AN L-CYSTEINE-PRODUCING BACTERIUM AND A METHOD FOR PRODUCING L-CYSTEINE BACKGROUND OF THE INVENTION Field of the Invention The present invention relates to a method for producing L-cysteine or related substances. Specifically, the present invention relates to a bacterium suitable for producing L-cysteine or related substances and a method for producing L-cystcine or related substances utilizing the bacterium. L-cysteine and L-cysteine-related substances are useful in the fields of drugs, cosmetics, and food. Background Art L-cystoino is obtained by extraction from keratin-containing substances such as hair, horns and feathers, or by the conversion of DL-2-aminothlazollne-4-c«rboxylic acid as a prouursor using a microbial enzyme. It is also planned to produce L-cysteInc in u large scale by an immobilized enzyme method utilizing a novel enzyme. Furthermore, it is also attempted to produce L-cysteine by fermentation utilizing a bacterium. For example, the inventors of the present invention disclosed a method for producing L-cysteine using an Escherichia bacterium having a suppressed L-cysteine decomposition system and a serine acetyltransferase (EC 2.3.1.30, henceforth also referred to as "SAT") in which feedback inhibition by L-cysteine is attenuated (Japanese Patent Laid-open (Kokai) No. 11-155571). Furthermore, as bacteria in which L-cysteine-producing ability is enhanced by suppressing the L-cysteine decomposition system, there are known corynefonn bacteria or Escherichia bacteria in which activity of cystathionine-P-lyase (Japanese Patent Laid-open No. 11-155571), tryptophanase (Japanese Patent Laid-open No. 2003-169668), or O-acetylserine sulfliydrylase B (Japanese Patent Laid-open No. 2005-2453 U) is attenuated or deleted. A method for producing L-cysteine by using a bacterium in which L-cysteine metabolism is decontrolled by using a DNA sequence coding for SAT that has a specific mutation for attenuating feedback inhibition by L-cysteine is also known (National Publication of Translated Version in Japan (Kohyo) No. 2000-504926). Furthermore, it is known that the ydeD gene which encodes the YdeD protein (Dabler et al., Mol, Microbiol., 36,1101 -1112 (2000)) and the>/: gene which encodes the YfiK protein (Japanese Patent Laid-open No. 2004-49237) participate in secretion of the metabolic products of the cysteine pathway. Furthermore, there are also known techniques of enhancing L-cysteine-producing ability by increasing expression of the mar-locus, acr-locus, cmr-locxxs, mex-genc, bmr-gene or qacA'gne, which encode proteins suitable for secreting a toxic substance from cells (U.S. Patent No. 5,972,663), or emrAB, emrKY.yoJIH, acrEF, bcr or cuxA gene (Japanese Patent Laid-open No. 2005-287333). As an L-cysteine-producing bacterium, there is also known Escherichia coli in which the activity of the positive transcription control factor of the cysteine regulon encoded by the cysB gene is increased (International Patent Publication WOOl/27307). Furthermore,, in the production of L-amino acids by fermentation, not only secretion of L-amino acids out of cells, but also uptake of L-amino acids is important. Microorganisms have an ability to take up many kinds of amino acids into cells from the environment, in which each microorganism grow, and use them. For example, Escherichia coli (E. coli) has a large number of transporters, and it is supposed that many of them participate in uptake of L-amino acids. There is even an estimation tliat among those predicted to be "membrane tranporter proteins" concerning the functions thereof, 14% of them participate in transport of amino acids (Paulsen et al., J. Mol. Biol., 277, 573-592 (1998)), However, there are generally a large number of various paralogues of transporters concerning substrate specificity, and many overlaps of functions (there are plural uptake systems for the same substrate). Therefore, it is extremely difficult to identify a function as a transporter (Hosie et al., Res. Microbiol,, 152, 259-270 (2001)). As described above, functions and physiological roles of transporters are very complicated. Therefore, characteristics of a transporter simply estimated on the basis of homology or phenotype may not reflect physiological functions as an actual transporter. For example, it cannot be expected transportation of which substrate is a physiologically important Hmclion among transportation of plural kinds of substrates, or the like. Furthermore, even if a certain substance is found to be transported, there may be plural factors which also transport that substance. Therefore, when a microorganism used for amino acid production by fermentation is modified, it is not easy to use a transporter as a target. I'urthermore, although there are .several llndings as described below concerning uptake systems of bacteria for cystine among L-cysteine and related compounds thereof, there is substantially no important finding about uptake of L-cystelne and S-sulfocysteine. It is expected that E. coli has at least two kinds of cystine uptake systems showing different kinetic characteristics (Berger et al., J. Biol. Chem., 247,7684-7694 (1972)). Fli Y has been demonstrated to bind to cystine in an in vitro experiment system (Butler et al.. Life Sci., 52,1209-1215 (1993)). and the///r gene is expected to form an operon together NiihyecC, yecS and yecO existing nearby, and function as an ABC transporter (Hosie et al., Res. Microbiol., 152,259-270 (2001)). However, it has not been experimentally demonstrated yet whether they function as a physiological cystine uptake system in E. coU. Although it is similarly expected that three cystine uptake systems of different kinetics are also present in Salmonella bacteria (Baptist et al., J. Bacteriol., 131,111 -1 ] 8 (1977)), proteins involved in them and genes coding for them have not been identified yet. Furthermore, three kinds of cystine uptake systems (YckKJI, YtmJKLMN, YhcL) have been reported for Bacillus sublilis, and if these three kinds of systems are deleted, the bacterium becomes unable to grow with cystine as a sole sulfur source (Burguiere et al., J. Bacteriol., 186,4875-4884 (2004)). Although it has been reported that YdjN of £. coli has a homology of 45% to TcyP, which is known to be involved in cystine uptake oi Bacillus subtilis (Burguiere et al., J. Bacteriol., 186,4875-4884 (2004)), it has not been confimied whether it actually has cystine uptake activity. It is known that in Lactobacillus fermentum BRl 1, bspA codes for a cystine uptake system (Turner et al., J. Bacteriol., 181,2192-2198 (1999)). Furthermore, although it is expected that there are two kinds of cysteine uptake systems showing diflferent kinetics in Legionella pneumophila (Ewatm et al., Appl. Environ. Microbiol., 72, 3993-4000 (2006)), neither genes nor proteins thereof have not been identified yet. SUMMARY OF THE INVENTION An object of the present invention is to develop novel techniques for improving bacterial production of L-cysteine, and thereby provide an L-cysteine-producing bacterium, as well as a method for producing L-cysteine, L-cystin«, a derivative or precursor thereof, or a mixture of these using such a bacterium. The inventors of the present invention conducted various researches in order to achieve the aforementioned object, and as a result, found that L-cysteine-producing ability of a bacterium could be improved by modifying the bacterium to decrease activity of a protein encoded by the >'cA'gene, and the L-cysteine-producing ability could be further improved by modifying the bacterium to decrease activity of a protein encoded by they//K gene in addition to the foregoing protein, and accomplished the present invention, The present invention thus provides the followings. (1) A bacterium belonging to the family Enterobacleriaceae, which is able to produce I,-cysteine, and lins been modified to decrease the activity of the YdjN protein, (2) Jhc bacterium as described above, wiierein the YdjN protein htis the amino iieiti sotiiience ol'.SBQ ID NO: 2 or 4, or a variant thereof. (3) The bacterium as described above, which has been further mocJIfieU to decrease the activity of the FliY protein. (4) The bacterium as described above, wherein the FHY protein has the amino acid sequence of SEQ ID NO: 6 or 8, or a variant tiiereof. (5) The bacterium as described above, wherein the activity of the YdjN or FliY protein is decreased by reducing expression of the j'tWor/f/J'gene coding for the protein, or by disrupting the gene. (6) The bacterium as described above, wherein the ydjN gene is selected from the group consisting of: (a) a DNA comprising the nucleotide sequence of SEQ ID NO: 1 or 3, (b) a DNA which is able to hybridize with a sequence complementary to the nucleotide sequence of SEQ ID NO: 1 or 3, or a probe which is prepared from the nucleotide sequence under stringent conditions, (c) a DNA which has a homology of 95% or more to the nucleotide sequence of SEQ ID NO: lor 3. (7) The bacterium as described above, wherein the/Kgene is selected from the group consisting of: . (d) a DNA comprising the nucleotide sequence of SEQ ID NO: 5 or 7, (e) a DNA which is able to hybridize with a sequence complementary to the nucleotide sequence of SEQ ID NO: 5 or 7, or a probe which is prepared from the nucleotide sequence under stringent conditions, (0 a DNA which has a homology of 95% or more to the nucleotide sequence of SEQ ID NO: 5 or 7. (8) The bacterium as described above, which ftirther has at least one of the following characteristics: i) it has been modified to increase serine acetyltransferase activity, ii) it has been modified to increase expression of the yeaS gene, iii) it has been modified to increase 3-phosphoglycerate dehydrogenase activity, iv) it has been modified to enhance activity of the sulfate/thiosulfate transport system. (9) The bacterium as described above, which is belongs to the genus Pantoea. (10) The bacterium as described above, which is Pantoea ananatis, (11) The bacterium as described above, which is Escherichia coli. (12) A method for producing L-cysteine, L-cystine, a derivative or precursor thereof, or a mixture of these, which comprises culturing the bacterium as described above in a medium and collecting L-cysteine, L-cystine, a derivative or precursor tliereof, or a mixture of these from the medium. According to the present invention, L-cysteine-producing ability of bacteria belonging to the family Enterobacteriaceae can be improved. Furthermore, according to the present invention, L-cysteine, L-cystine, derivatives and precursors thereof, and mixtures of them can be efficiently produced. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 shows uptake of S-suUbcysteine by TL coll MO 1655. Fig. 2 shows uptake of cystine by E. coll MGI655. Fig. 3 shows uptake of cysteine by E. coll MGI655. T''ig. 4 shows uptake of S-sulfocysteine by P. ananatixyjN, Fig. 5 shows uptake of cystine by/Z/K-dellcient E. coll Fig. 6 shows uptake of cystine by /7/K-enlianced E. coll. Fig. 7 shows uptake of cysteine by/f/l'-doficient E. coll. Fig. 8 shows the sequence of the promoter Pnlp (SEQ ID NO: 62), DF;TAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS <1> Bacterium The bacterium of the present invention belongs to the family Enterobacteriaceae, which is able to produce L-cysteine, and has been modified to decrease activity of the YdjN protein. A preferred embodiment of the bacterium of the present invention is the bacterium as described above, which has been further modified to decrease the activity of the Fli Y protein in addition to the YdjN protein. The YdjN and FliY proteins are encoded by ihefllYmdydjN genes, respectively. These proteins and genes will be explained later. The L-cysteine-producing ability refers to an ability of the bacterium to produce L-cysteine in a medium or cells and cause accumulation of L-cysteine in such an amount that L-cysteine can be collected from the medium or clls when the bacterium is cultured in the medium. Furthermore, a bacterium having L-cysteine-producing ability means a bacterium which can produce and cause accumulation of a larger amount L-cysteine in a medium or cells as compared with a wild-type, parent, or unmodified strain, preferably a microorganism which can produce and cause accumulation of L-cysteine in a medium in an amount of, preferably 0.3 g/L or more. more preferably 0.4 g/L or more, particularly preferably 0.5 g/L or more. A portion of the L-cysteine produced by the microorganism may be converted into L-cystine in the medium by the formation of a disulfide borid. Furthermore, as described below, S-sulfocysteine may be generated