Genetic determination of the effect of post-translational modification on the innate immune response to the 19 kDa lipoprotein of Mycobacterium tuberculosis
© Wilkinson et al; licensee BioMed Central Ltd. 2009
Received: 07 September 2008
Accepted: 14 May 2009
Published: 14 May 2009
The 19 kDa lipoprotein of Mycobacterium tuberculosis (MTB) is an important target of the innate immune response. To investigate the effect of post-translation modification of this protein on innate recognition in the context of the whole bacillus, we derived a recombinant M. tuberculosis H37Rv that lacked the 19 kDa gene (Δ19) and complemented this strain by reintroduction of the 19 kDa gene into the chromosome as a single copy to produce Δ19::19. We also reintroduced the 19 kDa gene in two modified forms that lacked motifs for acylation (Δ19::19NA) and O-glycosylation (Δ19::19NOG).
Both acylation and O-glycosylation were necessary for the protein to remain within the cell. IL-1 Beta secretion from human monocytes was significantly reduced by deletion of the 19 kDa gene (p < 0.02). Complementation by the wild type, but not the mutagenised gene reversed this phenotype. The effect of deletion and complementation on IL-12p40 and TNF secretion was less marked with no statistically significant differences between strains. Although deletion of the 19 kDa reduced apoptosis, an effect that could also only be reversed by complementation with the wild type gene, the results were variable between donors and did not achieve statistical significance.
These results confirm in the context of the whole bacillus an important role for post-translational modification of the 19 kDa on both the cellular location and immune response to this protein.
The vast increase in knowledge that has accompanied the discovery of microbial pattern recognition receptors has focussed research into the microbial ligands that initiate these cellular responses [1, 2] For example it is now known that bacterial LPS triggers responses via Toll like receptor (TLR) 4, and Flagellin via TLR5 [3, 4]. It is also increasingly appreciated that receptors may co-operate to recognise specific ligands . Thus triacylated lipopeptide is recognised by a heterodimer of TLR2 and 1, with diacylated lipopeptide being recognised by the TLR2/6 heterodimer . Many types of pathogens produce lipoproteins and are thus in part recognised by TLR2 [6–8].
Mycobacterium tuberculosis has over 100 probable or known lipoproteins, many of which are concentrated in the cell wall . Whilst a role has been assigned to some of these proteins (e.g. Phosphate binding and transport for the Pst S1-3 group ), most have not been assigned a function. They are characterised by an acylated N-terminus, processing of which is mediated by the consecutive activity of prolipoprotein diacylglyceryl transferase (Lgt) and lipoprotein signal peptidase (Lsp A) . Deletion of LspA reduces the virulence of M. tuberculosis. In addition many of the lipoproteins have been found to be targets of both the innate and acquired immune response. A prominent target of the innate response is the 19 kDa lipoprotein encoded by Rv3763. This molecule has been intensively researched because of its pleiotropic effects on the innate immune response that include induction of cytokine genes, bacterial killing, induction of apoptosis, and the downregulation of Interferon-γ (IFN-γ) induced MHC Class II expression [12–20]. More recently it has also been suggested that the 19 kDa protein acts an adhesin .
Many of the above studies of the 19 kDa were performed with purified or recombinant protein that may not fully reflect the role of the molecule in the context of natural infection. In particular expression in E. coli is unlikely to reproduce native patterns of post-translational modiufication. We have previously reported the effect of deletion and overexpression of the 19 kDa on the innate immune response . We found that the deletion mutant (Δ19) was moderately impaired in its ability to multiply in human monocyte-derived macrophages (MDM). Surface expression of MHC class II molecules was reduced in phagocytes infected with MTB; this effect was not seen in cells infected with Δ19. Δ19 induced lower IL-1β secretion from monocytes and MDM. Overexpression of the 19 kDa increased IL-1β, IL-12p40 and TNF-α secretion irrespective of phagocyte maturity. These findings confirmed the 19 kDa protein to be an important mediator of the innate immune response in the context of the whole bacillus.
In addition to being acylated, the 19 kDa protein is glycosylated [23, 24]. Earlier work in our laboratories established that poly threonine motifs towards the N-terminal of the molecule form a major glycosylation site [23, 24]. The aim of this study was therefore to evaluate the innate immune response to Δ19 mutants that had been complemented with a single copy of mutagenised 19 kDa molecules lacking the motifs for acylation and O-glycosylation respectively.
