Sequence and structural evolution of the KsgA/Dim1 methyltransferase family
BMC Research Notes volume 1, Article number: 108 (2008)
One of the 60 or so genes conserved in all domains of life is the ksgA/dim1 orthologous group. Enzymes from this family perform the same post-transcriptional nucleotide modification in ribosome biogenesis, irrespective of organism. Despite this common function, divergence has enabled some family members to adopt new and sometimes radically different functions. For example, in S. cerevisiae Dim1 performs two distinct functions in ribosome biogenesis, while human mtTFB is not only an rRNA methyltransferase in the mitochondria but also a mitochondrial transcription factor. Thus, these proteins offer an unprecedented opportunity to study evolutionary aspects of structure/function relationships, especially with respect to our recently published work on the binding mode of a KsgA family member to its 30S subunit substrate. Here we compare and contrast KsgA orthologs from bacteria, eukaryotes, and mitochondria as well as the paralogous ErmC enzyme.
By using structure and sequence comparisons in concert with a unified ribosome binding model, we have identified regions of the orthologs that are likely related to gains of function beyond the common methyltransferase function. There are core regions common to the entire enzyme class that are associated with ribosome binding, an event required in rRNA methylation activity, and regions that are conserved in subgroups that are presumably related to non-methyltransferase functions.
The ancient protein KsgA/Dim1 has adapted to cellular roles beyond that of merely an rRNA methyltransferase. These results provide a structural foundation for analysis of multiple aspects of ribosome biogenesis and mitochondrial transcription.
Ribosome biogenesis is a fundamental process in all cells, requiring the consumption of large quantities of cellular resources under the control of an extraordinary level of regulation. In comparing prokaryotic and eukaryotic ribosome biogenesis pathways, the conservation of the KsgA/Dim1 family is unique. The presence and function of this enzyme has been maintained in every evolutionary lineage, including eukaryotic organelles.
KsgA catalyzes the conversion of two adjacent adenosines in the small subunit rRNA (A1518 and A1519 of 16S rRNA, E. coli numbering) to N6,N6-dimethyladenosines . The KsgA family carries out this core methyltransferase function in all domains of life, including organelles, but has also added new roles as cellular organization became more complex. In eukaryotes, the KsgA ortholog Dim1 is essential for proper processing of the pre-18S small subunit rRNA , and Dim1 knockout is lethal. Pfc1, which is the KsgA ortholog found in chloroplasts of Arabidopsis thaliana, is important for chloroplast formation under chilling conditions .
A distinct eukaryotic ortholog, mtTFB, is transported into the mitochondria where, in addition to methylating the small subunit rRNA, it has adopted the ribosomally unrelated function of serving as a mitochondrial transcription factor [4, 5]. In some mitochondria, there are two separate mtTFB proteins, mtTFB1 and mtTFB2, which are proposed to have arisen from a gene duplication event . There is evidence that mtTFB1 has retained stronger methyltransferase activity, while mtTFB2 is more active as a transcription factor [5, 6]. The fungi have only a single mtTFB, suggesting either loss of one of the paralogs in this lineage, or that the duplication occurred later in evolution. In at least one case, S. cerevisiae, the single mtTFB protein (sc-mtTFB) serves as a transcription factor but has lost its methyltransferase activity entirely . sc-mtTFB lacks significant sequence homology to any of the KsgA/Dim1 enzymes; yeast mtTFBs are generally poorly conserved and difficult to identify via sequence homology .
Another important offshoot of the KsgA lineage is the Erm family of methyltransferases, which confer antibiotic resistance by methylating A2058 of the 23S rRNA . The present-day Erm family almost certainly resulted from one or more gene duplications of a KsgA gene and subsequent evolution to permit recognition of a distinct target base .
KsgA's remarkable degree of conservation, coupled with the adaptation of new cellular functions, give us a unique opportunity to look at structural/function evolution of a single protein lineage. Previous studies have established the structural similarity between some rRNA adenosine dimethyltransferases [11, 12]. Based on recent work in our groups we can now make predictions about how these proteins interact with their respective ribosomal targets. We also examine structural variations that may be important to the varied functions of KsgA orthologs and paralogs.
