Báo cáo khoa học: Cooperation of two carotene desaturases in the production of lycopene in Myxococcus xanthus pot

Cooperation of two carotene desaturases in the
production of lycopene in Myxococcus xanthus
Antonio A. Iniesta
1,2
, Marı
´a
Cervantes
1
and Francisco J. Murillo
1
1 Departamento de Gene
´
tica y Microbiologı
´
a, Facultad de Biologı
´
a, Universidad de Murcia, Spain
2 Department of Developmental Biology, Beckman Center, Stanford University School of Medicine, CA, USA
Carotenoids constitute one of the most widely distri-
buted and structurally diverse classes of natural pig-
ments, with important functions in photosynthesis,
nutrition, and protection against photooxidative dam-
age. Carotenoids are ubiquitously found in bacteria,
fungi, algae, and plants. Even though the end-products
of carotenoid biosynthesis are extremely diverse, a gen-
eral common pathway leading to the formation of
lycopene (red carotene), and cyclic b-carotene (yellow)
is observed in many organisms. However, the nature
of the involved enzymes varies among different organ-
isms [1]. Precursors for the synthesis of carotenoids are
derived from the general isoprenoid biosynthetic path-

way (along with a variety of other important natural
substances) [2], and start with the precursor farnesyl
diphosphate (Fig. 1). The condensation of two geranyl-
geranyl diphosphate (GGPP) molecules produces phy-
toene, mostly in the cis conformation. Generally,
phytoene is dehydrogenated in four desaturation
events, producing phytofluene, f-carotene, neurospo-
rene, and lycopene, respectively, in that order (Fig. 1).
On the basis of sequence homology, there are two
unrelated groups of phytoene desaturases, CrtI-like
and Pds-like. Noncyanobacterial bacteria and fungi
use a single CrtI-type phytoene desaturase to carry out
the four dehydrogenation steps producing lycopene.
The Pds-type phytoene desaturase is found in plants,
algae, and cyanobacteria, where it is known as CrtP.
Both Pds and CrtP converts phytoene into f-carotene
in two steps. The two remaining desaturation events
Keywords
carotenes; CrtI; dehydrogenation;
isomerization; phytoene
Correspondence
A. A. Iniesta, Beckman Center B355, 279
Campus Drive, Stanford, CA 94305, USA
Fax: +1 650 725 7739
Tel: +1 650 723 5685
E-mail:
F. J. Murillo, Facultad de Biologı
´
a,
Universidad de Murcia, Campus de

Espinardo, Murcia 30071, Spain
Fax: +34 957 355 039
Tel: +34 957 355 024
E-mail:
(Received 10 May 2007, revised 26 June
2007, accepted 28 June 2007)
doi:10.1111/j.1742-4658.2007.05960.x
In Myxococcus xanthus, all known carotenogenic genes are grouped
together in the gene cluster carB–carA, except for one, crtIb (previously
named carC). We show here that the first three genes of the carB operon,
crtE, crtIa, and crtB, encode a geranygeranyl synthase, a phytoene desatur-
ase, and a phytoene synthase, respectively. We demonstrate also that CrtIa
possesses cis-to-trans isomerase activity, and is able to dehydrogenate
phytoene, producing phytofluene and f-carotene. Unlike the majority of
CrtI-type phytoene desaturases, CrtIa is unable to perform the four dehy-
drogenation events involved in converting phytoene to lycopene. CrtIb, on
the other hand, is incapable of dehydrogenating phytoene and lacks cis-to-
trans isomerase activity. However, the presence of both CrtIa and CrtIb
allows the completion of the four desaturation steps that convert phytoene
to lycopene. Therefore, we report a unique mechanism where two distinct
CrtI-type desaturases cooperate to carry out the four desaturation steps
required for lycopene formation. In addition, we show that there is a
difference in substrate recognition between the two desaturases; CrtIa
dehydrogenates carotenes in the cis conformation, whereas CrtIb dehydro-
genates carotenes in the trans conformation.
Abbreviations
CTT, casitone ⁄ Tris; GGPP, geranylgeranyl diphosphate.
4306 FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS
are carried out by other desaturases, Zds in plants and
algae, and CrtQ in cyanobacteria, which are related to

the Pds-type desaturases. In organisms with a CrtI-
type desaturase, the cis-to-trans isomerization convert-
ing cis-phytoene into trans-phytoene is also performed
by CrtI. However, the Pds-type desaturases lack iso-
merase activity, and a CrtI-like enzyme (CrtISO in
plants and CrtH in cyanobacteria) is used for cis-to-
trans isomerization, converting cis-lycopene into trans-
lycopene [1]. A few exceptions to these generalizations
about the dehydrogenation or the isomerization pro-
cess have been reported [3–7].
Myxococcus xanthus is a Gram-negative bacterium
that produces carotenoids in response to blue light
[8,9] and the presence of copper [10]. M. xanthus accu-
mulates mainly esterified carotenoids such as ester of
myxobacton (final carotenoids), and all-trans-phytoene
(Table 1) [11,12]. In M. xanthus, all known and pre-
dicted carotenoid biosynthesis genes are grouped
together in the carB–carA gene cluster [13], except for
carC (hereafter renamed crtIb) (Fig. 2) [14]. The first
six ORFs of carB–carA are located in the carB operon,
and the rest are at the carA locus. We characterize
here the first three genes from the M. xanthus carB
operon, and show that their products, CrtE, CrtIa,
and CrtB, possess GGPP synthase, phytoene desatur-
ase and phytoene synthase activity, respectively. In
addition, we show that M. xanthus uses two desaturas-
es, CrtIa and CrtIb, to complete the four desaturation
processes required to transform phytoene into lyco-
pene. This is the first report of such unusual and
unique collaboration between two CrtI-like desaturas-

