are synthesized from geranylgeranyl pyrophosphate (Takaichi, 2013). In the first step, this compound is formed by geranylgeranyl pyrophosphate synthase (CrtE), which catalyzes the condensation of farnesyl pyrophosphate with an isopentyl pyrophosphate moiety. The second step, catalyzed by phytoene synthase (CrtB), involves the formation of the first C₄₀ carotenoid, phytoene from two molecules of geranylgeranyl diphosphate (GGPP). The colorless compound, phytoene, is converted to lycopene through a series of desaturation reactions catalyzed by a bacterial-type phytoene desaturase (CrtI) in bacteria other than cyanobacteria and green-sulfur bacteria (Takaichi and Mochimaru, 2007). Further downstream modification reactions, including the cyclization of lycopene, the addition of keto and/or hydroxy groups, and the introduction of 5-carbon (C₅) isoprene units, lead to the formation of different carotenoid products. Some carotenoid biosynthetic genes have been characterized in various bacteria and plants (Takaichi, 2013; Sandmann, 2021; Shimada and Takaichi, 2024).

Extremely halophilic archaea forming red-colored colonies produce acyclic C₅₀ bacterioruberin (BR) and its precursors (Dummer et al., 2011; Yang et al., 2015). These C₅₀ carotenoids increase membrane rigidity (Lazrak et al., 1988) and protect cells against UV light (Shahmohammadi et al., 1998). β-Carotene is the precursor of retinal (Peck et al., 2001). Retinal combines with bacterioopsin to generate bacteriorhodopsin, a light-induced proton pump (Oesterhelt and Stoeckenius, 1971).

The extremely halophilic archaeon Halobacterium salinarum produces BR and its precursors, such as iso­pentenyldehydrorhodopin (IDR), bisanhydrobacterioruberin (BABR), and monoanhydrobacterioruberin (MABR), as well as retinal, which constitutes bacteriorhodopsin (Oesterhelt and Stoeckenius, 1971; Dummer et al., 2011). Hb. salinarum synthesizes both BR and retinal from a common intermediate, lycopene. In retinal synthesis, lycopene is converted to β-carotene by lycopene β-cyclase (CrtY) (Peck et al., 2002), and β-carotene is then cleaved to form retinal (C₂₀) by β-carotene cleavage dioxygenase (Brp) (Peck et al., 2001).

Haloarcula japonica is a predominantly triangular disc-shaped extremely halophilic archaeon (Horikoshi et al., 1993). We previously reported that Ha. japonica produces some carotenoid, phytoene, lycopene, and BR and its precursors, similar to other extremely halophilic archaea (Yatsunami et al., 2014). In addition, Ha. japonica produces the retinal protein, cruxrhodopsin (Yatsunami et al., 1999). We demonstrated that three genes (c0507, c0506, and c0505) encode carotenoid 3,4-desaturase (CrtD), bifunctional lycopene elongase and 1,2-hydratase (LyeJ), and the C₅₀ carotenoid 2″,3″-hydratase (CruF), respectively (Fig. 1) (Yang et al., 2015). These three enzymes catalyze the conversion of lycopene to BR in Ha. japonica. Based on our previous study and the proposed biosynthetic pathways of BR in Hb. salinarum, BR and retinal biosynthetic pathways in Ha. japonica are shown in Fig. 2. However, the gene encoding phytoene desaturase in Ha. japonica and other extremely halophilic archaea remains unknown.

Fig. 1.

Organization of gene clusters for BR and retinal biosynthesis in Haloarcula japonica. The BR biosynthesis gene cluster consists of the crtI/crtD(c0507), lyeJ(c0506), and cruF(c0505) genes, whereas the retinal biosynthesis gene cluster consists of the crtI/crtD(d1086) and c1158 genes. Notably, the c1158 gene is homologous to the brp gene of Ha. japonica. Moreover, the c1219 gene, homologous to the crtY gene of Ha. japonica, is located in a distinct genomic region.

Fig. 2.

Main steps in BR and retinal biosynthetic pathways of Haloarcula japonica. Solid arrows indicate the main BR biosynthetic steps. Dashed arrows indicate other possible BR biosynthetic steps. Among the intermediates from phytoene to BR, the locations of structural changes in each step are circled in red. Asterisks indicate unidentified BR and retinal biosynthetic enzymes.

In the present study, we recharacterized C0507 (previously identified as CrtD) and then characterized D1086, a paralog of C0507. C0507 and D1086 were both found to be bifunctional enzymes, operating as both phytoene desaturases (CrtI) and 3,4-desaturases (CrtD) for BR biosynthesis. The present study is the first functional identification of these novel enzymes in archaea.

