| Human Molecular Genetics | Pages |
Co-existence of high levels of a cytochrome b mutation and of a tandem 200 bp duplication in the D-loop of muscle human mitochondrial DNA
Introduction
Results
Discussion
Materials And Methods
DNA sequencing
Single muscle fibre study
Acknowledgements
References
Co-existence of high levels of a cytochrome b mutation and of a tandem 200 bp duplication in the D-loop of muscle human mitochondrial DNA
Previous studies have suggested that some patients with large-scale mitochondrial DNA (mtDNA) deletions also presented a heteroplasmic 260 bp tandem duplication in the mtDNA D-loop region. Such duplications were observed not only in patients with mitochondrial pathology but also in aged subjects. However, the percentage of duplicated mtDNA did not exceed a few per cent of the total mtDNA, except in one example where it reached 30%. We report here another type of 200 bp duplication in the mtDNA D-loop region that, instead of being associated with a large-scale deletion, is correlated to the presence of a point mutation in the cytochrome b gene. The 200 bp duplication concerned up to 95% of the total mtDNA of some muscle mitochondria and was absent from the patient lymphocyte DNA. The percentages of the 200 bp duplication and that of the cytochrome b mutation were relatively close in whole muscle as well as in single muscle fibres, suggesting a correlation between the mutation and the duplication. This duplication could also be detected by PCR in two other patients with mitochondrial disorders but without known deletion or mtDNA mutation. These data suggest that the accumulation of these small duplications in the mtDNA D-loop could be indicative of the presence of other defects of the mtDNA which would damage the respiratory chain function. These deficiencies would induce the generation of small duplications in the D-loop.
INTRODUCTION
A heteroplasmic tandem duplication of low abundance with a size of ~260 bp in the D-loop of human mitochondrial DNA (mtDNA) was first observed by Brockington et al. (1). It was found in patients bearing large-scale deletions in their mtDNA. These deletions, which are the hallmark of some mitochondrial myopathies, are observed in diseases such as Kearns-Sayre syndrome or chronic progressive external ophthalmoplegia (2,3).
Torroni et al. (4) identified a tandem duplication similar to that of 260 bp in normal individuals belonging to a particular Caucasian mtDNA haplogroup (haplogroup I). These subjects harboured inserted stretches of 2-6 cytosines (Cs) between nucleotides 568 and 573 [mtDNA numbering being according to the Cambridge sequence (5)]. In these subjects, the percentage of duplicated mtDNA molecules always appeared to be low, and represented <2% of the total mtDNA. This heteroplasmic duplication was found in only one of the eight tested patients bearing large-scale mtDNA deletions (4). The patient concerned was also of haplogroup I. Therefore, this study suggested that the 260 bp tandem duplication correlated with haplogroup I and with the insertion of Cs, but not with large mtDNA deletions.
Similarly, none of the 30 patients bearing muscle mtDNA deletions exhibits the 260 bp duplication in the study of Manfredi et al. (6). Nevertheless, these authors identified this duplication in a 58-year-old man with a late-onset slowly progressive mitochondrial myopathy. In this case, ~32% of the patient's muscle mtDNA harboured the duplication. In addition, low levels of a triplication of this 260 bp mtDNA fragment could be detected by PCR in the patient's muscle. The histochemical analysis of the patient's muscle biopsy exhibited 10% of ragged red fibres and a partial decrease in cytochrome oxidase activity.
Lee et al. (7) have also detected a similar duplication of ~260 bp (named type I) and another of ~200 bp (type II) in the mtDNA D-loop of muscle biopsies from aged individuals. Both type I and type II duplications contained variable numbers of Cs at the junction site, as shown for haplotype I (4). The incidence of the type II duplication increased with age. In addition, the proportion of the type II duplicated mtDNA increased from 0.78% in the muscle of a 55-year-old individual to 3.1% in that of a 71-year-old subject. In that study, 90% of the subjects with duplicated mtDNA also had the 4977 bp mtDNA deletion present in low amounts. The authors concluded that tandem duplications often occurring in association with mtDNA deletions accumulate in an age-dependent manner causing synergistic deleterious effects on mitochondrial respiratory functions.