es, providing additional evidence for the wide plasticity
of carotenoid biosynthesis. Finally, we also show here
that CrtIa possesses cis-to-trans isomerase activity, and
recognizes substrates in the cis conformation, whereas
CrtIb has similar desaturase activity but recognizes
substrates in the trans conformation.
Results
Phytoene isomerization
In M. xanthus, a mutant with a transposon insertion
in the coding region of crtIb accumulates all-trans-phy-
toene (93%) and phytofluene (7%) [11]. Therefore,
CrtIb is required for phytoene dehydrogenation steps
producing lycopene, but is dispensable for the 15-cis-
phytoene to all-trans phytoene isomerization. This sug-
gests the existence of a second enzyme to carry out the
phytoene isomerization and, possibly, the first phyto-
ene dehydrogenation step leading to phytofluene. The
product encoded by crtIa showed high similarity to the
CrtI-type phytoene dehydrogenase from fungi and
noncyanobacterial bacteria [13], including the previ-
ously described CrtIb of M. xanthus [14]. An M. xan-
thus mutant with a transposon insertion in crtIa is
unable to produce carotenoids, indicating an early role
in carotenogenesis. The crtIa insertion could have a
polar effect on the expression of downstream genes
[12]. To clarify the possible function of crtIa in the
carotenoid synthesis pathway, we generated an
Fig. 1. Schematic of the initial carotenoid biosynthesis pathway in
M. xanthus. The addition of an isopentenyl diphosphate unit (IPP)
to farnesyl diphosphate generates a GGPP molecule. The conden-

sation of two GGPPs results in the synthesis of 15-cis-phytoene.
After its isomerization to the all-trans conformation, the phytoene
undergoes four dehydrogenation steps, producing phytofluene,
f-carotene, neurosporene, and lycopene, respectively. Dashed
circles represent the site where the desaturation events take place.
A. A. Iniesta et al. Phytoene dehydrogenation to lycopene in M. xanthus
FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS 4307
in-frame deletion of crtIa in M. xanthus strain MR151.
This strain contains a mutation, carR3, which renders
the expression of the carB operon independent of
external stimulation [10,15]. Colonies of the MR151-
derived strain with the crtIa deletion (MR841) com-
pletely lose the red color typical of colonies from the
parental strain, suggesting that the absence of CrtIa
blocks the synthesis of colored carotenoids. The possi-
ble accumulation of carotenoid precursors by MR841
was analyzed by carotene extraction and chromatogra-
phy on an alumina column. Only the presence of phy-
toene was detected (Fig. 3A and Table 1). The pattern
of absorbance of this carotene is similar to that shown
by 15-cis-phytoene and different from that shown by
all-trans-phytoene (Fig. 3B). Therefore, CrtIa is
required, at least, for the isomerization of 15-cis-phyto-
ene to all-trans-phytoene.
Two CrtI-type enzymes cooperate in phytoene
dehydrogenization to lycopene
In M. xanthus, a transposon insertion in crtE or in
crtB of the carB operon prevents the accumulation of
Table 1. Carotenoid content of several strains of M. xanthus and E. coli with plasmids bearing different carotenogenic genes. Mx,
M. xanthus; Ec, E. coli; ND, not detected.

Host Strain
Genotype (Mx)
Plasmids (Ec)
Carotenoid content (lgÆg
)1
of protein)
a
Phytoene Phytofluene f-carotene Neurosporene Lycopene
Final carotenoids,
esterified
carotenoids Ref.
Mx DK1050
b
Wild-type 1260 100 40 40 90 1470 [11]
Mx MR151 carR3 1260 170 30 100 240 3800 [11]
Mx MR841 carR3, DcrtIa 9500 ND ND ND ND ND This study
Mx MR728 carR3, DcrtIb 8000 250 5 ND ND ND This study
Ec FD6 pFD6 800 260 40 ND ND ND This study
Ec FD9 pFD9 ND ND ND ND 640 ND This study
Ec FD100 pACCRT-EBP ND ND 5 ND ND ND This study
Ec FD101 pACCRT-EBP pMAR183 ND 6 16 80 40 ND This study
Ec FD102 pACCRT-EBP pFD39 ND ND 3 3 2 ND This study
Ec MR2301 pACCRT-EBI ND ND ND 120 ND ND This study
Ec FD103 pACCRT-EBI pMAR183 ND ND ND 250 ND ND This study
Ec FD104 pACCRT-EBI pFD39 ND ND ND ND 130 ND This study
a
The average of three or more independent determinations is given.
b
In the presence of light.
Fig. 2. carB–carA gene cluster and crtIb. The carB operon is transcribed by the light-inducible promoter P

B
and contains the first six genes of the
carB–carA cluster. The carA operon includes the last five genes of the cluster, and its transcription is driven by a light-independent promoter, P
A
.
Fig. 3. The absence of CrtIa blocks the cis-to-trans isomerization
of phytoene. (A) Absorption spectra in hexane of carotenoids
extracted from strain MR841 (DcrtIa and carR3), showing only the
presence of cis-phytoene. (B) Absorption spectra in hexane of all-
trans-phytoene produced by M. xanthus (continuous line) and of
15-cis-phytoene produced by the fungus P. blakesleeanus (dashed
line), taken from Martinez-Laborda et al. [11]. All-trans-phytoene
presents three well-defined peaks (276 nm, 286 nm, and 297 nm),
unlike the spectrum of 15-cis-phytoene, which shows a maximum
at 286 nm and two inflections at 276 nm and 297 nm [46].
Phytoene dehydrogenation to lycopene in M. xanthus A. A. Iniesta et al.
4308 FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS
carotenoids or their C
40
precursors [11,12]. The amino
acid sequences of CrtE and CrtB showed significant
similarities, respectively, to the GGPP and phytoene
synthase from bacteria, fungi, and plants [13]. To
confirm the predicted enzyme activities of CrtE and
CrtB, the corresponding M. xanthus genes were both
expressed in Escherichia coli, which lacks carotenogenic
genes. E. coli was transformed with a plasmid
(pFD6) harboring crtE, crtIa and crtB from the carB
operon, to generate strain FD6. In pFD6, the expres-
sion of the crt genes was under the control of the arti-