Generation of recombinant strains of M. tuberculosis
The 19 kDa gene was deleted from M. tuberculosis (MTB) H37Rv to produce the Δ19 strain as previously described . Complementation of the Δ19 strain by the native and modified (non-acylated NA, and non-O-glycosylated NOG) 19 kDa genes led to the strains Δ19::19, Δ19::19NA and Δ19::19NOG. For complementation, the native sequence (including the entire intergenic region and part of upstream Rv3762 ORF) was amplified by PCR from H37Rv DNA. The site-directed mutagenised genes were amplified from previous episomal constructs [24, 25] engineered to come under the control of the endogenous 19 kDa promoter. Complementation was performed using the integrating vector pKINTA, based on the L5 phage integration system , which reintroduces a single copy of the 19 kDa gene into the chromsome under the control of its own promoter at the att B site . PCR was used to confirm deletion and insertion as previously described  and sequencing of the pKINTA insert confirmed nucleotide differences that would result in substitution of the N-terminal cysteine residue of the mature wild-type protein with alanine in Δ19::19NA; and substitution of two threonine clusters (5 amino acids in total) by valine residues in Δ19::19NOG. For Western blotting supernatants and sonicated preparations of wild-type M. tuberculosis H37Rv and the deleted and complemented strains were fractionated by SDS-PAGE and expression of the 19 kDa antigen compared by Western blot analysis using a polyclonal anti-19 kDa serum.
Isolation and culture of monocytes
Buffy coats from healthy donors were obtained from the National Blood Transfusion Service (Colindale, London, UK). Following dilution in RPMI (1/3 vol/vol), peripheral blood mononuclear cells (PBMC) were separated by centrifugation over Ficoll-Paque Plus (Pharmacia, Uppsala, Sweden). Cells were washed in RPMI and counted. Cells were suspended at 1.2 × 107/ml in RPMI/10% FCS medium and aliquots of 25 mls were added to 150 cm2 tissue culture flasks. Flasks were placed flat in a 5% CO2 incubator and monocytes allowed to adhere for 2 h at 37°C. Non-adherent cells were removed by washing 3 times with 10 mls of pre-warmed RPMI. Finally, 10 mls of ice-cold PBS was added and the flasks were incubated at 4°C for 20 mins. Using a scraper, monocytes were gently dislodged from the bottom of the flasks and pooled in a 50 ml Falcon tube to count. Cells were plated in RPMI containing 10% serum at 106/well in a 24-well tissue culture plate, and cultured overnight before infection.
Infection of cells
Bacilli used to infect cells were grown in Middlebrook 7H9 broth supplemented with ADC to mid-log phase (OD 0.4–0.8) then frozen in aliquots in 15% glycerol. The CFU content of aliquots was determined by serial dilution and plating on Middlebrook 7H11 agar supplemented with OADC. Monocytes were infected at a multiplicity of infection of 1:1 without removing non-phagocytosed bacteria. Culture duration was 72 hrs., at which time supernatants were aspirated, 0.22 μm filtered, and stored at -80°C pending analysis by ELISA.
Cytokine ELISA was performed using the DuoSet ELISA Development Systems (R&D Systems, Minneapolis, MN) following the manufacturer's recommendations. The sensitivity of the assays was 15 pg/ml for IL-12p40, 10 pg/ml for IL-1β and 50 pg/ml for TNF-α. Histone associated DNA fragments, released into tissue culture supernatant and interpreted as evidence of apoptotic cell death, were assayed by the cell death detection ELISA (Roche Applied Science, Lewes, Sussex, UK) according to the manufacturer's instructions.
Homologues of the M. tuberculosis 19 kDa gene LpqH were identified by Blast searches of sequenced genomes . Alignment of protein sequences was performed using Clustal W and results are displayed as a sequence pile-up and as a neighbour-joining tree. Strains and genome accession numbers: M. tuberculosis H37Rv, AL123456.2; M. smegmatis MC2155, CP000480.1; M. ulcerans Agy99, CP000325.1; M. marinum M, CP000854.1; M. leprae TN, AL450380.1; M. avium subsp. paratuberculosis K-10, AE016958.1; M. abscessus, CU458896.1; Nocardia farcinica IFM 10152, AP006618.1; Rhodococcus sp. RHA1, CP000431.1.