E. coli KsgA , human Dim1 (hDim1; A. Dong, H. Wu, H. Zeng, P. Loppnau, M. Sundstrom, C. Arrowsmith, A. Edwards, A. Bochkarev and A. Plotnikov, unpublished data), Plasmodium falciparum Dim1 , and sc-mtTFB  have been characterized structurally. Figure 1 shows the structures of KsgA, hDim1 and sc-mtTFB superimposed. The three structures share the same overall architecture, with a few notable differences (discussed below).
Divergent orthologues of bacterial KsgA, including Dim1 from the eukaryote S. cerevisiae and the archaeon Methanocaldococcus jannaschii as well as h-mtTFB1 and h-mtTFB2, can all complement for KsgA function in E. coli [5, 14, 15]. Therefore, it is expected that these functional orthologues all interact with substrate ribosome particles in a similar manner. We recently reported a mode of binding for KsgA onto 30S subunits . As was noted in this work, the binding site for KsgA on 30S involves protein interactions with 16S rRNA in helix 44 and in the 790 loop, but not in helix 45, the site of the methylated adenosines.
When 16S rRNA conservation is assessed across all three domains and the ribosome-containing organelles, regions that are important for translation show the highest evolutionary conservation . Included in the set of highly conserved nucleotides are residues at the base of helix 44 and the 790 loop which interact with KsgA . The high level of rRNA conservation at the KsgA binding site has likely constrained evolution of protein residues that interact with the rRNA, allowing evolutionarily divergent orthologues to recognize bacterial 30S subunits as a substrate. In order to assess KsgA conservation at the amino acid level, we aligned thirty-two KsgA, Dim1, mtTFB1 and Pfc1 proteins [see Additional file 1]. Since the mtTFB2 and mtTFB branches have become stronger transcription factors at the expense of the methyltransferase activity [5, 6], we omitted these branches from this analysis. We used the program Expresso to construct the alignment and assess the conservation at each amino acid position , and then mapped the results onto the KsgA structure. RNA conservation was taken from the minimal ribosome modeled by Mears et al . Figure 2 shows the patterns of conservation in KsgA and in helix 44 and the 790 loop. As expected, some of the most highly conserved residues in KsgA are found in the methyltransferase active site. Strikingly, there is also extensive conservation along the face of the protein that interacts with 16S rRNA, especially in the N-terminal domain. Conversely, the most poorly conserved residues are on the protein surface opposite its rRNA binding face.
The KsgA enzymes share the Rossman-like structural fold common to many SAM-dependent methyltransferases [12, 19]. This fold consists of a seven-stranded beta sheet surrounded by a variable number of alpha helices, with well-defined binding pockets for SAM and the target nucleotide. In this way, the methyltransferase function has a clear structural basis. However, the structural basis for the other functions of the KsgA/Dim1 enzymes is unknown. Extensive sequence alignments of KsgA, Dim1 and mtTFB enzymes identify elements unique to each group. The addition of specific inserts relative to KsgA likely has permitted the eukaryotic Dim1 and mtTFB enzymes to gain functions while maintaining their roles as methyltransferases.
Eukaryotic Dim1 has a large insert in the C-terminal domain compared to KsgA (Figure 3a). Sequence within this region is only sparsely conserved, but there is a well-conserved region of 20–40 amino acids [see Additional file 1]. To visualize what impacts this insert might have on ribosome subunit binding, we superimposed the structure of h-Dim1 onto KsgA in our docked KsgA/30S subunit complex (Figure 3b). When Dim1 is bound to the small ribosomal subunit we predict that this insert makes no direct contact with the small ribosomal subunit. Therefore, the question naturally becomes what is its role in ribosome biogenesis. It is tempting to speculate that Dim1's essential function in rRNA processing is mediated by this "extra" region.