ficial constitutive promoter Part-1-2. Strain FD6 was
grown in LB medium to stationary phase, and carote-
noids were purified from the cell extract and analyzed.
The absorption spectra of the extract showed the pres-
ence of all-trans-phytoene, phytofluene and f-carotene,
at decreasing concentrations (Table 1). Thus, CrtE,
CrtB and CrtIa are sufficient to carry out the synthesis
of phytoene, its isomerization, and the first two phyto-
ene dehydrogenation steps up to f-carotene produc-
tion. Similar results were obtained when crtIa was
coexpressed in E. coli with the crtE and crtB genes
from Rhodobacter (data not shown). This confirms that
CrtE and CrtB have GGPP and phytoene synthase
functions, respectively, and that CrtIa, besides its
isomerase activity, is responsible for the double dehy-
drogenation of phytoene up to f-carotene. CrtIa, how-
ever, seems unable to drive the rest of the desaturation
events that produce neurosporene and lycopene.
As mentioned above, crtIb is somehow required for
the dehydrogenation steps converting phytoene to lyco-
pene. However, the expression of crtIb in the E. coli
strain producing cis-phytoene, using the crtE and crtB
genes from Rhodobacter, did not transform the initial
cis-phytoene at all (data not shown). In order to deter-
mine the specific function of CrtIb, we analyzed the
carotenoids accumulated by an M. xanthus crtIb dele-
tion mutant, which also carries the carR3 mutation
(MR728) [16]. MR728 was shown to accumulate all-
trans-phytoene, phytofluene, and f-carotene, in decreas-
ing concentrations (Table 1). This pattern of carotene

accumulation, notably similar to that resulting from the
heterologous expression of crtIa in E. coli (strain FD6
in Table 1), indicates that CrtIb is acting at one or two
of the last dehydrogenation steps in lycopene produc-
tion, after the CrtIa isomerase and desaturase activities.
The low ratio of f-carotene to phytofluene found in
both the M. xanthus crtIb deletion mutant and the
E. coli strain expressing crtIa indicates a low efficiency
of desaturation by CrtIa in the absence of CrtIb. A
high relative accumulation of all-trans-phytoene is not
unusual, as it is also seen in an extract from an M. xan-
thus wild-type strain, which, however, produces only
traces of partially desaturated phytoene products
(Table 1). Altogether, the accumulated evidence sug-
gests novel cooperation between two CrtI-type desatu-
rases in the dehydrogenation of phytoene to lycopene.
In order to determine whether both CrtIa and CrtIb
are necessary and sufficient for the complete dehydro-
genation to lycopene, we generated E. coli strain FD9.
This strain contains plasmid pFD9, which bears the
M. xanthus genes crtE , crtIa, crtB and crtIb under the
control of the Part-1-2 promoter. Colonies from strain
FD9 developed a very strong red color on LB agar
plates. Carotenoid analysis showed that FD9 accumu-
lates high amounts of lycopene almost exclusively
(Table 1). This is also seen in an M. xanthus mutant
defective in lycopene cyclization, where lycopene is
more abundant than other desaturated precursors [12].
Therefore, in M. xanthus, CrtIa and CrtIb are both
required to complete the four phytoene dehydrogeniza-

tion steps necessary for lycopene formation. Moreover,
the exclusive accumulation of lycopene suggests some
kind of cooperation between both CrtI-type dehydro-
genases for the efficient processing of the partially
dehydrogenated intermediate substrates.
Different isomeric substrates for each
dehydrogenase
The requirement in M. xanthus for two CrtI-type pro-
teins to carry out the four dehydrogenation steps
involved in converting phytoene to lycopene is atypical
in the biogenesis of carotenoids. The precise dehydro-
genase activity of CrtIa and CrtIb could be due to dif-
ferential substrate recognition, based on the substrate
desaturation state, on its isomer conformation, or
both. To discriminate between these possibilities, we
expressed crtIa or crtIb in E. coli containing plasmid
pACCRT-EBP. This plasmid harbors the crtE and
crtB genes from Erwinia uredovora and the gene crtP
from the cyanobacterium Synechococcus PCC7942.
Carotene analysis of extract from E. coli with
pACCRT-EBP (FD100) identified cis-f-carotene as a
major carotene, and trace amounts of all-trans-f-caro-
tene and cis-phytoene [5]. The expression of crtIa in
this strain resulted in the accumulation of the four de-
hydrogenated phytoene derivatives phytofluene, f-caro-
tene, neurosporene, and lycopene (FD101 in Table 1).
On the other hand, the same strain expressing crtIb
produced f-carotene, neurosporene and lycopene, all in
low amounts, with no phytofluene being detected
(FD102 in Table 1). These data seem to indicate that