Paired and unpaired parametric variables were compared by student's t-test. Paired and unpaired non-parametric variables were compared by Wilcoxon signed rank or Mann Whitney U test respectively. Significance was inferred for p values ≤ 0.05.
Bioinformatic analysis of 19 kDa genes in various mycobacteria
Generation and characterization of recombinant M. tuberculosis strains
Induction of apoptosis
We have investigated deletion of the 19 kDa lipoprotein (Rv3763) from M. tuberculosis and chromosomal complementation of the deletion mutant by the wild type gene and site directed mutagenised variants lacking motifs for acylation and O-glycosylation. We have determined that both acylation and O-glycosylation are necessary for the protein to remain within the cell. Consistent with our previous findings, the 19 kDa is an important stimulus for the production of pro-inflammatory IL-1β, an effect that is dependent on acylation and O-glycosylation. The effect of deletion and complementation on IL-12p40 and TNF secretion was less marked with no statistically significant differences between strains. Although deletion of the 19 kDa reduced apoptosis, an effect that could also only be reversed by complementation with the wild type gene, the results were variable between donors and did not attain statistical significance.
An interesting finding was that 19 kDa protein was only detected in the supernatant of cultures of the non-acylated (NA) and non-O- glycosylated complemented strains, whereas the Δ19::19 strain expressed the molecule in both pellet and supernatant. This suggests that in order to be retained within the cell wall both acylation and glycosylation are necessary for anchoring within the cell wall. Whether this relates to a specific physicochemical interaction or merely reflects the recognised hydrophobicity of the mycobacterial cell membrane remains to be determined. Sartain and Belisle have recently shown that o- glycosylation affects the positioning in the cell wall but not the enzymatic activity of the superoxide dismuase sod C .
In a previous study overexpression of the 19 kDa in M. smegmatis reduced its capacity to induce the secretion of IL-12p40 and TNF. This effect was dependent on acylation and glycosylation, as tranformation of, M. smegmatis with NA and NOG variants of the 19 kDa did not reduce the secretion of these cytokines. By contrast overexpression of the native 19 kDa molecule in Δ19 strain of virulent M. tuberculosis had precisely the opposite effect, with the production of IL-12p40 and TNF increased irrespective of phagocyte maturity . In this study we reintroduced the 19 kDa gene as a single copy into the chromosome of H37Rv under the control of its own promoter. We precisely reproduced our previous findings with respect to the effect of deletion of the 19 kDa on the cytokine response of monocytes. We have shown that the 19 kDa mediated induction of IL-1β is dependent on acylation and glycosylation. Taken together these and other studies suggest a consistent effect of acylation and O-glycosylation on the cytokine response to the 19 kDa, but that the genetic background and level of expression are also important. Further evidence in favour of this hypothesis is our recent finding that a naturally occuring M. tuberculosis strain that lacks the 19 kDa gene does not have the same in vitro phenotype as the engineered knock out on the Rv background (data not shown). This potentially important finding requires further investigation as much of our knowledge about gene function in M. tuberculosis is inferred from studies of isogenic mutants on the H37Rv background.
Considerable evidence now points to the protective role of macrophage apoptosis in tuberculosis. Apoptosis may prevent the release of intracellular components and the spread of mycobacterial infection by sequestering the pathogens within apoptotic bodies [14, 31, 32]. In addition, uptake of apoptotic debris by competent phagocytes allows efficient cross-presentation of M. tuberculosis antigens . Thus, the avoidance of apoptosis may be considered a virulence mechanism and a recent study has in fact reported a inverse relationship between the intracellular growth rate and the ability of strains to induce apoptosis . Two previous studies have implicated the 19 kDa as pro-apoptotic [14, 17] and our results, although variable between donors tend to support this conclusion. However the dependence or otherwise on post-translation modification requires additional work as the findings of Lopez et al. suggested that this effect was acylation independent, whereas the trend in our study suggest acylation is necessary (Figure 6).