Materials and Methods

Microbial strains and growth conditions

All strains and plasmids used in this study are listed in Table 1. Escherichia coli strains were cultured in LB medium at 37°C (Sambrook et al., 1989). Ampicillin (50‍ ‍μg mL^–1) was added as required.

Table 1.

Microbial strains and plasmids used in this study

Strain or plasmid	Relevant property	Source or reference
Strains
E. coli JM109	Host used for gene cloning	Laboratory stock
E. coli JM110	Host used for preparing plasmids that are free of Dam and Dcm methylation	Laboratory stock
Ha. japonica	Wild-type strain of Ha. japonica (JCM 7785T)	Horikoshi et al., 1993
Ha. japonica Δd1086	d1086 gene disruptant of Ha. japonica	This work
Ha. japonica Δc0507	c0507 gene disruptant of Ha. japonica	Yang et al., 2015
Ha. japonica Δc0507Δd1086	Double disruptant of both the c0507 and d1086 genes of Ha. japonica	This work
Plasmids
pUC119	E. coli cloning vector; Apr	Laboratory stock
pWL102	E. coli-extremely halophilic archaea shuttle vector; Apr, Mevr	Lam and Doolittle, 1989
pJFZ33	pWL102 derivative carrying the Ha. japonica csg promoter and ftsZ2 structural gene; Apr, Mevr	Ozawa et al., 2005
pJc0507	pJFZ33 derivative in which the ftsZ2 structural gene is replaced by the c0507 structural gene; Apr, Mevr	Yang et al., 2015
pCRTI2	pUC119 derivative screened from the PstI sub-genomic library; Apr	This work
pWL102(Δori)-Δd1086	pWL102 derivative containing a disrupted fragment of the d1086 structural gene; Apr, Mevr	This work
pUC119-d1086	pUC119 derivative containing the d1086 structural gene; Apr	This work
pJd1086	pJFZ33 derivative in which the ftsZ2 structural gene is replaced by the d1086 structural gene; Apr, Mevr	This work

Ap^r, ampicillin resistance; Mev^r, mevinoline resistance.

Wild-type and genetically modified strains of Ha. japonica (JCM 7785^T) were grown in complex medium at 37°C in the dark, as previously described (Yatsunami et al., 2014). Medium was supplemented with 8‍ ‍μg mL^–1 of pravastatin (a gift of Daiichi Sankyo), instead of mevinoline, as required.

Isolation of genomic DNA and total RNA from Ha. japonica

Ha. japonica genomic DNA was isolated as previously described (Onodera et al., 2013). The total RNA of Ha. japonica was extracted using Sepasol RNA I (Nacalai Tesque) according to the manufacturer’s instructions, and was then treated with DNase I (GE Healthcare) to remove any genomic DNA contaminants.

Reverse transcription-polymerase chain reaction (RT-PCR) and PCR

Oligonucleotide primers for RT-PCR and PCR were purchased from Operon Biotechnologies. All primers are listed in Table 2. RT-PCR was performed as described below. Briefly, total RNA (4.5‍ ‍μg) was reverse-transcribed at 42°C for 60‍ ‍min in 50‍ ‍μL of the reaction buffer, which containing 20 pmol of each primer (c0507-A or d1086-F-A), 0.3‍ ‍mM (each) deoxynucleotide triphosphate, 2.5‍ ‍mM manganese (II) acetate, 20‍ ‍U RNase inhibitor (Toyobo), and 100‍ ‍U reverse transcriptase (ReverTra Ace; Toyobo). The cDNA generated was amplified by 40 cycles of PCR. PCR was performed using KOD-Plus DNA polymerase (Toyobo), according to the manufacturer’s instructions, except for the addition of 10% (v/v) dimethyl sulfoxide. The c0507-S/c0507-A and d1086-F-S/d1086-F-A primer sets were used to confirm the transcription of the c0507 and d1086 genes, respectively, in wild-type Ha. japonica.

Table 2.

Sequences of the primers used in this study

Primer	Sequence
c0507-S	5′-GGTTGGCCTCCAGCTCATTG-3′
c0507-A	5′-CTCGTGGTCCGGCAGGAGTTC-3′
d1086-F-S	5′-AACATGAACGTGTTCTACCC-3′
d1086-F-A	5′-AATATCTGCTCGAAGTGGTC-3′
d1086-S3	5′-GATTCCTATGGTGGGTATGCG-3′
d1086-A3	5′-GGGATCCTCATTAGTGGTGGTGGTGGTGGTGGTCCGCTTTCGAGGCCGAGC-3′
d1086-2F-3	5′-GTTCGAATCCTACGAGGATG-3′
d1086-2R-2	5′-GACGACCTCGCGAAGGAGTTC-3′

