CA2180024C - Dna chain useful for xanthophyll synthesis and process for producing xanthophylls - Google Patents

Abstract

Disclosed are the following DNA strands relating to the synthesis of keto group-containing xanthophylls such as astaxanthin and the like, and the techniques relating to the production of xanthophylls by genetic engineering: A DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting a methylene group at the 4-position of a .beta.- ionone ring into a keto group. A DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting a methylene group at the 4-position of a 3- hydroxy-.beta.-ionone ring into a keto group. A DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for adding a hydroxyl group to the 3-carbon of a 4-keto-.beta.-ionone ring. It is possible to produce a variety of xanthophylls such as canthaxanthin, astaxanthin and the like by introducing the DNA strands into an appropriate microorganism such as Escherichia coli and the like.

Description

i DNA STRANDS USEFUL FOR THE SYNTHESIS OF XANTHOPHYLLS
AND THE PROCESS FOR PRODUCING THE XANTHOPHYLLS
Technical Field The present invention relates to DNA strands useful for the synthesis of keto group-containing xanthophylls (ketocarotenoids) such as astaxanthin which are useful for heightening the color of cultured fishes and shellfishes such as sea breams, salmons, lobster and the like and is used for foods as a coloring agent and an antioxidant, and to a process for producing keto group-containing xanthophylls (ketocarotinoids) such as astaxanthin with use of a microorganism into which the DNA strands have been introduced.
Background Art The term xanthophylls mean carotenoid pigments having an oxygen-containing group such as a hydroxyl group , a keto group or an epoxy group. Carotenoids are synthesized by the isoprenoid biosynthetic process which is used in common halfway with steroids and other terpenoids with mevalonic acid as a starting material. C15 farnesyl pyrophosphate (FPP) resulting from isoprene basic biosynthetic pathway is condensed with C5 isopentenyl pyrophosphate (IPP) to give C20 geranylgeranyl pyrophosphate (GGPP). Two molecules of GGPP are condensed to synthesize a colorless phytoene as an initial carotenoid. The phytoene is converted into phytofluene, ~-carotene, neurosporene and then lycopene by a series of desaturation reactions, and lycopene is in turn converted into (3-carotene by the cyclization reaction. It is believed that a variety of xanthophylls are synthesized by introducing a hydroxyl group or a keto group into the ~i-carotene (See Britton, G., "Biosynthesis of Carotenoids";
Plant Pigments, Goodwin, T.W. ed., London, Academic Press, 1988, pp. 133-182).
The present inventors have recently made it possible to clone a carotenoid biosynthesis gene cluster from the epiphytic non-photosynthetic bacterium Erwinia uredovora in -2- 2~g0024 Escherichia coli with an index of the yellow tone of the bacterium, a variety of combinations of the genes being expressed in microorganisms such as Escherichia coli to produce phytoene, lycopene, ~i-carotene, and zeaxanthin which is a derivative of (3-carotene into which hydroxyl groups have been introduced (See Fig. 10; Misawa, N., Nakagawa, M., Kobayashi, K., Yamano, S., Izawa, Y., Nakamura, K., Harashima, K.; "Elucidation of the Erwinia uredovora Carotenoid biosynthetic Pathway by Functional Analysis of Gene Products Expressed in Escherichia coli", J. Bacteriol., 172, p.6704-6712, 1990; Misawa, N., Yamano, S., Ikenaga, H., "Production of (3-carotene in Zymomonas mobilis and Agrobacterium tumefaciencs by Introduction of the Biosynthesis Genes from Erwinia uredovora", Appl.
environ. Microbiol., 57, p. 1847-1849, 1991; and Japanese Patent Publication No. 58786/1991: "DNA Strands useful for the Synthesis of Carotenoids").
On the other hand, astaxanthin, a red xanthophyll, is a typical animal carotenoid which occurs particularly in a wide variety of marine animals including red fishes such as a sea bream and a salmon, and crustaceans such as a crab and a lobster. In general, animals cannot biosynthesize carotenoids, so that it is necessary for them to ingest carotenoids synthesized by microorganisms or plants from their environments. Thus, astaxanthin has hitherto been used widely for strengthening the color of cultured fishes and shellfishes such as a sea bream, a salmon, a lobster and the like. Moreover, astaxanthin has attracted attention not only as a coloring matter in foods but also as an anti-oxidant for removing active oxygen generated in bodies, which causes carcinoma (see Takao Matsuno ed., "Physiological Functions and Bioactivities of Carotenoids in Animals", Kagaku to Seibutsu, 28, p. 219-227, 1990). As the sources of astaxanthin, there have been known crustaceans such as a krill in the Antarctic Ocean, cultured products of a yeast Phaffia, cultured products of a green alga Haematococcus, and products obtained by the organic synthetic methods. However, when crustaceans such as a krill in the Antarctic Ocean or the like are used, it requires laborious works and much expenses for the isolation of astaxantin from contaminants such as lipids and the like during the harvesting and extraction of the krill. Moreover, in the case of the cultured product of the yeast Phaffia, a great deal of expenses are required for the gathering and extraction of astaxanthin, since the yeast has rigid cell walls and produces astaxanthin only in a low yield. Also, in the case of the cultured product of the green alga Haematococcus, not only a location for collecting sunlight or an investment of a culturing apparatus for supplying an artificial light is required in order to supply light which is essential to the synthesis of astaxantin, but also it is difficult to separate astaxanthin from fatty acid esters as by-products or chlorophylls present in the cultured products. For these reasons, astaxanthin produced from biological sources is in the present situation inferior to that obtained by the organic synthetic methods on the basis of cost. The organic synthetic methods however have a problem of by-products produced during the reactions in consideration of its use as a feed for fishes and shellfishes and an additive to foods, and the products obtained by the organic synthetic methods are opposed to the consumer's preference for natural products. Thus, it has been desired to supply an inexpensive astaxanthin which is safe and produced from biological sources and thus has a good image to consumers and to develop a process for producing the astaxanthin.
Disclosure of the Invention It would be considered very useful to find a group of genes for playing a role of the biosynthesis of astaxanthin, because it is possible to afford astaxanthin-producing ability to a microorganism optimum in safety as a food or in potentiality for producing astaxanthin, regardless of the presence of astaxanthin-producing ability, by introducing a gene cluster for astaxanthin biosynthesis into the microorganism. No problem of by-products as contaminants is caused in this case, so that it would be considered not so difficult to increase the production amount of astaxanthin with a recent advanced technique of gene manipulation to a level higher than that accomplished by the organic synthetic methods. However, the groups of genes for synthesizing zeaxanthin, one of the xanthophylls, have already been acquired by the present inventors as described above, while no genes encoding a keto group-introducing enzyme required for the synthesis of astaxanthin have not successfully obtained. The reasons for the failure in obtaining the genes include that the keto group-introducing enzyme is a membrane protein and loses its activity when isolated from the membrane, so that it was impossible to purify the enzyme or measure its activity and no information on the enzyme has been obtained. Thus, it has hitherto been impossible to produce astaxanthin in microorganisms by gene manipulation.
The object of the present invention is to provide DNA
strands which contain genes required for producing keto group-containing xanthophylls (ketocarotenoids) such as astaxanthin in microorganisms by obtaining such genes coding for enzymes such as a keto group-introducing enzyme required for producing keto group-containing xanthophylls (ketocarotenoids) such as astaxanthin, and to provide a process for producing keto group-containing xanthophylls (ketocarotenoids) such as astaxanthin with the microorganisms into which the DNA strands have been introduced.
The gene cloning method which is often used usually comprising purifying the aimed protein, partially determining the amino acid sequence and obtaining genes by a synthetic probe cannot be employed because of the purification of the astaxanthin synthetic enzyme being impossible as described above. Thus, the present inventors have paid attention to the fact that the cluster of carotenoid synthesis genes in non-photosynthetic bacterium (Erwinia) functions in Escherichia coli, in which lycopene and ~i-carotene which are believed to be intermediates for biosynthesis of astaxanthin are allowed to produce with combinations of the genes from the gene cluster, and have used Escherichia coli as a host for cloning of astaxanthin synthetic genes. The present inventors have also paid attention to the facts that some marine bacteria have an astaxanthin-producing ability (Yokoyama A., Izumida, H., Miki, W., "Marine bacteria produced astaxanthin", 10th International Symposium on Carotenoids, Abstract, CL11-3, 1993), that a series of related genes would constitute a cluster in the case of bacteria, and that the gene cluster would be expressed functionally in Escherichia coli in the case of bacteria. The present inventors have thus selected the marine bacteria as the gene sources. They have carried out researches with a combination of these two means and successfully obtained the gene group which is required for the biosynthesis of astaxanthin and the other keto group-containing xanthophylls from marine bacteria. They have thus accomplished the present invention. In addition, it has been first elucidated in the present invention that the astaxanthin synthesis gene cluster in marine bacteria constitutes a cluster and expresses its function in Escherichia coli, and these gene products can utilize ,Q-carotene or lycopene as a substrate.
The DNA strands according to the present invention are set forth as follows.
(1) A purified DNA strand derived from a marine 2~ 80024 bacterium and having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting the methylene group at the 4-position of the ~i-ionone ring into a keto group.
(2) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting the methylene group at the 4-position of the , ionone ring into a keto group and having an amino acid sequence substantially of amino acid Nos. 1-212 which is shown in the SEQ ID N0:1.
(3) A purified DNA strand hybridizing with the DNA
strand described in (2) and having a nucleotide sequence which encodes a polypeptide having the enzyme activity described in (2) .
(4) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting the methylene group at the 4-position of the ~i-ionone ring into a keto group and having an amino acid sequence substantially of amino acid Nos. 1-242 which is shown in the SEQ ID N0:5.
(5) A purified DNA strand hybridizing with the DNA
strand described in (4) and having a nucleotide sequence which encodes a polypeptide having the enzyme activity described in (4) .
(6) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting ,Q-carotene into canthaxanthin via echinenone and having an amino acid sequence substantially of amino acid .:?

-' - 2180x.24 Nos. 1-212 which is shown in the SEQ ID N0:1.

(7) A purified DNA strand hybridizing with the DNA
strand described in (6) and having a nucleotide sequence which encodes a polypeptide having the enzyme activity described in (6) .

(8) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting ~i-carotene into canthaxanthin via echinenone and having an amino acid sequence substantially of amino acid Nos. 1-242 which is shown in the SEQ ID N0:5.

(9) A purified DNA strand hybridizing with the DNA
strand described in (8) and having a nucleotide sequence which encodes a polypeptide having the enzyme activity described in (8) .

(10) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting the methylene group at the 4-position of the 3-hydroxy-~i-ionone ring into a keto group.

(11) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting the methylene group at the 4-position of the 3-hydroxy-~i-ionone ring into a keto group and having an amino acid sequence substantially of amino acid Nos. 1-212 which is shown in the SEQ ID NO:1.

(12) A purified DNA strand hybridizing with the DNA
strand described in (11) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (11).

_8_ (13) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting the methylene group at the 4-position of the 3-hydroxy-~i-ionone ring into a keto group and having an amino acid sequence substantially of amino acid Nos. 1-242 which is shown in the SEQ ID N0:5.

(14) A purified DNA strand hybridizing with the DNA
strand described in (13) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (13).

(15) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting zeaxanthin into astaxanthin by way of 4-ketozeaxanthin and having an amino acid sequence substantially of amino acid Nos. 1-212 which is shown in the SEQ ID NO:1.

(16) A purified DNA strand hybridizing with the DNA
strand described in (15) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (15).

(17) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting zeaxanthin into astaxanthin by way of 4-ketozeaxanthin and having an amino acid sequence substantially of amino acid Nos. 1-242 which is shown in the SEQ ID N0:5.

(18) A purified DNA strand hybridizing with the DNA
strand described in (17) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (17).

2~ 80024 _ g _ (19) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for adding a hydroxyl group to the 3-carbon of the 4-keto-,~-ionone ring.

(20) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for adding a hydroxyl group to position 3-carbon of the 4-keto-~i-ionone ring and having an amino acid sequence substantially of amino acid Nos. 1-162 which is shown in the SEQ ID N0:2.

(21) A purified DNA strand hybridizing with the DNA
strand described in (20) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (20).

(22) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for adding a hydroxyl group to position 3-carbon of the 4-keto-/3-ionone ring and having an amino acid sequence substantially of amino acid Nos. 1-162 which is shown in the SEQ ID N0:6.

(23) A purified DNA strand hybridizing with the DNA
strand described in (22) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (22).

(24) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting canthaxanthin into astaxanthin by way of phoenicoxanthin and having an amino acid sequence .,,, - 9a - 2 ~ 8 ~ ~ L
substantially of amino acid Nos. 1-162 which is shown in the SEQ ID N0:2.

(25) A purified DNA strand hybridizing with the DNA
strand described in (24) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (24).

(26) A purified DNA strand having a nucleotide sequence which encodes a polypeptide having an enzyme activity for converting canthaxanthin into astaxanthin by way of phoenicoxanthin and having an amino acid sequence substantially of amino acid Nos. 1-162 which is shown in the SEQ ID N0:6.