More recently, Wei et al. (8) identified a total of 10 types of such tandem duplications defined by their sequence and located in the mtDNA D-loop. Most of them (types I, II, V, VI and VIII) contained a variable number of inserted C residues. The types I, II, III, IV and X duplications occurred more frequently and abundantly than the others in ageing human tissues. Large-scale 4977 and 7436 bp mtDNA deletions also increased with age (9,10). However, there was no clear association between the percentage of either duplicated or deleted mtDNA molecules. In addition, the proportion of mtDNA with the tandem duplication remained very low. For example, the proportion of type IX duplicated mtDNA that could be detected in subjects between 30 and >70 years of age by a semi-quantitative PCR technique never exceeded 0.12% of the total mtDNA.
In the present study, we show that a 200 bp tandem duplication [type IX, according to Wei et al. (8)] can exist at a much higher percentage than previously expected. Indeed, >90% of duplicated mtDNA was present in the patient's muscle. The patient suffered from a mitochondrial myopathy with exercise intolerance caused by a decreased ubiquinol-cytochrome c reductase activity. This deficiency was due to a cytochrome b mutation in position 15 615, converting the highly conserved Gly290 to an Asp which decreased the complex III stability (11). The correlation between the formation of the mtDNA D-loop duplication and the respiratory chain deficiency due to a cytochrome b point mutation suggests a relationship between duplication formation and respiratory chain deficiency. This could apply to any type of respiratory chain deficiency whether the genetic origin of the deficiency be a deletion or a point mutation in the mtDNA.
RESULTS
Figure 1 shows the Southern blot obtained when the patient and control mtDNA fragments produced by restriction with MspI were hybridized with a 12S probe. The MspI fragments deduced from the Anderson's mtDNA sequence recognized by this probe were of 827 and 2146 bp. These two fragments were present in all the tested DNA controls. The patient DNA extracted either from the whole muscle or from the pellet obtained after two successive mitochondrial extractions from the muscle homogenate exhibited the expected band of ~2 kb. However, the band at 0.8 kb was very faint in the patient's muscle DNA and was a minor band in the case of the DNA from the pellet partially depleted of mitochondria. A new major band at ~1 kb was present in muscle and pellet DNA. A semi-quantitative estimation of the band intensities was performed. This indicated that the 0.8 kb band represented only ~5% of the 1 kb band in the muscle and ~25% of the 1 kb band in the pellet partially depleted of mitochondria. In addition, there was a faint band at ~1.2 kb for the muscle and pellet DNAs. Therefore, there is ~95 and 75% of duplication in the tested samples of muscle DNA and pellet DNA, respectively.
Figure
Normal mtDNA contains two MspI restriction sites at positions 104 and 931, normally producing the fragment of 827 bp which hybridizes with the 12S probe. However, several authors (12-14) have described a mtDNA polymorphism lacking the MspI site in position 104 due to a deletion of nucleotides 106-111. To confirm whether the unexpected bands observed by Southern blotting could correspond to an MspI polymorphism or to an insertion of ~200 bp, the patient and control mtDNAs were amplified from nucleotide 16 477 to 1325. The control DNA gave the expected fragment of 1430 bp for normal mtDNA. On the contrary, the patient DNA gave a fragment of ~1630 bp, confirming the insertion of ~200 bp in this part of the patient mtDNA (Fig. 2). To determine whether this insertion could correspond to one of the duplications previously characterized in the mtDNA D-loop (1,7,8), the patient mtDNA was amplified using various pairs of back-to-back primers that do not amplify normal mtDNA but can amplify duplicated mtDNA, as described (7). No amplification could be observed with the primers L336 and H335 known to reveal the 260 bp duplication (type I), the 200 bp duplication (type II) or the 150 bp duplication (type III) (8). On the contrary, a major 200 bp fragment and a minor 400 bp fragment were obtained (Fig. 3A) with the patient's mtDNA using the primers L467 and H466. These primers reveal type I (260 bp), type II (200 bp) or type IX (200 bp) duplications. As expected, there was no amplification for control mtDNA (not shown).