Plasmids

All plasmids used in this study are also listed in Table 1. pUC119 was used as the cloning vector for E. coli. pWL102 is an E. coli-extremely halophilic archaea shuttle vector (Lam and Doolittle, 1989). pJFZ33 is a recombinant plasmid containing the Ha. japonica cell surface glycoprotein (csg) gene promoter sequence, NdeI restriction site, and Ha. japonica ftsZ2 structural gene sequence, BamHI and NcoI restriction sites were inserted into pWL102 (Ozawa et al., 2005). Ha. japonica contains a large amount of glycoprotein (CSG) on its cell surface, indicating the significance of the csg gene promoter. pJc0507, the expression plasmid for the c0507 gene in Ha. japonica, was derived from pJFZ33 in which the ftsZ2 structural gene was replaced with the c0507 structural gene, as previously described (Yang et al., 2015).

Part of d1086 is upstream of the brp homolog (c1158) in Ha. japonica (Yatsunami et al., 2003). Therefore, we cloned the d1086 gene from the PstI subgenomic library by genome walking using the upstream region of the c1158 gene as a probe, and termed the plasmid containing the entire d1086 structural gene as pCRTI2. The PstI-HincII (containing a part of the d1086 gene from the start codon to the 474^th base) and RsaI-PstI (containing a part of the d1086 gene from the 1,187^th base to the stop codon) fragments of pCRTI2 were ligated to the large EcoRI-HindIII fragment of pWL102 via the linker sequences of pUC119 to construct the recombinant plasmid, pWL102(Δori)-Δd1086. This plasmid included a disrupted d1086 gene and lacked the replication origin necessary for replication in extremely halophilic archaea.

The d1086 structural gene was then amplified from Ha. japonica genomic DNA using the primer set, d1086-S3/d1086-A3. The PCR fragment was ligated to the SmaI site of pUC119, yielding the pUC119-d1086 plasmid. The NdeI-BamHI fragment of pUC119-d1086 was ligated to the NdeI/BamHI site of pJFZ33 to produce the pJd1086 plasmid. This plasmid contained a gene encoding C-terminally His-tagged D1086 under the control of the csg gene promoter and was used as an expression plasmid for d1086 in Ha. japonica.

The pWL102(Δori)-Δd1086, pJc0507, and pJd1086 plasmids were initially passed through E. coli JM110 to avoid restriction barrier formation in extremely halophilic archaea (Ozawa et al., 2005) and were then transformed in Ha. japonica.

Ha. japonica disruptants and their complemented strains

Ha. japonica disruptant strains with the deletion mutations were constructed by homologous recombination as previously described (Yang et al., 2015). The disruptant strain, Δd1086, was constructed by transforming pWL102(Δori)-Δd1086 into the Ha. japonica wild-type strain. We also used the previously established Δc0507 strain (Yang et al., 2015). Additionally, the double-disruptant strain, Δc0507Δd1086, was constructed by transforming pWL102(Δori)-Δd1086 into the Δc0507 strain. The disruption of the d1086 genes in the Δd1086 and Δc0507Δd1086 strains was confirmed by PCR using the d1086-2F-3/d1086-2R-2 primer set. Transformation was performed using a previously described standard method (Yang et al., 2015). The transformants were plated onto agar plates containing pravastatin. Pravastatin-resistant colonies were cultured in liquid medium without pravastatin for 96‍ ‍h and then plated on agar plates without pravastatin to isolate the recombinants in which the target gene was replaced with the corresponding disrupted gene. Gene disruption was confirmed by a PCR ana­lysis.

The Δc0507Δd1086(pJc0507) and Δc0507Δd1086(pJd1086)‍ ‍complemented strains were constructed by transforming pJc0507 and pJd1086, respectively, into the Δc0507Δd1086 strain.

Extraction and high-performance liquid chromatography (HPLC) and mass spectrometric ana­lyses of total carotenoids from Ha. japonica

Wild-type and modified Ha. japonica strains were pre-cultured at 37°C. The pre-inoculum (1‍ ‍mL) was then transferred to a 500-mL Erlenmeyer flask containing 100‍ ‍mL of liquid medium and cultured at 37°C for 240‍ ‍h in the dark. Cells were harvested by centrifugation (4,400×g, 4°C, 20‍ ‍min), disrupted by sonication in acetone/methanol [7:2 (v/v)], and centrifuged again (840×g, 4°C, 5‍ ‍min). Pigments were re-extracted from the pellets. The combined supernatants were evaporated to dryness under a vacuum. Dried carotenoid extracts were used for HPLC and purified carotenoids for mass spectrometric ana­lyses as previously described (Yang et al., 2015).