by the reaction of L-cysteine and thiosulfuric acid which are present in the medium (Szczepkowski T.W., Nature, vol, 182 (1958)). Furthermore, L-cysteine generated in bacterial cells may be condensed with a ketone, aldehyde, or, for example, pyruvic acid, which is present in the cells, to produce a thiazolidine derivative via a hemithioketal (refer to Japanese Patent No. 2992010). Thiazolidine derivative and hemithioketal may exist as an equilibrated mixture. Therefore, the ability to produce L-cysteine is not limited to the production of only L-cysteine in a medium or cells, but also includes the production of L-cystine or a derivative or precursor thereof, or a mixture of these, in addition to L-cysteine,. Examples of the aforementioned derivative of L-cysteine or L-cystine include, for example, S-sulfocysteine, thiazolidine derivatives, hemithioketals, and so forth. Examples of the precursor of L-cysteine or L-cystine include, for example, O-acetylserine, which is a precursor of L-cysteine, The precursors of L-cysteine or L-cystine also include derivatives of the precursors, and examples include, for example, N-acetylserine, which is a derivative of 0-acetylserine, and so forth. O-Acetylserine (OAS) is a precursor of L-cysteine biosynthesis. OAS is a metabolite of bacteria and plants, and is produced by acetylation of L-serine induced as an enzymatic reaction catalyzed by serine acetyltransferase (SAT). OAS is further converted into L-cysteine in cells. The ability to produce L-cysteine can be inherent to the bacterium, or it may be obtained by modifying a microorganism such as those described below by aiulageneiiis or a recombinant DNA technique. In the present invention, unless specially mentioned, the term L-cysteine may be used to refer to reduced type L-cystelne, L-cystinc, a derivative or precursor such as those mentioned above or a mixture thereof, The bacterium used for the present invention is not particularly limited so long as the bacterium belongs to the family Entemhacteriaceae such as those of the genera fischericlUa, EnUirohacter, Pmioea, Klebsiella, Serratia, Erwinla SalmonelUi and Morgamlla, and has L-cysteinc-producing ability, Specifically, (hose cUlssilled into the family Enterobacteriaceae according to the taxonomy used in the NCHf (National Center for Biotechnology Information) database (http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id«9l347) can be used. As a parent strain of the family Enterobacteriaceae used for the modification, a bacterium of the genus Escherichia, Enterobacter, Pantoea, Erwtnia, or Klebsiella can be used. Although the Escherichia bacteria are not particularly limited, specifically, those described in the work of Neidhardt et a(. (Backmann B.J., 1996, Derivations and Genotypes of some mutant derivatives oi Escherichia coli K-12, p.2460-2488. Table I, In F.D. Neidhardt (ed.), Escherichia coli and Salmonella Cellular and Molecular Biology/Second Edition, American Society for Microbiology Press, Washington, D.C.) can be used, Escherichia coli is preferable. Examples of Escherichia coli include bacteria derived from the prototype wild-type strain, K12 strain, such as Escherichia coll W3110 (ATCC 27325), Escherichia coli MG1655 (ATCC 47076) and so forth. These strains are available from, for example, the American Type Culture Collection (Address: 12301 Parklawn Drive, Rockville, Maryland 20852, P.O. Box 1549, Manassas, VA 20108, United States of America). That is, registration numbers are given to each of the strains, and the strains can be ordered by using these registration numbers (refer to http://viww.atcc.org/). The registration numbers of the strains are listed in the catalogue of the American Type Culture Collection. Examples of the Enterobacter bacteria include Enierobacter agglomerans, Enierobacter aerogenes and so forth, and examples of the Pantoea bacteria include Pantoea ananatis. Some strains of Enterobacter agglomerans were recently reclassified into Pantoea agglomerans, Pantoea ananatis, or Pantoea stewartii on the basis of nucleotide sequence analysis of 16S rRNA etc. A bacterium belonging to the genus Enterobacter or Pantoea may be used so long as it is classified into the family Enterobacter iaceae. In particular, Pantoea bacteria, Erwinia bacteria, and Enterobacter bacteria are classified as y-proteobacteria, and they are taxonomically very close to one another (J. Gen. Appl. Microbiol., 1997,43, 355-361; International Journal of Systematic Bacteriology, Oct. 1997, pp.1061-1067). In recent years, some bacteria belonging to the genus Enterobacter were reclassified as Pantoea agglomerans, Pantoea dispersa, or the like, on the basis of DNA-DNA hybridization experiments etc. (International Journal of Systematic Bacteriology, July 1989,39(3), pp.337-345). Furthermore, some bacteria belonging to the genus Erwinia were reclassified as Pantoea ananas or Pantoea stewartii (refer to International Journal of Systematic Bacteriology, Jan. 1993; 43(1), pp. 162-173). Examples of the Enterobacter bacteria include, but are not limited to, Enterobacter agglomerans, Enterobacter aerogenes, and so forth. Specifically, the strains exemplified in European Patent Publication No, 952221 can be used. A typical strain of the genus Enterobacter is the Enterobacter agglomeranses ATCC 12287 strain. Typical strains of tlie Pantoea bacteria incliKle, but are not limited to, J'cmioea iuumatis, ParHoaa .siewartii, Pantoea agglomerans, and Pantoea citmi, Specific exmiiples of Paniova ananatix include tlie Pantoea ananatis AJ13355 strain, SC i 7 strain, rind SC 17(0) strain. The SC 17 strain was selected as a low plilegm-firoducing miilant strain from the AJ13355 strain (PERM BP-6614) isolated iVom soil in Iwnta-Bhl, Sliizuoktt-ken, Japtin as a strain that can proliferate in a low pH ntedtum containing L-glutamic acid and a carbon source (U.S. PatpntNo. 6,596,517). The SC17(0) strain was constructed to be resistant to the X Red gene product for performing gene disruption in Pantoea ananatis (WO2008/075483). The SC17 strain was deposited at the National Institute of Advanced Industrial Science and Technology, International Patent Organism Depository (address: Tsukuba Central 6,1-1, Higashi 1-Chome, Tsukuba-shi, Ibaraki-ken, 305-8566, Japan) on February 4,2009, and assigned an accession number of PERM ABP-11091. The SCI 7(0) strain was deposited at the Russian National Collection of Industrial Microorganisms (VKPM), GNU Genetika (address: Russia, 117545 Moscow, 1 Dorozhny proezd. 1) on September 21,2005 with an accession number of VKPM B-9246. The Pantoea ananatis AJ13355 strain was deposited at the National Institute of Bioscience and Human-Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry (currently, the National Institute of Advanced Industrial Science and Technology, International Patent Organism Depositary, Address: Tsukuba Central 6,1-1, Higashi 1-Chome, Tsukuba-shi, Ibaraki-ken, 305-8566, Japan) on February 19,1998 and assigned an accession number of PERM P-16644, It was then converted to an International deposit under the provisions of Budapest Treaty on Januaiy 11,1999 and assigned an accession number of PERM BP-6614. This strain was identified as Enterobacter agglomerans when it was isolated and deposited as the Enierobacter agglomerans AJ13355 strain. However, it was recently reclassified as Pantoea ananatis on the basis of nucleotide sequencing of 16S rRNA and so forth. Examples of the Erwinia bacteria include, but are not limited to, Erwinia amylovora and Erwinia carotovora, and examples of tlie Klebsiella bacteria include Klebsiella planticola. Impartation or enhancement of L-cysteine-producing ability Hereinafter, methods for imparting L-cysteine-producing ability to bacteria belonging to Enterobacteriaceae, or methods for enhancing L-cysteine-producing ability of such bacteria, are described. To impart the ability to produce L-cysteine, methods conventionally employed in the breeding of coryneform bacteria or bacteria of the genus Eschericfiia (see "Amino Acid Fermentation", Gakkai Shuppan Center (Ltd.), 1st Edition, published May 30, 1986, pp. 77-100) can be used. Such methods include by acquiring the properties of an auxotrophic mutant, an analogue-resistant strain, or a metabolic regulation mutant, or by constructing a recombinant strain so that it overexpresses L-cysteine biosynthesis enzyme. Here, in the breeding of an L-cysteine-producing bacteria, one or more of the above described properties such as auxotrophy, analogue resistance, and metabolic regulation mutation may be imparted. The expression of L-cysteine biosynthesis enEyme(s) can be enhanced alone or in combinations of two or more. Furthermore, the methods of imparting properties such as an auxotrophy. analogue resistance, or metabolic regulation mutation may be combined with enhancement of the biosynthesis enzymes. An auxotrophic mutant strain, L-cysteine unalogue-resistant strain, or metabolic regulation mutant strain with the ability to produce L-cysteine can be obtained by subjecting a parent strain or wiUI-type strain to conventional mutatagenesis, such as exposure to X-rays or UV irradiation, or treatment with a mutagen such as N-melhyl-N'-nitro-N-nitrosoguanidine, ethyl methanesulfonate (EMS) etc, and thcii Hclecling (hose which exhibit uutotrophy, analogue resistance, or a metabolic reguUxlon muialion and which also have (he ability to produce L-cysteine iVom the oblainod nuaant strains. L-Cysteine-producing ability of a bacterium can be improved by enhancing activity of an enzyme of the L-cysteine biosynthesis pathway or an enzyme involved in production of a compound serving as a substrate of that pathway such as L-serine, for example, 3-phosphoglycerate dehydrogenase, serine acetyltransferase. and so forth. 3-Phosphoglycerate dehydrogenase is subject to feedback inhibition by serine, and therefore the enzymatic activity thereof can be enhanced by incorporating a mutant serA gene coding for a mutant 3-phosphoglycerate dehydrogenase for which the feedback inhibition is eliminated or attenuated into a bacterium. Furthermore, serine acetyltransferase is subject to feedback inhibition by L-cysteine. Therefore, the enzymatic activity thereof can be enhanced by incorporating a mutant cysE gene coding for a mutant serine acetyltransferase for which the feedback inhibition is eliminated or attenuated into a bacterium. The L-cysteine-producing ability can also be improved by enhancing the activity of the sulfate/thiosulfate transport system. The sulfatc/thiosulfate transport system protein group is encoded by the cysPTWAM gtnt cluster (Japanese Patent Laid-open No. 2005-137369, European Patent No. 1528108). The L-cysteine-producing ability of a bacterium can also be improved by increasing expression of the yeaS gene (European Patent Laid-open No. 1016710). The nucleotide sequence of the j/ea5gene and the amino acid sequence encoded by the gene are shown in SEQ ID NOS: 15 and 16, respectively. It is known that bacteria use various codons such as OTG besides ATG as the start codon (http://depts.washington.edu/agro/genoraes/students/stanstart.htm). Although the amino acid corresponding to the initial codon gtg is indicated as Val in SEQ ID NOS: 15 and 16, it is highly possible that it is actually Met. Specific examples of L-cysteine-producing bacteria include, but not limited to, E. coli JMl 5 transformed with multiple kinds of cysE gene alleles encoding serine acetyltransferase (SAT) resistant to feedback inhibition (U.S. Patent No. 6,218,168), E. coli W3110 in which a gene encoding a protein responsible for excretion of cytotoxic suKstances is overexpressed (U.S. Patent No. 5,972,663), E. coli strain having decreased cysteine desuUliydrase activity (Japanese Patent Laid-open No, 11-155571), and E. coli W3110 in which activity of the positive tran.scriptional control factor of the cysteine regulon encoded by the cysB gene is increased (WOOl/27307). For E. coli, proteins are knovm which have an activity of secreting L-cysteine, such as the protein encoded hyydeD (Japanese Patent Laid-open No. 2002-233384), the protein encoded by y/Ar (Japanese Patent Laid-open No. 2004-49237) and the proteins encoded by emrAB, emrKY,yojIH, acrEF, bcr, and cusA, respectively (Japanese Patent Laid-open No. 2005-287333) as described above. Activities of these L-cysteine secreting proteins can be increased. Hereafter, as the method for imparting an ability to produce L-cysteine, enhancing an activity of L-cysteine biosynthesis system enzyme will be described. Examples of the L-cysteine biosynthesis enzyme include, for example, serine acetyltransferase (SAT). The