Figure
Figure
Sequencing the 200 bp amplified fragment confirmed the presence of a 200 bp duplication located at the level of two 12 bp repeated sequences. The repeated sequences CCAAACCCCAAA are present in the Cambridge sequence (5) at positions 348-359 and 552-563. The sequence of the junction between the duplicated fragments is shown in Figure 3B. The 200 bp insertion corresponds to a repetition of the fragment located between these two sequences. Figure 3C represents a scheme of the D-loop containing the 200 bp duplication observed in the patient's muscle. Southern blotting of the patient's mtDNA was performed after digestion with BamHI, EcoRV or HpaI and hybridization with purified human mtDNA. The normal pattern of this Southern blot (not shown) indicated that the patient mtDNA did not contain any detectable large-scale deletion. The duplication could not be detected by PCR using back-to-back primers in the patient DNA extracted either from a myoblast culture obtained from the patient biopsy or from patient lymphocytes. The heteroplasmic cytochrome b mutation in position 15 615 was not present in either the patient's myoblast or lymphocyte DNA samples. Neither the cytochrome b mutation nor the 200 bp duplication could be detected in the DNA extracted from the lymphocytes of the patient's mother and sister. The co-existence of the 15 615 cytochrome b mutation and of the 200 bp duplication was studied in single muscle fibres isolated from two different 10 µm thick transverse muscle sections stained for cytochrome oxidase activity (11). Cytochrome b was amplified from single fibre DNA, using [[alpha]-32P]dCTP added to the PCR mixture during the last cycle. The percentage of mutated cytochrome b was tested by comparing the amount of cytochrome b cleaved by AatII with the total amount of amplified cytochrome b. The AatII restriction site was indeed created by the 15 615 mutation. The percentage of duplicated DNA was estimated by comparing the 622 bp band with the 422 bp band when the D-loop mtDNA fragment was amplified between nucleotides 244 and 665 of the Cambridge sequence. Figure 4 shows that, under these conditions, the presence of the duplication could barely be detected when the percentage of cytochrome b mutation was <20%. Beyond this threshold, the percentage of duplication increased more or less linearly with the percentage of mutated cytochrome b. In all the tested fibres with the exception of two of them, the percentage of the 200 bp duplication was lower than that of the cytochrome b mutation. In addition, there was no apparent difference between type I and type II fibres heavily or faintly stained for cytochrome oxidase respectively.
Figure
The 200 bp duplication was tested in the muscle DNA of several patients suffering from or suspected of having mitochondrial myopathies. It could not be detected either in five patients presenting large-scale mtDNA deletions [described in (15)] or in two patients with the 3243 MELAS mutation [described in (16)]. Seven other patients [described in (17)] presenting ragged red fibres in their muscle biopsy and respiratory chain deficiencies but no identified mtDNA deletion or mutation did not exhibit either the 200 or the 260 bp duplications. However, in two other patients' muscles with ragged red fibres, the 200 bp duplication could be amplified with the L467 and H466 back-to-back primers. One of them [subject A10 in (17)] was a 35-year-old male with a major complex III deficiency and a faint complex IV deficiency. He was suffering from a severe muscular exercise intolerance of unidentified sporadic genetic origin. The other patient [subject A16 in (17)] was a 42-year-old man with complex I and complex IV deficiencies. He was suffering from ataxia and lipomatosis of unidentified genetic origin but with a typical mitochondrial maternal inheritance. However, the amount of 200 bp duplicated mtDNA was very low for these two patients since it could be detected by ethidium bromide staining of the fragment amplified only when using the L467 and H466 back-to-back primers and not when the amplification was performed with the L244 and H665 primers. In the latter case, only the normal 422 bp fragment was visible. Using back-to-back primers, the presence of the 200 bp duplication could not be observed in the DNA extracted from the lymphocytes of this patient. The presence of the 260 bp duplication could not be detected in any of these patients.
DISCUSSION
In this study, we show that the mtDNA of a patient's muscle exhibiting a cytochrome b mutation also contained a 200 bp tandem duplication in the mtDNA D-loop. This duplication was similar to one of the small tandem duplications (type IX) described by Wei et al. (8) for aged subjects. This duplication is due to the repetition of a 12 bp sequence present in the D-loop: CCAAACCCCAAA found at positions 348-359 and 552-563 of the mtDNA. The percentage of the 200 bp mtDNA duplication observed in our study is much higher than that reported in other patients or in aged subjects. Indeed, the tandem duplications studied until now were generally limited to a few per cent of the total mtDNA (1,4,6-8) or to 32% in the case of a patient with ragged red fibres and cytochrome oxidase deficiency of unidentified genetic origin (6).