DNA sequence accession numbers

The DNA sequence data of d1086 and c0507 have been deposited in the DNA Data Bank of Japan, European Molecular Biology Laboratory, and GenBank nucleotide sequence databases. GenBank accession numbers for d1086 and c0507 are LC331100 and LC008542, respectively.

Results

Phytoene desaturase gene candidates in the Ha. japonica genome

By using the Basic Local Align Search Tool, two candidate genes encoding the unidentified phytoene desaturases, c0507 and d1086, were found from the Ha. japonica genome (Nakamura et al., 2011). Although C0507 was previously identified as carotenoid 3,4-desatirase (CrtD) (Yang et al., 2015), the enzyme showed 30% identity and E-values of 6e-70 at the amino acid level with Pantoea ananatis CrtI (PaCrtI), suggesting its role as a bifunctional enzyme of both 3,4-desaturase (CrtD) and phytoene desaturase (CrtI). Furthermore, D1086 exhibited 27% identity and E-values of 9e-52 with PaCrtI and 60% homology with C0507, suggesting its role also as a bifunctional enzyme.

Analysis of Δd1086 and Δc0507 strains

We constructed a d1086 gene disruptant Δd1086 strain. The Δd1086 strain showed the same growth rate as the wild-type strain. Carotenoids of the wild-type and Δd1086 strains were analyzed by HPLC. The wild-type strain exhibited five carotenoid peaks on HPLC for BR (Fig. 3A, peak 1), MABR (peak 2), BABR (peak 3), IDR (peak 4), and lycopene (peak 5) (Table 3). BR was the major carotenoid in the Δd1086 strain. Moreover, its carotenoid composition and HPLC elution profile were similar to those of the wild-type strain. Therefore, the Δd1086 strain, similar to the wild-type strain, exhibited both phytoene desaturase (CrtI) and 3,4-desaturase (CrtD) activities. C0507, which acts as CrtD (Yang et al., 2015), may also function as CrtI. Therefore, we redefined C0507 as a bifunctional phytoene desaturase and 3,4-desaturase, CrtI/CrtD(C0507). However, the specific functions of D1086 remain unclear.

Fig. 3.

Cell pellets of wild-type, disruptant, and complemented strains, and HPLC elution profiles of their carotenoids. (A) Wild-type strain. (B) Δc0507 strain. (C) Δc0507Δd1086 strain. (D) Δc0507Δd1086(pJc0507) strain., (E) Δc0507Δd1086(pJd1086) strain. Peak 1, BR; peak 2, MABR; peak 3, BABR; peak 4, IDR; peak 5, lycopene; peak 1’, TH-BABR; peak 2’, DH-IDR; peak 1’’, phytoene. The absorption spectra of the two peaks indicated by asterisks (* and **) differ from those of phytoene. The eluent was methanol-water [9:1 (v/v)] for the first 10‍ ‍min and was then changed to 100% methanol (1.5‍ ‍mL‍ ‍min^–1). All carotenoids, except for phytoene, were detected at 490‍ ‍nm, whereas phytoene was detected at 284‍ ‍nm.

Table 3.

Characteristics of carotenoids extracted from wild-type, disruptant, and complementation strains

Strain	Peak no.	Retention time (min)	λmax (nm) in HPLC eluent	Number of conjugated double bonds	Carotenoid
Wild-type	1	9.6	469, 492, 525	13	BR
	2	11.8	469, 491, 522	13	MABR
	3	13.8	467, 490, 521	13	BABR
	4	16.0	453, 479, 509	12	IDR
	5	19.4	442, 468, 499	11	Lycopene
Δc0507	1’	14.8	441, 466, 498	11	TH-BABR
	2’	16.7	440, 465, 495	11	DH-IDR
Δc0507Δd1086	1’’	21.3	273, 284, 297	3	Phytoene
Δc0507Δd1086(pJc0507)	1	9.8	461, 490, 520	13	BR
	2	11.9	460, 489, 520	13	MABR
	3	13.9	460, 489, 521	13	BABR
	4	16.2	452, 480, 509	12	IDR
	5	19.6	441, 468, 498	11	Lycopene
Δc0507Δd1086 (pJd1086)	1	9.8	460, 490, 522	13	BR
	2	11.9	462, 490, 523	13	MABR
	3	13.9	460, 489, 521	13	BABR
	4	16.2	452, 480, 511	12	IDR
	5	19.5	440, 468, 497	11	Lycopene

Carotenoids were analyzed using HPLC equipped with a μBondapak C₁₈ column (3.9×300‍ ‍mm; Waters) and eluted with methanol/water [9:1 (v/v)] for the first 10‍ ‍min and then with 100% methanol (1.5‍ ‍mL‍ ‍min^–1). The absorption spectra of carotenoids were recorded with a photodiode-array detector attached to the HPLC apparatus (Fig. 3).