A similar analysis can be performed by comparing KsgA and sc-mtTFB. mtTFB proteins have three inserts relative to KsgA proteins ; all of these inserts lie in the N-terminal domain (Figures 3c and 3d). We have superimposed sc-mtTFB onto the KsgA/30S model to evaluate what roles these inserts play, if any, in ribosome binding (Figure 3b). The first of these inserts is comprised of residues 116–130 of sc-mtTFB, and takes the form of a random coil leading into an extended beta strand; the second insert, residues 155–162, clusters together with insert 1 . These two inserts, which are located on the opposite side of the protein from the 30S subunit, are generally conserved in all mtTFB proteins [see Additional file 2]; however, insert 1 is shorter in mtTFB1 than in mtTFB and mtTFB2, which have evolved as transcription factors perhaps at the expense of their full function as methyltransferases. Given that these two inserts do not appear as though they make direct interactions with the small ribosomal subunit, we suggest that they might be important to mtTFB's role as a transcription factor, perhaps comprising a single binding surface involved in transcription. The third insert, consisting of residues 209–230, is on the other side of the protein from the first two, and consists of two alpha-helices which form a long loop on the protein surface. Residues in this loop are important for binding of sc-mtTFB to the S. cerevisiae mitochondrial RNA polymerase, and this insert may also contribute to the protein's role as a transcription factor . The presence of insert 3 is conserved in mtTFB and mtTFB2 proteins, but not in mtTFB1 proteins [see Additional file 2]. Notably, mtTFB insert 3 is in a position to interact with the 30S subunit, either in a positive or a negative manner.
sc-mtTFB, and possibly other fungal TFB proteins, have lost their methyltransferase activity during their evolution as transcription factors . Figure 4 highlights the structural explanation for this loss of function. KsgA and Dim1 both have a well-defined active site pocket, with clear binding sites for the SAM methyl donor and the target adenosine. The active site, which is essentially the same in the Erm enzymes [22–24], is homologous to other adenosine methyltransferases, and has been well-conserved. However, the area of the active site pocket has diverged in sc-mtTFB . sc-mtTFB retains residues important for SAM binding, including the canonical GXG motif, two acidic residues which stabilize SAM in the binding site, and an asparagine residue which interacts with SAM  (Figure 4a). However, the SAM binding pocket, which is clearly defined in KsgA (Figure 4b), is occluded in sc-mtTFB (Figure 4c), mainly by the side chains of residues Y22, Y54 and R79, which fill in the pocket. Similarly, the adenosine binding pocket in KsgA/Dim1 is not present in sc-mtTFB (Figure 4c), with that space being occupied by the side chains of residues K21, T139 and K240. Also, two aromatic residues that stabilize the target base, corresponding to residues Y116 and F181 in KsgA, are not found in the sc-mtTFB active site (Figure 4a).
Structural comparison of KsgA and ErmC' shows that the two proteins are very similar, with the most notable divergence being in the C-terminal domains (Figure 5a). These two enzymes perform the same chemistry, dimethylation of adenosine, and share an active site structure common to enzymes that catalyze this reaction [12, 23]. However, the Erm enzymes have evolved to recognize a distinct target, A2058 in 23S rRNA , and have thus been subject to different selective pressures in terms of substrate binding. While KsgA's RNA-binding surface extends along the length of the protein, ErmC' has been proposed to bind its target RNA across the cleft formed by the N- and C-terminal domains . We constructed an alignment of 31 Erm sequences [see Additional file 3] and mapped conservation onto the structure of ErmC' from B. subtilis ; the results are shown in Figure 5(b) and 5(c). Like KsgA, the active site contains some of the most highly conserved residues. However, the only other major conserved area is a helix in the N-terminal domain, along the cleft that forms the proposed RNA binding site. The Erm family's binding mode allows these enzymes to methylate a small RNA hairpin ; this lies in sharp contrast to KsgA's complex substrate requirements, which involve binding to an RNA helix distal from the site of its enzymatic activity . This dichotomy highlights the plasticity of the core KsgA structure, which has allowed orthologous and paralogous proteins to fill multiple and distinct roles with only modest adaptive changes.
To our knowledge, the KsgA proteins are unique in both their level of conservation and their functional flexibility. KsgA was one of the first rRNA methyltransferases to be identified and has received episodic scrutiny over the last thirty years. Despite a collection of important observations, a thorough understanding of this protein family's role in ribosome biogenesis has been elusive. Certainly, the KsgA/Dim1 family has significance beyond the methyltransferase activity that was first described. Sequence and structural comparisons of these proteins, in concert with data about the binding of KsgA to the 30S subunit, implicate certain regions that might be important to the various functions of KsgA orthologs. Further experiments will help define individual characteristics of this important family of proteins.
Helser TL, Davies JE, Dahlberg JE: Mechanism of kasugamycin resistance in Escherichia coli. Nat New Biol. 1972, 235: 6-9.
Lafontaine D, Vandenhaute J, Tollervey D: The 18S rRNA dimethylase Dim1p is required for pre-ribosomal RNA processing in yeast. Genes Dev. 1995, 9: 2470-2481. 10.1101/gad.9.20.2470.