It is interesting to compare the percentages of the 200 bp tandem duplication and of the cytochrome b mutation reported previously (11) in different tissue fractions. In a sample of whole muscle, they were of ~95 and 85% respectively. In single muscle fibres, these percentages could vary from 0 to 95%. On the contrary, they could not be detected in myoblasts or in lymphocytes. They could not be observed either in the mother's or sister's lymphocytes. Therefore, although the percentage of the 200 bp tandem duplication seemed to be higher than the percentage of cytochrome b mutation in whole muscle or in the mitochondria-depleted pellet, in single fibres, this percentage was generally slightly lower than that of the mutation. The difference between whole muscle data and single muscle fibre data is probably caused by a technical problem due to the fact that in the first study of cytochrome b mutation made with whole muscle (11), the percentage mutation was analysed by non-radioactive PCR followed by restriction analysis in ethidium bromide-stained gels. In the single cell muscle fibre study, the measurements were made after incorporation of radioactive nucleotide during the last PCR cycle. This technique prevents shuffling of normal and mutated DNA that could artificially decrease the amount of DNA cleaved by the restriction enzyme. The single fibre study clearly shows that, beyond a threshold of ~20% of mutated cytochrome b, the percentage of duplicated DNA increased with that of mutated cytochrome b. Therefore, there is a correlation between the presence of the two mtDNA modifications in the different tested samples. The fact that both the 200 bp duplication and the 15 615 cytochrome b mutation could not be found in cultured myoblasts or in the lymphocytes shows that both modifications only accumulate in post-mitotic tissues. There are probably variations between different muscle areas since in the sample used for single fibre studies, the mean percentage of duplication is lower than that used for Southern blotting analysis.
The very high percentage of the 200 bp duplication observed in the patient muscle has not been observed previously with other patients exhibiting respiratory chain deficiencies. One reason might be that the most frequently studied mtDNA deficiencies do not concern genes coding for proteins prone to increase free radical production. To our knowledge, the presence of the D-loop duplications has been observed mainly in patients with deleted mtDNA or in those with mutations in tRNAs. In both cases, the defect induces a decrease in the synthesis of all proteins of mitochondrial origin since tRNAs are missing. The electron transport is not blocked but is absent. Therefore, the free radical increase inside these fibres should be less important than when the electron transfer is blocked. The partial respiratory chain activity observed in these types of diseased muscles is probably due to the presence of intact fibres. Since, in the single fibre study, the percentage of 200 bp duplication is correlated to the percentage of mutation, it means that the effect is not propagated easily from fibre to fibre. It would be interesting to look for these types of duplications in muscle biopsies or in the post-mortem brain of patients suffering from Leigh syndrome with specific cytochrome oxidase deficiency. This should also prevent CoQH2 re-oxidation and induce duplication formation. However, since these patients generally die early in life, the duplication might not have enough time to accumulate.
The high percentage of 200 bp duplication observed here may be related to the fact that the deleterious mutation concerns the cytochrome b. Indeed, Guidarelli et al. (19) have inhibited electron transport in U937 cells by antimycin. This treatment appears to enhance the formation of DNA-damaging species generated by drugs such as organic hydroperoxides. This effect was specific for complex III inhibition. Blocking electron transfer through complex III by antimycin is likely to generate free radicals by incomplete Q-cycling. These free radicals would mediate the formation of single-strand breaks in mtDNA located in the vicinity of the inner mitochondrial membrane. We propose that these single-strand breaks would favour the duplication formation at the level of repeated sequences. In most other reports (1,4,6,7), the tandem duplications occurred at the level of hot spots containing poly(C) stretches. Slip mispairing (20) is the most frequently proposed mechanism to explain the generation of the duplications involving a variable number of Cs. Single-strand breaks combined with mispairing might better explain the generation of duplications at the level of repeated sequences which are relatively close to each other in the mtDNA D-loop. A better knowledge of the mtDNA repair mechanisms will probably shed some new light on the formation of these duplications. The problem which remains unexplained is the reason why the type IX 200 bp duplication is the only duplication found in this patient.
