共查询到20条相似文献,搜索用时 31 毫秒
1.
2.
The main features of translation are similar in all organisms on this planet and one important feature of it is the way the ribosome maintain the reading frame. We have earlier characterized several bacterial mutants defective in tRNA maturation and found that some of them correct a +1 frameshift mutation; i.e. such mutants possess an error in reading frame maintenance. Based on the analysis of the frameshifting phenotype of such mutants we proposed a pivotal role of the ribosomal grip of the peptidyl-tRNA to maintain the correct reading frame. To test the model in an unbiased way we first isolated many (467) independent mutants able to correct a +1 frameshift mutation and thereafter tested whether or not their frameshifting phenotypes were consistent with the model. These 467+1 frameshift suppressor mutants had alterations in 16 different loci of which 15 induced a defective tRNA by hypo- or hypermodifications or altering its primary sequence. All these alterations of tRNAs induce a frameshift error in the P-site to correct a +1 frameshift mutation consistent with the proposed model. Modifications next to and 3′ of the anticodon (position 37), like 1-methylguanosine, are important for proper reading frame maintenance due to their interactions with components of the ribosomal P-site. Interestingly, two mutants had a defect in a locus (rpsI), which encodes ribosomal protein S9. The C-terminal of this protein contacts position 32–34 of the peptidyl-tRNA and is thus part of the P-site environment. The two rpsI mutants had a C-terminal truncated ribosomal protein S9 that destroys its interaction with the peptidyl-tRNA resulting in +1 shift in the reading frame. The isolation and characterization of the S9 mutants gave strong support of our model that the ribosomal grip of the peptidyl-tRNA is pivotal for the reading frame maintenance. 相似文献
3.
4.
We developed a rapid mutagenesis method based on a modification of the QuikChange® system (Stratagene) to systemically replace endogenous gene sequences with a unique similar size sequence tag. The modifications are as follows: 1: the length of the anchoring homologous sequences of both mutagenesis primers were increased to 16 – 22 bp to achieve melting temperatures greater than 80°C. 2: the final concentrations of both primers were increased to 5–10 ng/µl and the final concentration of template to 1–2 ng/µl. 3: the annealing temperature was adjusted when necessary from 52°C to 58°C. We generated 25 sequential mutants in the cloned espD gene (1.2 kb), which encodes an essential component of the type III secretion translocon required for the pathogenesis of enteropathogenic E. coli (EPEC) infection. Each mutation consisted of the replacement of 15 codons (45 bp) with 8 codons representing a 24 bp sequence containing three unique restriction endonuclease sites (KpnI/MfeI/SpeI) starting from the second codon. The insertion of the restriction endonuclease sites provides a convenient method for further insertions of purification and/or epitope tags into permissive domains. This method is rapid, site-directed and allows for the systematic creation of mutants evenly distributed throughout the entire gene of interest. 相似文献
5.
Y Tsukamoto 《Journal of theoretical biology》1979,78(4):451-498
An information theory of the genetic code is given, which deals with the process by which template codes (nucleotides or codons) choose substrate codes (nucleotides or anti-codons) in accordance with the base-pairing rules in the chain elongation phase of polynucleotide or polypeptide synthesis. A definite period of recognition time (τ) required for a template code to discriminate a substrate code is proposed, and an experimental method for determining the time is suggested. A substrate word is defined to be the sequence of substrate codes which have appeared at a recognition site in turn before a substrate code complementary to a template code first appears, and the mean length of substrate words (F) is derived from the mole fractions of template codes and substrate codes. The chain elongation rate is greatest when the mole fractions of template codes is proportional to the square of those of substrate codes to minimize the mean recognition time per word (Fτ). The uncertainty of a template (G) and the uncertainty of a medium (M) respectively are derived from the minimum of the function F. The amount of genetic information contained in a template is measured by the function G. The unit of the amount of genetic information is termed “cit”. The function M, the ratio of the number of all binary collisions to the number of homogeneous binary collisions in a mixture of different molecules, may be the new other “entropy” which represents informational properties of the mixture not represented by thermodynamic entropy of mixing. Both functions (G and M) have maxima when all random variables are equal and they are multiplicative in nature in contrast to entropy which is additive. The multiplicativity of the function G may contribute to the enormous informational capacity of genes. 相似文献
6.