(27) A purified DNA strand hybridizing with the DNA
strand described in (26) and having a nucleotide sequence which encodes a polypeptide having an enzyme activity described in (26).
The present invention also relates to a process for producing xanthophylls.
That is, the process for producing xanthophylls according to the present invention is set forth below.
(1) A process for producing a xanthophyll comprising introducing the DNA strand described in any one of the above mentioned DNA strands (1)-(9) into a microorganism having a ~i-carotene-synthesizing ability, culturing the transformed microorganism in a culture medium, and obtaining canthaxanthin or echinenone from the cultured cells.
(2) A process for producing a xanthophyll comprising introducing the DNA strand described in any one of the above 2~~80024 - 9b -mentioned DNA strands (10)-(18) into a microorganism having a zeaxanthin-synthesizing ability, culturing the transformed microorganism in a culture medium, and obtaining astaxanthin or 4-ketozeaxanthin from the cultured cells.
(3) A process for producing a xanthophyll comprising introducing the DNA strand described in any one of the above mentioned DNA strands (19)-(27) into a microorganism having a canthaxanthin-synthesizing ability, ?.180024 culturing the transformed microorganism in a culture medium, and obtaining astaxanthin or phoenicoxanthin from the cultured cells.
( 4 ) A process for producing a xanthophyll according to any one of the above mentioned processes (1) - (3), wherein the microorganism is a bacterium or yeast.
Brief Description of the Drawings Fig. 1 illustrates diagrammatically the nucleotide sequence of the keto group-introducing enzyme gene (crt W
gene) of the marine bacterium Agrobacterium aurantiacus sp.
nov. MK1 and the amino acid sequence of a polypeptide to be encoded thereby.
Fig. 2 illustrates diagrammatically the nucleotide sequence of the hydroxyl group-introducing enzyme gene ( crt Z gene) of the marine bacterium Agrobacterium aurantiacus sp. nov. MK1 and the amino acid sequence of a polypeptide to be encoded thereby.
Fig. 3 illustrates diagrammatically the nucleotide sequence of the lycopene-cyclizing enzyme gene ( crt Y gene ) of the marine bacterium Agrobacterium aurantiacus sp. nov.
MKl and the amino acid sequence of a polypeptide to be encoded thereby.
Fig. 4 illustrates diagrammatically the continuation of the sequences following to those illustrated in Fig. 3.
Fig. 5 illustrates diagrammatically the nucleotide sequence of the xanthophyll synthesis gene cluster of the marine bacterium Agrobacterium aurantiacus sp. nov. MK1.
The letters A - F in Fig. 5 correspond to those in Figs. 1 - 4.
Fig. 6 illustrates diagramatically the continuation of the sequence following to that illustrated in Fig./5.
Fig. 7 illustrates diagrammatically the continuation of the sequence following to that illustrated in Fig. 6.
Fig. 8 illustrates diagrammatically the continuation of the sequence following to that illustrated in Fig. 7.
Fig. 9 illustrates diagrammatically the continuation of the sequence following to that illustrated in Fig. 8.
Fig. 10 illustrates diagrammatically the carotenoid biosynthetic route of the non-photosynthesis bacterium Erwinia uredovora and the functions of the carotenoid synthetic genes.
Fig. 11 illustrates diagrammatically the main xanthophyll biosynthetic routes of marine bacteria Agrobacterium aurantiacus sp. nov. MK1 and Alcaligenes sp.
PC-1 and the functions of the xanthophyll synthesis genes.
The function of crtY gene, however, has been confirmed only in the former bacterium.
Fig. 12 illustrates diagrammatically a variety of deletion plasmids containing the xanthophyll synthesis genes (cluster) of the marine bacterium Agrobacterium aurantiacus sp. nov. MKl.
The letter p represents the promoter of the lac of the vector pBluescript II SK. The positions of cutting with restriction enzymes are represented by abbreviations as follows: Sa, SacI; X, XbaI;.B, BamHI; P, PstI; E, EcoRI; S, SalI; A, Apal; K, Kpnl; St, StuI; N, NruI; Bg, BglII; Nc, NcoI; Hc, HincII.
Fig. 13 illustrates diagrammatically the nucleotide sequence of the keto group-introducing enzyme gene (crtW
gene) of the marine bacterium Alcaligenes sp. PC-1 and the amino acid sequence of a polypeptide to be encoded thereby.
Fig. 14 illustrates diagrammatically the continuation of the sequences following to those illustrated in Fig. 13.
Fig. 15 illustrates diagrammatically the nucleotide sequence of the hydroxyl group-introducing enzyme gene (crtZ gene) of the marine bacterium Alcaligenes sp. PC-1 and the amino acid sequence of a polypeptide to be encoded thereby.
Fig. 16 illustrates diagrammatically the nucleotide sequence of the xanthophyll synthetic gene cluster of the marine bacterium Alcaligenes sp. PC-1 and the amino acid sequence of a polypeptide to be encoded thereby. The - 12 - 2~
letters A -D in Fig. 16 correspond to those in Figs. 13 -15.
Fig. 17 illustrates diagrammatically the continuation of the sequences following to those illustrated in Fig. 16.
Fig. 18 illustrates diagrammatically the continuation of the sequences following to those illustrated in Fig. 17.
Fig. 19 illustrates diagrammatically a variety of deletion plasmids containing the xanthophyll synthetic genes (cluster) of the marine bacterium Alcaligenes sp. PC
1.
The letter p represents the lac promoter of the vector pBluescript II SK+, Fig. 20 illustrates diagrammatically xanthophyll biosynthetic routes containing miner biosynthetic routes in the marine bacteria Agrobacterium aurantiacus sp. no.
MK1 and Alcaligenes sp. PC-1 and the functions of the xanthophyll synthesis genes.
Miner biosynthetic routes are represented by dotted arrows.
Best Mode for carrying out the Invention The present invention is intended to provide DNA
strands which are useful for synthesizing a keto group-containing xanthophylls (ketocarotenoids) such as astaxanthin derived from a marine bacteria Agrobacterium aurantiacus sp. nov. MK1 and Alcaligenes sp. PC-1, and a process for producing keto group-containing xanthophylls (ketocarotenoids), i.e. astaxanthin, phoenicoxanthin, 4 ketozeaxanthin, canthaxanthin, and echinenone with use of a microorganism into which the DNA strands have been introduced.
The DNA strands according to the present invention are in principle illustrated generally by the aforementioned DNA strands (1), (10) and (19) from the standpoint of the fine chemical-generating reaction, and basically defined by the aforementioned DNA strands (2), (4), (11), (13), (20) and (22). The specific examples of the DNA strands (2) and (4) are the aforementioned DNA strands (6) and (8); the specific examples of the DNA strands (11) and (13) are the aforementioned DNA strands (15) and (17); and the specific examples of the DNA strands (20) and (22) are the aforementioned DNA strands (24) and (26). In this connection, the DNA strands ( 3 ) , ( 5 ) , ( 7 ) , ( 9 ) , ( 12 ) , ( 14 ) , (16), (18), (21), (23), (25) and (27) hybridize the DNA
strands (2), (4), (6), (8), (11), (13), (15), (17), (20), (22), (24) and (26), respectively, under a stringent condition.
The polypeptides encoded by the DNA strands according to the present invention have amino acid sequences substantially in a specific range as described above in SEQ
ID NOS: 1 - 2, and 5 - 6 (Figs. 1 2, and 13 - 15), e.g.
an amino acid sequence of amino acid Nos. 1 - 212 in SEQ ID
NOS: 1 (A - B in Fig. 1). In the present invention, four polypeptides encoded by these DNA strands, that is four enzymes participating in the xanthophyll-producing reaction) may be modified by deletion, substitution or addition in some of the amino acids provided that the polypeptides have the enzyme activities as described above (see Example 13). This corresponds to that "amino acid sequences ......substantially ..." For instance, an enzyme of which amino acid at the first position (Met) has been deleted is also involved in the polypeptide or enzyme obtained by the modification of the amino acid sequence.
In this connection, it is needless to say that the DNA
strands according to the present invention for encoding the polypeptides also include, in addition to those having nucleotide sequences in a specific range shown in SEQ ID
NOS: 1 - 2, and 13 - 15 (Figs. 1 - 2, and 13 - 15), degenerate isomers encoding the same polypeptides as above except degenerate codons.
Keto group-introducing enzyme gene (crtW) The DNA strands (1) - (18) are genes which encode the keto group-introducing enzymes (referred to hereinafter as crtW ) . Typical examples of the genes are crtW genes cloned from the marine bacteria Agrobacterium aurantiacus sp. nov.

MK1 or Alcaligenes sp. PC-1, which are the DNA strands comprising the nucleotide sequences encoding the polypeptides having the amino acid sequences A - B in Fig.
1 (amino acid Nos. 1 - 212 in SEQ ID NOS: 1) or A - B in Figs. 13 - 14 (amino acid Nos. 1 - 242 in SEQ ID NOS: 5).
The crtW gene product (also referred to hereinafter as CrtW ) has an enzyme activity for converting the 4-methylene group of the (3-ionone ring into a keto group, and one of the specific examples is an enzyme activity for synthesizing canthaxanthin with ~i-carotene as a substrate by way of echinenone (see Fig. 11). In addition, the crtW
gene product also has an enzyme activity for converting the 4-methylene group of the 3-hydroxy-~i-ionone ring into a keto group, and one of the specific examples is an enzyme activity for synthesizing astaxanthin with zeaxanthin as a substrate by way of 4-ketozeaxanthin (see Fig. 11). In this connection, the polypeptides having such enzyme activities and the DNA strands encoding the polypeptides have not hitherto been reported, and the polypeptides or the DNA strands encoding the polypeptides has no overall homology to polypeptides or DNA strands which have hitherto been reported. Moreover, no such information has been reported that one enzyme has an activity to convert directly a methylene group of not only the (3-ionone ring and the 3-hydroxy-~i-ionone ring but also the other compounds into a keto group. Moreover, a homology of CrtW
as high as 83% identity at an amino acid sequence level was shown between Agrobacterium and Alcaligenes.
On the other hand, it is possible to allow a microorganisms such as Escherichia coli or the like to produce ~3-carotene or zeaxanthin by using the carotenoid synthesis genes of the non-photosynthetic bacterium Erwinia, that is the crtE, crtB, crtl and crtY genes of Erwinia afford the microorganism such as Escherichia coli or the like the ~3-carotene-producing ability, and the crtE, crtB, crtI, crtY and crtZ genes of Erwinia afford the .microorganisms such as Escherichia coli or the like the zeaxanthin-producing ability (see Fig. 10 and Laid-Open Publication of W091/13078). Thus, the substrate of CrtW is supplied by the crt gene cluster of Erwinia, so that when additional crtW gene is introduced into the microorganism such as Escherichia coli or the like which contains the aforementioned crt gene cluster of Erwinia, the (3-carotene-producing microorganism will produce canthaxanthin by way of echinenone, and the zeaxanthin-producing microorganism will produce astaxanthin by way of 4-ketozeaxanthin.
Hydroxyl group-introducing enzyme gene (crtZ) The DNA strands (19) - (27) are genes encoding a hydroxyl group-introducing enzyme (referred to hereinafter as crtZ). Typical examples of 'the genes are crtZ genes cloned from the marine bacteria Agrobacterium aurantiacus sp. nov. MK1 or Alcaligenes sp. PC-l, which are the DNA
strands comprising the nucleotide sequences encoding the polypeptides having the amino acid sequences C - D in Fig.
2 (amino acid Nos. 1 - 162 in SEQ ID NOS: 2) or C - D in Figs. 15 (amino acid Nos. l - 162 in SEQ ID NOS: 6). The crtZ gene product (also referred to hereinafter as CrtZ) has an enzyme activity for adding a hydroxyl group to the 3-carbon atom of the (3-ionone ring, and one of the specific examples is an enzyme activity for synthesizing zeaxanthin with use of (3-carotene as a substrate by way of ~i-cryptoxanthin (see Fig. 11). In addition, the crtZ gene product also has an enzyme activity for adding a hydroxyl group to the 3-carbon atom of the 4-keto-(3-ionone ring, and one of the specific examples is an enzyme activity for synthesizing astaxanthin with canthaxanthin as a substrate by way of phoenicoxanthin (see Fig. 11). In this connection, the polypeptide having the latter enzyme activity and the DNA strand encoding the polypeptide have not hitherto been reported. Moreover, CrtZ of Agrobacterium and Alcaligenes showed a high homology with CrtZ of Erwinia uredovora (57% and 58% identity), respectively, at an amino acid sequence level. Also, a high homology of 90% identity at an amino acid sequence 2 ~ soo24 level was shown between the CrtZ of Agrobacterium and Alcaligenes.
It has been described above that it is possible to allow a microorganism such as Escherichia coli or the like to produce ~i-carotene by using the carotenoid synthetic genes of the non-photosynthetic bacterium Erwinia.
Moreover, it has been described above that it is possible to allow a microorganism such as Escherichia coli or the like to produce canthaxanthin by adding crtW thereto.
Thus, the substrate of CrtZ of Agrobacterium or Alcaligenes is supplied by the crtE, crtB, crtI and crtY genes of Erwinia (production of (3-carotene), and the crtW gene of Agrobacterium or Alcaligenes added thereto, so that when the crtZ gene of Agrobacterium or Alcaligenes is introduced into a microorganism such as Escherichia coli or the like containing the crt gene group, the ~i-carotene-producing microorganism will produce zeaxanthin by way of ~i cryptoxanthin, and the canthaxanthin-producing microorganism will produce astaxanthin by way of phoenicoxanthin.
Lycopene-cyclizing enzyme gene (crtY) The DNA strand encoding the amino acid sequence substantially from E to F of Figs. 3 and 4 (amino acid Nos.
1-386 in SEQ ID N0: 3) is a gene encoding a lycopene-cyclizing enzyme (referred to hereinafter as crtY). A
typical example of the gene is the crtY gene cloned from the marine bacterium Agrobacterium aurantiacus sp. nov.
MK1, which is the DNA strand comprising the nucleotide sequence encoding the polypeptide having the amino acid sequence E - F in Figs . 3 and 4 . The crtY gene product (also referred to hereinafter as CrtY) has an enzyme activity for synthesizing ~i-carotene with lycopene as a substrate (see Fig. 11). It is possible to allow a microorganism such as Escherichia coli or the like to produce lycopene by using a carotenoid biosynthesis genes of a non-photosynthetic bacterium Erwinia, that is the crtE, crtB and crtl genes of Erwinia give a microorganism such as Escherichia coli or the like a lycopene biosynthesis ability (see Fig. 10, and Laid-Open Publication of W091/1307$). Thus, the substrate of the CrtY of Agrobacterium is supplied by the crt gene group of Erwinia, so that when the crtY of Agrobacterium is introduced into a microorganism such as Escherichia coli or the like containing the crt gene group, it is possible to allow the microorganism to produce (3-carotene.
In this connection, the CrtY of Agrobacterium has a significant homology of 4~4.3o identity to the CrtY of Erwinia uredovora at the amino acid sequence level, and these CrtY enzymes also have the same enzymatic function (see Figs. 10 and 11).
Bacteriological properties of marine bacteria The marine bacteria Agrobacterium aurantiacus sp. nov.
MKl and Alcaligenes sp. PC-1 as the sources of the xanthophyll synthetic genes show the following bacteriological properties.
<Agrobacterium aurantiacus sp. nov. MKl>
(1) Morphology Form and size of bacterium: rod, 0.9 um x 1.2 um;
Motility: yes;
Flagellum: peripheric flagellum;
Polymorphism of cell: none;
Sporogenesis: none;
Gram staining: negative.
(2) Growths in culture media Broth agar plate culture: non-diffusive circular orange colonies having a gloss are formed.
Broth agar slant culture: a non-diffusive orange band having a gloss is formed.
Broth liquid culture: homogeneous growth all over the culture medium with a color in orange.
Broth gelatin stab culture: growth over the surface around the stab pore.
(3) Physiological properties Reduction of nitrate: positive:

- 1$ -Denitrification reaction: negative;
Formation of indole: negative;
Utilization of citric acid: negative;
Formation of pigments: fat-soluble reddish orange pigment;
Urease activity: negative;
Oxidase activity: positive;
Catalase activity: positive;
(3-Glucosidase activity (esculin degradability): positive;
(3-Galactosidase activity: positive;
Growth range: pH, 5 - 9; temperature, 10 - 40VC:
Behavior towards oxygen: aerobic;
Durability to seawater: positive;
O - F test: oxidation;
Anabolic ability of saccharides:
Positive: D-glucose, D-mannose, D-galactose, D-fructose, lactose, maltose, sucrose, glycogen, N-acetyl-D-glucosamine;
Negative: L-arabinose, D-mannitol, inositol, L-rhamnose, D-sorbitol;
Anabolic ability of organic acids:
Positive: lactate;
Negative: citrate, malate, gluconate, caprinate, succinate, adipate;
Anabolic ability of the other organic materials:
Positive: inosine, uridine, glucose-1-phosphate, glucose-6-phosphate;
Negative: gelatin, L-arginine, DNA, casein.
<Alcaligenes sp. PC-1>
(1) Morphology Form and size of bacterium: short rod, 1.4 um;
Motility: yes;
Flagellum: peripheric flagellum;
Polymorphism of cell: none;
Sporogenesis: none;
Gram staining: negative.
(2) Growths in culture media Broth agar plate culture: non-diffusive circular orange 218~0~~