SAT activity in cells of a bacterium belonging to the family Enterobacleriaceae can be enhanced by increasing the copy number of a gene coding for SAT, or modifying an expression control sequence sucli as promoter of the gone coding for SAT. For example, a recombinant DNA can be prepared by ligating a gene I'ragment coding for SAT with a vector, such us a multi-copy vector, which is able to l\mction in the chosen host bacterium belonging to the family Enterohacterlaceae to prepare it recombinant DNA. This recombinant DNA can then bo introduced into a host bacterium belonging to the family Enterobacleriaceae to transform it. Methods for enhancing expression of the SAT gene will be described below. Similar methods can also be applied to other L-cysteine biosynthesis systems enzyme genes, iXxQycaS gene, and genes of proteins having cysteine secretion activity. Modification for enhancing expression of the SAT gene can be attained by, for example, increasing copy number ofthe SAT gene in tlie cells by means of genetic recombination techniques. For example, a recombinant DNA can be prepared by ligating a DNA fragment containing the SAT gene with a vector, preferably a multi-copy vector, which is able to function in a host bacterium, and introduced into a bacterium to transform it. For example, the SAT gene of Escherichia coli can be obtained by PCR using chromosomal DNA of Escherichia coli as a template and primers prepared on the basis of the nucleotide sequence of SEQ ID NO: 9. The SAT genes of other bacteria can also be obtained from chromosomal DNA or chromosomal DNA library ofthe bacteria by hybridization using a probe prepared on the basis ofthe aforementioned sequence information. Copy number ofthe SAT gene can also be increased by introducing multiple copies ofthe SAT gene into a chromosomal DNA of a bacterium. To introduce multiple copies ofthe SAT gene into a chromosomal DNA of a bacterium, homologous recombination can be performed with targeting a sequence present on a chromosomal DNA in a multiple copy number. A repetitive DNA or inverted repeat present at the end of a transposable element can be used as the sequence present on a chromosomal DNA in a multiple copy number. Alternatively, as disclosed in Japanese Patent Laid-open No, 2-109985, multiple copies ofthe SAT gene can be introduced into a chjomosomal DNA by incorporating them into a transposon and transferring it. Furthermore, besides the amplification of copy number of gene described above, expression ofthe SAT gene can also be enhanced by replacing an expression regulatory sequence ofthe SAT gene .such as a promoter on a chromosomal DNA or a plasmid with a stronger promoter, by amplifying a regulator which increases expression of the SAT gene, or by deleting or attenuating a regulator which reduces expression of the SAT gene. As strong promoters, for example, lac promoter, irp promoter, ire promoter and so forth are known. Furthermore, a promoter ofthe SAT gene can also be modified to be stronger by introducing substitution of nucleotides or the like into the promoter region of the SAT gene. The aforementioned substitution or modification of the promoter enhances expression of the SAT gene. Examples of methods for evaluating strength of promoters and strong promoters are described in an article by Goldstein and Doi (Goldstein, M.A, and Doi R.H., 1995, Prokaryotic promoters in biotechnology, Biotechnol. Annu. Rev., 1,1 OS-128), and so forth. Modification of an expression regulatory sequence can be combined with the increasing copy number of the SAT gene. Furthermore, in order to enhance production ofthe SAT protein, a mutation can be itiiroduced near the translation initiation site of the SAT gene lo increfise translation elficienoy, and this can be combined with enhancement of expression of the SAT gene. Increase ol'expression of the SAT gene and increase of the SAT protein amount can Us conflrmod by qimntifyiiig mRNA or by Western blotting using an antibody, as the uonliimation ol'decreasc in transcription amount ofn target gene and the cortfimiation of decrease in a target protein amount described Inter. As the SAT gene, an SAT gene derived from Escherichia bacteria or an SAT gene derived from other organisms can be used. As the gene coding for SAT of Escherichia coU, cycE has been cloned from a wild-type strain and an L-cysteine excretion mutant strain, and the nucleotide sequence thereof has been elucidated (Denk, D. and Boeck. A.. J. General Microbiol., 133, 515-525 (1987)). The nucleotide sequence thereof and the amino acid sequence encoded by the nucleotide sequence arc shown in SEQ ID NOS: 9 and 10, respectively. A SAT gene can be obtained by PCR utilizing primers prepared based on the nucleotide sequence and chromosomal DNA of Escherichia bacterium as the template (refer to Japanese Patent Laid-open No. 11-155571). Genes coding for SAT of other organisms can also be obtained in a similar manner. Expression of the SAT gene as described above can be enhanced in the same manner as that for the cysE gene explained above. When a suppression mechanism such as "feedback inhibition by L-cysteine" exists for the expression of the SAT gene, expression of the SAT gene can also be enhanced by modifying an expression regulatory sequence or a gene involved in the suppression so that the expression of the SAT gene is insensitive to the suppression mechanism. For example, the SAT activity can be further increased by mutating the SAT so that the feedback inhibition by L-cysteine is reduced or eliminated in the bacterium (henceforth also referred to as "mutant SAT"). Examples of the mutant SAT include SAT having a mutation replacing an amino acid residue corresponding to the methionine residue at position 256 of a wild-type SAT (SEQ ID NO: 10) with an amino acid residue other than lysine residue and leucine residue, or a mutation deleting a C-terminus side region from an amino acid residue corresponding to the methionine residue as position 256. Examples of the amino acid residues other than lysine and leucine include the 17 amino acid residues which typically make up proteins except for methionine, lysine and leucine. Isoleucine and glutamic acid are more preferred examples. To introduce a desired mutation into a wild-type SAT gene, site-specific mutagenesis can be used. As a mutant SAT gene, a mutant cysE coding for a mutant SAT of Escherichia coU is known (refer to International Patent Publication W097/15673 and Japanese Patent Laid-open No. 11 -155571). Escherichia coli JM39-8 strain harboring a plasmid pCEM256E containing a mutant cysE coding for a mutant SAT in which methionine residue at position 256 is replaced with a glutamic acid residue (£. coli JM39-8(pCEM256E), private number: AJ13391) was deposited at the National Institute of Bioscience and Human-Technology, Agency of Industrial Science and Technology (Postal code: 305,1-3 Higashi l-Chome, Tsukuba-shi, Ibaraki-ken, Japan) on November 20,1997 and assigned an accession number of PERM P-16527. The deposit was then converted to an international deposit under the provisions of Budapest Treaty on July 8, 2002, and assigned an accession number of PERM BP-8112, Although a "SAT insensitive to feedback inhibition by L-cysteine" can be SAT which has been modified so that it is insensitive to the feedback inhibition by L-cysteine, it can be a SAT which in its native form is insensitive to the feedback inhibition by L-cysteine. Por example, SAT oi'Arahidopsls thciliana is known to be not subject to the feedback inhibition by L-cysteine and can be suitably used for the present invention, As a plasmid containing the SAT gene derived from Arabidopsis thaliana, pEAS-m is known (FEMS Microbiol. Lett., 179 (1999) 453-459). Purlhermore, the obility to produce L-cysteine can also be Improved by enhancing expression of the cysPTWAM cluster genes coding for the sulfato/thlosiiirulo Irnnspoit system proteins (Japanese Patent Laid-open No. 2005-137369, GP 1528108). Purthermore, a sulfide can be incorporated into O-acctyl-L-serine via a reaction catalyzed by the 0-acetylserine (thiol)-lyase A or B encoded by the cysK and cysM genes, respectively, to produce L-cysteine. Therefore, the ability to produce L-cysteine can also be improved by enliancing expression of the genes coding for these enzymes. Moreover, L-cysteine-producing ability can also be improved by suppressing the L-cysteine decomposition system. "L-cysteine decomposition system is suppressed" can mean that intracellular L-cysteine decomposition activity is decreased as compared to that of a non-modified strain such as a wild-type or parent strain. As proteins responsible for the L-cysteine decomposition system, cystathlonine-p-lyase (metC product, Japanese Patent Laid-open No, 11 -155571, Chandra et al., Biochemistry, 21 (1982) 3064-3069), tryptophanase QnaA product, Japanese Patent Laid-open No. 2003-169668, Austin Newton et al., J. Biol. Chem., 240 (1965) 1211-1218)), O-acetylserine sulfhydrylase B (cyA/gene product, Japanese Patent Laid-open No. 2005-245311) and the malYgene product (Japanese Patent Laid-open No. 2005-245311) are known. By decreasing the activities of these proteins, L-cysteine-producing ability can be improved. Modification for decreasing activity of a protein can be attained in the same manner as tliose for thefliYorydjN gene described later. The nucleotide sequence of the cy.sM gene of Escherichia coli and the amino acid sequence encoded by the gene are shown in SEQ ID NOS: 25 and 26, respectively. Decrease of activities of YdjN protein and FliY protein The bacterium of the present invention can be obtained by modifying such a bacterium belonging to the family Enterobacteriaceae and Itaving L-cysteine-producing ability as described above to decrease the activity of the YdjN protein or the activities of the YdjN protein and the FliY protein. After a bacterium is modified to decrease tlie activity of the YdjN protein or the activities of the YdjN and FliY proteins, L-cysteine-producing ability may be imparted to the bacterium. The YdjN protein and the FliY protein are proteins encoded by the ydjNgene and xhcfliYgene, lespectively. The activities of the YdjN and FliY proteins of a bacterium can be decreased by, for example, modifying the bacterium having they/ZK anAydjN genes to decrease the activities of FliY and YdjN encoded by these genes. In order to enhance the L-cysteine-producing ability, either the FliY activity or the YdjN activity may be decreased, but it is preferable to decrease the YdjN activity, and it is more preferable to decrease both the activities. In the present invention, "decrease" of activity include decrease of the activity of a modified strain to a level lower than that of a wild-type or a non-modified strain, and complete disappearance of the activity, unless especially specified. The inventors of the present invention found novel genes coding for proteins of which deletion from the cluomosomal DNA of Pctntoea amnalis enhanced the I,-cysteine-producing ability, and designated ihtmfliY mdyciJN, respectively, since they showed high homology XofliY and ydJN of E. coli (78% and 80%, respectively). In this specification, "homology" may means "identity". In the present invention, in addition to tho///)'and >*A genes o(E. coUJUY iu\i\ yclJN genes of Pantoett ammtis, and horaologue genes of those genes of other bacteria may also be calledy/iKgene and ydjN gene, respectively. Specific examples of thej/Kgene include a gene comprising the nucleotide sequence shown in SEQ ID NO: 5 or 7. Specific examples of the y'rf/Wgene of the Escherichia coli MG1655 strain is sliown in SEQ ID NO: I, and the aniino acid sequence encoded by the gene is shown in SEQ ID NO: 2. The nucleotide sequence of the>'<55Agene of the Pantoea ananatis SCI7 strain is shown in SEQ ID NO: 3, and the aniino acid sequence encoded by the gene is shown in SEQ ID NO: 4. The FliY and YdjN proteins are not limited to proteins having the aforementioned amino acid sequences and homologues thereof, and they may be a variant thereof Ths fliY ox yjN tna may be a gene coding for a variant of the FliY or YdjN protein. A variant of the FliY or the YdjN protein means a protein having the amino acid sequence of SEQ ID NO: 2,4,6 or 8 including substitutions, deletions, insertions or additions of one or several amino acid residues at one or several positions, and having the function of the FH Y or YdjN protein. Although the number meant by the aforementioned term "one or several" may differ depending on positions of amino acid residues in the three-dimensional structure of the protein or the types of amino acid