Translation rates and misreading characteristics of rpsD mutants in Escherichia coli 总被引:5,自引:0,他引:5
D I Andersson K Bohman L A Isaksson C G Kurland 《Molecular & general genetics : MGG》1982,187(3):467-472
Summary Three ribosomal ambiguity (Ram) mutants, changed in ribosomal protein S4, have been examined with respect to elongation rate and misreading of translation in vivo and in vitro. Ram mutants increase misreading of nonsense codons in vivo, compared to wild type, between 2–50 times depending on the nature of the nonsense codon, its position, and which rpsD allele is present. Ram ribosomes also show an increased error frequency in vitro. The elongation rate of translation does not seem to be significantly changed, neither in vivo nor in vitro, irrespective of which rpsD allele is present.We suggest that there exists no general relationship between the accuracy and the overall speed of translation in Ram strains.Abbreviations poly U
poly(uridylic acid)
- IPTG
isopropyl B-(scd)-thiogalactopyranoside
- ATP
adenosine (5) triphosphate
- GTP
guanosine (5) triphosphate
- ONPG
o-nitrophenyl-B-d-galactoside
- Phe
phenylalanine
- Leu
leucine
- EF-G
efongation factor G
- EF-Tu
elongation factor Tu
- EF-Ts
elongation factor Ts
- Tet-R
tetracycline resistance 相似文献
7.
8.
In Saccharomyces cerevisiae, 11 out of 42 tRNA species contain 5-methoxycarbonylmethyl-2-thiouridine (mcm5s2U), 5-methoxycarbonylmethyluridine (mcm5U), 5-carbamoylmethyluridine (ncm5U) or 5-carbamoylmethyl-2′-O-methyluridine (ncm5Um) nucleosides in the anticodon at the wobble position (U34). Earlier we showed that mutants unable to form the side chain at position 5 (ncm5 or mcm5) or lacking sulphur at position 2 (s2) of U34 result in pleiotropic phenotypes, which are all suppressed by overexpression of hypomodified tRNAs. This observation suggests that the observed phenotypes are due to inefficient reading of cognate codons or an increased frameshifting. The latter may be caused by a ternary complex (aminoacyl-tRNA*eEF1A*GTP) with a modification deficient tRNA inefficiently being accepted to the ribosomal A-site and thereby allowing an increased peptidyl-tRNA slippage and thus a frameshift error. In this study, we have investigated the role of wobble uridine modifications in reading frame maintenance, using either the Renilla/Firefly luciferase bicistronic reporter system or a modified Ty1 frameshifting site in a HIS4A::lacZ reporter system. We here show that the presence of mcm5 and s2 side groups at wobble uridines are important for reading frame maintenance and thus the aforementioned mutant phenotypes might partly be due to frameshift errors. 相似文献
9.
Diverse mitochondrial (mt) genetic systems have evolved independently of the more uniform nuclear system and often employ modified genetic codes. The organization and genetic system of dinoflagellate mt genomes are particularly unusual and remain an evolutionary enigma. We determined the sequence of full-length cytochrome c oxidase subunit 1 (cox1) mRNA of the earliest diverging dinoflagellate Perkinsus and show that this gene resides in the mt genome. Apparently, this mRNA is not translated in a single reading frame with standard codon usage. Our examination of the nucleotide sequence and three-frame translation of the mRNA suggest that the reading frame must be shifted 10 times, at every AGG and CCC codon, to yield a consensus COX1 protein. We suggest two possible mechanisms for these translational frameshifts: a ribosomal frameshift in which stalled ribosomes skip the first bases of these codons or specialized tRNAs recognizing non-triplet codons, AGGY and CCCCU. Regardless of the mechanism, active and efficient machinery would be required to tolerate the frameshifts predicted in Perkinsus mitochondria. To our knowledge, this is the first evidence of translational frameshifts in protist mitochondria and, by far, is the most extensive case in mitochondria. 相似文献
10.