Tokuhisa JG, Vijayan P, Feldmann KA, Browse JA: Chloroplast development at low temperatures requires a homolog of DIM1, a yeast gene encoding the 18S rRNA dimethylase. Plant Cell. 1998, 10: 699-711. 10.1105/tpc.10.5.699.
Lisowsky T, Michaelis G: Mutations in the genes for mitochondrial RNA polymerase and a second mitochondrial transcription factor of Saccharomyces cerevisiae. Mol Gen Genet. 1989, 219: 125-128. 10.1007/BF00261167.
Bonawitz ND, Clayton DA, Shadel GS: Initiation and beyond: multiple functions of the human mitochondrial transcription machinery. Mol Cell. 2006, 24: 813-825. 10.1016/j.molcel.2006.11.024.
Cotney J, Wang Z, Shadel GS: Relative abundance of the human mitochondrial transcription system and distinct roles for h-mtTFB1 and h-mtTFB2 in mitochondrial biogenesis and gene expression. Nucleic Acids Res. 2007, 35: 4042-4054. 10.1093/nar/gkm424.
Klootwijk J, Klein I, Grivell LA: Minimal post-transcriptional modification of yeast mitochondrial ribosomal RNA. J Mol Biol. 1975, 97: 337-350. 10.1016/S0022-2836(75)80044-1.
Carrodeguas JA, Yun S, Shadel GS, Clayton DA, Bogenhagen DF: Functional conservation of yeast mtTFB despite extensive sequence divergence. Gene Expr. 1996, 6: 219-230.
Weisblum B: Erythromycin resistance by ribosome modification. Antimicrob Agents Chemother. 1995, 39: 577-585.
Suvorov AN, van Gemen B, van Knippenberg PH: Increased kasugamycin sensitivity in Escherichia coli caused by the presence of an inducible erythromycin resistance (erm) gene of Streptococcus pyogenes. Mol Gen Genet. 1988, 215: 152-155. 10.1007/BF00331317.
Schubot FD, Chen CJ, Rose JP, Dailey TA, Dailey HA, Wang BC: Crystal structure of the transcription factor sc-mtTFB offers insights into mitochondrial transcription. Protein Sci. 2001, 10: 1980-1988. 10.1110/ps.11201.
O'Farrell HC, Scarsdale JN, Rife JP: Crystal structure of KsgA, a universally conserved rRNA adenine dimethyltransferase in Escherichia coli. J Mol Biol. 2004, 339: 337-353. 10.1016/j.jmb.2004.02.068.
Vedadi M, Lew J, Artz J, Amani M, Zhao Y, Dong A, Wasney GA, Gao M, Hills T, Brokx S, Qiu W, Sharma S, Diassiti A, Alam Z, Melone M, Mulichak A, Wernimont A, Bray J, Loppnau P, Plotnikova O, Newberry K, Sundararajan E, Houston S, Walker J, Tempel W, Bochkarev A, Kozieradzki I, Edwards A, Arrowsmith C, Roos D, Kain K, Hui R: Genome-scale protein expression and structural biology of Plasmodium falciparum and related Apicomplexan organisms. Mol Biochem Parasitol. 2007, 151: 100-110. 10.1016/j.molbiopara.2006.10.011.
Lafontaine D, Delcour J, Glasser AL, Desgres J, Vandenhaute J: The DIM1 gene responsible for the conserved m6(2)Am6(2)A dimethylation in the 3'-terminal loop of 18 S rRNA is essential in yeast. J Mol Biol. 1994, 241: 492-497. 10.1006/jmbi.1994.1525.
O'Farrell HC, Pulicherla N, Desai PM, Rife JP: Recognition of a complex substrate by the KsgA/Dim1 family of enzymes has been conserved throughout evolution. RNA. 2006, 12: 725-733. 10.1261/rna.2310406.
Xu Z, O'Farrell HC, Rife JP, Culver GM: A conserved rRNA methyltransferase regulates ribosome biogenesis. Nat Struct Mol Biol. 2008, 15: 534-536. 10.1038/nsmb.1408.
Mears JA, Cannone JJ, Stagg SM, Gutell RR, Agrawal RK, Harvey SC: Modeling a minimal ribosome based on comparative sequence analysis. J Mol Biol. 2002, 321: 215-234. 10.1016/S0022-2836(02)00568-5.