residues, specifically, it is preferably I to 20, more preferably 1 to 10, still more preferably 1 to 5. A.vfW gene-deficient strain shows decreased uptake of S-sulfocysteine and L-cystine as shown in Examples section. Therefore, it is estimated that the YdjN protein has a function to participate in uptake of S-sulfocysteine and L-cystine. On the other hand, although there is a reference pointing out the possibility of participation of FliY in uptake of L-cystine (Butler et al., Life Sci., 52,1209-1215 (1993); Hosie et al., Res. Microbiol., 152,259-270 (2001)), it is estimated that it does not participate in the uptake of L-cystine, or the activity thereof is lower than tliat of YdjN, if it participates in that uptake, as shown in Examples section. In any case, if both the ydjN and JliY genes are deleted, the L-cysleine-producing ability is markedly increased as compared with that obtainable by deletion of either one of them. Therefore, although the function of FliY is still indefinite, it is characterized that deletion thereof improves the L-cysteine-producing ability. The aforementioned substitutions, deletions, insertions, or additions of one or several amino acid residues are a conservative mutation that preserves the normal function of the protein. The conservative mutation is typically a conservative substitution. The conservative substitution is a mutation wherein substitution takes place mutually among Phe, Trp and Tyr, if the substitution site is an aromatic amino acid; among Leu, He and Val, if the substitution site is a hydrophobic amino acid; between Gin and Asn, if it is a polar amino acid; among Lys, Arg and His, if it is a basic amino acid; between Asp and Glu, if it is an acidic amino acid; and between Ser and Thr, if it is an amino acid having a hydroxyl group. Specific examples of conservative substitutions include: substitution of Ser or Thr for Ala; substitution of Oln, His or Lys for Arg; substitution of Glu, Gin, Lys, His or Asp i)r Asn; substitution of Asn, CJiu or Oln for Asp; substitution of Ser or Ala for Cys; substitution of Asn, Glu, Lys, His, Asp or Arg for Gin; substitution of Gly, Asn, Gin, Lys or Asp for Glu; substitution of Pro for Gly; substitution of Asn, Lys, Gin, Arg or Tyr for His; substitution of f.-cu, Met, Val or I'bi) lor lie; substitution of lie, Met, Val or Phe for I -eu; substitution of Asn, Glu, Gin, His or Arg for Lys; substitution of He, Leu, Val or Phe for Met; substitution of Trp, tyr. Met, He or Leu for Phe; substitution of Thr or Ala for Ser; substitution of Ser or Ala for Thr; substitution of Phe or 'lyr for Trp; substitution of His, Phe or Trp for Tyr; and substitution of Met, He or Leu for Val. The above-mentioned amino acid substitution, deletion, insertion, addition, inversion etc. can be the result of a naturally-occurring mutation (mutant or variant) due to an individual difference, a difference of species, or the like of a bacterium from which the gene is derived. Furthermore, the gene having such a conservative mutation as described above can be a gene encoding a protein showing a homology of 80% or more, preferably 90% or more, more preferably 95% or more, still more preferably 97% or more, particularly preferably 99% or more, to the entire encoded amino acid sequence. Sequence information of the genes coding for a protein homologous to such FliY or YdjN can be easily obtained from databases opened to public by BLAST searching or PASTA searching using the v/M-typcfliY or ydjN gene of the aforementioned Escherichia coil strain as a query sequence, and the genes can be obtained by using oligonucleotides produced based on such known gene sequences as primers. The///Tor YdjNgtne can be a gene which hybridizes with a sequence complementary to the aforementioned nucleotide sequences or a probe that can be prepared from the aforementioned nucleotide sequences under stringent conditions, so long as the function of the protein encoded by they7/Tor YdjNgetiQ is maintained. Examples of the "stringent conditions" include conditions of washing at 60°C, 1 SSC, 0.1% SDS, preferably eoC, O.lxSSC, 0.1% SDS, once or preferably twice or three times. The probe used for the aforementioned hybridization can have a partial sequence of a complementary sequence of the gene. Such a probe can be prepared by PCR using oligonucleotides prepared based on the known nucleotide sequences of the gene as primers, and a DNA fragment containing these sequences as the template. When a DNA fragment of a length of about 300 bp is used as the probe, the conditions of washing alter hybridization can be, for example, 50°C, 2xSSC, and 0.1% SDS, Methods for decreasing the activities of the FliY or YdjN protein will be explained below, The activities of the proteins of the L-cysteine decomposition system can also be decreased by the same methods. In the following descriptions, an objective protein of which activity is to be decreased is referred to as a "target protein", and a gene coding for the target protein is referred to as a "target gene". Activity of a target protein can be decreased by, for example, reducing expression of a target gene. Specifically, tor example, intracellular activity of the target protein can be reduced by deleting a part of, or the entire coding region of the target gene on a chromosome. For decrease of activity of a target protein, expression of the target gene can also be decreased by modifying an expression control sequence of the target gene such as promoter and Shine-Dalgarno (SD) sequence. Furthermore, the expression amount of the gene cun also be reduced by modification of a non-translation region otlier than the expression control sequence. Furthermore, the entire gene including the sequences on both sides of the gene on a chromosome can be deleted. Furthermore, the expression of the gene can also bo reduced by introducing a mutation for an amino acid substitution (missenise mutation), a stop codon (nonaenso mutation), or a frame shifi nuitatlon which adds or deletes one or two nucleotides into the coding regitm of llie target gene on a chromosome (Journal of Biological Chemisti'y, 272:8611-8617 (1997); Proceedings of the National Academy of Sciences, USA, 95 5511 -5515 {1998); .Journal ofBiological Chemistry, 266,2083.1-20839 (1991)). Activity of a target protein can also be decreased by enhancing activity of a regulator which down-regulates the target protein, or suppressing activity of a regulator which up-regulates the target protein. Activity of a target protein can also be decreased by adding a substance which down-regulates activity or expression of the target protein, or eliminating a substance which up-regulates activity or expression of the target protein. Furthermore, the modification can be a modification caused by a typical mutagenesis caused by X-ray or ultraviolet irradiation, or by use of a mutagen such as N-methyl-N'-nitro-N-nitrosoguanidine, so long as the modification results in a decrease of the activity of the target protein. Modification of an expression control sequence is performed for preferably one or more nucleotides, more preferably two or more nucleotides, particularly preferably three or more nucleotides. When a coding region is deleted, the region to be deleted can be an N-terminal region, an internal region or a C-terminal region, or even the entire coding region, so long as the function of the target protein is decreased or deleted. Deletion of a longer region can usually more surely inactivate a gene. Furthermore, reading frames upstream and downstream of the region tc be deleted can be the some or different. To inactivate a gene by inserting a sequence into the coding region of the gene, the sequence can be inserted into any part of the coding region of the gene. The longer the inserted sequence, the greater the likelihood of inactivating the gene. Reading frames located upstream and downstream of the insertion site can be the same or diflferent. The sequence to be inserted is not particularly limited so long as the insertion decreases or deletes the function of the encoded target protein, and examples include, for example, a transposon carrying an antibiotic resistance gene, a gene useful for L-cysteine production and so forth. A target gene on the chromosome can be modified as described above by, for example, preparing a deletion-type version of the gene in which a partial sequence of the gene is deleted so that the deletion-type version of the gene does not produce a target protein which normally functions, and transforming a bacterium with a DNA containing the deletion-type gene to cause homologous recombination between the deletion-type gene and the native gene on the chromosome, and thereby substitute the deletion-type gene for the gene on the genome. The target protein encoded by the deletion-type gene has a conformation different from that of the wild-type protein, if it is even produced, and thus the function is reduced or deleted. Such gene disruption based on gene substitution utilizing homologous recombination has been already established, and there are a method called Red-driven integration (Datsenko, K.A., and Wanner, B.L., Proc. Natl. Acad. Sci. USA, 97:6640-6645 (2000)), a method of using a linear DNA such as a method utilizing the Red driven integration in combination with an excision system derived from X phage (Cho, E.H., Oumport, R.I., Gardner, J.F., J. Bacteriol., 184:5200-5203 (2002)) (refer to WO2005/010175), a method of using a plasmid containing a temperature sensitive replication origin or a plasmid capable of conjugative transfer, a method of utilizing a suicide vector not having replication origin in a host (U.S. Patent No. 6,303,383, Japanese Patent Laid-open No. 05-007491), and so forth. Decrease of transcription level of a target gene can be confirmed by comparing amount olmRNA transcribed from the gene with that of the wild-type or non-modified strain. Examples of the method for confirming mRNA amount include Northern hybridization. RT-PCR (Molecular Cloning, Cold Spring Harbor Laboratory Press, Cold Spring Harbor (USA), 2001), and so forth. Pocrcaso of amount of a target protein can also be confirmed by Western blotting using tin antibody (Molecular Cloning, Cold Spring Harbor Labomtory Pre»a, Cold Spring Harbor (USA), 2001). Furthermore, when the target protein is the YdjN protein, decrease of the amount of the protein can also be confirmed by measuring activity to take up S-sulfocysteine or L-cystine of the cell. Whether a protein has the activity to take up the aforementioned compound can be confirmed by preparing a bacterium in which expression of a gene coding for the protein is increased from a wild strain or a parent strain, culturing this strain in a medium, and quantifying amount of L-cysteine, L-cystine, a derivative or precursor thereof, or a mixture of them accumulated in the medium. Alternatively, the activity can also be confirmed by preparing a bacterium in which expression of a gene coding for the protein is decreased or deleted from a wild-type or parent strain, culturing the strain in a medium containing S-sulfocysteine or L-cystine, and confirming decrease of the decreased amount of the compound added to the medium. Specific examples are described in Examples section. When thQ fliY or ydJN gene of Escherichia coli is used as i\\e fltY or ydJN gene, thefliY oxydjN gene can be obtained by PCR using chromosomal DNA of Escherichia coli as a template and primers prepared on the basis of the nucleotide sequence of SEQ ID NO: 5 or 1. Similarly, UnsfliYotydiN&nt of Pantoea ananatis can be obtained by PCR using chromosomal DNA of Pantoea ananatis as a template and primers prepared on the basis of the nucleotide sequence of SEQ ID NO: 7 or 3. The JJiYorydJN gene of other bacteria can also be obtained from chromosomes or chromosomal DNA library of the bacteria by hybridization or PCR using a probe or primers prepared on the basis of the aforementioned sequence information. When it is necessary to increase expression of tht fliY orydjN gene in order to confirm whether the FHY or YdjN protein has the activity to take up S-sulfocysteine, L-cystine or L-cysteine, multiple copies of the gene can be introduced into a bacterium. In order to introduce multiple copies of they7/y or ydjN gene into a bacterium, the method of using a multi-copy type vector, the method of introducing multiple copies of genes into chromosomal DNA by homologous recombination, and so forth can be used as described for the SAT gene. <2> Method for producing L-cysteine, L-cystine, derivative or precursor thereof or mixture thereof of the present invention These compounds