In conclusion we have presented further evidence of the role of the 19 kDa as a key modulator of the human innate immune response. There is considerable evidence that the protein downregulates IFN-γ induced macrophage activation, an effect that will tend to favour bacillary survival during the development of an acquired immune response. On the other hand the molecule will tend to give away the presence of bacilli to the innate system early in infection, perhaps teleologically explaining why it is not upregulated early after infection . In addition, this work provides further evidence of the utility of defined mutants to delineate key determinants of the innate immune response in the context of whole bacilli.
This work was supported by the Wellcome Trust (Refs. 064261, 060079 and 038997).
- Gordon S: Pattern recognition receptors: doubling up for the innate immune response. Cell. 2002, 111 (7): 927-930. 10.1016/S0092-8674(02)01201-1.PubMedView ArticleGoogle Scholar
- Takeda K, Kaisho T, Akira S: Toll-like receptors. Annu Rev Immunol. 2003, 21: 335-376. 10.1146/annurev.immunol.21.120601.141126.PubMedView ArticleGoogle Scholar
- Hawn TR, Verbon A, Lettinga KD, Zhao LP, Li SS, Laws RJ, Skerrett SJ, Beutler B, Schroeder L, Nachman A: A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to legionnaires' disease. J Exp Med. 2003, 198 (10): 1563-1572. 10.1084/jem.20031220.PubMed CentralPubMedView ArticleGoogle Scholar
- Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, Du X, Birdwell D, Alejos E, Silva M, Galanos C: Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science. 1998, 282 (5396): 2085-2088. 10.1126/science.282.5396.2085.PubMedView ArticleGoogle Scholar
- Ozinsky A, Underhill DM, Fontenot JD, Hajjar AM, Smith KD, Wilson CB, Schroeder L, Aderem A: The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors. Proc Natl Acad Sci USA. 2000, 97 (25): 13766-13771. 10.1073/pnas.250476497.PubMed CentralPubMedView ArticleGoogle Scholar
- Seya T, Matsumoto M: A lipoprotein family from Mycoplasma fermentans confers host immune activation through Toll-like receptor 2. Int J Biochem Cell Biol. 2002, 34 (8): 901-906. 10.1016/S1357-2725(01)00164-9.PubMedView ArticleGoogle Scholar
- Alexopoulou L, Thomas V, Schnare M, Lobet Y, Anguita J, Schoen RT, Medzhitov R, Fikrig E, Flavell RA: Hyporesponsiveness to vaccination with Borrelia burgdorferi OspA in humans and in TLR1- and TLR2-deficient mice. Nat Med. 2002, 8 (8): 878-884.PubMedGoogle Scholar
- Darrah PA, Monaco MC, Jain S, Hondalus MK, Golenbock DT, Mosser DM: Innate immune responses to Rhodococcus equi. J Immunol. 2004, 173 (3): 1914-1924.PubMedView ArticleGoogle Scholar
- Young DB, Garbe T: Lipoprotein antigens of M. tuberculosis. Res Microbiol. 1991, 142: 55-65. 10.1016/0923-2508(91)90097-T.PubMedView ArticleGoogle Scholar
- Peirs P, Lefevre P, Boarbi S, Wang XM, Denis O, Braibant M, Pethe K, Locht C, Huygen K, Content J: Mycobacterium tuberculosis with disruption in genes encoding the phosphate binding proteins PstS1 and PstS2 is deficient in phosphate uptake and demonstrates reduced in vivo virulence. Infect Immun. 2005, 73 (3): 1898-1902. 10.1128/IAI.73.3.1898-1902.2005.PubMed CentralPubMedView ArticleGoogle Scholar
- Sander P, Rezwan M, Walker B, Rampini SK, Kroppenstedt RM, Ehlers S, Keller C, Keeble JR, Hagemeier M, Colston MJ: Lipoprotein processing is required for virulence of Mycobacterium tuberculosis. Mol Microbiol. 2004, 52 (6): 1543-1552. 10.1111/j.1365-2958.2004.04041.x.PubMedView ArticleGoogle Scholar