CrtIa is able to carry out two consecutive dehydro-
genation events on carotenes in the cis conformation:
the conversion of cis-phytoene into phytofluene and
A. A. Iniesta et al. Phytoene dehydrogenation to lycopene in M. xanthus
FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS 4309
f-carotene, and that of cis-f-carotene into neurospo-
rene and lycopene. However, CrtIb appears to only
transform the small amounts of trans-f-carotene into
neurosporene and lycopene.
To further confirm the hypothesis of the substrate
isomer specificity of CrtIa and CrtIb, we expressed
crtIa or crtIb in an E. coli strain producing trans-neu-
rosporene [5]. This strain (MR2301) contains plasmid
pACCRT-EBI bearing the crtE and crtB genes from
Er. uredovora and crtI from Rhodobacter capsulatus,
which encodes a dehydrogenase that is capable of
transforming cis-phytoene into trans-neurosporene but
is unable to catalyze efficiently the fourth desaturation
step to produce lycopene. The expression of crtIa
in the trans-neurosporene-producing strain did not
change the dehydrogenation state of the accumulated
carotenes (FD103 in Table 1). However, the expression
of crtIb caused the almost complete transformation of
trans-neurosporene into lycopene (FD104 in Table 1).
All of these results suggest a scenario where CrtIa and
CrtIb specific substrate recognition properties depend
on the cis–trans conformation of the substrate, rather
than on its desaturation state.
Discussion
A variety of isoprenoid compounds, such as choles-

terol, dolichol, ubiquinone, coenzyme Q, isoprenoid
quinines, sugar carrier lipids, and carotenoids, are syn-
thesized by polyprenyl synthases in eukaryotic and
prokaryotic organisms. Two distinct types of evolu-
tionarily conserved prenyltransferases, CrtE and CrtB,
catalyze the early reactions of carotenoid biosynthesis
from farnesyl diphosphate to phytoene [2]. As pre-
dicted from sequence alignments [13], we report here
that the crtE and crtB genes from the M. xanthus carB
operon encode enzymes with GGPP and phytoene syn-
thase activity, respectively. After phytoene synthesis,
this carotene undergoes several desaturation events
(Fig. 4). In M. xanthus, an enzyme similar in sequence
to the CrtI-type phytoene dehydrogenases, previously
called CarC [11,14] and referred to here as CrtIb, was
shown to be involved in carotenoid biosynthesis. The
crtIb gene is not linked to the carotenogenic carB
operon, which contains a gene predicted to encode a
second phytoene dehydrogenase [13], referred here as
CrtIa. Interestingly, CrtIa is unable to catalyze the
four desaturations necessary for lycopene production
in the absence of CrtIb, and instead it leads to the
accumulation of the intermediates phytofluene and
f-carotene in decreasing amounts. On the other hand,
CrtIb is itself incapable of introducing any double
bonds into phytoene. We have demonstrated that a
unique collaboration between CrtIa and CrtIb is used
to successfully introduce four double bonds into phy-
toene. To our knowledge, this is the first case reported
where two CrtI-type desaturases function together to

generate lycopene.
Although changes of CrtI-type enzymes producing
loss or gain of dehydrogenation activities are not
frequent, some cases have been reported. In the bacteria
R. capsulatus and R. sphaeroides, CrtI introduces three
double bonds into phytoene, producing neurosporene,
but lacks the capacity to introduce the fourth double
bond needed to produce lycopene [17,18]. A fifth
dehydrogenation step is carried out by a CrtI-type
desaturase from the fungus Neurospora crassa, produc-
ing 3,4-dihydrolycopene [19]. Other variations in the
lycopene biosynthesis pahtway have been reported
in some cyanobacteria species. The cyanobacteria
Anabaena PCC7120 converts the f-carotene produced
by CrtP into lycopene using a CrtI-type desaturase
instead of the typical cyanobacterial CrtQ [5]. Like
M. xanthus CrtIb, the Anabaena CrtI-type desaturase is
unable to introduce any double bonds into phytoene. In
the cyanobacterium Gloebacter, a single CrtI-type desat-
urase is responsible for the four dehydrogenation steps
required for lycopene formation, and homologs of crtP
and crtQ are not found in its genome [6,7]. It has been
proposed that in the course of evolution, cyanobacteria
acquired a gene encoding an unrelated CrtI-type desat-
urase, which was duplicated, resulting in crtP and crtQ.
These two genes would be the ancestors of pds and zds
from algae and plants [1]. The lack of CrtP and CrtQ in
Gloebacter may be related to its evolutionary distance
from other groups of cyanobacteria [20]. This organism
Fig. 4. General representation of the pathway for synthesis of

trans-lycopene from cis-phytoene and the enzymes involved in dif-
ferent organisms, including M. xanthus.
Phytoene dehydrogenation to lycopene in M. xanthus A. A. Iniesta et al.
4310 FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS
is thought to retain traces of the ancestral properties of
cyanobacteria [4].
The formation of b-carotene is one of the most com-
mon steps in the synthesis of carotenoids. It requires the
cyclization of lycopene to ionone end-groups. Lycopene
in the cis conformation cannot be cyclized, due to its
steric arrangement, and therefore it must be synthesized
in the all-trans configuration, or be converted to that
form [1,21]. In organisms that use Pds ⁄ CrtP-type and
Zds ⁄ CrtQ-type desaturases, where the dehydrogenation
steps are performed on carotenes in the cis confor-
mation, the final isomerization from cis-lycopene to
all-trans-lycopene is performed by a CrtI-type enzyme
called CrtISO in algae and plants, and CrtH in cyano-
bacteria [21–24] (Fig. 4). cis-to-trans isomerization can
be also enhanced by light [5,25,26]. However, in all
organisms that use a CrtI-type desturase, the cis-to-
trans isomerization is associated with the desaturation
processes producing trans-phytoene [6,27,28] (Fig. 4).
In the case of Anabaena, the cis-to-trans isomerization
is carried out on f-carotene, instead of on phytoene, by
its CrtI-type f-carotene dehydrogenase [5]. It is not clear
why two different biosynthetic pathways for lycopene
exist in nature. The discovery of the biosynthetic
enzymes CrtP, CrtQ and CrtH in the green sulfur bacte-
rium Clorobium tepidum, an obligate photoautotroph,