can be produced by culturing the bacterium of the present invention obtained as described above in a medium, and collecting L-cysteine, L-cystine, a derivative or precursor thereof or a mixture thereof from the medium. Examples of the derivative or precursor of I,-cysteine include S-sulfocysteine, a thiazolidine derivative, a hemithioketal coitesponding tlie thiazolidine derivative mentioned above, and so fortii. Examples of the medium used for the culture can include ordinary media containing a carbon source, nitrogen source, sulftir source, inorganic ions, and other organic components as required, As the carbon source, saccharides such as glucose, fructose, sucrose, molasses and starch hydrolysate, and organic acids such as fumarlc acid, citric acid and suocltilc acid can be used. As the nitrogen source, inorganic ammoniuni salts such as ammonium sulfate, ammonium chloride and ammonium phosphate, organic nitrogen such as soybean hydrolysate, ammonia gas, aqueous ammonia and so forth can be used. As the sulfur source, inorganic sulfur compounds, such as sulfates, sulfites, sulfides, hyposulfites and thiosulfates can be used. As organic trace amount nutrients, it is desirable to add required substances such as vitamin B|, yeast extract and so forth in appropriate amounts. Other than these, potassium phosphate, magnesium sulfate, iron ions, manganese ions and so forth are added in small amounts. The culture is preferably performed under aerobic conditions for 30 to 90 hours. The culture temperature is preferably controlled to be at 25°C to 37*0, and pH is preferably controlled to be 5 to 8 during the culture. For pH adjustment, inorganic or organic acidic or alkaline substances, ammonia gas and so forth can be used. Collection of L-cysteine from the culture can be attained by, for example, any combination of known ion exchange resin methods, precipitation and other known methods, L'Cysteine obtained as described above can be used tor production of L-cysteine derivatives. The cysteine derivatives include methylcysteine, ethylcysteine, carbocysteine, sulfocysteine, acetylcysteine, and so forth. Furthemtore, when a thiazolidine derivative of L-cysteine is accumulated in the medium, L-cysteine can be produced by collecting the thiazolidine derivative from the medium to break the reaction equilibrium between the thiazolidine derivative and L-cysteine so that L-cysteine is excessively produced. Furthermore, when S-sulfocysteine is accumulated in the medium, it can be converted into L-cysteine by reduction with a reducing agent such as dithiothreitol. L-Cysteine, a derivative thereof, and so forth collected in the present invention may contain cells of microorganism, medium components, moisture, and by-products of microbial metabolism in addition to the objective compound. Purity of the collected objective compound is 50% or higher, preferably 85% or higher, particularly preferably 95% or higher. Examples Hereinafter, the present invention will be explained more specifically. In the followfing descriptions, cysteine means L-cysteine. Example I: Identification of protein having activity to take up cysteine or cystine (1) Acquisition of mutant strain unable to utilize S-sulfocysteine as sole cysteine source (1-1) Acquisition of cysE gene-deficient strain from E. coli MG1655 strain (ATCC No. 47076) Tthe cysE gene was deleted by the method called "Red-driven integration" developed by Datsenko. Wamter el al. (Proc. Natl. Acad. Sci. USA, 2000, vol. 97. No. 12, pp.6640-6645) and the excisive system derived from X, phage (J. Bacteriol., 2000, 184, 5200-5203 (2002)). According to the Red-driven integration, a gene-disrupted strain can be constructed in one step by using a PCR product obtained with synthetic oligonucleotides designed so as to have a part of an objective gene on the 5' side, and a part of an antibiotic resi.*iitance gene on the .3' side. 13y further combining the excisive system derived IVom X,-phage, the antibiotic resistance gene incorptirated into the geno-disruptud strain can be eliminated. Methods for deleting a gene of £ coll using this Reci-drlven integration and the excisive system derived from X-phago are described in dettill in Japanese Patent Laid-open No. 2005-058227, WO2007/! 19880, and so forth. A tysE gene-deCicient strain was also obtained in Ihc same manner. A DNA fragment containing an antibiotic resistance gene (kanamycin resistance gene (Km')) between sequences homologous to the both ends of the cysE gene was obtained by PCR. Specific experimental procedure and experimental materials were the same as those described in Japanese Patent Laid-open No. 2005-058227, except that DcysE(Ec)-F (ccggcccgcg cagaacgggc cggtcattat ctcatcgtgt ggagtaagca tgaagcctgc ttttttatac taagttggca, SEQ ID NO: 50), and DcysE(Ec)-R (actgtaggcc ggatagatga ttacatcgca tccggcacga tcacaggaca cgctcaaglt agtataaaaa agctgaacga, SEQ ID NO: 51) were used as primers, and pMWll8-(AattL-Km'-XattR) (WO2006/093322A2) was used as a template. The obtained deficient strain was designated MG1655AcysE. (1-2) Preparation of transposon-mutatcd strain library from the M01655AcysE strain By using EZ-Tn5Tnp Transposome Kit (EPICENTRE), a libraiy of mutant strains in which I'nS was randomly inserted was prepared from the MGl 655AcysE strain. As for specific experimental procedure, the experiment was performed according to the instruction attached to the kit. (1-3) Screening of mutant strain library for mutant strain unable to utilize S-sulfocysteine as sole cysteine source The aforementioned library was screened for a mutant strain unable to utilize S-sulfocysteine as a sole cysteine source. The "cysteine source" herein refers to a substrate which is taken up into cells and used for the production of cysteine. When cysteine cannot be synthesized in cells, cysteine itself is encompassed by the cysteine source. The mutant strains were each spotted with a toothpick on eiich of the M9 agar medium (Sambrook and Russell, Molecular Cloning: A Laboratory Manual (Third Edition), Cold Spring Harbor Laboratory Press) containing 20 [xM cysteine and the M9 agar medium containing 20 iM S-sulfocysteine (cat# C2196, SIGMA) to perform screening for a mutant strain able to grow on a cysteine-containing medium, but unable to grow on an S-sulfocysteine-containing medium. One strain was obtained as the target mutant strain from about 1000 strains. When the genome region of inserted Tn5 was identified, it was found that Tn5 was inserted into the ydJN gene. (1 -4) Analysis of S-sulfocysteine-assimiiating ability ofydjN gene-deficient strain In order to elucidate whether the phenotype of the mutant strain obtained by insertion of Tn5 was due to functional deficiency of the>'af/gene, &ydjN gene-deficient strain was constructed from the MGl655AcysE strain by the Red-driven integration described above. For this construction, primers DydjN(Ec)-F (cactatgact gctacgcagt gatagaaata ataagatcag gagaacgggg tgaagcctgc ttttttatac taagttggca, SEQ ID NO: 52), and DydjN(Ec)-R (aaagtaaggc aacggcccct atacaaaacg gaccgttgcc agcataagaa cgctcaagtt agtataaaaa agctgaacga, SEQ ID NO; 53) were used. The constructed deficient strain was designated MG1655AcysEAydjN::Kni strain, MGI655Ayd)N::Km strain was also obtained from MGl655 by the same method. Growth of the strains on the M9 agar medium (provided that MgCh was used instead of MgSO at the same concentration as the medium component) containing 50 (.iM cysteine, 50 jiM cystine or 50 \xM S-sulfocysteine is shown In Table 1. Although any sulfur source was not added to the M9 agar medium in this experiment (M9 w/o sulfur), the cv.v£-non-dismpted strains (MG1655 strain, M01655AydjN::Km strain etc,) can grow with a trace amount of contaminating sulfur compounds. Therefore, in order to investigate whether use of S-sulfocysteine as a cysteine source is possible or not, the bacicground of the cysE deficiency should be determined. That is, the MG1655 AcysE strain cannot grow without cysteine source (w/o sulfur), but it can grow by using cysteine, cystine or S-sulfobysteine as a cysteine source, if they exist. It was found that the >'cf/Adeficient strain could grow with cysteine or cystine as a cysteine source, but it could not grow with S-sulfocysteine as a cysteine source (Table 1, MG 1655AcysEAydjN:: Km strain). From this result, it was found that ydjN was a gene indispensable to the assimilation of S-sulfocysteine. On the other hand, even if j'' was deleted, cysteine and cystine could be assimilated. (2) Functional analysis of ydjN and fliY (2-1) Cloning of>i"jVgene from E. coli MG1655 strain and P. anmafis SC17 strain When xhsydjN gene was cloned from the £'. coli MG1655 strain and the Panloea ananatis SCI7 strain (U.S. Patent No, 6,596,517), oqiression vectors based on pMlV-Pnlp8 and pMIV-PnlpO were used. The potent nlp8 promoter (or nlpO promoter) and an rmB terminator were integrated into these expression vectors, and by inserting a target gene between the promoter and the terminator, it can be functioned as an expression unit. "PnlpO" indicates a promoter of a wild-type nlpD gene, and "Pnlp8" indicates a mutant promoter of the nlpD gene. The details of the construction of diese expression vectors are described in Example 3 as construction of pMIV-Pnlp8-YeaS7 and pMIV-PnlpO-YeaS3. In pMIV-Pnlp8-yeaS7 and pMIV-Pnlp0.yeaS3, the yeaS gene is cloned between the nlp8 or nlpO promoter and the rmB tenminator using the Sail and Xbal sites. If the Sail and Xbal sites are designed in primers beforehand, the ydjN gene can also be inserted into those vectors in the same manner as that fox yeaS. That is, the expression plasmids to be constructed correspond to expression plasmids having structures of pMIV-Pnlp8-yeaS7 and pMIV-l'nlpO-yeaS3 mentioned later in which the>'c'a.y gene is replaced by the yclJN gene. The yd/N gene of £'. coli was amplified by using the genomic DNA of the MG1655 strain as a template as well as;'£i)W(Ec)-SallFW2 (acgcgtcgac atgaactttc caltaattgc gaacatcgtg gtg, SEQ ID NO: 54) and;£///V(Kc)-xbaIRV2 (ctagtctaga Uaatggtgt gccagttcgg cglcg, SEQ ID NO: 55) as primers with a PCR cycle consisting of Q'C for 5 minutes, followed by 30 cycles of 98°C for 5 seconds, 55''C for 5 seconds and 72°C for 90 seconds, and final incubation at 4''C. In the case of P. amnalis, the yc/'N gene was lunpllficd by using the genomic DNA of the SC17 strain as a template m well as ydjN2(l>a)-SalII< W (acgcgtcgac atggalattc ctcttacgu. SEQ ID NO: 56) and ydjN2(Pa)-xbalRV (tgctctagal tagctgtgct ctaattcac, SEQ ID NO: 57) as primers with a PCR cycle consisting of 94"C for 5 minutes, followed by 30 cycles of PSC for 5 seconds, 55°C for 5 seconds and 72''C for 2 minutes, and final incubation at 4°C. Sail and Xhal sites were designed at the both ends in all the primers. The amplified fragments were each integrated into the pMIV-PnlpO vector, and the constructed plasmids were designated according to the origin of the gene (E. coli (Ec) or P. ananatis (Pa)) as pMIV-PnlpO-ydjN(Ec) and pMIV-PnlpO-ydjN(Pa), respectively. Furthermore, as an empty vector corresponding to these plasmids for control, pMIV-5JS (Japanese Patent Laid-open No. 2008-99668) was used. (2-2) Functional analysis ofycffN When ydjN was deleted, the .strains could not grow with S-sulfocysteine as a sole cysteine source. Therefore, it was suspected that the yd/N gene might code for a transporter (uptake factor) of S-sulfocysteine. Therefore, it was examined whether there was any difference in the ability to take up S-sulfbcysteine between the ydjN' deficient strain of the MG1655 strain (MG1655AydjN::Km strain described above) and >'£?