Translational readthrough of nonsense codons is seen not only in organisms possessing one or more tRNA suppressors but also in strains lacking suppressors. Amber suppressor tRNAs have been reported to suppress only amber nonsense mutations, unlike ochre suppressors, which can suppress both amber and ochre mutations, essentially due to wobble base pairing. In an Escherichia coli strain carrying the lacZU118 episome (an ochre mutation in the lacZ gene) and harboring the supE44 allele, suppression of the ochre mutation was observed after 7 days of incubation. The presence of the supE44 lesion in the relevant strains was confirmed by sequencing, and it was found to be in the duplicate copy of the glnV tRNA gene, glnX. To investigate this further, an in vivo luciferase assay developed by D. W. Schultz and M. Yarus (J. Bacteriol. 172:595-602, 1990) was employed to evaluate the efficiency of suppression of amber (UAG), ochre (UAA), and opal (UGA) mutations by supE44. We have shown here that supE44 suppresses ochre as well as opal nonsense mutations, with comparable efficiencies. The readthrough of nonsense mutations in a wild-type E. coli strain was much lower than that in a supE44 strain when measured by the luciferase assay. Increased suppression of nonsense mutations, especially ochre and opal, by supE44 was found to be growth phase dependent, as this phenomenon was only observed in stationary phase and not in logarithmic phase. These results have implications for the decoding accuracy of the translational machinery, particularly in stationary growth phase.Translation termination is mediated by one of the three stop codons (UAA, UAG, or UGA). When such stop codons arise in coding sequences due to mutations, referred to as nonsense mutations, they lead to abrupt arrest of the translation process. However, the termination efficiency of such nonsense codons is not 100%, as certain tRNAs have the ability to read these nonsense codons. Genetic code ambiguity is seen in several organisms. Stop codons have been shown to have alternate roles apart from translation termination. In organisms from all three domains of life, UGA encodes selenocysteine through a specialized mechanism. In Methanosarcinaceae, UAG encodes pyrrolysine (3). UAA and UAG are read as glutamine codons in some green algae and ciliates such as Tetrahymena and Diplomonads (24), and UAG alone encodes glutamine in Moloney murine leukemia virus (32). UGA encodes cysteine in Euplotes; tryptophan in some ciliates, Mycoplasma species, Spiroplasma citri, Bacillus, and tobacco rattle virus; and an unidentified amino acid in Pseudomicrothorax dubius and Nyctotherus ovalis (30). In certain cases the context of the stop codon in translational readthrough has been shown to play a role; for example, it has been reported that in vitro in tobacco mosaic virus, UAG and UAA are misread by tRNATyr in a highly context-dependent manner (34, 9).Termination suppressors are of three types, i.e., amber, ochre, and opal suppressors, which are named based on their ability to suppress the three stop codons. Amber suppressors can suppress only amber codons, whereas ochre suppressors can suppress ochre codons (by normal base pairing) as well as amber codons (by wobbling) and opal suppressors can read opal and UGG tryptophan codon in certain cases. As described by Sambrook et al. (27), a few amber suppressors can also suppress ochre mutations by wobbling. The suppression efficiency varies among these suppressors, with amber suppressors generally showing increased efficiency over ochre and opal suppressors. supE44, an amber suppressor tRNA, is an allele of and is found in many commonly used strains of Escherichia coli K-12. Earlier studies have shown that supE44 is a weak amber suppressor and that its efficiency varies up to 35-fold depending on the reading context of the stop codon (8).Translational accuracy depends on several factors, which include charging of tRNAs with specific amino acids, mRNA decoding, and the presence of antibiotics such as streptomycin and mutations in ribosomal proteins which modulate the fidelity of the translational machinery. Among these, mRNA decoding errors have been reported to occur at a frequency ranging from about 10−3 to 10−4 per codon. Translational misreading errors also largely depend on the competition between cognate and near-cognate tRNA species. Poor availability of cognate tRNAs increases misreading (18).Several studies with E. coli and Saccharomyces cerevisiae have shown the readthrough of nonsense codons in suppressor-free cells. In a suppressor-free E. coli strain, it has been shown in vitro that glutamine is incorporated at the nonsense codons UAG and UAA (26). It has been reported that overexpression of wild-type tRNAGln in yeast suppresses amber as well as ochre mutations (25). In this study, we have confirmed the presence of an amber suppressor mutation in the glnX gene in a supE44 strain by sequence analysis. This was done essentially because we observed that supE44 could also suppress lacZ ochre mutations, albeit inefficiently. On further investigation using an in vivo luciferase reporter assay system for tRNA-mediated nonsense suppression (28), we found that the efficiency of suppression of amber lesion by supE44 is significantly higher than that reported previously in the literature. An increased ability to suppress ochre and opal nonsense mutations was observed in cells bearing supE44 compared to in the wild type. Such an effect was observed only in the stationary phase and was abolished in logarithmic phase. 相似文献
11.