colonies having a gloss are formed.
Broth agar slant culture: a non-diffusive orange band having a gloss is formed.
Broth liquid culture: homogeneous growth all over the culture medium with a color in orange.
Broth gelatin stab culture: growth over the surface around the stab pore.
(3) Physiological properties Formation of pigments: fat-soluble reddish orange pigment;
Oxidase activity: positive;
Catalase activity: positive;
Growth range: pH, 5 - 9; temperature, 10 - 40vC:
Behavior towards oxygen: aerobic;
Durability to seawater: positive;
O - F test: oxidation;
Degradability of gelatin: negative.
Xanthophyll synthetic gene cluster of the other marine i~,n..to,-; ~
It has hitherto been reported that 16 marine bacteria have an ability to synthesize ketocarotenoids such as astaxanthin and the like (Yokoyama, A., Izumida, H., Miki, W., "Marine bacteria produced astaxanthin", 10th International Symposium on Carotenoids, Abstract, CL11-3, 1993). If either of the crt genes of the aforementioned marine bacteria Agrobacterium aurantiacus sp. nov. MK-1 or Alcaligenes sp. PC-1 is used as a probe, the gene cluster playing a role of the biosynthesis of ketocarotenoids such as astaxanthin and the like should be obtained from the other astaxanthin producing marine bacteria by using the homology of the genes. In fact, the present inventors have successfully obtained the crtW and crtZ genes as the strongly hybridizing DNA fragments from the chromosomal DNA
of Alcaligenes PC-1 with use of a DNA fragment containing crtW and crtZ of Ag. aurantiacus sp. nov. MKl as a probe (see Examples as for the details). Furthermore, when Alteromonas SD-402 was selected from the remaining 14 marine bacteria having an astaxanthin synthetic ability and a chromosomal DNA was prepared therewith and subjected to the Southern hybridization experiment with a DNA fragment containing crtW and crtZ of Ag. aurantiacus sp. nov. MK1, the probe hybridized with the bands derived from the chromosomal DNA of the marine bacteria. The DNA strands according to the present invention also includes a DNA
strand which hybridizes with the DNA strands ( 2 ) , ( 4 ) , ( 6 ) , (8), (11), (13), (15), (17), (20), (22), (24) and (26).
Acquisition of DNA strands Although one of the methods for obtaining the DNA
strand having a nucleotide sequence which encodes the amino acid sequence of each enzyme described above is to chemically synthesize at least a part of the strand length according to the method for synthesizing a nucleic acid, it is believed more preferable than the chemical synthetic method to obtain the DNA strand by using the total DNA
having been digested with an appropriate restriction enzyme to prepare a library in Escherichia coli, from which library the DNA strand is obtained by the methods conventionally used in the art of genetic engineering such as a hybridization method with an appropriate probe (see the xanthophyll synthetic gene cluster of the other marine bacteria).
Transformation of an microorganism such as Escherichia coli and gene expression A variety of xanthophylls can be prepared by introducing the present DNA strands described above into appropriate microorganisms such as bacteria, for example Escherichia coli, Zymomonas mobilis and Agrobacterium tumefaciens, and yeasts, for example Saccharomyces cerivisiae.
The outline for introducing an foreign gene into a preferred microorganism is described below.
The procedure or method for introducing and expressing the foreign gene in a microorganism such as Escherichia coli or the like comprises the ones usually used in the art of genetic engineering in addition to those described below in the present invention and may be carried out according to the procedure or method ( see, e. g. , "Vectors for Cloning Genes", Methods in Enzymology, 216, p. 469-631, 1992, Academic Press, and "Other Bacterial Systems", Methods in Enzymology, 204, p. 305-636, 1991, Academic Press).
<Escherichia coli>
The method for introducing foreign genes into Escherichia coli includes several efficient methods such as the Hanahan's method and the rubidium method, and the foreign genes may be introduced according to these methods (see, for example, Sambrook, J., Fritsch, E.F., Maniatis, T., "Molecular Cloning - A Laboratory Manual", Cold Spring Harbor Laboratory Press, 1989). While foreign genes in Escherichia coli may be expressed according to the conventional methods (see, for example, "Molecular Cloning - A Laboratory Manual"), the expression can be carried out for example with a vector for Escherichia coli having a lac promoter in the pUC or pBluescript series. The present inventors have used a vector pBluescrip II SK or KS for Escherichia coli having a lac promoter and the like to insert the crtW, crtZ and crtY genes of Agrobacterium aurantiacus sp. nov. MK1 and the crtW and crtZ genes of Alcaligenes sp. PC-1 and allowed to express these genes in Escherichia coli.
<Yeast>
The method for introducing foreign genes into yeast Saccharomyces cerivisiae includes the methods which have already been established such as the lithium method and the like, and the introduction may be carried out according to these methods (see, for example, Ed. Yuichi Akiyama, compiled by Bio-industry Association,"New Biotechnology of Yeast", published by IGAKU SHUPPAN CENTER). Foreign genes can be expressed in yeast by using a promoter and a terminator such as PGK and GPD to construct an expression cassette in which the foreign gene is inserted between the promoter and the terminator so that transcription is led through, and inserting the expression cassette into a vector such as the YRp system which is a multi-copy vector for yeast having the ARS sequence of the yeast chromosome as the replication origin, the YEp system which is a multi-copy vector for yeast having the replication origin of the 2 um DNA of yeast, and the YIp system which is a vector for integrating a yeast chromosome having no replication origin of yeast (see "New Biotechnology of Yeast", published by IGAKU SHUPPAN CENTER, ibid.; NIPPON NOGEI-KAGAKU KAI ABC
Series "Genetic Engineering for Producing Materials", published by ASAKURA SHOTEN; and Yamano, S., Ishii, T., Nakagawa, M., Ikenaga, H., Misawa, N., "Metabolic Engineering for Production of (3-carotene and lycopene in Saccharomyces cerevisiae", Biosci. Biotech. Biochem., 58, p. 1112-1114, 1994).
<Zymomonas mobilis>
Foreign genes can be introduced into an ethanol-producing bacterium Zymomonas mobilis by the conjugal transfer method which is common to Gram-negative bacteria, and the foreign genes can be expressed by using a vector pZA22 for Zymomonas mobilis (see Katsumi Nakamura, "Molecular Breeding of Zymomonas mobilis", Nippon Nogei-Kagaku Kaishi, 63, p. 1016-1018, 1989; and Misawa, N., Yamano, S., Ikanaga, H., "Production of ~i-Carotene in Zymomonas mobilis and Agrobacterium tumefaciens by Introduction of the Biosynthesis Genes from Erwinia uredovora", Appl. Environ. Microbiol., 57, p.1847-1849, 1991).
<Agrobacterium tumefaciens>
Foreign genes can be introduced into a plant pathogenic bacterium Agrobacterium tumefaciens by the conjugal transfer method which is common to Gram-negative bacteria, and the foreign genes can be expressed by using a vector pBI121 for a bacterium such as Agrobacterium tumefaciens (see Misawa, N., Yamano, S., Ikenaga, H., "Production of (3 Carotene in Zymomonas mobilis and Agrobacterium tumefaciens by Introduction of the Biosynthesis Genes from Erwinia uredovora", Appl. Environ. Microbiol., 57, p. 1847-1849, 1991).
Production of xanthophylls by microorganisms The gene cluster for the synthesis of ketocarotenoids such as astaxanthin derived from a marine bacterium can be introduced and expressed by the procedure or method described above for introducing and expressing an foreign gene in a microorganism.
Farnesyl pyrophosphate (FPP) is a substrate which is common not only to carotenoids but also to other terpenoids such as sesquiterpenes, triterpenes, sterols, hopanols and the like. In general, microorganisms synthesize terpenoids even if they cannot synthesize carotenoids, so that all of the microorganisms should basically have FPP as an intermediate metabolite. Furthermore, the carotenoid synthesis gene cluster of a non-photosynthetic bacterium Erwinia has an ability to synthesize the substrates of the crt gene products of Agrobacterium aurantiacus sp. nov. MK1 or Alcaligenes sp. PC-1 by using FPP as a substrate (see Fig. 10). The present inventors have already confirmed that when the group of crt genes of Erwinia is introduced into not only Escherichia coli but also the aforementioned microorganisms, that is the yeast Saccharomyces cerevisiae, the ethanol producing bacterium Zymomonas mobilis, or the plant pathogenic bacterium Agrobacterium tumefaciens, carotenoids such as ~i-carotene and the like can be produced, as was expected, by these microorganisms (Yamano, S., Ishii, T., Nakagawa, M., Ikenaga, H., Misawa, N., "Metabolic Engineering for Production of ~i-Carotene and Lycopene in Saccharomyces cerevisiae", Biosci. Biotech.
Biochem., 58, p. 1112-1114, 1994; Misawa, N., Yamano, S., Ikenaga, H. , "Production of ~3-Carotene in Zymomonas mobilis and Agrobacterium tumefaciens by Introduction of the Biosynthetic Genes from Erwinia uredovora", Appl. Environ.
Microbiol., 57, p. 1847-1849, 1991; and Japanese Patent Publication No. 58786/1991 by the present inventors: "DNA Strands useful for the Synthesis of Carotenoids").