In the D-loop, the 200 bp duplication modifies the number of promoters and of conserved sequence boxes (CSB) able to bind regulatory factors. It has been suggested previously that this could modify the regulation of mtDNA replication or transcription. Recently, Hao et al. (18) isolated several transmitochondrial cybrid lines harbouring the 260 bp duplication, one of which was essentially homoplasmic for the duplication. They showed that oxidative phosphorylation was normal in this homoplasmic clone, suggesting that this duplication was not pathogenic perse. In addition, the mtDNA copy number, the steady-state level of heavy and light strand transcripts and the steady-state levels of RNAs made from the two promoters were similar in clones containing almost no or 100% 260 bp duplication. Therefore, twice the number of promoters did not affect mtDNA replication or its transcription. This is probably because two RNA polymerase molecules were not likely to bind simultaneously to the same mtDNA molecule. The 200 bp duplication observed in this study started at the repeated sequence spanning nucleotides 348-359 and ended at that spanning nucleotides 552-563. The 260 bp duplication started at a poly(C) stretch comprising nucleotides 302-308 and ended at another poly(C) stretch from 567 to 573. The same regulatory elements are present in both types of duplications, except CSBII which is not duplicated in the 200 bp duplication (see Fig. 3C). In spite of this difference, it is very unlikely thatthe 200 bp duplication by itself would behave differently from the 260 bp duplication on oxidative phosphorylation or on mtDNA replication or transcription. Therefore, the 200 bp duplication would not be more pathogenic per se than the 260 bp duplication. The generation of the 200 bp duplication might be a consequence of the cytochrome b mutation that decreases electron transfer through complex III, as discussed above.
Two of the studied patients known until now with mitochondrial dysfunction but without mtDNA modification also presented the 200 bp duplication. However, the percentage of duplicated DNA was very low in these patients, as observed previously for aged patients (8). This suggests that these patients would have some deleterious mtDNA modification which remains to be identified. The presence of such a type of duplication which is easy to identify by PCR using back-to-back primers, might be an additional criterion to select the patients susceptible to present new mtDNA mutations.
MATERIALS AND METHODS
All experiments were performed with the informed consent of the patients.
Chemicals were supplied either by Sigma or Boehringer. Restriction enzymes were from New England Laboratories or Boehringer. Taq DNA polymerase was either from Life Technologies or from Sigma (experiments with single fibres).
Previously described procedures were used to collect and study the muscle biopsies by histochemical methods (15), to prepare the mitochondria from the muscle and to analyse their oxidative phosphorylation capacities by oxypolarography and spectrophotometric methods (17), to establish myoblast cultures (16) and to prepare DNA and amplify it by PCR (11), unless otherwise indicated in the figure legends. The 12S probe was prepared by PCR using a hybridization temperature of 61°C and primers corresponding to the Cambridge mtDNA sequence (5) from nucleotide 534 to 552 (H534) and from 1696 to 1679 (L1696) (H for heavy mtDNA strand and L for light mtDNA strand). This probe corresponds to the end of the D-loop, the 12S mt rRNA and the beginning of the 16S mt rRNA. Two couples of back-to-back primers were used to detect the duplications. The first one corresponding to the Cambridge sequence (5) from nucleotide 467 to 486 (L467) and from 466 to 447 (H466) was used at a hybridization temperature of 55°C. The second primer pair L336-H335 was used as described (6). The primers L467-H466 could amplify the 260 bp (type I) or 200 bp (types II and IX) duplications, while the primers L336-H335 could amplify type I, II or III (150 bp) duplications (8). MtDNA fragments encompassing the 200 bp duplication were also amplified using a hybridization temperature of 55°C and two primers corresponding to the mtDNA sequence either from nucleotide 16 477 to 16 496 (L16477) and from 1344 to 1325 (H1344), or from nucleotide 244 to 261 (L244) and from 665 to 646 (H665). Cytochrome b was amplified from nucleotide 15 437 to 15 777 and digested with ApaII as described previously (11).