12.
The mitochondrial genome of the stramenopile alga Chrysodidymus synuroideus. Complete sequence, gene content and genome organization
下载免费PDF全文
![点击此处可从《Nucleic acids research》网站下载免费的PDF全文](/ch/ext_images/free.gif)
Chesnick JM Goff M Graham J Ocampo C Lang BF Seif E Burger G 《Nucleic acids research》2000,28(13):2512-2518
This is the first report of a complete mitochondrial genome sequence from a photosynthetic member of the stramenopiles, the chrysophyte alga Chrysodidymus synuroideus. The circular-mapping mitochondrial DNA (mtDNA) of 34 119 bp contains 58 densely packed genes (all without introns) and five unique open reading frames (ORFs). Protein genes code for components of respiratory chain complexes, ATP synthase and the mitoribosome, as well as one product of unknown function, encoded in many other protist mtDNAs (YMF16). In addition to small and large subunit ribosomal RNAs, 23 tRNAs are mtDNA-encoded, permitting translation of all codons present in protein-coding genes except ACN (Thr) and CGN (Arg). The missing tRNAs are assumed to be imported from the cytosol. Comparison of the C.synuroideus mtDNA with that of other stramenopiles allowed us to draw conclusions about mitochondrial genome organization, expression and evolution. First, we provide evidence that mitochondrial ORFs code for highly derived, unrecognizable versions of ribosomal or respiratory genes otherwise ‘missing’ in a particular mtDNA. Secondly, the observed constraints in mitochondrial genome rearrangements suggest operon-based, co-ordinated expression of genes functioning in common biological processes. Finally, stramenopile mtDNAs reveal an unexpectedly low variability in genome size and gene complement, testifying to substantial differences in the tempo of mtDNA evolution between major eukaryotic lineages. 相似文献
13.
Michael E.S. Hudspeth W.Michael Ainley Deborah S. Shumard Ronald A. Butow Lawrence I. Grossman 《Cell》1982,30(2):617-626
Alleles of the var1 locus on yeast mitochondrial DNA specify the size of var1 ribosomal protein. We report the nucleotide sequence of a var1 allele that determines the smallest var1 protein. It contains an open reading frame of 396 codons, which we identify as the structural gene for var1 protein. The var1 protein specified by this allele has an amino acid composition in close agreement with that predicted by the DNA sequence. The var1 coding region is highly unusual: it is 89.6% AT and contains a 46 bp GC-rich palindromic cluster that accounts for 38% of the total GC residues. Our results strongly suggest that like mammalian mitochondria but unlike those from Neurospora, yeast mitochondria use AUA as a methionine codon. Comparison with the sequence of a var1 allele specifying a larger protein suggests that some size polymorphism of var1 protein results from in-frame insertions of a variable number of AAT (Asn) codons. 相似文献
14.
Prokaryotic toxin–antitoxin loci encode mRNA cleaving enzymes that inhibit translation. Two types are known: those that cleave mRNA codons at the ribosomal A site and those that cleave any RNA site specifically. RelE of Escherichia coli cleaves mRNA at the ribosomal A site in vivo and in vitro but does not cleave pure RNA in vitro. RelE exhibits an incomplete RNase fold that may explain why RelE requires its substrate mRNA to presented by the ribosome. In contrast, RelE homologue YoeB has a complete RNase fold and cleaves RNA independently of ribosomes in vitro. Here, we show that YoeB cleavage of mRNA is strictly dependent on translation of the mRNA in vivo. Non-translated model mRNAs were not cleaved whereas the corresponding wild-type mRNAs were cleaved efficiently. Model mRNAs carrying frameshift mutations exhibited a YoeB-mediated cleavage pattern consistent with the reading frameshift thus giving strong evidence that YoeB cleavage specificity was determined by the translational reading frame. In contrast, site-specific mRNA cleavage by MazF occurred independently of translation. In one case, translation seriously influenced MazF cleavage efficiency, thus solving a previous apparent paradox. We propose that translation enhances MazF-mediated cleavage of mRNA by destabilization of the mRNA secondary structure. 相似文献
15.