Thus, it should be possible in principle to allow all of the microorganisms, in which the gene introduction and expression system has been established, to produce ketocarotenoids such as astaxanthin and the like by introducing the combination of the carotenoid synthesis gene cluster derived from Erwinia and the DNA strands according to the present invention (typically the carotenoid synthesis gene cluster derived from Agrobacterium aurantiacus sp. nov. MK1 or Alcaligenes sp. PC-1) at the same time into the same microorganism. The process for producing a variety of ketocarotenoids in microorganisms are described below.
<Production of canthaxanthin and echinenone>
It is possible to produce canthaxanthin as a final product and echinenone as an intermediate metabolite by introducing into a microorganism such as Escherichia coli and expressing the crtE, crtB, crtI and crtY genes of Erwinia uredovora required for the synthesis of ,Q-carotene and any one of the DNA strands of the present invention (1)-(9) which is a keto group-introducing enzyme gene (typically, the crtW gene of Agrobacterium aurantiacus sp. nov. MK1 or Alcaligenes PC-1). The yields or the ratio of canthaxanthin and echinenone can be changed by controlling the expression level of the DNA
strand (crtW gene) or examining the culturing conditions of a microorganism having the DNA strand. Two embodiments in Escherichia coli are described below, and more details will be illustrated in Examples.
The plasmid pACCARI6ncrtX formed by inserting a fragment containing the crtE, crtB, crtI and crtY genes of Erwinia uredovora into the Escherichia coli vector pACYC184 and the plasmid pAK916 formed by inserting a fragment containing the crtW gene of Ag~robacterium aurantiacus sp. nov.
MK1 into the Escherichia coli vector pBluescript II SK- were introduced into Escherichia coli JM101 and the resulting E.
coli was cultured to the stationary phase, and then bacterial cells were collected and carotenoid pigments were extracted.
The extracted pigments comprised 94% of canthaxanthin and 6%
of echinenone. Also, canthaxanthin was obtained in a yield of 3 mg starting from 2 liters of the culture solution.
The plasmid pACCARI6ocrtX formed by inserting a fragment containing the crtE, crtB, crtI and crtY genes of Erwinia uredovora into the Escherichia coli vector pACYCl84 and the plasmid pPCl7-3 formed by inserting a fragment containing the crtW gene of Alcalig~enes PC-1 into the Escherichia coli vector pBluescript II SK+ were introduced into Escherichia coli JM101 and the resulting E. coli was cultured to the stationary phase, then bacterial cells were collected and carotenoid pigments were extracted. The extracted pigments comprised 40% of canthaxanthin and 50% of echinenone. The remainder comprised 10% of unreacted ~i-carotene.
<Production of astaxanthin and 4-ketozeaxanthin>
It is possible to produce astaxanthin as a final product and 4-ketozeaxanthin as an intermediate metabolite by introducing into a microorganism such as Escherichia coli or the like and expressing the crtE, crtB, crtI, crtY and crtZ
genes of Erwinia uredovora required for the synthesis of zeaxanthin and any one of the DNA strands of the present invention (10)-(18) which is a keto group-introducing enzyme gene (typically, the crtW gene of Agrobacterium aurantiacus sp. nov. MK1 or Alcalig~enes PC-1). The yields or the ratio of astaxanthin and 4-ketozeoxanthin can be changed by controlling the expression level of the DNA strand (crtW gene) or examining the culturing conditions of a microorganism having the DNA strand.
Two embodiments in Escherichia coli are described below, and more details will be illustrated in Examples.
The plasmid pACCAR25ncrtX formed by inserting a fragment containing the crtE, crtB, crtI, crtY and crtZ genes of Erwinia uredovora into the Escherichia coli vector pACYC184 and the plasmid pAK916 formed by inserting a fragment containing the crtW gene of Aq. aurantiacus sp. nov. MK1 into the Escherichia coli vector pBluescript II SK- were introduced into Escherichia coli JM101 and cultured to the stationary phase, and then bacterial cells were collected and carotenoid pigments were extracted. The yield of the extracted pigments was 1.7 mg of astaxanthin and 1.5 mg of 4-ketozeaxanthin based on 2 liters of the culture solution.
The plasmid pACCAR25ocrtX formed by inserting a fragment containing the crtE, crtB, crtI, crtY and crtZ genes of Erwinia uredovora into the Escherichia coli vector pACYC184 and the plasmid pPCl7-3 formed by inserting a fragment containing the crtW gene of Alcalig~enes PC-1 into the Escherichia coli vector pBluescript II SK+ were introduced into Escherichia coli JM101 and the resulting E. coli was cultured to the stationary phase, and then bacterial cells were collected and carotenoid pigments were extracted. The yield of the extracted pigments was about 1 mg of astaxanthin and 4-ketozeaxanthin, respectively based on 2 liters of the culture solution.
<Production of astaxanthin and phoenicoxanthin>
It is possible to produce astaxanthin as a final product and phoenicoxanthin as an intermediate metabolite by introducing into a microorganism such as Escherichia coli or the like and expressing the crtE, crtB, crtI and crtY genes of Erwinia uredovora required for the synthesis of ~3-carotene, any one of the DNA strands of the present invention (1)-(9) which is a keto group-introducing enzyme gene (typically, the crtw gene of Agrobacterium aurantiacus sp. nov. MK1 or Alcalig~enes PC-1), and any one of the DNA strands of the present invention (19)-(27) which is a hydroxyl group-introducing enzyme gene (typically the crtZ gene of Aq.
aurantiacus sp. nov. MK1 or Alcaligenes PC-1). The yields or the ratio of astaxanthin and phoenicoxanthin can be changed by controlling the expression level of the DNA strand (crtLV and crtZ genes) or examining the culturing conditions of a microorganism having the DNA strands. An embodiment in Escherichia coli is described below, and more details will be illustrated in Examples.
The plasmid pACCARI6ncrtX formed by inserting a fragment containing the crtE, crtB, crtI and crtY genes of Erwinia uredovora into the Escherichia coli vector pACYC184 and the plasmid pAK96K formed by inserting a fragment - 27a -containing the crtW and crtZ genes of Act. aurantiacus sp. nov.
MK1 into the Escherichia coli vector pBluescript II SK- were introduced into Escherichia coli JM101 and the resulting E.
coli was cultured to the stationary phase, and then bacterial cells were collected and carotenoid pigments were extracted.
The yield of the extracted pigments comprised was 3 mg of astaxanthin and 2 mg of phoenicoxanthin starting from 4 liters of the culture solution.
Deposition of microorganisms Microorganisms as the gene sources of the DNA
strands of the present invention and Escherichia coli carrying the isolated genes (the DNA strands of the present invention) have been deposited to National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology.
(i) AQrobacterium aurantiacus sp. nov. MK1 Deposition No: FERM BP-4506 Entrusted Date: December 20, 1993 (ii) Escherichia coli JM101 (pAccrt-EIB, pAK92) Deposition No: FERM BP-4505 Entrusted Date: December 20, 1993 (iii) Alcali eq nes sp. PC-1 Deposition No: FERM BP-4760 Entrusted Date: July 27, 1994 (iv) Escherichia coli Vii- pPCl7 Deposition No: FERM BP-4761 Entrusted Date: July 27, 1994 -2~b- 280024 Examples The present invention is further described more specifically with reference to the following examples without restriction of the invention. In addition, the ordinary experiments of gene manipulation employed herein is based on the standard methods (Sambrook, J., Fritsch, E.F., Maniatis, T., "Molecular Cloning - A Laboratory - 2s - 2~ 80024 Manual" , Cold Spring Harbor Laboratory Press, 1989 ) , unless otherwise specified.
Example 1: Preparation of chromosomal DNA
Chromosomal DNAs were prepared from three marine bacterial strains, i.e. Agrobacterium aurantiacus sp. nov.
MK1, Alcaligenes sp. PC-1, and Alteromonas SD-402 (Yokoyama, A., Izumida, H., Miki, W., "Marine bacteria produced astaxanthin", 10th International Symposium on Carotenoids, Abstract, CL11-3, 1993). After each of these marine bacteria was grown in 200 ml of a culture medium (a culture medium prepared according to the instruction of "Marine Broth" manufactured by DIFCO) at 25yC for 4 days to the stationary phase, the bacterial cells were collected, washed with a TES buffer ( 20 mM Tris, 10 mM EDTA, 0. 1 M
NaCl, pH 8), subjected to heat treatment at 68'C for 15 minutes, and suspended into the solution I (50 mM glucose, mM Tris, 10 mM EDTA, pH 8) containing 5 mg/ml of lysozyme ( manufactured by SEIKAGAKU KOGYO ) and 100 ug/ml of RNase A (manufactured by Sigma). After incubation of the 20 suspension at 37'C for 1 hour, Proteinase K (manufactured by Boehringer-Mannheim) was added and the mixture was incubated at 37'C for 10 minutes. After SARCOSIL (N-lauroylsarcosine Na, manufactured by Sigma) was then added at the final concentration of 1% and the mixture was 25 sufficiently mixed, it was incubated at 37~C for several hours. The mixture was extracted several times with phenol/chloroform, and ethanol in a two-time amount was added slowly. Chromosomal DNA thus deposited was wound around a glass rod, rinsed with 70% ethanol and dissolved in 2 ml of a TE buffer ( 10 mM Tris, 1 mM EDTA, pH 8 ) to prepare a chromosomal DNA solution.
Example 2: Preparation of hosts for a cosmid library (1) Preparation of phytoene-producing Escherichia coli After the removal of the BstEII (1235) - Eco521 (4926) fragment from a plasmid pCARl6 having a carotenoid synthesis gene cluster except the crtZ gene of Erwinia uredovora (Misawa, N.,Nakagawa, M., Kobayashi, K., Yamano, Trade-mark 2'18002, S., Izawa, Y., Nakamura, K., Harashima, K., "Elucidation of the Erwinia uredovora Carotenoid Biosynthetic Pathway by Functional Analysis of Gene Porducts expressed in Escherichia coli", J. Hacteriol., 172, p. 6704-6712, 1990;
and Japanese Patent Publication No. 58786/1991:
"DNA Strands useful for the Synthesis of Carotenoids"), a 2.3 kb Asp718 (KpnI) -EcoRI fragment containing the crtE and crtB genes required for the production of phytoenes was cut out. This fragment was then inserted into the EcoRV site of the E . coli vector pACYC184 to give an aimed plasmid (pACCRT-EB). The bacterium E. coli containing pACCRT-EH exhibits resistance to an antibiotic chloramphenicol (Cmr) and produces phytoenes (Linden, H., Misawa, N., Chamovitz, D., Pecker, I., Hirschberg, J., Sandmann, G., "Functional Complementation in Escherichia coli of Different Phytoene Desaturase Genes and Analysis of Accumulated Carotenes", Z.
Naturforsch., 46c, 1045-1051, 1991).
(2) Preparation of lycopene-producing Escherichia coli After the removal of the BstEII (1235) - SnaBI (3497) fragment from a plasmid pCARl6 having a carotenoid synthesis gene cluster except the crtZ gene of Erwinia uredovora, a 3.75 kb Asp718 (KpnI) - EcoRI fragment containing the crtE, crtl and crtB genes required for the production of lycopene was cut out. This fragment was then inserted into the EcoRV site of the E. coli vector pACYC184 to give an aimed plasmid (pACCRT-EIB). The bacterium E.
coli containing pACCRT-EIH exhibits Cmr and produces lycopene (Cunningham Jr, F.X., Chamovitz, D., Misawa, N., Gatt, E., Hirschberg, J., "Cloning and Functional Expression in Escherichia coli of Cyanobacterial Gene for Lycopene Cyclase, the Enzyme that catalyzes the Biosynthesis of ~i-Carotenes", FEBS Lett., 328, 130-138, 1993).
(3) Preparation of R-carotene-producing Escherichia coli After the crtX gene was inactivated by subjecting a plasmid pCARl6 having a carotenoid synthesis gene cluster except the crtZ gene of Erwinia uredovora to digestion with restriction enzyme BstEII, the Klenow fragment treatment and the ligation reaction, a 6.0 kb Asp718 (KpnI) - EcoRI
fragment containing crtE, crtY, crtI and crtB genes required for the production of ~3-carotene was cut out.
This fragment was then inserted into the EcoRV site of the E. coli vector pACYC184 to give an aimed plasmid (referred to hereinafter as pACCAR160crtX). The bacterium E. coli containing pACCARI6dcrtX exhibits Cmr and produces (3-carotene. In this connection, the restriction enzyme and enzymes used for genetic manipulation have been purchased from TAKARA SHUZO (K. K.) or Boehringer-Mannheim.
Example 3: Preparation of a cosmid library and acquisition of Escherichia coli which exhibits orange in color After the restriction enzyme Sau3AI was added in an amount of one unit to 25 ug of the chromosomal DNA of Agrobacterium aurantiacus sp. nov. MK1, the mixture was incubated at 37'C for 15 minutes and heat treated at 68'C
for 10 minutes to inactivate the restriction enzyme. Under the condition, many partially digested fragments with Sau3AI were obtained at about 40 kb. The cosmid vector pJBB (resistant to ampicillin (Apr)) which had been subjected to BamHI digestion and alkaline phosphatase treatment and the right arm (shorter fragment) of pJBB
which had been digested with Sall/BamHI and then recovered from the gel were mixed with a part of the above Sau3AI
partial fragments, and ligated at l2yC overnight. In this connection, pJBB has been purchased from Amersham.
Phage particles were obtained in an amount sufficient for preparing a cosmid library by the in vitro packaging with a Gigapack Gold* (manufactured by Stratagene; available from Funakoshi) using the DNA above ligated.
After Escherichia coli DH1 (ATCC33849) and Escherichia coli DH1, each of which has one of the three plasmids prepared in Example 2, were infected with the phage Trade-mark particles, these bacteria were diluted so that 100 - 300 colonies were found on a plate, plated on LB containing appropriate antibiotics (lo trypton, 0.5o yeast extract, to NaCl), and cultured at 37~C or room temperature for a period of overnight to several days.
As a result, in cosmid libraries having the simple Escherichia coli (beige) or the phytoene-producing Escherichia coli (beige) with pACCRT-EB as a host, no colonies with changed color were obtained notwithstanding the screening of a ten thousand or more of the colonies for respective libraries. On the other hand, in cosmid libraries having the lycopene-producing Escherichia coli (light red) with pACCRT-EIB or the (3-carotene-producing Escherichia coli (yellow) with pACCARI6~crtX as a host, colonies exhibiting orange have appeared in a proportion of one strain to several hundred colonies, respectively. Most of these transformed Escherichia coli strains which exhibits orange contained plasmid pJB8 in which about 40 kb partially digested Sau3AI .-fragments were cloned. It is also understood from the fact that no colonies with changed color appeared in cosmid libraries having the simple Escherichia coli or the phytoene-producing Escherichia coli with pACCRT-EB as a host, that Escherichia coli having an ability of producing a carotenoid synthetic intermediate of the later steps of at least phytoene should be used as a host for the purpose of expression-cloning the xanthophyll synthesis gene cluster from the chromosomal DNA of Agrobacterium aurantiacus sp. nov. MK1.
Example 4: Localization of a fragment containing an orange pigment synthesis gene cluster When individual several ten colonies out of the orange colonies obtained in cosmid libraries having the lycopene-producing Escherichia coli (light red) with pACCRT-EIB or the ~3-carotene-producing Escherichia coli (yellow) with pACCARI6~crtX as a host were selected to analyze the plasmids, 33 kb - 47 kb fragments partially digested with Sau3AI were inserted in vector pJB8 in all of the colonies 2 l 80024 except one strain. The remaining one strain (lycopene-producing Escherichia coli as a host) contains a plasmid, in which a 3.9 kb fragment partially digested with Sau3AI
was inserted in pJBB (referred to hereinafter as plasmid pAK9). This was considered to be the one formed by the in vivo deletion of the inserted fragment after the infection to Escherichia coli. The same pigment (identified as astaxanthin in Example 6) as that in the orange colonies obtained from the other cosmid libraries was successfully synthesized with the lycopene-producing Escherichia coli having pAK9, pAK9 was used as a material in the following analyses.
Example 5: Determination of the nucleotide sequence in the orange pigment synthesis gene cluster A 3.9 kb EcoRI inserted fragment prepared from pAK9 was inserted into the EcoRI site of the Escherichia coli vector pBluescrip II SK+ to give two plasmids (pAK91 and pAK92) with the opposite directions of the fragment to the vector.
The restriction enzyme map.af one of the plasmids (pAK92) is illustrated in Fig. 12. When pAK92 was introduced into the lycopene-producing Escherichia coli, orange colonies were obtained as a result of the synthesis of astaxanthin (Example 6). However, no ability for synthesizing new pigments was afforded even if pAK91 was introduced into the lycopene-producing Escherichia coli. It was thus considered that the pigment synthesis gene cluster in the plasmid pAK92 has the same direction as that of the lac promoter of the vector. Next, each of a 2.7 kb PstI
fragment obtained by the PstI digestion of pAK9l, a 2.9 kb BamHI fragment obtained by the BamHI digestion of pAK92, and 2.3 kb and 1.6 kb Sall fragments obtained by the SalI
digestion of pAK92 was cloned into the vector pBluescrip II
SK-. The restriction maps of plasmids referred to as pAK94, pAK96, pAK98, pAK910, pAK93, and pAK95 are illustrated in Fig. 12. The plasmids pAK94, pAK96, pAK98 and pAK910 have the pigment synthesis gene cluster in the same direction as that of the lac promoter of the vector, ~ ~ 80024 while the plasmids pAK93 and pAK95 have the pigment synthesis gene cluster in the opposite direction to that of the promoter.
It was found that when the plasmid pAK96 having a 2.9 kb BamHI fragment was introduced into the lycopene producing Escherichia coli, the transformant also synthesized astaxanthin as in the case when the plasmid pAK92 having a 3.9 kb EcoRI fragment was introduced (Example 6), so that the DNA sequence of the 2.9 kb BamHI
fragment was determined.
The DNA sequence was determined by preparing deletion mutants of the 2.9 kb BamHI fragment from the normal and opposite directions and determining the sequence using clones having various lengths of deletions. The deletion mutants were prepared from the four plasmids pAK96, pAK98, pAK93 and pAK95 according to the following procedure: Each of the plasmids, 10 ug, was decomposed with SacI and XbaI
and extracted with phenol/chloroform to recover DNA by ethanol precipitation. Each of DNA was dissolved in 100 ul of ExoIII buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM MgClz, 10 mM 2-mercaptoethanol, pH 8.0), 180 units of ExoIII
nuclease was added, and the mixture was maintained at 37VC.
A 10 ul portion was sampled at every 1 minute, and two samples were transferred into a tube in which 20 ul of MB
buffer (40 mM sodium acetate, 100 mM NaCl, 2 mM ZnClz, 10%
glycerol, pH 4.5) is contained and which is placed on ice.
After completion of the sampling, five tubes thus obtained were maintained at 65°C for 10 minutes to inactivate the enzyme, five units of mung bean nuclease were added, and the mixtures were maintained at 37'C for 30 minutes. After the reaction, five DNA fragments different from each other in the degrees of deletion were recovered for each plasmid by agarose gel electrophoresis. The DNA fragments thus recovered was blunt ended with the Klenow fragment, subjected to the ligation reaction at 16~C overnight, and Escherichia coli JM109 was transformed. A single stranded DNA was prepared from each of various clones thus obtained with a helper phage M13K07, and subjected to the sequence reaction with a fluorescent primer cycle-sequence kit available from Applied Biosystem (K.K.), and the DNA
sequence was determined with an automatic sequencer.
The DNA sequence comprising 2886 base pairs (bp) thus obtained is illustrated in Figs. 5 - 9 (SEQ ID NO: 4). As a result of examining an open reading frame having a ribosome binding site in front of the initiation codon, three open reading frames which can encode the corresponding proteins (A - B (nucleotide positions 229 -864 of SEQ ID N0: 4 ) , C - D ( nucleotide positions 864 -1349), E - F (nucleotide positions 1349 - 2506) in Figs. 5 - 9) were found at the positions where the three xanthophyll synthesis genes crtW, crtZ and crtY are expected to be present. For the two open reading frames of A - B and E - F, the initiating codon is GTG, and for the remaining open reading frame C - D, it is ATG.
Example 6: Identification of the orange pigment The lycopene-producing Escherichia coli JM101 having pAK92 or pAK96 introduced thereinto (Escherichia coli (pACCRT-EIB, pAK92 or pAK96); exhibiting orange) or the (3-carotene-producing Escherichia coli JM101 having pAK94 or pAK96K (Fig. 12) introduced thereinto (Escherichia coli (pACCAR160crtX, pAK94 or pAK96K); exhibiting orange) was cultured in 4 liters of a 2YT culture medium ( 1. 6 o trypton, 1% yeast extract, 0.5% NaCl) containing 150 ~ag/ml of ampicillin (Ap, manufactured by Meiji Seika) and 30 ug/ml of chloramphenicol ( Cm, manufactured by Sankyo ) at 37yC for 18 hours. Bacterial cells collected from the culture solution was extracted with 600 ml of acetone, concentrated, extracted twice with 400 ml of chloroform/methanol (9/1), and concentrated to dryness.
Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with - 35 - 2 ~ a 0 0 2 chloroform/methanol (15/1). The original orange pigment was separated into three spots at the Rf values of 0.72, 0.82 and 0.91 by TLC. The pigment of the darkest spot at Rf 0.72 corresponding to 50% of the total amount of orange pigment and the pigment of secondly darker spot at Rf 0.82 were scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or methanol, and chromatographed on a Sephadex*LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or methanol to give purified materials in a yield of 3 mg (Rf 0.72) and 2 mg (Rf 0.82), respectively.
It has been elucidated from the results of the UV-visible, 1H-NMR and FD-MS (m/e 596) spectra that the pigment at Rf 0.72 has the same planar structure as that of astaxanthin. When the pigment was dissolved in diethyl ether . 2-propanol . ethanol (5 . 5 . 2) to measure the CD
spectrum, it was proved to have stereochemical configuration of 3S, 3'S, and thus identified as astaxanthin; see Fig. 11 for the structural formula).
Also, the pigment at Rf 0.82 was identified as phoenicoxanthin (see Fig. 11 for the structural formula) from the results of its UV-visible, 1H-NMR and FD-MS (m/e 580) spectra. In addition, the pigment at 0.91 was canthaxanthin (Example 7(2)).
Example 7: Identification of metabolic intermediates of xanthophyll (1) Identification of 4-ketozeaxanthin The zeaxanthin producing Escherichia coli was prepared according to the following procedure. That is to say, the plasmid pCAR25 having total carotenoid synthesis gene cluster of Er. uredorora (Misawa, N., Nakagawa, M., Kobayashi, K., Yamano, S., Izawa, Y., Nakamura, K., Harashima, K., "Elucidation of the Erwinia uredovora Carotenoid Biosynthetic Pathway by Functional Analysis of Gene Products expressed in Escherichia coli", J.
Bacteriol., 172, p. 6704-6712, 1990; and Japanese Patent Trade-mark --~ - 36 - 2~ soa2~
Publication No. 58786/1991~
"DNA Strands useful for the Synthesis of Carotenoids") was digested with restriction enzyme BstEII, and subjected to the Klenow fragment treatment and ligation reation to inactivate the crtX gene by reading frame shift, and then a 6 . 5 kb Asp718 ( KpnI ) - EcoRI fragment containing the crtE, crtY, crtI, crtB and crtZ genes required for producing zeaxanthin was cut out. This fragment was then inserted into the EcoRV site of the Escherichia coli vector pACYC184 to give the aimed plasmid ( referred to hereinafter as pACCAR250crtX).
The zeaxanthin-producing Escherichia coli JM101 having pAK910 or pAK916 (Fig. 12) introduced thereinto (Escherichia coli (pACCAR25dcrtX, pAK910 or pAK916);
exhibiting orange) was cultured in 2 liters of a 2YT
culture medium containing 150 ug/ml of Ap and 30 ug/ml of Cm at 37VC for 18 hours. Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, extracted twice with 200 ml of chloroform/methanol (9/1), and concentrated to dryness.
Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with chloroform/methanol (15/1). The original orange pigment was separated into three spots at the Rf values of 0.54 (46%), 0.72 (53%) and 0.91 (lo) by TLC. The pigment at Rf 0.54 was scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or methanol, and chromatographed on a Sephadex LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or methanol to give a purified material in a yield of 1.5 mg.
This material was identified as 4-ketozeaxanthin (see Fig. 11 for the structural formula) since its W-visible spectrum, FD-MS spectrum (m/e 582) and mobility in silica gel TLC (developed with chloroform/methanol (15/1)) accorded perfectly with those of the standard sample of 4-2 '~ 8 0 0 2 ~4 ketozeaxanthin (purified from Agrobacterium aurantiacus sp.
nov. MK1; Japanese Patent Publication No. 165684/1994). In addition, the pigments at Rf 0.72 and 0.91 are astaxanthin (Example 6) and canthaxanthin (Example 7 (2)), respectively.
(2) Identification of canthaxanthin The (3-carotene producing Escherichia coli JM101 having pAK910 or pAK916 introduced thereinto (Escherichia coli (pACCAR16c1crtX, pAK910 or pAK916); exhibiting orange) was cultured in 2 liters of a 2YT culture medium containing 150 ug/ml of Ap and 30 ug/ml of Cm at 37'C for 18 hours.
Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, extracted twice with 200 ml of chloroform/methanol (9/1), and concentrated to dryness. Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with chloroform/methanol (50/1). The pigment of the darkest spot corresponding to 940 of the total amount of orange pigments was scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or chloroform/methanol (1/1), and chromatographed on a Sephadex LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or chloroform/methanol (1/1) to give a purified material in a yield of 3 mg.
This material was identified as canthaxanthin ( see Fig.
11 for the structural formula) since its W-visible, 1H-NMR, FD-MS (m/e 564) spectra and mobility in silica gel TLC
(Rf 0.53 on developing with chloroform/methanol (50/1)) accorded perfectly with those of the standard sample of canthaxanthin (manufactured by BASF). In addition, the pigment corresponding to 60 of the total orange pigments found in the initial extract was considered echinenone ( see Fig. 11 for the structural formula ) on the basis of its W-visible spectrum, mobility in silica gel TLC (Rf 0.78 on Trade-mark _. - 38 - 21 8002 -developing with chloroform/methanol (50/1)), and mobility in HPLC with NOVA PACK HR 6p C18 (3.9 x 300 mm;
manufactured by Waters) (RT 16 minutes on developing at a flow rate of 1.0 ml/min with acetonitrile/methanol/2 propanol (90/6/4)).
(3) Identification of zeaxanthin The ~i-carotene-producing Escherichia coli JM101 having pAK96NK introduced thereinto (Escherichia coli (pACCARI6~crtX, pAK96NK); exhibiting yellow) was cultured in 2 liters of a 2YT culture medium containing 150 ug/ml of Ap and 30 ug/ml of Cm at 37yC for 18 hours. Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, extracted twice with 200 ml of chloroform/methanol (9/1), and concentrated to dryness. Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with chloroform/methanol (9/1). The pigment of the darkest spot corresponding to 870 of the total amount of yellow pigments was scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or methanol, and chromatographed on a Sephadex LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or methanol to give a purified material in a yield of 3 mg.
It has been elucidated that this material has the same planar structure as that of zeaxanthin since its W-visible, 1H-NMR, FD-MS (m/e 568) spectra and mobility in silica gel TLC (Rf 0.59 on developing with chloroform/methanol (9/1)) accorded perfectly with those of the standard sample of zeaxanthin (manufactured by BASF).
When the pigment was dissolved in diethyl ether . 2-propanol . ethanol (5 . 5 . 2) to measure the CD spectrum, it was proved to have a stereochemical configuration of 3R, 3'R, and thus identified as zeaxanthin (see Fig. 11 for the structural formula). Also, the pigment corresponding to Trade-mark 2~80C24 13% of the total yellow pigments found in the initial extract was considered (3-cryptoxanthin ( see Fig. 11 for the structural formula) on the basis of its UV-visible spectrum, mobility in silica gel TLC ( Rf 0. 80 on developing with chloroform/methanol (9/1)), and mobility in HPLC with NOVA PACK HR 6u C18 (3.9 x 300 mm; manufactured by Waters) (RT 19 minutes on developing at a flow rate of 1.0 ml/min with acetonitrile/methanol/2-propanol (90/6/4)).
(4) Identification of ~i-carotene The lycopene-producing Escherichia coli JM101 having pAK98 introduced thereinto (Escherichia coli (pACCRT-EIB, pAK98); exhibiting yellow) was cultured in 2 liters of a 2YT culture medium containing 150 ~ag/ml of Ap and 30 ug/ml of Cm at 37°C for 18 hours. Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, and extracted twice with 200 ml of hexane.
The hexane layer was concentrated and chromatographed on a silica gel column (15 x 300 mm) with an eluent of hexane/ethyl acetate (50/1-) to give 3 mg of a purified material.
The material was identified as (3-carotene (see Fig. 11 for the structural formula), since all of the data of its UV-visible, FD-MS spectrum (m/e 536) and mobility in HPLC
with NOVA PACK HR 6u C18 ( 3. 9 x 300 mm; manufactured by Waters) (RT 62 minutes on developing at a flow rate of 1.0 ml/min with acetonitrile/methanol/2-propanol (90/6/4)) accorded with those of the standard sample of (3-carotene (all trans type; manufactured by Sigma).
Example 8: Identification of xanthophyll synthesis gene cluster (1) Identification of a keto group-introducing enzyme gene It is apparent from the results of Example 6 that among the 3 . 9 kb fragment contained in pAK9 ( Example 4 ) or pAK92, all of the genes required for the synthesis of astaxanthin from lycopene is contained in the 2.9 kb BamHI fragment at the right side (pAK96, Fig. 12). Thus, the 1.0 kb fragment at the left side is not needed. Unique NcoI and KpnI sites are present within the 2.9 kb BamHI fragment of pAK96. It is found from the results of Example 7 (3) that the 1.4 kb fragment (pAK96NK) between the NcoI and KpnI sites has a hydroxyl group-introducing enzyme activity but has no keto group-introducing enzyme activity. Canthaxanthin can also be synthesized from (3-carotene with the 2.9 kb BamHI
fragment from which a fragment of the right side from unique SalI site between the NcoI and KpnI sites had been removed ( pAK910 ) or with the 2 . 9 kb BamHI fragment from which a fragment of the right side from the HincII site positioned at the left side of the SalI site had been removed (pAK916), but activity for. synthesizing canthaxanthin from ~i-carotene disappeared in the 2.9 kb BamHI fragment of pAK96 from which a fragment of the right side from the Ncol site left of the HincII site had been removed. On the other hand, even if a fragment of the left side from unique BglII site which is present leftward within the 0.9 kb BamHI - HincII fragment of pAK916 was removed, similar activity_to that of the aforementioned BamHI - HincII fragment (pAK916) was observed. It is thus considered that a gene encoding a keto group-introducing enzyme having an enzyme activity for synthesizing canthaxanthin from ~i-carotene as a substrate is present within the 0.74 kb BglII - HincII fragment of pAK916, and the aforementioned NcoI site is present within this gene.
As a result of determining the nucleotide sequence, an open reading frame which corresponds to the gene and has a ribosome binding site just in front of the initiation codon was successfully detected, it was then designated as the crtW gene. The nucleotide sequence of the crtW gene and the encoded amino acid sequence are illustrated in Fig. 1 (SEQ ID N0: 1).
The crtW gene product (CrtW) of Agrobacterium aurantiacus sp. nov. MK1 has an enzyme activity for converting a methylene group at the 4-position of a (3 ionone ring into a keto group, and one of the specific examples is an enzyme activity for synthesizing canthaxanthin from ~i-carotene as a substrate by way of echinenone (Example 7 (2); see Fig. 11). Furthermore, the crtW gene product also has an enzyme activity for converting a methylene group at the 4-position of a 3-hydroxy-(3-ionone ring into a keto group, and one of the specific examples is an enzyme activity for synthesizing astaxanthin from zeaxanthin as a substrate by way of 4-ketozeaxanthin (Example 7 (1); see Fig. 11). In addition, polypeptides having such enzyme activities and DNA strands encoding these polypeptides have not hitherto been known, and the polypeptides and the DNA strands encoding these polypeptides have no overall homology to any polypeptides or DNA strands having been hitherto known. Also, no such informations have hitherto been described that a methylene group of not only a ~3-ionone ring and a 3-hydroxy-~i-ionone ring but also the other compounds is directly converted into a keto group with an enzyme.
(2) Identification of a hydroxyl group-introducing enzyme gene -Unique SalI site is present within the 2.9 kb BamHI
fragment of pAK96. When the 2.9 kb BamHI fragment is cut into two fragments at the Sall site, these two fragments (pAK910 and pAK98) have no hydroxyl group-introducing activity. That is to say, the left fragment (pAK910) has only a keto group-introducing enzyme activity (Example 7 (2)), and the right fragment (pAK98) has only a lycopene-cyclizing enzyme activity (Example 7 (4)). On the other hand, when a 1.4 kb Ncol - KpnI fragment (pAK96NK) containing the aforementioned Sall site is introduced into a ~i-carotene-producing Escherichia coli, zeaxanthin is synthesized by way of (3-cryptoxanthin (Example 7 (3)). It is thus considered that a gene encoding a hydroxyl group-introducing enzyme which has an enzyme activity for synthesizing zeaxanthin from ~i-carotene as a substrate is present within the 1.4 kb Ncol - Kpnl fragment of pAK96NK, and the aforementioned SalI site is present within this gene. As a result of determining the nucleotide sequence, an open reading frame which corresponds to the gene and has a ribosome binding site just in front of the initiation codon was successfully detected, it was then referred to as the crtZ gene. The nucleotide sequence of the crtZ gene and the encoded amino acid sequence are illustrated in Fig .
2 (SEQ ID N0: 2).
The crtZ gene product (CrtZ) of Agrobacterium aurantiacus sp. nov. MK1 has an enzyme activity for adding a hydroxyl group to the 3-carbon of a ~i-ionone ring, and one of the specific examples is an enzyme activity for synthesizing zeaxanthin from ~i-carotene as a substrate by way of (3-cryptoxanthin (Example 7 (3); see Fig. 11).
Furthermore, the crtZ gene product also has an enzyme activity for adding a hydroxyl group to the 3-carbon of a 4-keto-(3-ionone ring, and one of the specific examples is an enzyme activity for synthesizing astaxanthin from canthaxanthin as a substrate by way of phoenicoxanthin ( Example 6; see Fig. 11 ) . In addition, polypeptides having the latter enzyme activity. and DNA strands encoding these polypeptides have not hitherto been known. Also, the CrtZ
of Agrobacterium showed significant homology to the CrtZ of Erwinia uredovora (identity of 57%) at the level of amino acid sequence.
(3) Identification of a lycopene cyclase gene Astaxanthin can be synthesized from (3-carotene with the 2.9 kb BamHI fragment from which a fragment of the right side from a KpnI site had been removed ( pAK96K ) or with the 2.9 kb BamHI fragment from which a fragment right from the Pstl site which is placed further right of the KpnI site had been removed (pAK94) (Example 6), but astaxanthin cannot be synthesized from lycopene. On the other hand, when a 1. 6 kb SalI fragment ( pAK98 ) , which contains a right fragment from unique SalI site present further left than the aforementioned KpnI site within the 2.9 kb BamHI
fragment, was introduced into lycopene-producing Escherichia coli, (3-carotene was synthesized (Example 7 (4)). It is thus considered that a gene encoding lycopene cyclase that has an enzyme activity for synthesizing [i-carotene from lycopene as a substrate is present within the 1.6 kb SalI fragment of pAK98, and this gene is present over a range of the Kpnl site and the PstI site. As a result of determining the nucleotide sequence, an open reading frame which corresponds to the gene and has a ribosome binding site just in front of the initiation codon was successfully detected, it was then referred to as the crtY gene. The nucleotide sequence of the crtY gene and the amino acid sequence to be encoded are illustrated in Figs. 3 - 4 (SEQ ID N0: 3).
The crtY gene product (CrtY) of Agrobacterium aurantiacus sp. nov. MKl has significant homology to the CrtY of Erwinia uredovora (identity of 44.3%) at the level of amino acid sequence, and the functions of both enzymes are the same.
Example 9: Southern blotting analysis with the chromosomal DNA of the other marine bacteria Examination was conducted whether a region exhibiting homology with the isolated crtW and crtZ is obtained from a chromosomal DNAs of the other marine microorganisms. The chromosomal DNAs of Alcaligenes sp. PC-1 and Alteromonas sp. SD-402 prepared in Example 1 were digested with restriction enzymes BamHI and PstI, and separated by agarose gel electrophoresis. All of the DNA fragments thus separated were denaturated with an alkali solution of 0.5 N NaOH and 1.5 M NaCl, and transferred on a nylon membrane filter over an overnight period. The nylon membrane filter on which DNAs had been adsorbed was dipped in a hybridization solution (6 x Denhardt, 5 x SSC, 100 ug/ml ssDNA), and pre-hybridization was conducted at 60JC for 2 hours. Next, the 1.5 kb DNA fragment cut out from pAK96K
with Ball, which contains crtW and crtY, was labelled with a Mega primeTM DNA labelling systems (Amersham) and [a-32P]dCTP (~110TBq/mmol) and added to the aforementioned prehybridization solution to conduct hybridization at 60vC