/-enhanced strain of the MG1655 strain (MG1655 strain transformed with pMIV-PnlpO-ydjN(Ec)), Similar investigation was also conducted for cystine and cysteine as related compounds similar to S-sulfocysteine. The S-sulfocysteine uptake experiment was performed as follows. First, the MG1655AydjN::Km strain and a control strain therefor, MG1655 strain, as well as the MG1655/pMIV-PnlpO-ydjN(Ec) strain and a control strain therefor, MG1655/pMIV-5JS strain, were cultured overnight In the LB liquid medium (3-ml test tube, 37°C, shaking culture). The cells were collected from the culture medium, washed twice with the M9 minimal medium containing 0.4% glucose, and then suspended in the M9 minimal medium containing 0.4% glucose at a density twice higher than that of the original culture medium. The cell suspension of each strain prepared as described above was inoculated in a voliune of 40 \i\ to 4 ml of the M9 minimal medium containing 0.4% glucose* and culture was performed at 37°C with shaking by using an automatically OD measuring culture apparatus, BIO-PHOTORECORDER TN-1506 (ADVANTEC). When OD reached around 0.3 (culture for about 5 hours), 20 \il of 100 aiM S-sultbcysteine was added (final concentration: 0.5 mM), and the medium was sampled (0.2 ml of the culture medium was taken and mixed with 0.8 ml of 1 N hydrochloric acid) over time for 2 hours after the addition of S-sulfocysteine (0 hour). Amino acid analysis of the sample of each time point was performed with an amino acid analyzer (L-8900, Hitachi), and S-sulfocysteine concentration in the medium was determined by comparison with a 0.4 mM standard sample similarly prepared with I N hydrochloric acid. In the culture, 25 mg/L of chloramphenicol was added to the medium for all the plasm id-harboring strains, Change of S-sulfocysteine concentration in the culture medium of each strain is shown in Fig. 1. In the graph, the MGl655/pMlV-PnlpO-ydjN(Ec), MGI655/pMIV-5JS strain, and MG1655AydjN::Km strains are abbreviated as M01655/ydjN(Ec)-plasmid, MO 1655/vector and MO 1655 delta-ydjN, respectively. It was found that the S-.«iuIfocysteine concentration in the medium gradually decreased with the wild-type strain, but S-sulfocysteine did not decrease at all with the ciV-doflcient strain, and decrease of S-sulfocysteine was accelerated with t\\e ydJN-evi\mcwi strain as comparetl with the control strain. Furthermore, when YdJN was analyzed with a membrune protein prediction program SOSUI (http;//bp.nunp.nagoya-u.acjp/sosui/), ten tniUHmembraiio domains were found in YdjN, and therefore il WHS* O.\ peeled to be a membrane protein. The results described above strongly suggested a poauibility that ycflN coded for a transporter (uptake factor) of S-sulfocysteine. Furthermore, since >'<;//W deficiency invited incapability of assimilating S-sulfocysteine (Table l),>'c/7Vis considered to be a sole S-sulfocysteine transporter in E. colt. Furthermore, uptake of cystine or cysteine by YdjN was also examined by using the same experiment system and adding cystine or cysteine as a substrate instead of S-sulfocysteine, The results are shown in Figs. 2 and 3. The strain names in the graphs are the same as those used in Fig. 1. As shown in Fig. 2, the same results as those for S-sulfocysteinc were obtained when cystine was used as the substrate, and therefore it was suggested that YdjN had the activity to take up cystine. In the >'£W-deficient strain, uptake of cystine markedly decreased, and the residual activity to take up cystine became extremely weak. Therefore, it is considered that YdjN is a major transporter of cystine in the E. coli MCil655 strain. However, since the yf/yW-deficient strain could also grow with cystine as the cysteine source (Table 1), it was considered that there was an active transporter of cysteine other than at least YdjN. On the other hand, when cysteine was used as the substrate, the uptake was not promoted even when >'c//?/was enhanced as shown in Fig. 3, and therefore it was suspected that it might not participate in the uptake of cysteine. Furthermore, the plasmid pMrV-PnlpO-ydjN(Pa) expressing j/iTV derived from P. ananatis was introduced into E. coli and P. ananatis to enhance ><;SAf, and uptake of S-sulfocysteine under such enhancement of yc/N was examined. The results are shown in Fig. 4. In the graph, pMIV-PnlpO-ydjN(Pa) and pMIV-5JS are abbreviated as ydjN(Pa)-pIasmid and vector, respectively. It was confirmed that YdjN of P ananatis also had the activity to take up S-sulfocysteine, like YdjN ofE. coli. Furthermore, the amino acid sequences of YdjN of P. ananatis and YdjN of £ coli show a homology of 80%. (2-3) Functional analysis 6{fliY Then, in order to examine whetheryj/T participated in uptake of cystine and cysteine, rT-deficient and enhanced strains were constructed first. Deletion of the/y/K gene was attained by using the aforementioned Red-driven integration and exclsive system derived from X-phage. DfliY(Ec)-FW (atgaaattag cacatctggg acgtcaggca ttgatgggtg tgatggccgt tgaagcctgc ttttttatac taagttggca, SiiQ ID NO: 58) and DfliY(Ec)'<:A gene into pMIV-PnlpO mentioned above, but in this experiment, for amplification of theZ/Kgene, fliY(Ec)SalI-F (acgcgtcgac atgaaattag cacatctggg acg, SEQ ID NO: 60) and niY(Ec)XbaI-R (ctagtctaga ttatttggtc acatcagcac c, SEQ tD NO: 61) were used as primers. First, in order to investigate the effect of///K deficiency on the cystine uptake, an uptake experiment was performed by using cystine as a substrate with 4 strains, the Mai655, MG1655AydjN::Km, M01655AfliY, and MG1655Ay4iN!;KmAniY::Cm stniins. The results are shown in Fig, 5, hi the graph, "WT" represents the MG1655 strain, "dcUa-ydiN" represents the M01655AydiN::Km strain, "delta fliY" represents the MGl655AfliY strain, and "delta-ydjN.fllY" represents the MG1655AydjN::KmAfliY::Cm strain. As a result, the cystine uptake rate of the MG1655AfliY strain, in which only the///T gene was deleted, was not different from that of the MO 1655 strain (Fig. 5). Furthermore, although »t was observed that the cystine uptake rate of the MG1655AydjNAfli Y strain, which was nfliV and ydjN gene-double deficient strain, seemed to decrease as compared with that of the MG1655AydjN strain, in which only ydJN was deficient, the difference was very small, and therefore it was unclear whether it was a significant difference induced by they/ZK deficiency (Fig. 5). Furthermore, uptake of cystine by they7/y-enhanced £. coli MGl 655 strain was also investigated. The results are shown in Fig. 6, In the graph, "fliY(Ec)-ptasmid" represents pMIV-I*nlpO-fliY(Ec). and "vector" represents pMlV-5JS. Also when./7/r was enhanced, significant difference in uptake of cystine was not observed (Fig. 6). Although there were references suggesting a possibility of participation of fliY in uptake of cysteine (Butler et al., Life Sci., 52,1209-1215 (1993) etc.), there were no findings directly proving it through experiments, in fact, results showing participation of FliY in uptake of cysteine were not obtained also in this experiment. Furthermore, from the results of this experiment, it was predicted that even if FliY participated in uptake of cystine, it was not a highly active transporter, like YdjN. In addition, since the yd/N and//iT double deficient strain could grow with cystine as a sole cysteine source, it is considered that a transporter of cystine exists besides these two kinds of transporters. Furthermore, although the cysteine uptake activity of the /7/K-deficient strain was also examined, such a significant difference as seen {oxy'Nv/aB not seen compared with the non-deficient strain, and it was considered that FliY did not participate in the uptake of cysteine (Fig. 7). On the other hand, since cysteine in the medium gradually decreased (Fig. 7), uptake of cysteine into cells was expected, and it is supposed that a certain cysteine transporter (uptake system) exists in E. coli. Example 2: Cysteine production by >'2 (agctgagcut gctlccaact gcgctaatga cgc, SEQ ID NO: 31) as primers, a DNA fragment containing a promoter region of the nIpD gene (PnlpO) of about 300 bp was obtained. At the 5' and 3' ends of the aforementioned primers, sites for the restriction enzymes .SV»/1 and Pael were designed, respectively. The PCR cycle was as follows: 95°C for 3 minutes, then 2 cycles of 95°C for 60 seconds, SCC for 30 seconds, and 72°C for 40 seconds, 25 cycles of 940 for 20 seconds, SSC for 20 seconds, and 72°C for 15 seconds, and 72°C for 5 minutes as the final cycle. The obtained fragment was treated with Sail and Pael, and inserted into pMIV-5JS (Japanese Patent Laid-open No. 2008-99668) at the Sall-Pael site to obtain a plasmid pMlV-PnlpO, The nucleotide sequence of the Pael-SaU fragment of the PnlpO promoter inserted into this pMIV-PnlpO plasmid was as shown in SEQ ID NO: 27. Then, by PCR using the chromosomal DNA of M01655 as a template, as well as P3 (agctgatcta gaaaacagaa tttgcctggc ggc, SEQ ID NO: 32) and P4 (agctgaggat ccaggaagag tttgtagaaa cgc, SEQ ID NO: 33) as primers, a DNA fragment containing a tenninator region of the rrnB gene of about 300 bp was obtained. At the 5' ends of the aforementioned primers, sites for the restriction enzymes Xbdl and BamHl were designed, respectively. The PCR cycle was as follows: 95"'C for 3 minutes, then 2 cycles of 95°C for 60 seconds, 50°C for 30 seconds, and 72'*C for 40 seconds, 25 cycles of 94''C for 20 seconds, 59°C for 20 seconds, and 72°C for 15 seconds, and 72°C for 5 minutes as the final cycle. The obtained fragment was treated with Xbal and BamHl, and inserted into pMIV-PnlpO at the Xbal-BamUl site to obtain a plasmid pMIV-PnlpO-ter. Then, by PCR using the chromosomal DNA of the M01655 strain as a template, as well as P5 (agctgagtcg acgtgttcgc tgaatacggg gt, SEQ ID NO: 34) and P6 (agctgatcta gagaaagcat caggattgca gc, SEQ ID NO: 35) as primers, a DNA fragment of about 700 bp containing the yeaS gene was obtained. At the 5' ends of the aforementioned primers, sites for the restriction enzymes SafL and Xbal were designed, respectively. The PCR cycle was as follows: 95°C for 3 minutes, then 2 cycles of 95 for 60 seconds, SOC for 30 seconds, and 72''C for 40 seconds, 25 cycles of 94°C for 20 seconds, 55°C for 20 seconds, and 72''C for 15 seconds, and 72°C for 5 minutes as the final cycle. The obtained fragment was treated with Sail and Xhal, and inserted i nto pMIV-PnlpO-ter at the Sall-Mal site to obtain a plasmid pMIV-PnlpO-YeaS3. As described above, &yeaS expression unit comprising the pMIV-5JS vector on which the nlpD promoter, the yeaS gene, and the rmB terminator were ligated in this order was constructed. In order to modify the -10 region of the nlpD promoter to make it a stronger promoter, the -10 region was randomized by the following method. The nlpD promoter region contains two of regions presumed to function as promoters (Fig, 8), and they are indicated as pnlpl and pnlp2, respectively, in the drawing. By PCR using the plasmid pMIV-PnlpO as a template as well as PI and P7 (atcgtgaaga tcttttccag tgttnannag ggtgccttgc acggtnatna ngtcactgg ("n" means that the corresponding residue can be any of a, t, g and c), SEQ ID NO: 36) as primers, a DNA fragment in which the -10 region contained in the 3' end sequence of the nIpD promoter (referred to as -10(l*nlpl)) was randomized was obtained, The PCR cycle was as follows: 95°C for 3 minutes, then 2 cycles of 95°C for 60 seconds, .50°C for 30 seconds, and 72°C for 40 seconds. 25 cycles of 94''C for 20 seconds, 60%: for 20 seconds, and 72''C for 15 seconds, and 72''(2 for 5 minutes as the final cycle, rurthermore, by PCR using the plasmid pMlV-PnIpO as a template as well as P2 and PK (tgguaitagat cttcannnnn cgctgacctg eg ("n" means that Iho corresponding residue can be any of a, t, g and c), SEQ ID NO: 37) as primers, a DNA fragment in which the -10 region contained in the 5' end sequence of the nlpD promoter (referred to as -10(Pnlp2)) was randomized was similarly obtained (Fig, I). The PCR cycle was as follows: 95°C for 3 minutes, then 2 cycles of 95°C for 60 seconds, SCC for 30 seconds, and 72''C for 40 seconds, 25 cycles of 94'*C for 20 seconds, eCC for 20 seconds, and 72°C for 15 seconds, and 72°C for 5 minutes as the final cycle. The obtained 3' and 5' end fragments could be ligated using the BglU sites designed in the primers P7 and P8, and the full length of the nlpD promoter in which two -10 regions were randomized could be constructed by such ligation. By PCR using this fragment as a template as well as PI and P2 as primers, a DNA fragment corresponding to a modified type nlpD promoter of the full length was obtained. The PCR cycle was as follows: 95''C for 3 minutes, then 2 cycles of 95''C for 60 seconds, 50°C for 30 seconds, and 72 for 40 seconds, 12 cycles of 94°C for 20 seconds, 60°C for 20 seconds, and 72°C for 15 seconds, and 72°C for 5 minutes as the final cycle. The amplified fragment was treated with the restriction enzymes Sail and Pael, for which sites were designed in the 5' ends of the primers, and inserted into the plasmid pMlV-PnlpO-YeaS3 similarly treated with Sail and Pael to substitute the mutant Pnlp for the wild-type nlpD promoter region (PnlpO) on the plasmid. From such plasmids, one having the promoter sequence (PnlpS) shown in SEQ ID NO: 28 was selected, and designated pMIV-Pnlp8-YeaS7 (the nucleotide sequence of the Pael-Sall fragment of the PnlpS promoter inserted into this plasmid was as shown in SEQ ID NO: 28), In the same manner, a DNA fragment of the nlpD promoter region containing a mutation was inserted into tiie plasmid pMlV-PnlpO-ter treated with Sal] and Pael to substitute the mutant Pnlp for the nlpD promoter region (region of PnlpO) on the plasmid. One of them was designated pMIV-Pnlp23-ter. The nucleotide sequence of the Pael-Sall fragment of the Pnlp23 promoter inserted into this plasmid was as shown in SEQ ID NO: 29. Then, from pMW-Pomp-cysE5 (WO2005/007841), the Pomp-cysE5 cassette portion was excised with Pael and Sad, and inserted into the same site of pMlV-5JS to construct pMIV-Pomp-CysE5. pMW-Pomp-cysE5 was a plasmid obtained by inserting the cysES gene coding for the mutant SAT Ugated with the ompC gene promoter into pMWllS. From pACYC184 (GcnBank/EMBLaccession number X06403, available from NIPPON GENE), the tetracycline resistence gene was excised with Xbal and £co88I, and this gene fragment was treated with the Klenow fragment, and then inserted into pMIV-Pomp-CysE5 at the Pvul site to construct pMT-Pomp-CysES. Then, pMIV-Pnlp8-YeaS7 was digested with Hindlll, blunt-ended with the Klenow fragment, and then digested with Ncol to excise a fragment containing the cassette of the Pntp8-YeaS-rmB terminator and the chloramphenicol resistance marker. This fragment was ligated with a Sntal and Ncol digestion firagment of pMT-Pomp-CysE5 similarly having pMIV-5JS as the backbone to construct pMT-EY2. pMT-EY2 is a plasmid having the Pnlp8-YeaS-r/*«B terminator cassette and the Pomp-CysE5 cassette on one plasmid. pMT-EY2 described above has the attachment sites of Mu phage originated from pMlV-5JS (Japanese Patent Laid-open No. 2008-99668). By allowing this plasmid to coexist with the helper plasmid pMHI 0 having Mu tronsposaso (Zitrienkov I), et HI„ Biotcchnologiya and (in Russian). 