We used the previously constructed c0507 gene disruptant strain of Ha. japonica, the Δc0507 strain (Yang et al., 2015). The Δc0507 strain did not accumulate phytoene, but accumulated tetrahydro-bisanhydrobacterioruberin (TH-BABR) and dihydro-isopentenyldehydrorhodopin (DH-IDR) (Fig. 3B and Table 3), indicating that it exhibited CrtI, but not CrtD activity. Therefore, in the Δc0507 strain, TH-BABR and DH-IDR were not desaturated, leading to their accumulation. Furthermore, unlike C0507, D1086 functioned as CrtI.

Analysis of the Δc0507Δd1086 strain

A double-disruptant of both the c0507 and d1086 genes, the Δc0507Δd1086 strain, was also constructed. The cell pellet of the Δc0507Δd1086 strain was colorless (Fig. 3C). The carotenoids of this strain were analyzed by HPLC, and only one carotenoid was detected (Fig. 3C, peak 1’’). Based on its HPLC retention time, absorption spectrum, and mole­cular mass of 544, the carotenoid was identified as phytoene (Table 3). Therefore, the Δc0507Δd1086 strain did not exhibit CrtI activity, suggesting that C0507 and D1086 cooperatively catalyzed phytoene desaturation as CrtIs.

Complementation study on the Δc0507Δd1086 strain

An in vivo complementation study was performed using the Δc0507Δd1086 strain. The Δc0507Δd1086 strain was independently transformed using c0507 and d1086 gene expression plasmids (pJc0507 and pJd1086, respectively) to yield the Δc0507Δd1086(pJc0507) and Δc0507Δd1086(pJd1086) strains, respectively.

The expression of C0507 in the Δc0507Δd1086 strain complemented BR biosynthesis (Fig. 3D and Table 3). In addition to BR, BABR was detected as the primary product. In the Δc0507 strain, the transcription levels of both the c0506 and c0505 genes were reduced, indicating that the disruption of the c0507 gene affected the expression of the downstream c0506 and c0505 genes (Yang et al., 2015). The Δc0507Δd1086 strain was then constructed by introducing the disrupted d1086 gene using the Δc0507 strain as the host. The transcription levels of the c0506 and c0505 genes were lower in the Δc0507Δd1086 and Δc0507Δd1086(pJc0507) strains than in the wild-type strain, leading to the accumulation of intermediates (MABR, BABR, IDR, and lycopene) and a decrease in the percentage of the final product (BR) in the c0507Δd1086(pJc0507) strain.

The expression of D1086 in the Δc0507Δd1086 strain also complemented BR biosynthesis (Fig. 3E and Table 3), suggesting that D1086 also functioned as CrtD. This result indicates that D1086 participated in the bifunctional conversion desaturation of phytoene to BR, similar to CrtI and CrtD. Therefore, we defined D1086 as a bifunctional phytoene desaturase and 3,4-desaturase, CrtI/CrtD(D1086), similar to C0507. The Δc0507Δd1086(pJd1086) strain synthesized BR; however, the major carotenoid was BABR. Similar to the Δc0507Δd1086(pJc0507) strain, the transcription levels of the c0506 and c0505 genes were low, leading to the accumulation of intermediates (MABR, BABR, IDR, and lycopene) and a decrease in the percentage of the final product (BR) in the c0507Δd1086(pJd1086) strain.

Transcriptional ana­lysis of c0507 and d1086

C0507 and D1086 both act as bifunctional CrtI/CrtDs. As shown above, the Δc0507 strain accumulated TH-BABR and DH-IDR, whereas the Δd1086 strain produced BR similar to the wild-type strain. This discrepancy may be due to differences in the transcription levels of the c0507 and d1086 genes in Ha. japonica. To verify this, the transcription levels of both genes in the wild-type strain were assessed by RT-PCR. The transcription level of the d1086 gene was markedly lower than that of the c0507 gene; however, the d1086 and c0507 genes were both transcribed (Fig. 4). This result indicates that the transcription level of the d1086 gene was extremely low in the Δc0507 strain and also that the amount of D1086 was not sufficient to catalyze the desaturation of phytoene and the formation of double bonds at C-3,4 and C-3′,4′ in lycopene derivatives. Therefore, D1086 only exhibited phytoene desaturase (CrtI) activity in the Δc0507 strain. In complementation experiments, unlike the Δc0507 strain, the Δc0507Δd1086(pJd1086) strain produced BR without the accumulation of intermediates, such as TH-BABR and DH-IDR. The pJc0507 and pJd1086 expression plasmids contained a powerful promoter of the Ha. japonica csg gene. Therefore, the transcription level of the d1086 gene was higher in the Δc0507Δd1086(pJd1086) strain than in the wild-type strain, and the overexpression of D1086 led to BR production in the Δc0507Δd1086(pJd1086) strain

Fig. 4.