The wobble hypothesis revisited: uridine-5-oxyacetic acid is critical for reading of G-ending codons
According to Crick's wobble hypothesis, tRNAs with uridine at the wobble position (position 34) recognize A- and G-, but not U- or C-ending codons. However, U in the wobble position is almost always modified, and Salmonella enterica tRNAs containing the modified nucleoside uridine-5-oxyacetic acid (cmo5U34) at this position are predicted to recognize U- (but not C-) ending codons, in addition to A- and G-ending codons. We have constructed a set of S. enterica mutants with only the cmo5U-containing tRNA left to read all four codons in the proline, alanine, valine, and threonine family codon boxes. From the phenotypes of these mutants, we deduce that the proline, alanine, and valine tRNAs containing cmo5U read all four codons including the C-ending codons, while the corresponding threonine tRNA does not. A cmoB mutation, leading to cmo5U deficiency in tRNA, was introduced. Monitoring A-site selection rates in vivo revealed that the presence of cmo5U34 stimulated the reading of CCU and CCC (Pro), GCU (Ala), and GUC (Val) codons. Unexpectedly, cmo5U is critical for efficient decoding of G-ending Pro, Ala, and Val codons. Apparently, whereas G34 pairs with U in mRNA, the reverse pairing (U34-G) requires a modification of U34. 相似文献
16.
Takahito Mukai Atsushi Yamaguchi Kazumasa Ohtake Mihoko Takahashi Akiko Hayashi Fumie Iraha Satoshi Kira Tatsuo Yanagisawa Shigeyuki Yokoyama Hiroko Hoshi Takatsugu Kobayashi Kensaku Sakamoto 《Nucleic acids research》2015,43(16):8111-8122
The immutability of the genetic code has been challenged with the successful reassignment of the UAG stop codon to non-natural amino acids in Escherichia coli. In the present study, we demonstrated the in vivo reassignment of the AGG sense codon from arginine to l-homoarginine. As the first step, we engineered a novel variant of the archaeal pyrrolysyl-tRNA synthetase (PylRS) able to recognize l-homoarginine and l-N6-(1-iminoethyl)lysine (l-NIL). When this PylRS variant or HarRS was expressed in E. coli, together with the AGG-reading tRNAPylCCU molecule, these arginine analogs were efficiently incorporated into proteins in response to AGG. Next, some or all of the AGG codons in the essential genes were eliminated by their synonymous replacements with other arginine codons, whereas the majority of the AGG codons remained in the genome. The bacterial host''s ability to translate AGG into arginine was then restricted in a temperature-dependent manner. The temperature sensitivity caused by this restriction was rescued by the translation of AGG to l-homoarginine or l-NIL. The assignment of AGG to l-homoarginine in the cells was confirmed by mass spectrometric analyses. The results showed the feasibility of breaking the degeneracy of sense codons to enhance the amino-acid diversity in the genetic code. 相似文献
17.
Laura E. Holberger Christopher S. Hayes 《The Journal of biological chemistry》2009,284(46):32188-32200
Translational pausing in Escherichia coli can lead to mRNA cleavage within the ribosomal A-site. A-site mRNA cleavage is thought to facilitate transfer-messenger RNA (tmRNA)·SmpB- mediated recycling of stalled ribosome complexes. Here, we demonstrate that the aminoglycosides paromomycin and streptomycin inhibit A-site cleavage of stop codons during inefficient translation termination. Aminoglycosides also induced stop codon read-through, suggesting that these antibiotics alleviate ribosome pausing during termination. Streptomycin did not inhibit A-site cleavage in rpsL mutants, which express streptomycin-resistant variants of ribosomal protein S12. However, rpsL strains exhibited reduced A-site mRNA cleavage compared with rpsL+ cells. Additionally, tmRNA·SmpB-mediated SsrA peptide tagging was significantly reduced in several rpsL strains but could be fully restored in a subset of mutants when treated with streptomycin. The streptomycin-dependent rpsL(P90K) mutant also showed significantly lower levels of A-site cleavage and tmRNA·SmpB activity. Mutations in rpsD (encoding ribosomal protein S4), which suppressed streptomycin dependence, were able to partially restore A-site cleavage to rpsL(P90K) cells but failed to increase tmRNA·SmpB activity. Taken together, these results show that perturbations to A-site structure and function modulate A-site mRNA cleavage and tmRNA·SmpB activity. We propose that tmRNA·SmpB binds to streptomycin-resistant rpsL ribosomes less efficiently, leading to a partial loss of ribosome rescue function in these mutants. 相似文献
18.