?_ ~ 80024 for 16 hours.
After hybridization, the filter was washed with 2 x SSC
containing 0.1% SDS at 60vC for 1 hour, and subjected to the detection of signals showing homology by autoradiography. As a result, strong signals were obtained at about 13 kb in the product digested with BamHI and at 2.35 kb in the product digested with PstI in the case of Alcaligenes sp. PC-1, and strong signals were obtained at about 5.6 kb in the product digested with BamHI and at 20 kb or more in the product digested with PstI in the case of Alteromonas sp. SD-4.
Example 10: Acquisition of a xanthophyll synthesis gene cluster from the other marine bacterium As it was found from the results of Example 9 that the PstI digest of the chromosomal DNA of Alcaligenes sp. PC-1 has a region of about 2.35 kb hybridizing with a DNA
fragment containing the crtW and crtZ genes of Agrobacterium aurantiacus sp. nov. MK1, the chromosomal DNA
of Alcaligenes was digested with Pstl, and then DNA
fragments of 2 - 3.5 kb in size was recovered by agarose gel electrophoresis. The DNA fragments thus collected were inserted into the Pstl site of a vector pBluescript II SK+, and introduced into Escherichia coli DHSa to prepare a partial library of Alcaligenes. When the partial library was subjected to colony hybridization with a 1.5 kb DNA
fragment containing the crtW and crtZ genes of Agrobacterium as a probe, a positive colony was isolated from about 5,000 colonies. In this case, colony hybridization was conducted under the same condition as in the Southern blotting analysis shown in Example 9. When plasmid DNA was isolated from the colony thus obtained, and digested with Pstl to examine the size of the integrated DNA fragments, it was found that the plasmid contained three different fragments. Thus, a 2.35 kb fragment to be hybridized was selected from the three different DNA
fragments by the Southern blotting analysis described in Example 9, the 2.35 kb PstI fragment was recovered by e.. 2180024 agarose gel electrophoresis and inserted again into the PstI site of pBluescript II SK+ to prepare the plasmids pPCll and pPCl2. In pPCll and pPCl2, the aforementioned 2.35 kb PstI fragment was inserted into the Pstl site of pBluescript II SK+ in an opposite direction to each other.
The restriction enzyme map of pPCll is illustrated in Fig.
19.
Example 11: Determination of nucleotide sequence of xanthophyll synthesis gene cluster in Alcaligenes When each of pPCl1 and pPCl2 was introduced into (3-carotene-producing Escherichia coli, orange colonies were obtained due to the synthesis of astaxanthin (Example 12) in the former, but no other pigments were newly synthesized in the latter. It was thus considered that the direction of the astaxanthin synthesis gene cluster in the plasmid pPCll was the same as that of the vector lac promoter. It was also found that pPCll contained no lycopene cyclizing enzyme genes, since no other pigments were newly produced even if pPCll was introduced into the lycopene-producing Escherichia coli.
It was found that even if a plasmid having a 0.72 kb BstEII - EcoRV fragment positioned at the right side of the PstI fragment had been removed (referred to as pPCl7, Fig.
19) was introduced into the ~i-carotene-producing Escherichia coli, the transformant of Escherichia coli synthesized astaxanthin and the like (Example 12), same as in the case of E. coli into which pPCll was introduced, so that the nucleotide sequence of the 1.63 kb PstI - BstEII
fragment in pPCl7 was determined.
Deletion mutants were prepared with pPCl7 and pPCl2 according to the following procedure. A 10 ug portion of each of pPCl7 and pPCl2 was digested with KpnI and HindIII
or KpnI and EcoRI, extracted with phenol/chloroform, and DNA was recovered by precipitation with ethanol. Each of DNAs was dissolved in 100 ul of ExoIII buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM MgCl~, 10 mM 2-mercaptoethanol, pH