suggests that these enzymes originated from a common
ancestor of modern-day green sulfur bacteria and
cyanobacteria [3]. The fact that organisms with
CrtP ⁄ CrtQ ⁄ CrtH also contain type I photosynthetic
reaction centers, and that cis carotenoids appear to
perform important functions in these reaction centers,
suggests a link between photosynthesis and the presence
of cis carotenoids [3,29]. We show here that in
M. xanthus, the cis-to-trans isomerization is catalyzed
by CrtIa. CrtIa recognizes phytoene and also f-carotene
in the cis conformation. In addition, CrtIa is unable
to dehydrogenate trans-neurosporene, indicating its
preference for carotenes in the cis conformation.
However, CrtIb seems unable to recognize substrates
in the cis conformation, but can transform trans-
neurosporene into lycopene. Therefore, we propose
that, in M. xanthus, the whole biosynthetic process
from cis-phytoene to trans-lycopene is carried out by
the cooperation of two CrtI-type desaturases, CrtIa and
CrtIb. They may form a protein complex, where CrtIa
recognizes cis-phytoene, isomerizes it to trans-phytoene,
and dehydrogenates it twice to produce trans- f-caro-
tene. This last carotene would be then transferred
directly from CrtIa to CrtIb, and this desaturase would
introduce two new double bonds, forming trans-
lycopene. The high water insolubility of carotenoids,
and the improvement of the desaturase efficiency of
CrtIa in the presence of CrtIb, suggest a mechanism for
the direct transfer of substrates from CrtIa to CrtIb.
In the absence of CrtIb, CrtIa would be blocked when

bound to f-carotene. The idea of a linear assembly
chain for carotenoid synthesis was proposed years ago,
on the basis of work with the fungus Phycomyces
blakesleeanus [30,31].
Why M. xanthus uses two CrtI-type desaturases for
the dehydrogenation of phytoene to lycopene is cer-
tainly an unanswered question. One possibility is that
having two unlinked desaturase genes provides more
regulatory options. The crtIa gene is inserted in the
carB operon, which is driven by a light-activated pro-
moter [15]. The expression of crtIb is also activated by
light, but through a tight mechanism that operates
only when the cells have reached the stationary phase
or are starved of a carbon source [14]. This may have
advantages if carotenoids are synthesized only when
needed, in stationary phase but not before, leaving the
isoprenoid components for metabolic uses in the
growth phase. In the presence of light, the carotenoid
biosynthetic machinery would be present but blocked,
waiting for the last enzymatic element, CtrIb, which
reaches a very high level soon after the cell’s entrance
into the stationary phase [14]. This scenario would be
possible if, in the course of the evolution, CrtIa lost its
capacity to perform the two final desaturation steps of
the four catalyzed by a typical CrtI-type enzyme, and
these activities were taken over by a second desaturase,
CrtIb. An obvious idea is that the crtIb gene arose by
duplication of the original, single M. xanthus crtI gene.
However, the CrtIb protein is more closely related
(46% identity) to the f-carotene desaturase from

Anabaena PCC7120 [5]. Therefore, a possible horizon-
tal gene transfer event from a cyanobacterium to
myxobacteria cannot be ruled out.
Experimental procedures
Bacterial strains, media, and transducing phages
E. coli DH5a [32] was used for cloning and carotenoid pro-
duction, and E. coli MC1061 [33] for transducing plasmids
into M. xanthus with the coliphage P1clr100Cm (hereafter
called P1) [34]. M. xanthus strains, derived from MR151
[35], were grown in casitone ⁄ Tris (CTT) rich medium [36],
and E. coli was grown in LB rich medium [37]. When nec-
essary, 40 lgÆmL
)1
kanamycin was added to CTT medium.
Plasmid and strain construction
Standard protocols were followed for DNA manipulation
[37]. PCR-derived clones were generated using Pfu DNA
A. A. Iniesta et al. Phytoene dehydrogenation to lycopene in M. xanthus
FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS 4311
polymerase, and sequenced to verify the absence of PCR-
generated mutations.
To generate an M. xanthus crtIa deletion mutant, we
digested plasmid pMAR161 with MluI, and the biggest
fragment was autoligated, resulting in plasmid pMAR162.
Plasmid pMAR161 is a pUC9 vector [38] that contains a
4.3 kb fragment including the first three genes of the carB
operon (Fig. 2) and a part of the fourth gene. Plasmid
pMAR162 is similar to pMAR161 but with deletion of a
fragment encoding 160 amino acids of the coding region of
crtIa. The M. xanthus DNA fragment from pMAR162 was