Agarose gel electrophoresis of RT-PCR products using c0507-S/c0507-A and d1086-S/d1086-A primers. Lanes 1 and 4: positive control reaction products using Haloarcula japonica genomic DNA as the template; lanes 2 and 5: RT-PCR products with total RNA as the template; lanes 3 and 6: negative control reaction products without reverse transcription.

Discussion

In the present study, we identified two enzymes with bifunctional phytoene desaturase (CrtI) and 3,4-desaturase (CrtD) activities in the extremely halophilic archaeon, Ha. japonica. The first phytoene desaturase (CrtI) was C0507, previously identified as 3,4-desaturase (CrtD) (Yang et al., 2015). The second phytoene desaturase (CrtI) was D1086; similar to C0507, it also functioned as a 3,4-desaturase (CrtD). C0507 and D1086 complemented each other during carotenoid biosynthesis in Ha. japonica. To the best of our knowledge, this study marks the first identification of two functional CrtIs in archaea. Furthermore, functional enzymes exhibiting both phytoene desaturase and 3,4-desaturase activities have not been previously reported, suggesting the novelty of the enzymes identified herein.

Fig. 5 shows the evolutionary phylogenetic tree of the previously identified functional CrtIs and CrtDs as well as‍ ‍Ha. japonica CrtI/CrtD(C0507) and CrtI/CrtD(D1086) reported in this study. The tree includes sequences annotated as CrtI for BR- and/or BR analog-producing microbes. CrtIs and CrtDs are desaturases that share significant homology with each other. A phylogenetic tree ana­lysis revealed that‍ ‍CrtDs originated from CrtIs via gene duplication at an early stage, resulting in the divergence of CrtIs and CrtDs with unique functions. The tree shows that CrtI/CrtD(C0507) and CrtI/CrtD(D1086) were closely related members of the CrtI family. The Ha. japonica crtI/crtD(d1086) gene was located upstream of c1158, which is homologous to the Hb. salinarum brp gene, in opposite directions (Fig. 1). Consequently, CrtI/CrtD(D1086) may be involved in retinal biosynthesis and CrtI/CrtD(C0507) in BR biosynthesis. In BR biosynthesis, not only CrtI activity, but also CrtD activity is essential. Therefore, the ancestral CrtI(C0507) in the Ha. japonica genome may have underwent the expansion of its substrate specificity during evolution, subsequently acquiring CrtD activity and evolving into bifunctional CrtI/CrtD(C0507). As shown in Fig. 5, Ha. japonica was slower to acquire CrtI/CrtD(D1086) than CrtI/CrtD(C0507). Therefore, Ha. japonica acquired the retinal biosynthetic pathway via gene duplication of the already acquired BR biosynthetic genes. Since CrtI(C0507) evolved into bifunctional CrtI/CrtD(C0507), CrtI/CrtD(D1086) inevitably exhibited dual functionality. Hb. salinarum also produces both BR and retinal and possesses two CrtIs [HsCrtI(1) and HsCrtI(2)]. Closely related Holoferax volcanii produces BR, but not retinal, with only one CrtI (HvCrtI). CrtI and CrtD activities are both essential for BR biosynthesis, and CrtIs in these archaea may also be bifunctional enzymes. Microbes possessing bifunctional CrtI(s) are not necessarily limited to extremely halophilic archaea, such as Ha. japonica. Two bacterial CrtIs, RrCrtI from Rubrobacter radiotolerans and CfCrt from Cutobacterium flaccumfaciens, may also act as bifunctional enzymes because R. radiotolerans and C. flaccumfaciens both only possess one copy of CrtI and produce BR.

Fig. 5.