21$0024 8.0), 180 units of ExoIII nuclease was added, and the mixture was maintained at 37JC. A 10 ul portion was sampled at every 1 minute, and two samples were transferred into a tube in which 20 ul of an MB buffer (40 mM sodium acetate, 100 mM NaCl, 2 mM ZnCl2, loo glycerol, pH 4.5) is contained and which is placed on ice. After~completion of the sampling, five tubes thus obtained were maintained at 65~C for 10 minutes to inactivate the enzyme, five units of mung bean nuclease were added, and the mixture was maintained at 37~C for 30 minutes. After the reaction, ten DNA fragments different from each other in the degrees of deletion were recovered for each plasmid by agarose gel electrophoresis. The DNA fragments thus recovered were blunt ended with the Klenow fragment, subjected to the ligation reaction at l6eC overnight, and Escherichia coli JM109 was transformed. A single stranded DNA was prepared from each of various clones thus obtained with a helper phage M13K07, and subjected to the sequence reaction with a fluorescent primer cycle-sequence kit available from Applied Biosystem (K.K.), and the DNA sequence was determined with an automatic sequencer.
The DNA sequence comprising 1631 base pairs (bp) thus obtained is illustrated in Figs. 16 - 18 (SEQ ID N0: 7).
As a result of examining an open reading frame having a ribosome binding site in front of the initiating codon, two open reading frames which can encode the corresponding proteins (A - B (nucleotide positions 99 - 824 of SEQ ID
NO: 7 ) , C - D ( nucleotide positions 824 - 1309 ) in Figs . 16 - 18 were found at the positions where the two xanthophyll synthesis genes crtW and crtZ were expected to be present.
Example 12: Identification of pigments produced by Escherichia coli having an Alcaligenes xanthophyll synthesis gene cluster (1) Identification of astaxanthin and 4-ketozeaxanthin A deletion plasmid ( having only crtW ) having a deletion from the right BstEI I to the nucleotide position 1162 ( Fig .

.. 2180024 17) (nucleotide position 1162 of SEQ ID NO: 7) among the deletion plasmids from pPCl7 prepared in Example 11 was referred to as pPCl7-3 (Fig. 19).
The zeaxanthin-producing Escherichia coli JM101 (Example 7 (1)) having pPCl7-3 introduced thereinto (Escherichia coli (pACCAR25~crtX, pPCl7-3); exhibiting orange) was cultured in 2 liters of 2YT culture medium containing 150 ug/ml of Ap and 30 ug/ml of Cm at 37vC for 18 hours. Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, extracted twice with 200 ml of chloroform/methanol (9/1), and concentrated to dryness.
Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with chloroform/methanol (15/1). The original orange pigment was separated into three spots at the Rf values of 0.54 ( ca. 25 0 ) , 0. 72 ( ca. 30 0 ) and 0 . 91 ( ca. 25% ) . The pigments at the Rf values of 0.54 and 0.72 were scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or methanol, and chromatographed on a Sephadex LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or methanol to give purified materials in a yield of about 1 mg, respectively.
The materials were identified as 4-ketozeaxanthin (Rf 0.54) and astaxanthin (Rf 0.72), since all of the data of their UV-visible, FD-MS spectra and mobility in TLC
(developed with chloroform/methanol (15/1)) accorded with those of the standard samples of 4-ketozeaxanthin and astaxanthin. In addition, the pigment at the Rf value of 0.91 was canthaxanthin (Example 12 (2)).
It was also confirmed by the similar analytical procedures that the ~i-carotene-producing Escherichia coli JM101 having pPCll or pPCl7 introduced thereinto (Escherichia coli (pACCARI6~crtX, pPCll or pPCl7) (exhibiting orange) produces astaxanthin, 4-ketozeaxanthin 48 - 2180~2~
and canthaxanthin. Furthermore, it was also confirmed with the authentic sample of phoenicoxanthin obtained in Example 6 that these E. coli transformants produce a trace amount of phoenicoxanthin.
(2) Identification of canthaxanthin The ~i-carotene-producing Escherichia coli JM101 having pPCl7-3 introduced thereinto (Escherichia coli (pACCARI6dcrtX, pPCl7-3); exhibiting orange) was cultured in 2 liters of 2YT culture medium containing 150 ug/ml of Ap and 30 ug/ml of Cm at 37yC for 18 hours. Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, extracted twice with 200 ml of chloroform/methanol (9/1), and concentrated to dryness. Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with chloroform/methanol (50/1). The darkest pigment corresponding to 40% of the total amount of orange pigments was scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or chloroform/methanol (1/1), and chromatographed on a Sephadex LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or chloroform/methanol (1/1) to give a purified material in a yield of 2 mg.
The material was identified as canthaxanthin, since all of the data of its UV-visible, FD-MS (m/e 564) spectra and mobility in TLC (developed with chloroform/methanol (50/1)) accorded with those of the standard sample of canthaxanthin (manufactured by HASF). In addition, the pigment of which amount corresponds to 50% of the total amount of the orange pigments observed in the initial extract was considered to be echinenone from its UV-visible spectrum, mobility in silica gel TLC (developed with chloroform/methanol (50/1)), and mobility in HPLC with NOVA PACK*HR 6u C18 (3.9 x 300 nun; manufactured by Waters) (developed with acetonitrile/methanol/2-propanol (90/6/4)) (Example 7 (2)).
Trade-mark . 2180024 In addition, the balance of the extracted pigments, 10%, was unreacted ~i-carotene.
(3) Identification of zeaxanthin A plasmid having a 1.15 kb SalI fragment within pPCll inserted in the same direction as the plasmid pPCll into the SalI site of pBluescript II SK+ was prepared (referred to as pPCl3, see Fig. 19).
The (3-carotene-producing Escherichia coli JM101 having pPCl3 introduced thereinto (Escherichia coli (pACCAR160crtX, pPCl3); exhibiting yellow) was cultured in 2 liters of 2YT culture medium containing 150 ug/ml of Ap and 30 ug/ml of Cm at 37yC for 18 hours. Bacterial cells collected from the culture solution was extracted with 300 ml of acetone, concentrated, extracted twice with 200 ml of chloroform/methanol (9/1), and concentrated to dryness.
Then, thin layer chromatography (TLC) was conducted by dissolving the residue in a small amount of chloroform/methanol (9/1) and developing on a silica gel plate for preparative TLC manufactured by Merck with chloroform/methanol (9/1). The darkest pigment corresponding to 90 0 of the total amount of orange pigments was scratched off from the TLC plate, dissolved in a small amount of chloroform/methanol (9/1) or methanol, and chromatographed on a Sephadex LH-20 column (15 x 300 mm) with an eluent of chloroform/methanol (9/1) or methanol to give a purified material in a yield of 3 mg.
The material was identified as zeaxanthin, since all of the data of its UV-visible, FD-MS (m/e 568) spectra and mobility in TLC (developed with chloroform/methanol (9/1)) accorded with those of the standard sample of zeaxanthin (Example 7 (3)). In addition, the pigment of which amount corresponds to l00 of the total amount of the orange pigments observed in the initial extract was considered to be (3-cryptoxanthin from its UV-visible spectrum, mobility in silica gel TLC (developed with chloroform/methanol (9/1)), and mobility in HPLC with NOVA PACK HR 6u C18 (3.9 x 300 mm; manufactured by Waters) (developed with acetonitrile/methanol/2-propanol (90/6/4)) (Example 7 (3)).
Example 13: Identification of the Alcaligenes xanthophyll synthesis gene cluster (1) Identification of a keto group-introducing enzyme gene It is apparent from the results of Examples 11 and 12 (1) that all of the genes required for the synthesis of astaxanthin from ~i-carotene among the 2.35 kb Pstl fragment contained in pPCll is contained in the 1.63 kb PstI
BstEII fragment (pPCl7, Fig. 19) in the left side. Thus, the 0.72 kb HstEII - PstI fragment in the right side is not needed. Unique SmaI and SalI sites are present within the 1.63 kb PstI - BstEII fragment of pPCl7 (Fig. 19). It is confirmed by the pigment analysis with a ~i-carotene-producing Escherichia coli having the deletion plasmids introduced thereinto that the keto group-introducing enzyme activity was lost when the 0.65 kb and 0.69 kb fragments at the left side from SmaI and SalI sites were removed. It.
was also confirmed by the pigment analysis with a (3-carotene-producing Escheri_chia coli having the plasmid introduced thereinto that the plasmid having a 0.69 kb PstI
- SalI fragment positioned at the left side of the 1.63 kb PstI - BstEII fragment inserted into the PstI - SalI site of pBluescript SK+ has no keto group-introducing enzyme activity. On the other hand, the deletion plasmid pPCl7-3 (Fig. 19) in which deletion from the BstEII end at the right end to the nucleotide No. 1162 (nucleotide position 1162 in SEQ ID NO: 7 ) occurred has a keto group-introducing enzyme activity (Example 12 (1), (2)), so that it is considered a gene encoding a keto group-introducing enzyme having an enzyme activity for synthesizing canthaxanthin or astaxanthin with a substrate of (3-carotene or zeaxanthin is present in the 1162 by fragment in pPCl7-3, and the aforementioned SmaI and SalI sites are present within this gene. As a result of determining the nucleotide sequence, an open reading frame which corresponds to the gene and has a ribosome binding site just in front of the initiation codon was successfully detected, so that it was referred to as the crtW gene. The nucleotide sequence of the crtW gene and the encoded amino acid sequence are illustrated in Figs. 13 - 14 (SEQ ID NO: 5).
The crtW gene product (CrtW) of Alcaligenes sp. PC-1 has an enzyme activity for converting a methylene group at the 4-position of a ~3-ionone ring into a keto group, and one of the specific examples is an enzyme .activity for synthesizing canthaxanthin from (3-carotene as a substrate by way of echinenone (Example 12 (2); see Fig. 11).
Furthermore, the crtW gene product also has an enzyme activity for converting a methylene group at the 4-position of a 3-hydroxy-~i-ionone ring into a keto group, and one of the specific examples is an enzyme activity for synthesizing astaxanthin from zeaxanthin as a substrate by way of 4-ketozeaxanthin (Example 12 (1); see Fig. 11). In addition, polypeptides having such enzyme activities and DNA strands encoding these polypeptides have not hitherto been known, and the polypeptides and the DNA strands encoding these polypeptides have no total homology to any polypeptides or DNA strands having been hitherto known.
Also, the crtW gene products (CrtW) of Agrobacterium aurantiacus sp. nov. MK1 and Alcaligenes sp. PC-1 share high homology (identity of 83%) at the level of amino acid sequence, and the functions of both enzymes are the same.
The amino acid sequence in the region of 17o having no identity among these amino acid sequences is considered not so significant to the functions of the enzyme. It is thus considered particularly in this region that a little amount of substitution by the other amino acids, deletion, or addition of the other amino acids will not afftect the enzyme activity.
It can be said the keto group-introducing enzyme gene crtW of marine bacteria encodes the ~i-ionone or 3-hydroxy-~i-ionone ring ketolase which converts directly the methylene group at the 4-position into a keto group irrelative to whether a hydroxyl group is added to the 3-position or not. In addition, no such informations have 218fl024 hitherto been described that a methylene group of not only a a-ionone ring and a 3-hydroxy-~i-ionone ring but also the other compounds is directly converted into a keto group with one enzyme.
(2) Identification of a hydroxyl group-introducing enzyme gene All of the genes rerquired for the synthesis of astaxanthin from ~i-carotene is contained in the 1.63 kb PstI - BstEII fragment (Fig. 19) of pPCl7. One SalI site is present within the 1.63 kb PstI - BstEII fragment of pPCl7. It is apparent from the results of Example 12 (3) that a hydroxyl group-introducing enzyme activity is present in a fragment at the right side from the SalI site.
It is thus understood that the hydroxyl group-introducing enzyme activity is present in the 0.94 kb SalI - BstEII
fragment which is the right fragment in the 1.63 kb PstI -BstEII fragment. As a result of determining the nucleotide sequence, an open reading frame which corresponds to the gene and has a ribosome binding site just in front of the initiation codon was successfully detected, it was referred to as the crtZ gene. The nucleotide sequence of the crtZ
gene and the encoded amino acid sequence are illustrated in Fig. 15 (SEQ ID N0: 6).
The crtZ gene product (CrtZ) of Alcaligenes sp. PC-1 has an enzyme activity for adding a hydroxyl group to the 3-carbon of a (3-ionone ring, and one of the specific examples is an enzyme activity for synthesizing zeaxanthin from ~i-carotene as a substrate by way of ~i-cryptoxanthin (Example 12 (3); see Fig. 11). Furthermore, the crtZ gene product also has an enzyme activity for adding a hydroxyl group to the 3-carbon of a 4-keto-~3-ionone ring, and one of the specific examples is an enzyme activity for synthesizing astaxanthin from canthaxanthin as a substrate by way of phoenicoxanthin (Example 12 (1); see Fig. 11).
In addition, polypeptides having the latter enzyme activity and DNA strands encoding these polypeptides have not hitherto been known. Also, the CrtZ of Alcaligenes sp. PC-1 showed significant homology to the CrtZ of Erwinia uredovora (identity of 58%) at the level of amino acid sequence. In addition, the crtZ gene products (CrtZ) of Agrobacterium aurantiacus sp. nov. MKl and Alcaligenes sp.
PC-1 have high homology (identity of 90%) at the level of amino acid sequence, and the functions of both enzymes are the same. The amino acid sequence in the region of 10%
having no identity among these amino acid sequences is considered not so significant to the functions of the enzyme. It is thus considered particularly in this region that a little amount of substitution by the other amino acids, deletion, or addition of the other amino acids will not afftect the enzyme activity.
(3) Consideration on minor biosynthetic pathways of xanthophylls It has been elucidated by our studies with carotenoid synthesis genes of the epiphytic bacterium Erwinia or the photosynthetic bacterium Rhodobacter that carotenoid biosynthesis enzymes generally act by recognizing the half of a carotenoid molecule as a substrate. By way of example, the lycopene cyclase gene of Erwinia, crtY, recognizes the halves of the lycopene molecule to cyclize it. When the phytoene desaturase gene crtl of Rhodobacter was used for the synthesis of neurosporene in place of lycopene in Escherichia coli and crtY of Erwinia was allowed to work on it, the crtY gene product recognizes the half molecular structure common to lycopene to produce a half cyclized (3-zeacarotene (Linden, H., Misawa, N., Chamovits, D., Pecher, I., Hirschberg, J., Sandmann, G., "Functional Complementation in Escherichia coli of Different Phytoene Desaturase Genes and Analysis of Accumulated Carotenes", Z. Naturforsch., 46c, p. 1045-1051, 1991). Also, in the present invention, when CrtW is allowed to work on (3-carotene or zeaxanthin, echinenone or 4-ketozeaxanthin in which one keto group has been introduced is first synthesized, and when CrtZ is allowed to work on (3-carotene or canthaxanthin, (3-cryptoxanthin or --, 2180024 phoenicoxanthin in which one hydroxyl group .has been introduced is first synthesized. It can be considered because these enzymes recognize the half molecule of the substrate. Thus, while Escherichia coli having the crtE, crtB, crtI and crtY genes of Erwinia and the crtZ gene of a marine bacterium produces zeaxanthin as described above, (3-cryptoxanthin which is ~i-carotene having one hydroxyl group introduced thereinto can be detected as an intermediate metabolite. It can be thus considered that if CrtW is present, 3'-hydroxyechinenone or 3-hydroxyechinenone can be synthesized from ~i-cryptoxanthin as a substrate, and that phoenicoxanthin can be further synthesized by the action of CrtW on these intermediates.
The present inventors have not identified these ketocarotenoids in the culture solutions, and the reason is considered to be that only a trace amount of these compounds is present under the conditions carried out in the present experiments. In fact, it was described that 3-hydroxyechinenone or 3'-hydroxyechinenone was detected as a minor intermediate metabolite of astaxanthin in a marine bacterium Agrobacterium aurantiacus sp. nov. MKl as a gene source (Akihiro Yokoyama ed., "For the biosynthesis of astaxanthin in marine bacteria", Nippon Suisan Gakkai, Spring Symposium, 1994, Abstract, p. 252, 1994). It can be considered from the above descriptions that minor metabolic pathways shown in Fig. 20 are also present in addition to the main metabolic pathways of astaxanthin shown in Fig. 11.
Industrial Applicability According to the present invention, the gene clusters required for the biosynthesis of keto group-containing xanthophylls such as astaxanthin, phoenicoxanthin, 4-ketozeaxanthin, canthaxanthin and echinenone have successfully been obtained from marine bacteria, and their structures, nucleotide sequences, and functions have been elucidated. The DNA strands according to the present invention are useful as genes capable of affording the ability of biosynthesis of keto group-containing xanthophylls such as astaxanthin to microorganisms such as F~~hPrichia coli and the like.