6, I -22 (2004)) in the same cell, the cassette oJ'PompC-cy8B5-Pnlp8-YcaS-rmB terminator including the chloramphenicol resistance murker locutud between the attachment sites of Mu phage on this pMT-EY2 plasmid can bo Inserted into the chromosome of the P ammtis SCI7 strain (U.S. Patent No. 6,596,517). Furthermore, since the chloramphenicol resistance marker located on the pM'r-EY2 plasmid exists between two attachment sites of X phage (UttR and XattL), the chloramphenicol resistance marker can be excised and removed by the method described later. First, an SCI7 strain introduced with pMHlO by electroporatlon was selected by overnight culture at 30°C on the LB agar medium containing 20 mg/L of kanamycin. The obtained transformant was cultured at 30*C, and pMT-EY2 was further introduced into this strain by electroporation. This strain transformed with both pMHI 0 and pMT-EY2 was given a heat shock with the conditions of 42°C for 20 minutes, and colonies of chloramphenicol resistant strains were selected on the LB jjgar medium containing 20 mg/L of chloramphenicol. The culture temperature for this selection was 39°C. As described above, about 50 clones were obtained, and curing of pMHlO and pMT-EY2 was performed by culturing each clone at 39'C for 48 hours on the LB agar medium. A strain showing chloramphenicol resistance due to the insertion of the cassette un the chromosome and showing kanamycin and ampicillin sensitivities due to the curing of both the plasmids was obtained. Furthermore, it was confirmed that tlie objective cassette was inserted in the chromosome of the obtained strain by PCR using the chromosomal DNA of this strain as a template as well as PI and P6 as primers. All the obtained clones were designated EYOl to EY50, respectively, and L-cysteine production culture was performed by using the EYOl to EY50 strains as described below. The EY19 strain was selected, which was a clone that produced L-c>steine in the largest amount as a result of the culture. An L-cysteine production medium (composition: 15 g/L of ammonium sulfate, 1.5 g/L of potassium dihydrogenphosphate, 1 g/L of magnesium sulfate heptahydrate, 0.1 g/L of tryptone, 0.05 g/L of yeast extract, 0.1 g/L sodium chloride, 20 g/L of calcium carbonate, 40 g/L of glucose, and 20 mg/L of tetracycline) was used for the culture. The L-cysteine production culture was performed by the following procedure. The SC17/pMT-PompCysE5 strain and SC17/pMT-EY2 strain were each applied on the LB agar medium to perform preculture overnight at 34°C, then cells corresponding to 1/8 of the plate were scraped with an inoculation loop, inoculated into 2 ml of the L-cysteine production medium contained in a large test tube (internal diameter: 23 mm, length: 20 cm), and cultured at 32''C with shaking at 220 to 230 rpm, and the culture was terminated after two days. The chloramphenicol resistance marker introduced into the EY19 strain was removed with an excision system derived from X, phage. Specifically, the EY19 strain was transformed with pMT-Int-Xis2 (WO2005/010175) carrying the Inl-Xis gene of phage, and an EY19(s) strain showing chloramphenicol sensitivity was obtained from the obtained transformants. Examples of removal of a marker using the excision system derived from l phage are described in detail in Japanese Patent Laid-open No. 2005-058227, WO2007/119880, and so forth. (1-2) Preparation of cysPTWA gene expression-enhanced strain from EY19(8) strain Then, in order to enhance expression of the cysPTWA gene, the promoter located upstream of the cysPTWA gene cluster on the chromosome was replaced with the aforementioned potent promoter Pnlp8. A UNA fragment containing the nlp8 |)iiiniolcr ornliinil 300 bp was obtained by PCR using pMIV-PnlpH-YottS7 us u lomplato as well as PI and P2 as primers. The PCR cycle was as follows: 95°C for 3 minutes, then 2 cycles or95''C for 60 seconds, SCC for 30 seconds, and IT'C for 40 seconds, 20 cycles of 94''C Ibr 20 seconds, 59°C for 20 seconds, and 72°C for 15 seconds, and 72''C for 5 minutes as the final cycle. The amplified DNA fragment containing nlp8 promoter was treated with the Klenow fragment, inserted into the plasmid pMW118-(XattL-KmR-XattR) (WO2006/093322A2) digested with A*al and then treated with the Klenow fragment to obtain a plasmid pMW-Km-Pnlp8, By PCR using pMW-K.m-Pnlp8 as a template as well as P9 (tccgctcacg attttlttca tcgctggtaa ggtcatttat cccccaggaa aaattggtta, SEQ ID NO: 38) and PI 0 (tttcacaccg ctcaaccgca gggcataacc ggcccttgaa gcctgctttt ttatactaag ttg, SEQ ID NO: 39) as primers, a DNA fragment of about 1,6 kb containing the Km-Pnlp8 cassette was amplified. The PCR cycle for this amplification was as follows: 95°C for 3 minutes, then 2 cycles of 95°C for 60 seconds, 50°C for 30 seconds, and 72"'C for 40 seconds, 30 cycles of 94°C for 20 seconds, 54''C for 20 seconds, and 72°C for 90 seconds, and 72°C for 5 minutes as the final cycle. On both the primers, a sequence serving as a target on the chromosome for inserting an objective fragment by X- , dependent integration (the method called "Red-driven integration" (Proc. Natl. Acad. Sci. USA, 2000, vol, 97, No, 12, pp.6640-6645)) (in this case, a sequence near the promoter of cysPTWA) was designed. Therefore, if the obtained DNA fragment is inserted into the objective strain by this X-dependent integration, there is provided a structure that Km-Pnlp8 is inserted immediately before the cysPTWA gene on the chromosome, and the cysPTWA gene is ligated with the nlp8 promoter. The nucleotide sequence of the cysPTWA gene cluster is shown in SEQ ID NO: 19, and the amino acid sequences encoded by the cysP, cysTmd cysW genes are shown in SEQ ID NOS; 20 to 22, respectively. The nucleotide sequence of the cysA gene and the amino acid sequence encoded by tliis gene are shown in SEQ ID NOS: 23 and 24, respectively. The P ammtis SC17(0)/RSF-Red-TER strain is a host strain for efficiently performing the X-dependent integration, and it is a strain obtained by introducing the helper plasmid RSF-Red-TER which expresses gam, bet and exo genes of X (henceforth referred to as "X Red genes") into the SCI7(0) strain, which is a X Red gene product-resistant P ananatis strain (WO2008/075483). A method for constructing the RSF-Red-TER plasmid is disclosed in detail in WO2008/075483. The aforementioned SC17(0)/RSF-Red-TER strain was cultured with addition of IPTG for inducing expression of X Red genes to prepare cells for electroporation. The aforementioned objective DNA fragment was introduced into these cells by electroporation, and a recombinant strain in which the nlp8 promoter was inserted upstream of the cysPTWA gene by -dependent integration was obtained by using the kanamycin resistance as a marker. By PCR using tlie cliromosomal DNA of the obtained strain as a template, as well as P11 (ctttgtccct ttagtgaagg, SEQ ID NO: 40) and PI 2 (agctgatcta gaagctgact cgagttaatg gcctcccaga cgac. SHQ ID NO: 41) as primers, it was confirmed that the objective structure, Km-VvXi-cysPTWA, was fonned, and this strain was designated SC17(0)-Pnlp8-PTWA strain. Then, the chromosomal DNA of the SC 17(0)-Pnlp8-PTWA stt»in was purified, and 10 ng of this chromosomal DNA was introduced into the EY19(s) strain by electroporation to obtain a kanamycin resistant strain, Amplification was performed by PCR using the chromosomal DNA of the obtained strain as a template as well as PI 1 and PI2 as primers to confirm that the structure of Km-PnIpS-cysPTWA had been introduced into the chromosome of the BY! 9(s) strain. The strain obtained as described above was designated EYP197 .strain, Furthermore, the kanamycin rcsislnnce marker was removed from the chromosome by using pM 1-Int-Xi82 m described above, and the strain that became kanamycin sensitive was designated B YP197(s) strain. (1-3) Preparation of mutant 3-phosphoglycerate dehydrogenase (tierA348) gene-carrying strain of EYP197(s) strain As a gene of 3-phosphoglycerate dehydrogenase to be introduped into the L-cysteine-producing bacterium, the serA348 gene which is a gene coding for 3-phosphoglycerate dehydrogenase of Pantoea ananatis and coding for a mutant enzyme including a mutation for substitution of an alanine residue tor the asparagine residue at the 348th position (N348A) (J. Biol. Chem., 1996,271 (38):23235-8) was constructed by the following method. The sequence of the wild-type serA gene derived from Pantoea ananatis and the amino acid sequence are shown in SEQ ID NOS: 17 and 18, respectively. In order to obtain a 3'-end side DNA fragment of the serA gene into which the aforementioned mutation was introduced, PCR was performed by using the chromosomal DNA of the SC 17 strain as a template as well as P13 (agctgagtcg acatggcaaa ggtatcactg gaa. SEQ ID NO: 42) and P14 (gagaacgccc gggcgggctt cgtgaatatg cage, SEQ ID NO: 43) as primers (95*C for 3 minutes, then 2 cycles of 95°C for 60 seconds, SCC for 30 seconds, and 72°C for 40 seconds, 25 cycles of 94°C for 20 seconds, dCC for 20 seconds, and 72°C for 60 seconds, and 72°C for 5 minutes as the final cycle). Then, in order to obtain a 5'-end side DNA fragment into which the mutation was introduced, PCR was performed in the same manner by using the chromosomal DNA of the SCI7 strain as a template as well as PI5 (agctgatcta gacgtgggat cagtaaagca gg, SEQ ID NO: 44) and Pi 6 (aaaaccgccc gggcgttctc ac, SEQ ID NO; 45) as primers (95°C for 3 minutes, then 2 cycles of 95°C for 60 seconds, 50°C for 30 seconds, and 72°C for 40 seconds, 20 cycles of 94''C for 20 seconds, 60T for 20 seconds, and 72''C for 20 seconds, and 72°C for 5 minutes as the final cycle). Both the obtained PCR fragments were treated with the restriction enzyme Smal, and ligated by using a DNA llgase to obtain a DNA fragment corresponding to a ftill length mutant serA gene including the objective mutation (N348A). This DNA fragment was amplified by PCR using it as a template as well as PI3 and PI 5 as primers (95''C for 3 minutes, then 2 cycles of 95''G for 60 seconds, 50°C tor 30 seconds, and 72''C for 40 seconds. 