Specific induction of transitions and transversions of G-C base pairs by 4-nitroquinoline-1-oxide in iso-1-cytochrome c mutants of yeast 总被引:7,自引:0,他引:7
The base-pair changes induced by the highly carcinogenic agent, 4-nitroquinoline-1-oxide, have been determined from the reversion rates of defined tester strains and from the amino acid replacements of revertant iso-1-cytochromes c. The mutant codons and the base-pair changes of reverse mutations of 14 cyc1 mutants were previously determined from alterations of iso-1-cytochromes c in intragenic revertants. These 14 cyc1 mutants, which were used as tester strains, included nine mutants with altered AUG initiation codons, an ochre (UAA) mutant, an amber (UAG) mutant and three frameshift mutants (Stewart et al., 1971,1972; Stewart &; Sherman, 1972,1974; Sherman &; Stewart, 1973). NQO2 induced a high rate of reversion in the initiation mutant cyc1-131, the only mutant in the group which reverts to normal iso-1-cytochrome c by a G · C → A · T transition. In addition, NQO produces a significant rate of reversion of all cyc1 mutants which revert by G · C transversions, e.g. the amber (UAG) mutant and the initiation mutants containing AGG, and probably CUG mutant codons. It did not revert the ochre mutant which contains no G · C base pairs. Ten NQO-induced revertants of the amber mutant cyc1-179 contained the expected replacements of residues of tyrosine, and ten NQO-induced revertants of each of the cyc1-131 and cyc1-133 initiation mutants all contained the expected normal iso-1-cytochrome c. The structures of these iso-1-cytochromes c and the pattern of reversion of the tester strains indicate that base-pair substitutions arise at G · C base pairs which are the site of NQO attack. Thus NQO induces G · C → A · T transitions, G · C → T · A transversions and possibly G · C → C · G transversions. Because of its mode of action, NQO may be useful in less-defined systems for identifying G · C base pairs in mutant codons. 相似文献
19.
S. Joakim Näsvall 《Journal of molecular biology》2009,385(2):350-31112
If a ribosome shifts to an alternative reading frame during translation, the information in the message is usually lost. We have selected mutants of Salmonella typhimurium with alterations in tRNAcmo5UGGPro that cause increased frameshifting when present in the ribosomal P-site. In 108 such mutants, two parts of the tRNA molecule are altered: the anticodon stem and the D-arm, including its tertiary interactions with the variable arm. Some of these alterations in tRNAcmo5UGGPro are in close proximity to ribosomal components in the P-site. The crystal structure of the 30S subunit suggests that the C-terminal end of ribosomal protein S9 contacts nucleotides 32-34 of peptidyl-tRNA. We have isolated mutants with defects in the C-terminus of S9 that induce + 1 frameshifting. Combinations of changes in tRNAcmo5UGGPro and S9 suggest that an interaction occurs between position 32 of the peptidyl-tRNA and the C-terminal end of S9. Together, our results suggest that the cause of frameshifting is an aberrant interaction between the peptidyl-tRNA and the P-site environment. We suggest that the “ribosomal grip” of the peptidyl-tRNA is pivotal for maintaining the reading frame. 相似文献
20.
G. A. Zhouravleva S. E. Moskalenko S. V. Chabelskaya M. Philippe S. G. Inge-Vechtomov 《Molecular Biology》2006,40(4):647-653
We have earlier characterized Saccharomyces cerevisiae strains with mutations of essential SUP45 and SUP35, which code for translation termination factors eRF1 and eRF3, respectively. In this work, the sup45 and sup35 nonsense mutants were compared with respect to the levels of eight tRNAs: tRNATyr, tRNAGln, tRNATrp, tRNALeu, tRNAArg (described as potential suppressor tRNAs), tRNAPro, tRNAHis, and tRNAGly. The mutants did not display a selective increase in tRNAs, capable of a noncanonical read-through at stop codons. Most of the mutations increased the level of all tRNAs under study. The mechanisms providing for the viability of the sup45 and sup35 nonsense mutants are discussed. 相似文献