56 21 8 0 0 2 ~
SEQUENCE LISTING
SEQ ID NO: 1 SEQUENCE LENGTH: 639 SEQUENCE TYPE: Nucleotide with corresponding protein STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: A_grobacterium aurantiacus STRAIN: sp. nov. MK1 SEQUENCE

Met His Ala Leu Trp Phe Leu Asp Ala Ala Ala His Pro Ile Leu Ala Ile Ala Asn Phe Leu Gly Leu Thr Trp Leu Ser Val Gly Leu Phe Ile Ile Ala His Asp Ala Met His Gly Ser Val Val Pro Gly Arg Pro Arg Ala Asn Ala Ala Met Gly Gln Leu Val Leu Trp Leu Tyr Ala Gly Phe r 2~8~024 SerTrpArg LysMetIleVal LysHisMet AlaHisHis ArgHisAla 6a 70 7S 80 GlyThrAsp AspAspProAsp PheAspHis GlyGlyPro YalArgTrp TyrAlaArg PheIleGlyThr TyrPheGly TrpArgGlu GlyLeuLeu CTGCCCGTC 100GTGACGGT'CTATGCGCTG ATCCTTGGG GATCGCTGG 384 ATC

LeuProVal IleValThrVal TyrAlaLeu IleLeuGly AspArgTrp MetTyrYal ValPheTrpPro LeuPro5er IleLeuAla SerIleGln TTC TGG

LeuPheYal PheGlyThrTrp LeuProHis ArgProGly HisAspAla PheProAsp ArgHisAsnAla ArgSerSer ArgIleSer AspProVal 165 170 1~5 SerLeuLeu ThrCysPheHis PheGlyGly TyrHisHis GluHisHis CTG CAC CCG ACG GTG CCG TGG TGG CGC CTG CCC AGC ACC CGC ACC AAG 62~
Leu His Pro Thr Yal Pro Trp Trp Arg Leu Pro Ser Thr Arg Thr Lps Glp Asp Thr Ala ~~~

2180p24 SEQ ID NO: 2 SEQUENCE LENGTH: 489 SEQUENCE TYPE: Necleotide with corresponding protein STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: Agrobacterium aurantiacus STRAIN: sp. nov. MKl SEQUENCE

Met Thr Asn Phe Leu Ile Val Val Ala Thr Val Leu Val Met Glu Leu Thr Ala Tyr Ser Val His Arg Trp Ile Met His Gly Pro Leu Gly Trp 20 Gly Trp His Lys Ser His His Glu Glu His Asp His Ala Leu Glu Lys Asn Asp Leu Tyr Gly Leu Val Phe Ala Val Ile Ala Thr Val Leu Phe X

Thr Val Gly Trp Ile Trp Ala Pro Val Leu Tip Trp Ile Ala Leu Gly Met Thr Val Tyr Gly Leu Ile Tyr Phe Yal Leu His Asp Giy Leu Yal His Gln Arg Trp Pro Phe Arg Tyr Ile Pro Arg Lys Gly Tyr Ala Arg 100 lOS 110 Arg Leu Tyr Gln Ala His Arg Leu His His Ala Val Glu Gly Arg Asp CAT TGC GTC AGC TTC GGC TTC ATC TAT GCG CCC CCG GTC GAC AAG CTG d32 His Cys Val Ser Phe Gly Phe lle Tyr Ala Pro Pro Val Asp Lys Leu 130 13S . 140 Lys G(n Asp Leu Lys Met Ser Gly Val Leu Arg Ala Glu Ala Gln Glu , Arg Thr ---SEQ ID NO: 3 SEQUENCE LENGTH: 1161 SEQUENCE TYPE: Nucleotide with corresponding protein STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: Agrobacterium aurantiacus STRAIN: sp. nov. MK1 SEQUENCE

Met Thr His Asp Val Leu Leu Ala Gly Ala Gly Leu Ala Asn Gly Leu Ile Ala Leu Ala Leu Arg Ala Ala Arg Pro Asp Leu Arg Val Leu Leu 20 Leu Asp His Ala Ala Gly Pro Ser Asp Gly His Thr Trp Ser Cys His Asp Pro Asp Leu Ser Pro Asp Trp Leu Ala Arg Leu Lys Pro Leu Arg xi Atg Ala Asn Trp Pro Asp Gln Gln Val Arg Phe Pro Arg His Ala Arg Arg Leu Ala Thr Gly Tyr Gly Ser Leu Asp Gly Ala Ala Leu Ala Asp Ala Yal Yal Arg Ser Gly Ala Glu Ile Arg Trp Asp Ser Asp Ile Ala Leu Leu A.sp Ala Gln Gly Ala Thr Leu Ser Cys Gly Thr Arg lle Glu 115 120 125.

Ala Gly Ala Val Leu Asp Gly Arg GIy Ala Gln Pro Ser Arg His Leu Thr Yal Gly Phe Gln Lys Phe Yal Gly Yal Glu Ile Glu Thr Asp Arg 145 1S0 155 ,160 Pro His Gly Val Pro Arg Pro Met Ile Met Asp Ala Thr Yal Thr Gln Gln Asp Gly Tyr Arg Phe Ile Tyr Leu Leu Pro Phe Ser Pro Thr Arg Ile Leu Ile Glu Asp Thr Arg Tyr Ser Asp Gly Gly Asp Leu Asp Asp 195 200 205 .
GAC GCG CTG GCG GCG GCG TCC CAC GAC TAT GCC CGC~CAG CAG GGC TGG 672 Asp Ala Leu Ala Ala Ala Ser His Asp Tyr Ala Arg Gln Gln Gly Trp Tht Gly Ala Glu Yal Arg Arg Glu Atg Gly Ile Leu Pro Ile Ala Leu Ala His Asp Ala Ala Gly Phe Trp Ala Asp His Ala Ala Gly Pro Val Pro Val Gly Leu Arg Ala Gly Phe Phe His Pro Val Thr Gly Tpr Ser Leu Pro Tyr Ala Ala Gln Yal Ala Asp Val Val Ala Gly Leu Ser Gly ,' Pro Pro Gly Thr Asp Ala Leu Arg Gly Ala Ile Arg Asp Tyr Ala Ile Asp Arg Ala Arg Arg Asp Arg Phe Leu Arg Leu Leu Asn Arg ~fet Leu 2~g0024 Phe Arg Gly Cys Ala Pro Asp Arg Arg Tyr Thi Leu Leu Gln Arg Phe Tyr Arg Met Pro His Gly Leu Ile Glu Arg Phe Tyr Ala Gly Arg Leu Ser Yal Ala Asp Gln Leu Arg Ile Val Thr Gly Lys Pro Pro Ile Pro Leu Gly Thr Ala Ile Arg Cys Leu Pro Glu Arg Pro Leu Leu Lys Glu Asn Ala ~~~

SEQ ID NO: 4 SEQUENCE LENGTH: 2886 SEQUENCE TYPE: Nucleotide STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: Agrobacterium aurantiacus STRAIN: sp. nov. MK1 SEQUENCE

CCTAGGCCGC TGGAACGCCG CGACGCGGCG CGCGGAAACG ACCACGGACC CGGCCCACCG

GTTACCAGCG TTCGTTGCCC CTACCTTTGG CCGCTACGCC CTGACATCAG ACGCGCCTAG

CGGCCAGGCC CCCTGTTCTA CTCGCGTGTA CGGGACGGGT TCCGTCTAGA CTGGCGGTGG

TCGGACTAGC AGAGCCCGCC GTAGTAGCGG CGAACCGACC GGGACGTACA CGTACGCGAC

ACCAAAGACC TGCGTCGCCG CGTAGGGTAG GACCGCTAGC GTTTAAAGGA CCCCGACTGG
x'~

ACCGACAGCC AGCCTA,4CAA GTAGTAGCGC GTACTGCGCT ACCTGCCCAG CCACCACGGC

CCCGCAGGCG CGCGGTTACG CCGCTACCCG GTCGAACAGG ACACCGACAT ACGGCCTAAA

AGCACCGCGT TCTACTAGCA GTTCGTGTAC CGGGTAGTGG CGGTACGGCC TTGGCTGCTG

CTGGGGCTAA AGCTGGTACC GCCGGGCCAG GCGACCATGC GGGCGAAGTA GCCGTGGATA

AAGCCGACCG CGCTCCCCGA CGACGACGGG CAGTAGCACT GCCAGATACG CGACTAGGAA

CCCCTAGCGA CCTACATGCA CCAGAAGACC GGCGACGGCA GCTAGGACCG CAGCTAGGTC

GACAAGCACA AGCCGTGGAC CGACGGCGTG GCGGGGCCGG TGCTGCGCAA GGGCCTGGCG

GTGTTACGCG~CCAGCAGCGC CTAGTCGCTG GGGCACAGCG ACGACTGGAC GAAAGTGAAA

CCGCCAATAG TAGTGCTTGT GGTGGACGTG GGCTGCCACG GCACCACCGC GGACGGGTCG

TGGGCGTGGT TCCCCCTGTG GCGTACTGGT TAAAGGACTA GCAGCAGCGG TGGCACGACC

ACTACCTCAA CTGCCGGATA AGGCAGGTGG CGACCTAGTA CGTGCCGGGG GACCCGACCC

CGACCGTGTT CAGGGTGGTG CTCCTTGTGC TGGTGCGCGA CCTTTTCTTG CTGGACATGC

CGGACCAGAA ACGCCACTAG CGGTGCCACG ACAAGTGCCA CCCGACCTAG ACCCGCGGCC

AGGACACCAC CTAGCGGAAC CCGTACTGAC AGATACCCGA CTAGATAAAG CAGGACGTAC

TGCCCGACCA CGTAGTCGCG ACCGGCAAGG CAATATAGGG CGCGTTCCCG ATACGGTCTG

CGGACATAGT CCGGGTGGCG GACGTGGTAC GCCAGCTCCC CGCGCTGGTA ACGCAGTCGA

AGCCGAAGTA GATACGCGGG GGCCAGCTGT TCGACTTCGT CCTGGACTTC TACAGCCCGC

ACGACGCCCG GCTCCGCGTC CTCGCGTGCA CTGGGTACTG CACGACGACC GTCCCCGCCC
CCTTGCCAAC GGGCTGATCG CCCTGGCGCT GCGCGCGGCG CGGCCCGACC TGCGCGTGCT 14"0 GGAACGGTTG CCCGACTAGC GGGACCGCGA CGCGCGCCGC GCCGGGCTGG ACGCGCACGA

CGACGACCTG GTACGGCGTC CTGGCAGTCT GCCGGTGTGG ACCAGGACGG TGCTGGGGCT

GGACAGCGGC CTGACCGACC GCGCCGACTT CGGGGACGCG GCGCGGTTGA CCGGGCTGGT

CCTCCACGCG AAAGGGGCGG TACGGGCCGC CGACCGGTGG CCAATGCCCA GCGACCTGCC
GGCGGCGCTG GCGGATGCGG TGGTCCGGTC GGGCGCCGAG ATCCGCTGGG ACAGCGACAT 16$0 CCGCCGCGAC CGCCTACGCC ACCAGGCCAG CCCGCGGCTC TAGGCGACCC TGTCGCTGTA

GCGGGACGAC CTACGCGTCC CCCGCTGCGA CAGGACGCCG TGGGCCTAGC TCCGCCCGCG

CCAGGACCTG CCCGCCCCGC GCGTCGGCAG CGCCGTAGAC TGGCACCCAA AGGTCTTTAA

GCACCCACAG CTCTAGCTCT GGCTGGCGGG GGTGCCGCAC GGGGCGGGCT ACTAGTACCT

GCGCTGGCAG TGGGTCGTCC TGCCCATGGC GAAGTAGATA GACGACGGGA AGAGAGGCTG

CGCGTAGGAC TAGCTCCTGT GCGCGATAAG GCTACCGCCG CTAGACCTGC TGCTGCGCGA

CCGCCGCCGC AGGGTGCTGA TACGGGCGGT CGTCCCGACC TGGCCCCGGC TCCAGGCCGC

GCTTGCGCCG TAGGAAGGGT AGCGCGACCG GGTACTACGC CGCCCGAAGA CCCGGCTAGT

GCGCCGCCCC GGACAAGGGC ACCCTGACGC GCGCCCCAAG AAAGTAGGCC AGTGGCCGAT

AAGCGACGGG ATACGCCGTG TCCACCGCCT GCACCACCGC CCAGACAGGC CCGGCGGGCC

... 2180024 GTGGCTGCGC GACGCGCCGC GGTAGGCGCT AATGCGCTAG CTGGCCCGCG CGGCGCTGGC

GAAAGACGCG GAAAACTTGG CCTACGACAA GGCGCCGACG CGCGGGCTGG CCGCGATATG

GGACGACGTC GCCAAGATGG CGTACGGCGT ACCTGACTAG CTTGCCAAGA TACGGCCGGC

CGACTCGCAC CGCCTAGTCG ACGCGTAGCA CTGGCCGTTC GGAGGGTAAG GGGAACCGTG

CCGGTAGGCG ACGGACGGGC TTGCAGGGGA CGACTTCCTT TTGCGTACTT GCGGGTAAGC
CCCGCGGCCA AG:4CCGCCAT CGTGATCGGC GCAGGCTTTG GCGGGCTGGC CCTGGCCATC 2580 GGGCGCCGGT TCTGGCGGTA GCACTAGCCG CGTCCGAAAC CGCCCGACCG GGACCGGTAG