- Brightbill HD, Libraty DH, Krutzik SR, Yang RB, Belisle JT, Bleharski JR, Maitland M, Norgard MV, Plevy SE, Smale ST: Host defense mechanisms triggered by microbial lipoproteins through toll-like receptors. Science. 1999, 285 (5428): 732-736. 10.1126/science.285.5428.732.PubMedView ArticleGoogle Scholar
- Noss EH, Pai RK, Sellati TJ, Radolf JD, Belisle J, Golenbock DT, Boom WH, Harding CV: Toll-like receptor 2-dependent inhibition of macrophage class II MHC expression and antigen processing by 19-kDa lipoprotein of Mycobacterium tuberculosis. J Immunol. 2001, 167 (2): 910-918.PubMedView ArticleGoogle Scholar
- Lopez M, Sly LM, Luu Y, Young D, Cooper H, Reiner NE: The 19-kDa Mycobacterium tuberculosis protein induces macrophage apoptosis through Toll-like receptor-2. J Immunol. 2003, 170 (5): 2409-2416.PubMedView ArticleGoogle Scholar
- Fortune SM, Solache A, Jaeger A, Hill PJ, Belisle JT, Bloom BR, Rubin EJ, Ernst JD: Mycobacterium tuberculosis inhibits macrophage responses to IFN-gamma through myeloid differentiation factor 88-dependent and -independent mechanisms. J Immunol. 2004, 172 (10): 6272-6280.PubMedView ArticleGoogle Scholar
- Thoma-Uszynski S, Stenger S, Takeuchi O, Ochoa MT, Engele M, Sieling PA, Barnes PF, Rollinghoff M, Bolcskei PL, Wagner M: Induction of direct antimicrobial activity through mammalian toll-like receptors. Science. 2001, 291 (5508): 1544-1547. 10.1126/science.291.5508.1544.PubMedView ArticleGoogle Scholar
- Ciaramella A, Cavone A, Santucci MB, Garg SK, Sanarico N, Bocchino M, Galati D, Martino A, Auricchio G, D'Orazio M: Induction of Apoptosis and Release of Interleukin-1 beta by Cell Wall-Associated 19-kDa Lipoprotein during the Course of Mycobacterial Infection. J Infect Dis. 2004, 190 (6): 1167-1176. 10.1086/423850.PubMedView ArticleGoogle Scholar
- Post FA, Manca C, Neyrolles O, Ryffel B, Young DB, Kaplan G: The 19 kDa lipoprotein of Mycobacterium tuberculosis inhibits Mycobacterium smegmatis induced cytokine production by human macrophages in vitro. Infect Immun. 2001, 69: 1433-1439. 10.1128/IAI.69.3.1433-1439.2001.PubMed CentralPubMedView ArticleGoogle Scholar
- Pai RK, Convery M, Hamilton TA, Boom WH, Harding CV: Inhibition of IFN-gamma-induced class II transactivator expression by a 19-kDa lipoprotein from Mycobacterium tuberculosis: a potential mechanism for immune evasion. J Immunol. 2003, 171 (1): 175-184.PubMedView ArticleGoogle Scholar
- Tobian AA, Potter NS, Ramachandra L, Pai RK, Convery M, Boom WH, Harding CV: Alternate class I MHC antigen processing is inhibited by Toll-like receptor signaling pathogen-associated molecular patterns: Mycobacterium tuberculosis 19-kDa lipoprotein, CpG DNA, and lipopolysaccharide. J Immunol. 2003, 171 (3): 1413-1422.PubMedView ArticleGoogle Scholar
- Diaz-Silvestre H, Espinosa-Cueto P, Sanchez-Gonzalez A, Esparza-Ceron MA, Pereira-Suarez AL, Bernal-Fernandez G, Espitia C, Mancilla R: The 19-kDa antigen of Mycobacterium tuberculosis is a major adhesin that binds the mannose receptor of THP-1 monocytic cells and promotes phagocytosis of mycobacteria. Microb Pathog. 2005, 39 (3): 97-107. 10.1016/j.micpath.2005.06.002.PubMedView ArticleGoogle Scholar