Southern blot analysis was performed according to the method of Southern (21). Patient and control DNA (5-15 µg) were digested for 2 h at 37°C with 10-30 U of MspI, BamHI, EcoRVor HpaI according to the manufacturer's instructions. After electrophoresis and transfer to a Nylon membrane, it was hybridized with a 32P-labelled fragment corresponding either to mtDNA purified from human placenta or to the 12S probe. The probes were labelled with [32P]dCTP using the Mega-Prime kit (Amersham). The proportion of normal and duplicated mtDNA was estimated by comparing band intensities of the Southern blot of mtDNA fragments digested with MspI,following hybridization with the 32P-labelled 12S probe. The analysis was performed with the GS 250 Molecular Imager and using the Molecular analyst software (Biorad).
DNA sequencing
PCR fragments were amplified with the back-to-back primers H466 and L467. DNA fragments were purified by electrophoresis on a 1.5% agarose gel, extracted by freeze-squeeze (22) and sequenced with the Thermo Sequenase kit in the presence of [[alpha]-33P]ddNTP (Amersham), using the primers L467 or H466.
Single muscle fibre study
Transverse muscle sections (10 µm thick) stained for cytochrome oxidase (11,15) were treated with toluene to remove the upper glass cover and rehydrated successively with 70% ethanol and water. Individual fibres were dissected and aspirated from the glass slide by gentle suction into glass capillaries attached to an Axiovert 135 micromanipulator (Zeiss, Germany) under an inverted microscope. Each fibre was collected into 20 µl of water and centrifuged for 10 min at 8000 g. Ten µl of 10× PCR buffer (Sigma) and 10 µg of proteinase K were added to each tube and the tubes were incubated for 3 h at 55°C and put into a boiling water bath for 10 min in order to inactivate the proteinase K. Amplification of the D-loop fragment containing the 200 bp duplication or of the cytochrome b fragment containing the15 615 mutation was performed using the primers H244 andL665 or H15437 and L15777, respectively. The PCR was run as described (11) for 30 cycles. An additional cycle was performed after addition of 50 pmol of each primer, 0.2 units of Taq DNA polymerase and 2 µCi of [[alpha]-32P]dCTP (3000 Ci/mmol; ICN). This last cycle was performed at 94°C for 5 min, 55°C for 1 min and 72°C for 10 min. The amplified cytochrome b fragment was digested with 2 U of AatIIfor 2 h at 37°C. The fragments were separated by electrophoresis on a 5% non-denaturing polyacrylamide gel in Tris-borate/EDTA buffer and visualized after overnight exposure of the dried gels in a BioRad phosphorimager screen. Quantification was performed with the Molecular Analyst software (BioRad).
ACKNOWLEDGEMENTS
We would like to thank Dr H. Carrier for providing us with the muscle biopsies from patients, for performing the histochemical analysis and for being continuously interested in this work. We are also indebted to Dr J. M. Collombet for his gift of DNA from the patient's myoblasts and to Dr B. Mousson for the patient's and his mother's and sister's lymphocytes. The authors also acknowledge Drs P. Guibaud, B. Bady, C. Vial and F. Flocard for referring the patients. This work was supported by grants from the Centre National de la Recherche Scientifique (CNRS), the French Ministry of Education and Scientific Research (MERS), the `Association Française contre les Myopathies (AFM)' and the `Région Rhône-Alpes'. M.F.B. was supported by the Moroccan government and A.P. by the ARC (`Association pour la Recherche contre le Cancer').
REFERENCES
*To whom correspondence should be addressed. Tel: +33 4 72 44 83 56; Fax: +33 4 72 44 05 55; Email: godinot@univ-lyon1.fr
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Feb 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  W.-Y. Hung, J.-C. Lin, L.-M. Lee, C.-W. Wu, L.-M. Tseng, P.-H. Yin, C.-W. Chi, and H.-C. Lee Tandem duplication/triplication correlated with poly-cytosine stretch variation in human mitochondrial DNA D-loop region Mutagenesis, March 1, 2008; 23(2): 137 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Mancuso, C Vives-Bauza, M Filosto, R Marti, A Solano, J Montoya, J Gamez, S DiMauro, and A L Andreu A mitochondrial DNA duplication as a marker of skeletal muscle specific mutations in the mitochondrial genome J. Med. Genet., June 1, 2004; 41(6): e73 - e73. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||