GCGGACGTCA GGCGCCCGTA GCGGTGGTGG GACCAGCTCC GGGCCCTGTT CGGGCCGCCC

GCGCGGATAC AGACCGTGCT AGTCCCGGTA GAGAAGCTGC GCCCGGGCTG GCAGTAGTGG

z i sooz4 CTGGGGCTAC GCGACTTTCT CGACACCCGG GACTGGCCCG TCCTGTACCG CGCGCTGCAC

TGCGACTACG GCCAGAGCGG GAAGATAGCC GACTACACCG GCCCGCCCTT CCAGAAGCTA

ATGCACTTGC TCCGGCTAGG TCCCAGACCC AGAACGGCAC GGTCCACTTC GACAACGGCA

CCTAGG

72 21 $002~t -SEQ ID NO: 5 SEQUENCE LENGTH: 729 SEQUENCE TYPE: Nucleotide with corresponding protein STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: Alcaligenes STRAIN: sp. PC-1 SEQUENCE

Met Ser Gly Arg Lys Pro Gly Thr Thr Gly Asp Thr Ile Val Asn Leu Gly Leu Thr Ala Ala Ile Leu Leu Cys Trp Leu Val Leu His Ala Phe 20 Thr Leu Trp Leu Leu Asp Ala Ala Ala His Pro Leu Leu Ala Val Leu TGC CTG GC'T GGG CTG ACC TGG CTG TCG GTC GGG CTG TTC ATC ATC GCG 192 Cys Leu Ala Gly Leu Thr Trp Leu Ser Val Gly Leu Phe Ile Ile Ala Xi z ~ sooz4 HisAsp AlaMetHis GlySerValYal ProGlyArg ProArgAla Asn 65 70 7h 80 AlaAla lleG!yGln LeuAlaLeuTrp LeuTyrAla GlyPheSer Trp ProLys LeuIleAla LysHisMetThr HisHisArg HisAlaGly Thr CCC

AspAsn AspProAsp PheGlyHisGly GlyProVal ArgTrpTyr Gly SerPhe ValSerThr TyrPheGlyTrp ArgGluGly LeuLeuLeu Pto ATC

YalIle YalThrThr TyrAlaLeuIle LeuGlyAsp ArgTrpMet Tyr 1~5 150 155 160 ValIle PheTrpPro ValProAlaVal LeuAlaSer IleGlnIle Phe ValPhe GlyThrTrp LeuProHisArg ProGlyHis AspAspPhe Pro AspArg HisAsnAla ArgSerThrGly IleGly AspProLeu SerLeu CTGACC TGCTTCCAT TTCGGCGGCTAT CACCAC GAACATCAC CTGCAT 6~2 LeuThr CysPheHis PheGlyGlyTyr HisHis GluHisHis LeuHis CAT

ProHis YalProTrp TrpArgLeuPro ArgThr ArgLysThr GlyGly CGCGCA TGA

ArgAla~~~

SEQ ID NO: 6 SEQUENCE LENGTH: 489 SEQUENCE TYPE: Nucleotide with corresponding protein STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: Alcaligenes STRAIN: sp. PC-1 Met Thr Gln Phe Leu Ile Val Val Ala Thr Val Leu Val Met Glu Leu Thr Ala Tyr Ser Val His Arg Trp Ile Met His Gly Pro Leu Gly Trp 20 Gly Trp His Lys Ser His His Glu Glu His Asp His Ala Leu Glu Lys Asn Asp Leu Tyr Gly Val Val Phe Ala Val Leu Ala Thr Ile Leu Phe d GGC TAT

Thr Val Ala TyrTrpTrpPro YalLeuTrp TrpIleAla LeuGly Gly bletThrValTyr GlyLeuIleTyr PheIleLeu HisAspGly LeuVal His GlnArgTrp ProPheArgTyr IleProArg ArgGlyTyr PheArg Arg LeuTyrGln AlaHisArgLeu HisHisAla YalGluGly ArgAsp His CpsValSer PheGlyPhelle TyrAlaPro ProValAsp LysLeu CAG TCG

Lys GlnAspLeu LysArgSerGly YalLeuArg ProGlnAsp GluArg CCG TGA ~
TCG

Pro Ser~~~

2~gpp~4 SEQ ID NO: 7 SEQUENCE LENGTH: 1631 SEQUENCE TYPE: Nucleotide STRANDEDNESS: double TOPOLOGY: linear MOLECULE TYPE: genomic DNA
ORIGINAL SOURCE:
ORGANISM: Alcaligenes STRAIN: sp. PC-1 SEQUENCE

GACGTCCGGC CCGGGCCACC GGTTACCAGC GTTGGCCGTC CTGACCTTGT CCTGCCGCCC

GGCCAGATCC GACAGCGGGA TGCGTCGTCC TCAAAGCCTA CAGGCCTGCC TTCGGACCGT

GTTGACCGCT GTGCTAGCAG TTAGAGCCAG ACTGGCGGCG CTAGGACGAC ACGACCGACC

AGGACGTGCG GAAATGCGAT ACCAACGATC TACGCCGGCG CGTAGGCGAC GAACGGCACG
t ACACGGACCG ACCCGACTGG ACCGACAGCC AGCCCGACAA GTAGTAGCGC GTACTGCGTT

ACGTGCCCAG GCACCACGGC CCCGCCGGCG CGCGGTTACG CCGCTAGCCC GTTGACCGCG

ACACCGAGAT ACGCCCCAAG AGCACCGGGT TCGACTAGCG GTTCGTGTAC TGCGTAGTGG

CCGTGCGGCC GTGGCTGTTG CTAGGGCTAA AGCCAGTGCC TCCCGGGCAC GCGACCATGC

CGTCGAAGCA GAGGTGGATA AAGCCGACCG CTCTCCCTGA CGACGATGGC CACTAGCAGT

GGTGGATACG CGACTAGGAC CCGCTAGCGA CCTACATACA GTAGAAGACC GGCCAGGGCC

GGCAAGACCG CAGCTAGGTC TAAAAGCAGA AGCCTTGAAC CGACGGGGTG GCGGGCCCTG

TACTGCTAAA AGGGCTGGCC GTGTTGCGCT CCAGCTGGCC GTAGCCGCTG GGCAACAGTG

... 2 ~ 80024 _ 79 _ ATGACTGGAC GAAGGTAAAG CCGCCGATAG TGGTGCTTGT AGTGGACGTA GGCGTACACG

GCACCACCGC GGACGGAGCA TGTGCGTTCT GGCCTCCGGC GCGTACTGCG TTAAGGAGTA

ACAGCACCGC TGTCAGGAGC ACTACCTCGA CTGGCGGATA AGGCAGGTGG CGACCTAATA

CGTGCCGGGG GATCCGACCC CGACCGTGTT CAGGGTAGTG CTTCTCGTGC TGGTGCGCAA

CCTCTTCTTG CTGGAGATGC CGCAGCAGAA GCGCCACGAC CGCTGCTAGG AGAAGTGGC,4 CCCGCGGATA ACCACCGGCC ACGACACCAC CTAGCGGGAC CCGTACTGCC AGATACCCAA

CTAGATAAAG TAGGACGTGC TGCCCGAACA CGTAGTTGCG ACCGGCAAAG CCATATAAGG

CGCCGCCCCG ATAAAGGCGT CCGAGATGGT TCGAGTAGCG GACGTGGTGC GCCAGCTCCC

CGCCCTGGTG ACGCAGTCGA AGCCGAAGTA GATACGGGGT GGGCACCTGT TCGACTTCGT

CCTAGACTTC GCCAGCCCAC AGGACGCGGG GGTCCTGCTC GCAGGCAGCA CTAGAGACTA

GGGCCGCACC GGCGTACTTT AGGCTGCACG ACGACCGTCC CCGGCCGGAA CGGTTGCCTG

ACTAGCGCGA CCGCTAGGCG TTCCGCGCCG GGCTGGAAGC GCACGACGAC GACCTGGCAC

GCCGCCCGCG GAGCCTGCCC GTATGAACCA GGACGGTGCT GTGGCTAAAC CGCGGCGTGA

CCGACCTGGC GGACTTCGGC TAGTCCGCAC CGCTGACCGG GCTAGTCCTC CACGCCAAGG

GTCTGGTAAG CGCTTCCGAG GCCCGGCCTA TACCGAGCTA GCTGCCCGCC CCCGACTACG

CACGCCACTG G

Claims (30)

CLAIMS:

1, A purified DNA strand comprising a nucleotide sequence which is derived from an organism selected from the group consisting of Agrobacterium aurantiacus and Alcaligenes and which codes for a polypeptide having an enzyme activity to convert the methylene group at position 4 of a .beta.-ionone ring or at position 4 of a 3-hydroxy-.beta.-ionone ring contained in a carotenoid compound to a keto group, wherein the nucleotide sequence is a member selected from the group consisting of:
(1) a nucleotide sequence coding for an amino acid sequence substantially identical to amino acids Nos. 1-212 shown in SEQ ID NO: 1;
(2) a nucleotide sequence coding for an amino acid sequence substantially identical to amino acids Nos. 1-242 shown in SEQ ID NO: 5; and (3) a nucleotide sequence which hybridizes with a complement of the nucleotide sequence (1) or (2) in a hybridization solution containing 6xDenhardt, 5xSSC, 0.2%SDS, 100 µg/ml ssDNA at 55ÀC followed by washing with 2xSSC, 0.1%SDS
at 60ÀC.

2. A purified DNA strand comprising a nucleotide sequence coding for a polypeptide having an enzyme activity to convert the methylene group at position 4 of a .beta.-ionone ring contained in a carotenoid compound to a keto group and having an amino acid sequence substantially identical to amino acid Nos. 1-212 shown in SEQ ID NO: 1.

3. A purified DNA strand comprising a nucleotide sequence which hybridizes with a complement of the DNA strand of claim 2 in a hybridization solution containing 6xDenhardt, 5xSSC, 0.2%SDS, 100 µg/ml ssDNA at 55ÀC followed by washing with 2xSSC, 0.1%SDS at 60ÀC and which codes for a polypeptide having an enzyme activity to convert the methylene group at position 4 of a .beta.-ionone ring contained in a carotenoid compound to a keto group.

4, A purified DNA strand comprising a nucleotide sequence coding for a polypeptide having an enzyme activity to convert the methylene group at position 4 of a .beta.-ionone ring contained in a carotenoid compound to a keto group and having an amino acid sequence substantially identical to amino acid Nos. 1-242 shown in SEQ ID NO: 5.

5. A purified DNA strand comprising a nucleotide sequence which hybridizes with a complement of the DNA strand according to claim 4, in a hybridization solution containing 6xDenhardt, 5xSSC, 0.2%SDS, 100 µg/ml ssDNA at 55ÀC followed by washing with 2xSSC, 0.1%SDS at 60ÀC and which codes for a polypeptide having the enzyme activity according to claim 4.

6. The DNA strand according to claim 2, wherein the polypeptide has an amino acid sequence of amino acids Nos. 1-212 shown in SEQ ID NO: 1.

7. The DNA strand according to claim 6, wherein the nucleotide sequence has the sequence shown in the top row of SEQ ID NO: 1.

8. The DNA strand according to claim 4, wherein the polypeptide has an amino acid sequence of amino acids Nos. 1-242 shown in SEQ ID NO: 5.

9. The DNA strand according to claim 8, wherein the nucleotide sequence has the sequence shown in the top row of SEQ ID NO: 5.

10. A purified DNA strand comprising a nucleotide sequence which codes for a polypeptide having an enzyme activity for adding a hydroxyl group at position 3 of a 4-keto-.beta.-ionone ring or at position 3 of a .beta.-ionone ring contained in a carotenoid compound, wherein the polypeptide has an amino acid sequence selected from the group consisting of:
(1) an amino acid sequence having amino acids Nos. 1-162 shown in SEQ ID NO: 2;
(2) an amino acid sequence having amino acids Nos. 1-162 shown in SEQ ID NO: 6; and (3) an amino acid sequence at least 90% identical to (1) or (2).

11. A purified DNA strand comprising a nucleotide sequence coding for a polypeptide having an enzyme activity to add a hydroxyl group to the 3-carbon of a 4-keto-.beta.-ionine ring in a carotenoid compound and having an amino acid sequence substantially identical to amino acid Nos. 1-162 which is shown in SEQ ID NO: 2.

12. A purified DNA strand comprising a nucleotide sequence coding for a polypeptide having an enzyme activity to add a hydroxyl group to the 3-carbon of a 4-keto-.beta.-ionone ring in a carotenoid compound and having an amino acid sequence substantially identical to amino acid Nos. 1-162 which is shown in SEQ ID NO: 6.

13. The DNA strand according to claim 11, wherein the polypeptide has an amino acid sequence of amino acids Nos. 1-162 shown in SEQ ID NO: 2.

14. The DNA strand according to claim 13, wherein the nucleotide sequence has the sequence shown in the top row of SEQ ID NO: 2.

15. The DNA strand according to claim 12, wherein the polypeptide has an amine acid sequence of amino acids Nos. 1-162 shown in SEQ ID NO: 6.

16. The DNA strand according to claim 15, wherein the nucleotide sequence has the sequence shown in the top row of SEQ ID NO: 6.

17. A microorganism transformed with the DNA strand as defined in any one of claims 1 to 9.

18. The microorganism according to claim 17, which has therein the crtE, crtB, crtI and crtY genes.

19. The microorganism according to claim 18, wherein the crtE, crtB, crtI and crtY genes are of Erwinia uredovora.

20. The microorganism according to claim 17, which has therein the crtE, crtB, crtI, crtY and crtZ genes.

21. The microorganism according to claim 20, wherein the crtE, crtB, crtI, crtY and crtZ genes are of Erwinia uredovora.

22. A microorganism transformed with the DNA strand as defined in any one of claims 10 to 16.

23. A microorganism transformed with the DNA strand as defined in any one of claims 1 to 9 and the DNA strand as defined in any one of claims 10 to 16.

24. The microorganism according to claim 23, which has therein the crtE, crtB, crtI and crtY genes.

25. The microorganism according to claim 24, wherein the crtE, crtB, crtI and crtY genes are of Erwinia uredovora.

26. A method for producing a xanthophyll, which comprises:

introducing the DNA strand according to any one of claims 1 to 9 into a microorganism having a .beta.-carotene-synthesizing ability, culturing the transformed microorganism in a culture medium, and obtaining cantaxanthin or echinenone from the culture.

27. A method for producing a xanthophyll, which comprises:
introducing the DNA strand according to any one of claims 1 to 9 into a microorganism having a zeaxanthin-synthesizing ability, culturing the transformed microorganism in a culture medium, and obtaining astaxanthin or 4-ketozeaxanthin from the culture.

28. A method for producing a xanthophyll, which comprises:
introducing the DNA strand according to any one of claims 10 to 16 into a microorganism having a canthaxanthin-synthesizing ability, culturing the transformed microorganism in a culture medium, and obtaining astaxanthin or phoenicoxanthin from the culture.

29. A method for producing a xanthophyll, which comprises:

introducing the DNA strand according to any one of claims 1 to 9 and the DNA strand according to any one of claims to 16 into a microorganism having a .beta.-carotene-synthesizing ability, culturing the transformed microorganism in a culture medium, and obtaining astaxanthin or phoenicoxanthin from the culture.

30. A method according to any one of claims 26 to 29, wherein the microorganism is a bacterium or yeast.