15 cycles of 94T for 20 seconds, eOC for 20 seconds, and 72°C for 75 seconds, and 72''C for 5 minutes as the final cycle). The Sail and the Xbal restriction enzyme sites designed in the P13 and PIS primers were treated with Sail and Xbal, and the fragment was inserted into pMIV-Pii|p8-ter similarly treated with Sah and Xbal to prepare pMIV-Pnlp8-serA348. The constructed pMIV-Pnlp8-serA348 carried the attachment site of Mu originating in pMIV-5JS (Japanese Patent Laid-open No. 2008-99668). By using this plasmid together with the helper ptasmid pMHlO having Mu transposase, the cassette of Pnlp8-serA348-rmB terminator including the chloramphenicol resistance marker can be inserted into the cliromosome of the P. ananatls SCI 7 strain, as described above. The pMlV-Pnlp8-serA348 plasmid and pMHlO were introduced into the SCI7(0) strain to obtain a strain in which the cassette of Pnlp8-serA348-rmB terminator was inserted into the chromosome. By PCR using the primers PI and PI 5, it was confirmed that the objective cassette existed in the cells. The 3-phosphoglycerate dehydrogenase activity in cell extracts of about 50 of the obtained clones was measured, and a strain which showed the highest activity was selected, and designated SC17int-8erA348 strain. Then, 10 jig of the chromosomal DNA of the SC17int-serA348 strain was Introduced into the RYPI97(8) strain by clectroporation to obtain a chloramphenicol roslslunt strain, and by PCR using the primers PI and Pl5, It wos confirmed that the structure of Pnlp8-HorA348 had been introduced together with the chloramphenicol resistance miuker into the chromosome of the EYPl 97(s) strain. The strain obtained as described above was designated EYPS1976 strain. By the aforementioned method for removing marker using pMT-Int-Xis2, the chloramphenicol resistance marker was removed, and the strain that became chloramphenicol sensitive was designated EYPSl976(s) strain. (2) Construction of/lA/and/or>7/K-deficient cysteine-producing bacteria Strains deficient in ydjN and/or///K were constructed from the EYJ'S 1976(s) strain. Kydjhl gene-deficient strain and a,///)'region-deficient strain were prepared by X-dependent integration using the aforementioned E mamtis SCI 7(0)/RSF-Red-TER strain as a host bacterium. The yecS gene (nucleotide sequence: SEQ ID NO: 11, amino acid sequence: SEQ ID NO: 12) and eyecC gene (nucleotide sequence: SEQ ID NO: 13, amino acid sequence: SEQ ID NO: 14) are present downstream from the/Kgene, and these may possibly form an operon and function as an ABC transporter (information that;ec.S and yecC may form one transcription unit is mentioned in a database open to public on an internet website (http://ecocyc.org/)). Therefore, for the7?/K deficiency, it was decided to delete all three of these genes (/7JT flanking regions). To obtain a DNA fragment for eLydjN gene-deficient strain, a primer DydjN(Pa)-F (acctctgctg ctctcctgac cagggaatgc tgcattacat cggagttgct tgaagcctgc ttttttatac taagltggca, SEQ ID NO: 46) and a primer DydjN(Pa)-R (agacaaaaac agagagaaag acctggcggt gtacgccagg tctggcgtga cgctcaagtt agtataaaaa agctgaacga, SEQ ID NO: 47) were used, and to obtain a DNA fi-agment for afliY region-deficient strain, a primer DfliY-FW (atggctttct cacagattcg tcgccaggtg gtgacgggaa tgatggcggt tgaagcctgc ttttttatac taagttggca, SEQ ID NO: 48) and a primer DyecC-RV (ttacgccgcc aacttctggc ggcaccgggt ttattgatta agaaatttat cgctcaagtt agtataaaaa agctgaacga. SEQ ID NO: 49) were used. As a template, pMW 118-(XattL-Km'-XattR) (WO2006/093322A2) mentioned above was used, and PCR was performed at 94''C for 5 minutes, followed by 30 cycles of 94''C for 30 seconds, 55*C for 30 seconds and 72°C for 2 minutes and 30 seconds to obtain a DNA fragment comprising Km between the homologous sequences each used for the recombination. Each of the obtained DNA fi-agments was introduced into P. ananatis SCI 7(0)/RSF-Red-TER strain by electroporation to construct SCI 7(0)AydjN::Km strain (the XattL-Kmr-XattR fragment was inserted into the ''A gene region) and SCI 7(0)AfliY::Km strain (the XattL-Kmr-A.attR fragment was inserted into they7/Tgene region). Then, the EYPS1976(s) strain was transformed with the chromosomal DNAs prepared from the SCI 7(0)AydjN::Km strain and the SC17(0)AfliY::Km strain to obtain strains deficient in each gene, EYPSAydjN::Km strain and EYPSAfliY::Kra strain, fi'om the EYPS1976(s) strain. Furthermore. pMT-Int-Xis2 (WO2005/010175) mentioned above was introduced into the EYPSAniY:;Km strain, and the kanamycin resistance gene was excised by using the excisive system derived from "K phage to obtain kanamycin-sonsitive EYPSAfliY strain. Then, the EYPSAfliY strain was transformed with chromosomal DNA prepared the from the SC17(0)AydjN;:Km strain to obtain a dDiible deficient strain, nYPSAfliYAydjM:;Km strnin, In such a manner, the yc(fN gene-deficient strain, the.///K region-deficient strains, tuid the double dellclont strain for thcHc genos biwed on the cysteinc-prodiicing bacterium, BYPS197ft(H) strain, were conHlructcd. (3) Investigation of effect of >'fA'deficiency and,///)'deficiency in P. ananatis on cysteine production Culture for fermentative production was performed with the yrf/A gene-deficient strain, the .///I'region-deficient strains, and the double deficient strain for these genes obtained above, and production amounts of the cysteine related compounds were compared. For the culture, a P. ananatis cysteine production medium having tive following composition was used. R ananatis cysteine production medium (concentration of the components are final concentrations): Component J: (NH<,)2S04 15 g/L ICH2PO4 1.5 g/L MgS04-7H20 Ig/L Thiamine hydrochloride O.I mg/L Component 2: FeS04-7H20 l.Tmg/l. Na2Mo04-2H20 0.15 mg/L C0CI26H2O 0.7 mg/L MnCIHzO 1.6 mg/L ZnS04-7H20 0.3 mg/L CuS04-5H20 0.25 mg/L Components: Tryptone 0.6 g/L Yeast extract 0.3 g/L NaCl 0.6 g/L Component 4: Calcium carbonate 20 g/L Component 5: L-Histidine monohydrochloride monohydrate 135 mg/L Component 6: Sodium thiosulfate 6 g/L Component?: Pyridoxine hydrochloride 2 mg/L Component 8: Glucose 40 g/L For these components, 10-fold (Component 1), 1000-fold (Component 2), 100/6-fold (Component 3), 100-fold (Component 5), 350/6-fold (Component 6), 1000-fold (Component 7) and 10-foId (Component 8) concentration stock solutions were prepared, and mixed upon use, and the volume of the mixture was adjusted to a predetermined volume with sterilized water to obtain the final concentrations. Sterilization was performed by autoclaving at ri0"C for 30 minutes (Components 1,2, 3,5 and 8), hot air sterilization at ISO'C for 5 hours or longer (Component 4), and filter .sterilization (Components 6 and 7). The cysteine production culture was performed according to the following procedures. The cysleine-producing strains were each applied and spread on the LB agar medium, and precultured overnight at 34''C. Then, the cells corresponding to about 7 em on the plate were scraped twice with an inoculating loop of 10 |.il size (Blue l,oop, NUNC), and inoculated into 2 ml of the aforementioned P. imanatis cysteine production medium contained in a large test tube (Internal diameter: 23 mm, length: 20 cm). The amounts of inoculated cells were adjusted so that the cell amounts at the time of the start of the culture should be substantially the same. The culture was performed at 32''C with shaking, and terminated after 43 hours. At this time, complete consumption of glucose in the medium was confirmed. The quantification of cysteine accumulated in the medium was performed by the method described by Gaitonde, M.K. (Biochem. J., 1967 Aug., 104(2):627-33). The experiment was performed in hexaplicate for each of the strains, and averages and standard deviations of the results are shown in Table 3. For the sUrain decient in only JliY, increase of the cysteine related compounds was not observed, and for the strain deficient in only ydjN, the cysteine related compounds slightly increased. Furthermore, lor the double deficient strain for fliYmdydjN, increase of the cysteine related compounds was observed in a much higher degree compared with that obtainable with deficiency of one of the genes. It was found that ycfy'N deficiency in P. ananads was effective for the production of the cysteine related compounds, and fiirther combination of/7/K deficiency exerted a synergistic effect. Explanation of Sequence Listing SEQ ID NO: 1: Nucleotide sequence of ydj'N gene of Escherichia coli SEQ ID NO: 2: Amino acid sequence of YdjN of Escherichia coli SEQ ID NO: 3: Nucleotide sequence otydjN gene otPantoea atutmlis SEQ ID NO; 4: Amino acid sequence of YdjN of Pantoea ananatis SEQ ID NO: 5: Nucleotide sequence of fliY gene of Escherichia coli SEQ ID NO: 6: Amino acid sequence of FliY of Escherichia coli SEQ ID NO; 7: Nucleotide sequence of fliY gene of Pantoea ananatis SEQ ID NO: 8; Amino acid sequence of FliY of Pantoea ananatis SEQ ID NO: 9: Nucleotide sequence of cysE gene of Escherichia coli SEQ ID NO: 10: Amino acid sequence of SAT encoded by cysE gene of Escherichia coli SEQ ID NO: 11: Nucleotide sequence of yecS gene of Pantoea ananatis SEQ ID NO: 12: Amino acid sequence encoded hy yecS gene of Pantoea ananatis SEQ ID NO: 13: Nucleotide sequence of yecC gene of Pantoea ananatis SEQ ID NO: 14: Amino acid sequence encoded by yecC gene of Pantoea ananatis SEQ ID NO: 15: Nucleotide sequence of yeaS gene of Escherichia coli SEQ ID NO; 16; Amino acid sequence encoded by yeaS gene of Escherichia coll SEQ ID NO: 17: Nucleotide sequence of wM gone of Pantoea ananatis SEQ ID NO: 18: Amino acid sequence encoded by serA gene otPantoea ananatis SEQ ID NO; 19; Nucleotide sequence of cysPTWA gene cluster SfiQ ID NO: 20: Amino acid sequence encoded by cysP gene SEQ ID NO: 21: Amino acid sequence encoded by cysT gene SEQ ID NO: 22: Amino acid sequence encoded by cysW gene SEQ ID NO: 23; Nucleotide sequence ofcysA gene SEQ ID NO: 24: Amino acid sequence encoded by cysA gene SEQ ID NO: 25; Nucleotide sequence of cy-yA/gene Sl'IQ ID NO: 26: Amino acid sequence encoded by cysM gene SI'.Q 113 NO: 27: Nucleotide sequence of PnlpO SJ;Q ID NO: 28: Nucleotide sequence of Pnlp8 SlQ ID NO: 29: Nucleotide sequence of Pnlp23 SHQ ID NOS: 30 to 45: Primers PI to P16 SEQ ID NO: 46: Nucleotide sequence of primer DydjN(Pa)-F SEQ ID NO: 47: Nucleotide sequence of primer DydjN(Pa)-R SEQ ID NO: 48: Nucleotide sequence of primer Dfli Y-FW . SEQ ID NO: 49: Nucleotide sequence of primer DyecC-RV SEQ ID NO: 50: Primer DcysE(Ec)-F SEQ ID NO: 51: Primer DcysE(Ec)-R SEQ ID NO: 52: Primer DydjN(Ec)-F SEQ ID NO: 53: Primer DydjN(Ec)-R SEQ ID NO: 54: Primer ydjN(Ec)-SalIFW2 SEQ ID NO: 55; Primer ydjN(Ec)-xbaIRV2 SEQ ID NO: 56: Primer ydjN2(Pa)-SallFW SEQ ID NO: 57: Primer ydjN2(Pa)-xbaIRV SEQ ID NO: 58: Primer DfliY(Ec)-FW SEQ ID NO: 59: Primer DniY(Ec)-RV SEQ ID NO: 60: Primer fliY(Ec)SalI-F SEQ ID NO: 61: Primer niY(Ec)XbaI-R SEQ ID NO: 62: Nucleotide sequence of the promoter Pnlp. WHAT IS CLAIMED IS: 1. A bacterium belonging to the family Enterobacteriaceae, which is able to produce 1,-cysteine, and has been modified to decrease the activity of the YdjN protein. 2. The bacterium according to claim 1, wherein the YdjN protein has the amino acid sequence of SEQ ID NO: 2 or 4, or a variant thereof. 3. The bacterium according to claim 1 or 2, which has been further modified to decrease the activity of the FliY protein. 4. The bacterium according to claim 3, wherein the FliY protein has the amino acid sequence of SEQ ID NO: 6 or 8, or a viiriant thereof. 5. The bacterium according to any one of claims 1 to 4, wherein the activity of the Y4iN or FliY protein has been decreased by reducing expression of the,V(iJ)'Aor fUY gene coding for the protein, or by disrupting the gene. 6. The bacterium according to claim 5, wherein the yc\iN gene is selected IVoru the group consisting of: (u) a l)N A comprising the nucleotide sequence of SEQ ID NO; 1 or 3, (b) a DNA which is able to hybridize with a sequence complementary to the nucleotide secjuence of SEQ ID NO: 1 or 3, or a probe which is prepared from the nucleotide sequence under stringent conditions, (c) a DNA which has a homology of 95% or more to the nucleotide sequence of SEQ ID NO: 1 or 3. 7. The bacterium according to claim 5 or 6, wherein WxfliY gene is selected from the group consisting of: (d) a DNA comprising the nucleotide sequence of SEQ ID NO: 5 or 7, (e) a DNA which is able to hybridize with a sequence complementary to the nucleotide sequence of SEQ ID NO: 5 or 7, or a probe which is prepared from the nucleotide sequence under stringent conditions, (f) a DNA which has a homology of 95% or more to the nucleotide sequence of SEQ ID NO: 5 or 7. 8. The bacterium according to any one of claims 1 to 7, which further has at least one of the following characteristics: i) it has been modified to increase serine acetyltransferase activity, ii) it has been modified to increase expression of i\\e yeaS %Qne, iii) it has been modified to increase 3-phosphoglycerate dehydrogenase activity, iv) it has been modified to enhance activity of the sulfate/thiosulfate transport system. 9. The bacterium according to any one of claims 1 to 8, which belongs to the genus Pantoea. \ 0. The bacterium according to claim 9, which is Pantoea ananatis, 11. The bacterium according to any one of claims 1 to 8, which is Escherichia coli. 12. A method for producing L-cysteine, L-cystine, a derivative or precursor thereof, or a mixture of these, which comprises culturing the bacterium according to any one of claims 1 to 11 in a medium and collecting L-cysteine, L-cystine, a derivative or precursor thereof, or a mixture of these from the medium.