Phylogenetic tree of CrtI and CrtD homologues. The abbreviations of enzymes and the accession numbers of genomes are as follows: SgCrtI, Streptomyces griseus CrtI (AAA91950.1); SsCrtI, Saccharolobus solfataricus P2 CrtI (AAK43016.1); HsCrtI(1) and HsCrtI(2), Halobacterium salinarum NRC-1 CrtIs (AAG19932.1, and DAC78712.1, respectively); HvCrtI, Haloferax volcanii DS2 CrtI (ADE03359.1); MlCrtI, Micrococcus luteus CrtI (QZY83930.1); CgCrtI, Corynebacterium glutamicum CrtI (AAK64299.1); CfCrtI, Cutobacterium flaccumfaciens CrtI (WP_111022696.1); RrCrtI, Rubrobacter radiotolerans CrtI (AHY46015.1); RgCrtI, Rubrivivax gelatinosus CrtI (TCP03938.1); RcCrtI, Rhodobacter capsulatus CrtI (WER10061.1); DrCrtI, Deinococcus radiodurans CrtI (WP_010887507.1); SeCrtI, Sphingomonas elodea ATCC 31461 CrtI (AEP37354.1); PaCrtI, Pantoea ananatis CrtI (PZD69089.1); PhCrtI, Paracoccus haeundaensis CrtI (AAY28420.1); DrCrtD, Deinococcus radiodurans CrtD (AAF11796.1); RcCrtD, Rhodobacter capsulatus (CAA77544.1); RgCrtD, Rubrivivax gelatinosus CrtD (BAL96698.1); TrCrtD, Thiocapsa roseopersicina CrtD (AAP59036.1). RrCrtD, Rhodospirillum rubrum CrtD (AAN75037.1). Proteins annotated as Crts with unknown functions are indicated by asterisks (*). CrtIs from microorganisms synthesizing BR and/or BR analogs are indicated in red. Sequences were aligned using CLUSTALW, and a phylogenetic tree was constructed using the neighbor-joining (NJ) method. Branch length labels are illustrated using bars, and each bootstrap value is annotated with a numerical label.

To examine the functions of CrtI/CrtD(C0507) and CrtI/CrtD(D1086) in more detail, we expressed crtI/crtD(c0507) and crtI/crtD(d1086) in E. coli BL21(DE3) cells containing pAC-EB1 (Furubayashi et al., 2015) to facilitate phytoene production using the T7 promoter. Carotenoids were not produced, except for phytoene, in E. coli; however, the expression of CrtI/CrtD(C0507) and CrtI/CrtD(D1086) was detected by Western blotting of His-tagged CrtI/CrtD(C0507) and CrtI/CrtD(D1086) probed with an anti-His-tag antibody (data not shown). Since enzymes from extremely halophilic archaea generally require high salt concentrations for their activity and stability, E. coli-produced CrtI/CrtD(C0507) and CrtI/CrtD(D1086) in the present study may not function under physiological conditions. Further studies are needed to characterize these enzymes at the mole­cular level.

Citation

Yatsunami, R., Ando, A., Miyoko, N., Yang, Y., Takaichi, S., and Nakamura, S. (2024) Two Distinct Enzymes Have Both Phytoene Desaturase and 3,4-Desaturase Activities Involved in Carotenoid Biosynthesis by the Extremely Halophilic Archaeon Haloarcula japonica. Microbes Environ 39: ME24004.

https://doi.org/10.1264/jsme2.ME24004

Acknowledgements

The authors thank the Open Research Facilities for Life Science and Technology, Tokyo Institute of Technology for the DNA sequence ana­lysis.

This study was partially supported by a JSBBA Female Scientist Grant for Challenging Project, Kobayashi Foundation, and KOSÉ Cosmetology Research Foundation (to R.Y.).