cloned into plasmid pDAH160 [39], which carries a kana-
mycin resistance gene and the incompatibility region of P1
for transferring the plasmid from E. coli to M. xanthus
by P1-specialized transduction. The resulting plasmid,
pMAR164, was transduced into M. xanthus MR151, where
it integrated by homologous recombination to generate a
kanamycin-resistant merodiploid. We grew this merodiploid
in CTT without kanamycin to allow a second recombina-
tion event that causes the loss of the kanamycin resistance
marker, generating kanamycin-sensitive colonies, either with
a wild-type crtIa or with the crtIa deletion. The presence of
this deletion was confirmed by Southern blot analysis using
as a template a 4.3 kb RcaI-digested fragment from
pMAR161, and M. xanthus genomic DNA digested with
NcoI. The strain with the crtIa deletion was named
MR841.
To make plasmid pFD3, a DNA fragment bearing crtE,
crtIa, and crtB, which also includes the ribosomal-binding
site upstream of crtE, was PCR-amplified using pMAR161
as template and the oligonucleotides ORF1-3 (5¢-GGT
TCTTCGGAGGAAAGACATATGGCACTCACGCTTCC
C-3¢) and ORF3-2 (5¢-CCGAAGCTCCGTCTAGATTCC
CTCGCCACGC-3¢) as primers. The fragment was digested
by NdeI and XbaI, and cloned into the expression vector
pUC19 [40]. An artificial constitutive-expression promoter,
Part-1-2, was inserted just before crtE in plasmid pFD3,
generating plasmid pFD6 and strain FD6. This promoter
was generated by hybridization of two complementary oli-
gonucleotides, Part-1 (5¢-AGCTTGACAGGCCGGAATAT
TTCCCTATAATGCGCTGCA-3¢) and Part-2 (5¢-GCG

CATTATAGGGAAATATTCCGGCCTGTCA-3¢), which
contain the E. coli RNA polymerase r
70
consensus binding
site, TTGACA, in position ) 35 and TATAAT in position
) 10 [41], and was cloned in vectors digested with HindIII
and PstI. The sequence between the ) 35 and ) 10 positions
was based on the highly expressed promoter 1 of Es. coli
rrnA, which encodes ribosomal RNA [42].
A DNA fragment containing the crtIb coding region plus
12 and 26 additional bp upstream and downstream, respec-
tively, was PCR-amplified using pMAR202 as a template
[14], and CRTI-1a (5¢-GTGGGATTCCGTTCATCTAGAT
ACCGGAGGGCCTTGGC-3¢) and CRTI-2 (5¢-GAGCGC
GCCACTGGATCCCGCGGCGCTCACC-3¢) as primers.
The fragment obtained was cloned, after digestion with
XbaI and BamHI, into vector pUC19, resulting in plasmid
pFD1. To generate plasmid pFD9 (present in E. coli strain
FD9), plasmid pFD1 was digested with XbaI and BamHI,
and the crtIb fragment was cloned into pFD6.
To generate plasmid pMAR183, a DNA fragment was
amplified by PCR using pMR161 as a template, and
ORF2-1 (5¢-ACCGCGCCGCCTGCAGATCCCATGAGT
GCATCG-3¢) and ORF2-2 (5¢-ACCAGCGCCTTGTCGA
CAGGCGGGC-3¢) as primers. This fragment was digested
with PstI and SalI to generate a product containing the crtIa
coding region plus 2 and 14 bp upstream and downstream,
respectively; this was then cloned into vector pUC9.
To generate plasmid pFD39, the same crtIb fragment
used for pFD9 was PCR-amplified using pMAR202 as a

template, and CRTI-3 (5¢-GTGGGATTCCGTTCATCGA
TATACCGGAGGGCC-3¢) and CRTI-2 as primers. After
digestion by ClaI and BamHI, this fragment was cloned
into vector pACYC184 [43], to create plasmid pFD24. An
artificial constitutive-expression promoter, Par-5-6, was
inserted just before crtIb in plasmid pFD24, generating
plasmid pFD26. The Part-5-6 promoter was generated
by hybridization of two complementary oligonucleotides,
Part-5 (5¢-CTAGATTGACAGGCCGGAATATTTCCCTA
TAATGCGCAT-3¢) and Part-6 (5¢-CGATGCGCATTAT
AGGGAAATATTCCGGCCTGTCAAT-3¢), which con-
tain, like the Part-1-2 promoter, the E. coli RNA polymer-
ase r
70
consensus binding site, and was cloned in vectors
digested with XbaI and BamHI. Plasmid pFD26 was
digested by XbaI and BamHI, and the Part-5-6-crtIb frag-
ment was cloned into vector pBJ114 [44], resulting in plas-
mid pFD39.
Carotenoid extraction and analysis
E. coli was grown in LB medium to stationary phase, and
1 mL of this culture was inoculated into 100 mL of LB
medium and incubated at 37 °C for 12 h. FD100, FD101
and FD102 E. coli strains were supplemented with isopro-
pyl thio-b-d-galactoside (0.5 mm) to increase expression of
crtP, and incubated at 28 °C for 48 h [5]. In the case of
M. xanthus, 100 mL of CTT was inoculated with 1 mL of
culture in stationary phase, and incubated at 33 °C until
this culture reached stationary phase. Additional experi-
mental procedures were identical for all cultures. A 1.5 mL

aliquot was taken for protein assay [15], and the remaining
volume was centrifuged at 15 700 g using an Eppendorf
5415D 24-place rotor for 1.5–2.0 ml tubes (Eppendorf AG,
Hamburg, Germany) to pellet the cells. The pellet was
stored at ) 20 °C until analysis.
The pellet was resuspended in 20 mL of methanol by
vigorous shaking for 15 min, and then centrifuged at
27 200 g using a Beckman J2-21 centrifuge and JA-20
rotor (GMI Inc., Albertville, MN, USA) and re-extracted
until the sample was totally colorless. An equal volume of
Phytoene dehydrogenation to lycopene in M. xanthus A. A. Iniesta et al.
4312 FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS
light petroleum (40–60 °C) was added to the methanol
extract. The two-phase sample was vigorously shaken for
5 min, and the upper, light petroleum phase was removed.
Three further extractions with light petroleum were per-
formed, and the fractions were pooled and concentrated
under vacuum. Carotenes were resuspended in 1–10 mL of
hexane, and identified by their absorption spectra and
quantified by their extinction coefficient [45]. In some
cases, the concentrated extract was resuspended in 1 mL
of light petroleum and chromatographed on a Brockman
grade III deactivated alumina column [45], which was
developed using light petroleum with increasing quantities
of acetone. When chromatographed in activated alumina,
the cis-phytoene accumulated by M. xanthus strain
MR841 (Fig. 3A) behaves like 15-cis-phytoene extracted
from the fungus P. blakesleeanus, as described in detail in
Martinez–Laborda et al. [11]. The separated carotenes
were collected, concentrated and resuspended in hexane