- Stewart GR, Wilkinson KA, Newton SM, Sullivan SM, Neyrolles O, Wain JR, Patel J, Pool KL, Young DB, Wilkinson RJ: Effect of Deletion or Overexpression of the 19-Kilodalton Lipoprotein Rv3763 on the Innate Response to Mycobacterium tuberculosis . Infect Immun. 2005, 73 (10): 6831-6837. 10.1128/IAI.73.10.6831-6837.2005.PubMed CentralPubMedView ArticleGoogle Scholar
- Herrmann JL, Delahay R, Gallagher A, Robertson B, Young D: Analysis of post-translational modification of mycobacterial proteins using a cassette expression system. FEBS Lett. 2000, 473 (3): 358-362. 10.1016/S0014-5793(00)01553-2.PubMedView ArticleGoogle Scholar
- Herrmann JL, O'Gaora P, Gallagher A, Thole JE, Young DB: Bacterial glycoproteins: a link between glycosylation and proteolytic cleavage of a 19 kDa antigen from Mycobacterium tuberculosis. EMBO J. 1996, 15 (14): 3547-3554.PubMed CentralPubMedGoogle Scholar
- Neyrolles O, Gould K, Gares M-P, Brett S, Janssen R, O'Gaora P, Herrmann J-L, Prévost M-C, Perret E, Thole J: Lipoprotein access to MHC Class I presentation during infection of murine macrophages with live mycobacteria. J Immunol. 2001, 166: 447-457.PubMedView ArticleGoogle Scholar
- Lee MH, Pascopella L, Jacobs WR, Hatfull GF: Site-specific integration of mycobacteriophage L5: integration-proficient vectors for Mycobacterium smegmatis, Mycobacterium tuberculosis, and bacille Calmette-Guerin. Proc Natl Acad Sci USA. 1991, 88 (8): 3111-3115. 10.1073/pnas.88.8.3111.PubMed CentralPubMedView ArticleGoogle Scholar
- Stewart GR, Newton SM, Wilkinson KA, Humphreys IR, Murphy HN, Robertson BD, Wilkinson RJ, Young DB: The stress-responsive chaperone alpha-crystallin 2 is required for pathogenesis of Mycobacterium tuberculosis . Mol Microbiol. 2005, 55 (4): 1127-1137. 10.1111/j.1365-2958.2004.04450.x.PubMedView ArticleGoogle Scholar
- Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ: Basic local alignment search tool. J Mol Biol. 1990, 215 (3): 403-410.PubMedView ArticleGoogle Scholar
- Babu MM, Priya ML, Selvan AT, Madera M, Gough J, Aravind L, Sankaran K: A database of bacterial lipoproteins (DOLOP) with functional assignments to predicted lipoproteins. J Bacteriol. 2006, 188 (8): 2761-2773. 10.1128/JB.188.8.2761-2773.2006.PubMed CentralPubMedView ArticleGoogle Scholar
- Sartain MJ, Belisle JT: N-Terminal clustering of the O-glycosylation sites in the Mycobacterium tuberculosis lipoprotein SodC. Glycobiology. 2009, 19 (1): 38-51. 10.1093/glycob/cwn102.PubMed CentralPubMedView ArticleGoogle Scholar
- Balcewicz-Sablinska MK, Keane J, Kornfeld H, Remold HG: Pathogenic Mycobacterium tuberculosis evades apoptosis of host macrophages by release of TNF-R2, resulting in inactivation of TNF- alpha. J Immunol. 1998, 161 (5): 2636-2641.PubMedGoogle Scholar
- Fratazzi C, Arbeit RD, Carini C, Balcewicz-Sablinska MK, Keane J, Kornfeld H, Remold HG: Macrophage apoptosis in mycobacterial infections. J Leukoc Biol. 1999, 66 (5): 763-764.PubMedGoogle Scholar
- Winau F, Weber S, Sad S, de Diego J, Hoops SL, Breiden B, Sandhoff K, Brinkmann V, Kaufmann SH, Schaible UE: Apoptotic Vesicles Crossprime CD8 T Cells and Protect against Tuberculosis. Immunity. 2006, 24 (1): 105-117. 10.1016/j.immuni.2005.12.001.PubMedView ArticleGoogle Scholar
- Park JS, Tamayo MH, Gonzalez-Juarrero M, Orme IM, Ordway DJ: Virulent clinical isolates of Mycobacterium tuberculosis grow rapidly and induce cellular necrosis but minimal apoptosis in murine macrophages. J Leukoc Biol. 2005, 79 (1): 80-6. 10.1189/jlb.0505250.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.