References

•  Dummer,  A.M.,  Bonsall,  J.C.,  Cihla,  J.B.,  Lawry,  S.M.,  Johnson,  G.C., and  Peck,  R.F. (2011) Bacterioopsin-mediated regulation of bacterioruberin biosynthesis in Halobacterium salinarum. J Bacteriol 193: 5658–5667.
•  Dundas,  I.D., and  Larsen,  H. (1963) A study on the killing by light of photosensitized cells of Halobacterium salinarium. Arch Mikrobiol 46: 19–28.
•  Furubayashi,  M.,  Ikezumi,  M.,  Takaichi,  S.,  Maoka,  T.,  Hemmi,  H.,  Ogawa,  T., et al. (2015) A highly selective biosynthetic pathway to non-natural C₅₀ carotenoids assembled from moderately selective enzymes. Nat Commun 6: 7534.
•  Holt,  N.E.,  Zigmantas,  D.,  Valkunas,  L.,  Li,  X.P.,  Niyogi,  K.K., and  Fleming,  G.R. (2005) Carotenoid cation formation and the regulation of photosynthetic light harvesting. Science 307: 433–436.
•  Horikoshi,  K.,  Aono,  R., and  Nakamura,  S. (1993) The triangular halophilic archaebacterium Haloarcula japonica strain TR-1. Experientia 49: 497–502.
•  Lam,  W.L., and  Doolittle,  W.F. (1989) Shuttle vectors for the archaebacterium Halobacterium volcanii. Proc Natl Acad Sci U S A 86: 5478–5482.
•  Lazrak,  T.,  Wolff,  G.,  Albrecht,  A.M.,  Nakatani,  Y.,  Ourisson,  G., and  Kates,  M. (1988) Bacterioruberins reinforce reconstituted Halobacterium lipid membranes. Biochim Biophys Acta Biomembr 939: 160–162.
•  Miller,  N.J.,  Sampson,  J.,  Candeias,  L.P.,  Brameley,  P.M., and  Rice-Evans,  C.A. (1996) Antioxidant activities of carotenes and xanthophylls. FEBS Lett 384: 240–242.
•  Naguib,  Y.M. (2000) Antioxidant activities of astaxanthin and related carotenoids. J Agric Food Chem 48: 1150–1154.
• Nakamura, S., Nakasone, K., and Takashina, T. (2011) Genetics and genomics of triangular disc-shaped halophilic archaeon Haloarcula japonica TR-1. In Extremophiles Handbook, vol 2. Horikoshi, K., Antranikian, G., Bull, A.T., Robb. F.T., and Stetter, K.O. (eds). Heidelberg: Springer, pp. 364–381.
•  Oesterhelt,  D., and  Stoeckenius,  W. (1971) Rhodopsin-like protein from the purple membrane of Halobacterium halobium. Nat New Biol 233: 149–152.
•  Onodera,  M.,  Yatsunami,  R.,  Tsukimura,  W.,  Fukui,  T.,  Nakasone,  K.,  Takashina,  T., et al. (2013) Gene ana­lysis, expression, and characterization of an intracellular amylase from the extremely halophilic archaeon Haloarcula japonica. Biosci Biotechnol Biochem 77: 281–288.
•  Ozawa,  K.,  Harashina,  T.,  Yatsunami,  R., and  Nakamura,  S. (2005) Gene cloning, expression and partial characterization of cell division protein FtsZ1 from extremely halophilic archaeon Haloarcula japonica strain TR-1. Extremophiles 9: 281–288.
•  Peck,  R.F.,  Echavarri-Erasun,  C.,  Johnson,  E.A.,  Ng,  W.V.,  Kennedy,  S.P.,  Hood,  L., et al. (2001) Brp and blh are required for synthesis of the retinal cofactor of bacteriorhodopsin in Halobacterium salinarum. J Biol Chem 276: 5739–5744.
•  Peck,  R.F.,  Johonson,  E.A., and  Krebs,  M.P. (2002) Identification of a lycopene β-cyclase required for bacteriorhodopsin biogenesis in the archaeon Halobacterium salinarum. J Bacteriol 184: 2889–2897.
• Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Springer Harbor Laboratory Press.
•  Sandmann,  G. (2021) Diversity and origin of carotenoid biosynthesis: its history of coevolution towards plant photosynthesis. New Phytol 232: 479–493.
•  Shahmohammadi,  H.R.,  Asgarani,  E.,  Terato,  H.,  Saito,  T.,  Ohyama,  Y.,  Gekko,  K., et al. (1998) Protective roles of bacterioruberin and intracellular KCl in the resistance of Halobacterium salinarium against DNA damaging agents. J Radiat Res 39: 251–262.
• Shimada, K., and Takaichi, S. (eds). (2024) Anoxygenic Phototrophic Bacteria. Cambridge, MA: Academic Press.
•  Takaichi,  S., and  Mochimaru,  M. (2007) Carotenoids and carotenogenesis in cyanobacteria: Unique ketocarotenids and carotenoid glycosides. Cell Mol Life Sci 64: 2607–2619.
• Takaichi, S. (2013) Tetraterpenes: Carotenoids. In Natural Products: Phytochemistry, Botany and Metabolism of Alkaloids, Phenolics and Terpenes. Ramawat, K.G., and Mérillon, J.-M. (eds). New York, NY: Springer, pp. 3251–3283.
•  Yang,  Y.,  Yatsunami,  R.,  Ando,  A.,  Miyoko,  N.,  Fukui,  T.,  Takaichi,  S., et al. (2015) Complete biosynthetic pathway of the C₅₀ carotenoid bacterioruberin from lycopene in the extremely halophilic archaeon Haloarcula japonica. J Bacteriol 197: 1614–1623.
•  Yatsunami,  R.,  Kawakami,  T.,  Ohtani,  H., and  Nakamura,  S. (1999) Transcriptional regulation of cruxrhodopsin gene from extremely halophilic archaeon Haloarcula japonica strain TR-1. Nucleic Acids Symp Ser 42: 73–74.
•  Yatsunami,  R.,  Iwamoto,  M.,  Ito,  S., and  Nakamura,  S. (2003) A brp homolog of Haloarcula japonica: gene cloning and transcriptional ana­lysis. Nucleic Acids Res Suppl 3: 267–268.
•  Yatsunami,  R.,  Ando,  A.,  Yang,  Y.,  Takaichi,  S.,  Kohno,  M.,  Matsumura,  Y., et al. (2014) Identification of carotenoids from the extremely halophilic archaeon Haloarcula japonica. Front Microbiol 5: 100–105.