for their analysis. All manipulations of carotenoids were
carried out in the dark at 4 °C. The same carotenes were
always detected in various independent analyses of the
same strain, although some quantitative differences were
observed, particularly in the heterologous expression
experiments.
Acknowledgements
We thank Jose
´
A. Madrid for technical assistance, and
Dr Gerhard Sandmann and Dr Agustı
´
n Vioque for
providing plasmids and strains. This work was
supported by the Spanish Ministerio de Educacio
´
n
y Cultura (grant PB96-1096 and fellowship to
M. Cervantes), Ministerio de Ciencia y Tecnologı
´
a
(grant BMC2000-1006), and Fundacio
´
nSe
´
neca (fellow-
ship to M. Cervantes).
References
1 Sandmann G (2002) Molecular evolution of carotenoid
biosynthesis from bacteria to plants. Physiol Plant 116,

431–440.
2 Armstrong GA (1994) Eubacteria show their true col-
ors: genetics of carotenoid pigment biosynthesis from
microbes to plants. J Bacteriol 176, 4795–4802.
3 Frigaard NU, Maresca JA, Yunker CE, Jones AD &
Bryant DA (2004) Genetic manipulation of carotenoid
biosynthesis in the green sulfur bacterium Chlorobium
tepidum. J Bacteriol 186, 5210–5220.
4 Liang C, Zhao F, Wei W, Wen Z & Qin S (2006)
Carotenoid biosynthesis in cyanobacteria: structural and
evolutionary scenarios based on comparative genomics.
Int J Biol Sci 2, 197–207.
5 Linden H, Vioque A & Sandmann G (1993) Isolation of
a carotenoid biosynthesis gene coding for f-carotene
desaturase from Anabaena PCC 7120 by heterologous
complementation. FEMS Microbiol Lett 106, 99–103.
6 Steiger S, Jackisch Y & Sandmann G (2005) Caroten-
oid biosynthesis in Gloeobacter violaceus PCC4721
involves a single CrtI-type phytoene desaturase instead
of typical cyanobacterial enzymes. Arch Microbiol 184,
207–214.
7 Tsuchiya T, Takaichi S, Misawa N, Maoka T,
Miyashita H & Mimuro M (2005) The cyanobacterium
Gloeobacter violaceus PCC 7421 uses bacterial-type
phytoene desaturase in carotenoid biosynthesis.
FEBS Lett 579, 2125–2129.
8 Burchard RP & Dworkin M (1966) Light-induced lysis
and carotenogenesis in Myxococcus xanthus. J Bacteriol
91, 535–545.
9 Burchard RP & Hendricks SB (1969) Action spectrum

for carotenogenesis in Myxococcus xanthus. J Bacteriol
97, 1165–1168.
10 Moraleda-Munoz A, Perez J, Fontes M, Murillo FJ &
Munoz-Dorado J (2005) Copper induction of carotenoid
synthesis in the bacterium Myxococcus xanthus. Mol
Microbiol 56, 1159–1168.
11 Martinez-Laborda A, Balsalobre JM, Fontes M &
Murillo FJ (1990) Accumulation of carotenoids in struc-
tural and regulatory mutants of the bacterium Myxococ-
cus xanthus. Mol Gen Genet 223, 205–210.
12 Ruiz-Vazquez R, Fontes M & Murillo FJ (1993) Clustering
and co-ordinated activation of carotenoid genes in Myxo-
coccus xanthus by blue light. Mol Microbiol 10, 25–34.
13 Botella JA, Murillo FJ & Ruiz-Vazquez R (1995) A
cluster of structural and regulatory genes for light-
induced carotenogenesis in Myxococcus xanthus. Eur
J Biochem 233, 238–248.
14 Fontes M, Ruiz-Vazquez R & Murillo FJ (1993)
Growth phase dependence of the activation of a bacte-
rial gene for carotenoid synthesis by blue light. EMBO
J 12, 1265–1275.
15 Balsalobre JM, Ruiz-Vazquez RM & Murillo FJ (1987)
Light induction of gene expression in Myxococcus
xanthus. Proc Natl Acad Sci USA 84, 2359–2362.
16 Martinez-Argudo I, Ruiz-Vazquez RM & Murillo FJ
(1998) The structure of an ECF-sigma-dependent, light-
inducible promoter from the bacterium Myxococcus
xanthus. Mol Microbiol 30 , 883–893.
17 Armstrong GA (1997) Genetics of eubacterial caroten-
oid biosynthesis: a colorful tale. Annu Rev Microbiol 51,

629–659.
18 Wang CW & Liao JC (2001) Alteration of product spec-
ificity of Rhodobacter sphaeroides phytoene desaturase
by directed evolution. J Biol Chem 276, 41161–41164.
19 Hausmann A & Sandmann G (2000) A single five-step
desaturase is involved in the carotenoid biosynthesis
pathway to beta-carotene and torulene in Neurospora
crassa. Fungal Genet Biol 30, 147–153.
A. A. Iniesta et al. Phytoene dehydrogenation to lycopene in M. xanthus
FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS 4313
20 Honda D, Yokota A & Sugiyama J (1999) Detection of
seven major evolutionary lineages in cyanobacteria
based on the 16S rRNA gene sequence analysis with
new sequences of five marine Synechococcus strains.
J Mol Evol 48, 723–739.
21 Isaacson T, Ronen G, Zamir D & Hirschberg J (2002)
Cloning of tangerine from tomato reveals a carotenoid
isomerase essential for the production of beta-carotene
and xanthophylls in plants. Plant Cell 14, 333–342.
22 Breitenbach J, Vioque A & Sandmann G (2001) Gene
sll0033 from Synechocystis 6803 encodes a carotene
isomerase involved in the biosynthesis of all-E lycopene.
Z Naturforsch [C] 56, 915–917.
23 Masamoto K, Wada H, Kaneko T & Takaichi S (2001)
Identification of a gene required for cis-to-trans carotene
isomerization in carotenogenesis of the cyanobacterium
Synechocystis sp. PCC 6803. Plant Cell Physiol 42,
1398–1402.
24 Park H, Kreunen SS, Cuttriss AJ, DellaPenna D &
Pogson BJ (2002) Identification of the carotenoid

isomerase provides insight into carotenoid biosynthesis,
prolamellar body formation, and photomorphogenesis.
Plant Cell 14, 321–332.
25 Breitenbach J & Sandmann G (2005) zeta-carotene cis
isomers as products and substrates in the plant poly-cis
carotenoid biosynthetic pathway to lycopene. Planta
220, 785–793.
26 Giuliano G, Giliberto L & Rosati C (2002) Carotenoid
isomerase: a tale of light and isomers. Trends Plant Sci
7, 427–429.
27 Breitenbach J, Braun G, Steiger S & Sandmann G
(2001) Chromatographic performance on a C30-bonded
stationary phase of monohydroxycarotenoids with
variable chain length or degree of desaturation and of
lycopene isomers synthesized by various carotene
desaturases. J Chromatogr A 936, 59–69.
28 Sandmann G (2001) Carotenoid biosynthesis and
biotechnological application. Arch Biochem Biophys 385,
4–12.
29 Heathcote P (2001) Type I photosynthetic reaction
centres. Biochim Biophys Acta 1507, 1–2.
30 Aragon CM, Murillo FJ, de la Guardia MD &
Cerda-Olmedo E (1976) An enzyme complex for the
dehydrogenation of phytoene in Phycomyces. Eur J
Biochem 63, 71–75.
31 De la Guardia MD, Aragon CM, Murillo FJ & Cerda-
Olmedo E (1971) A carotenogenic enzyme aggregate in
Phycomyces: evidence from quantitive complementation.
Proc Natl Acad Sci USA 68, 2012–2015.
32 Hanahan D (1983) Studies on transformation of

Escherichia coli with plasmids. J Mol Biol 166, 557–580.
33 Casadaban MJ & Cohen SN (1980) Analysis of gene
control signals by DNA fusion and cloning in
Escherichia coli. J Mol Biol 138, 179–207.
34 Gill RE, Cull MG & Fly S (1988) Genetic identification
and cloning of a gene required for developmental cell
interactions in Myxococcus xanthus. J Bacteriol 170,
5279–5288.
35 Martinez-Laborda A, Elias M, Ruiz-Vazquez R &
Murillo FJ (1986) Insertions of Tn5 linked to mutations
affecting carotenoid synthesis in Myxococcus xanthus.
Mol Gen Genet 205, 107–114.
36 Bretscher AP & Kaiser D (1978) Nutrition of Myxococ-
cus xanthus, a fruiting myxobacterium. J Bacteriol 133,
763–768.
37 Sambrook J & Russel DW (2000) Molecular Cloning:
a Laboratory Manual, 3rd edn. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY.
38 Vieira J & Messing J (1982) The pUC plasmids, an
M13mp7-derived system for insertion mutagenesis and
sequencing with synthetic universal primers. Gene 19,
259–268.
39 Hodgson DA (1993) Light-induced carotenogenesis in
Myxococcus xanthus: genetic analysis of the carR
region. Mol Microbiol 7, 471–488.
40 Yanisch-Perron C, Vieira J & Messing J (1985)
Improved M13 phage cloning vectors and host strains:
nucleotide sequences of the M13mp18 and pUC19
vectors. Gene 33, 103–119.
41 Hawley DK & McClure WR (1983) Compilation and

analysis of Escherichia coli promoter DNA sequences.
Nucleic Acids Res 11, 2237–2255.
42 Glaser G, Sarmientos P & Cashel M (1983) Functional
interrelationship between two tandem E. coli ribosomal
RNA promoters. Nature 302, 74–76.
43 Chang AC & Cohen SN (1978) Construction and
characterization of amplifiable multicopy DNA cloning
vehicles derived from the P15A cryptic miniplasmid.
J Bacteriol 134, 1141–1156.
44 Julien B, Kaiser AD & Garza A (2000) Spatial control
of cell differentiation in Myxococcus xanthus. Proc Natl
Acad Sci USA 97, 9098–9103.
45 Davies BH (1976) Carotenoids. In Chemistry and
Biochemistry of Plant Pigments, 2nd edn (Goodwin TW,
ed.), pp. 38–165. Academic Press, London.
46 Herber R, Maudinas B, Villoutreix J & Granger P
(1972) The formation of all-trans-phytoene in Mucor
hiemalis. Biochim Biophys Acta 280, 194–202.
Phytoene dehydrogenation to lycopene in M. xanthus A. A. Iniesta et al.
4314 FEBS Journal 274 (2007) 4306–4314 ª 2007 The Authors Journal compilation ª 2007 FEBS