Armougom F, Moretti S, Poirot O, Audic S, Dumas P, Schaeli B, Keduas V, Notredame C: Expresso: automatic incorporation of structural information in multiple sequence alignments using 3D-Coffee. Nucleic Acids Res. 2006, 34: W604-W608. 10.1093/nar/gkl092.
Fauman EB, Blumenthal RM, Cheng X: Structure and evolution of AdoMet-dependent methyltransferases. S-Adenosylmethionine-Dependent Methyltransferases: Structures and Functions. Edited by: Cheng S, Blumenthal RM. 1998, Singapore: World Scientific Publishing Co. Pte. Ltd, 1-38.
Shutt TE, Gray MW: Homologs of mitochondrial transcription factor B, sparsely distributed within the eukaryotic radiation, are likely derived from the dimethyladenosine methyltransferase of the mitochondrial endosymbiont. Mol Biol Evol. 2006, 23: 1169-1179. 10.1093/molbev/msk001.
Cliften PF, Park JY, Davis BP, Jang SH, Jaehning JA: Identification of three regions essential for interaction between a sigma-like factor and core RNA polymerase. Genes Dev. 1997, 11: 2897-2909. 10.1101/gad.11.21.2897.
Bussiere DE, Muchmore SW, Dealwis CG, Schluckebier G, Nienaber VL, Edalji RP, Walter KA, Ladror US, Holzman TF, Abad-Zapatero C: Crystal structure of ErmC', an rRNA methyltransferase which mediates antibiotic resistance in bacteria. Biochemistry. 1998, 37: 7103-7112. 10.1021/bi973113c.
Schluckebier G, Zhong P, Stewart KD, Kavanaugh TJ, Abad-Zapatero C: The 2.2 A structure of the rRNA methyltransferase ErmC' and its complexes with cofactor and cofactor analogs: implications for the reaction mechanism. J Mol Biol. 1999, 289: 277-291. 10.1006/jmbi.1999.2788.
Yu L, Petros AM, Schnuchel A, Zhong P, Severin JM, Walter K, Holzman TF, Fesik SW: Solution structure of an rRNA methyltransferase (ErmAM) that confers macrolide-lincosamide-streptogramin antibiotic resistance. Nat Struct Biol. 1997, 4: 483-489. 10.1038/nsb0697-483.
Maravic G, Bujnicki JM, Feder M, Pongor S, Flogel M: Alanine-scanning mutagenesis of the predicted rRNA-binding domain of ErmC' redefines the substrate-binding site and suggests a model for protein-RNA interactions. Nucleic Acids Res. 2003, 31: 4941-4949. 10.1093/nar/gkg666.
Goedecke K, Pignot M, Goody RS, Scheidig AJ, Weinhold E: Structure of the N6-adenine DNA methyltransferase M. TaqI in complex with DNA and a cofactor analog. Nat Struct Biol. 2001, 8: 121-125. 10.1038/84104.
Nomenclature Center for MLS Genes. [http://faculty.washington.edu/marilynr]
Liu M, Kirpekar F, Van Wezel GP, Douthwaite S: The tylosin resistance gene tlrB of Streptomyces fradiae encodes a methyltransferase that targets G748 in 23S rRNA. Mol Microbiol. 2000, 37: 811-820. 10.1046/j.1365-2958.2000.02046.x.
This work was supported by the U.S. National Institutes of Health grants GM066900 to JPR and GM62432 to GMC.
The authors declare that they have no competing interests.
All authors discussed results and contributed to the manuscript. HCO compiled sequences and prepared alignments. JPR docked proteins onto the 30S subunit structure.
Electronic supplementary material
Additional file 1: Sequence alignment of KsgA orthologs. Archaeal, eukaryotic, and mitochondrial KsgA orthologues were identified by performing a genomic BLAST search using the E. coli protein sequence (accession number P06992) as the query sequence. Organisms were chosen to represent a broad evolutionary diversity of species. The structure-based sequence alignment was perfomed using the program Expresso . Structures used for the alignment were 1QYR, 1ZQ9 (A. Dong, H. Wu, H. Zeng, P. Loppnau, M. Sundstrom, C. Arrowsmith, A. Edwards, A. Bochkarev and A. Plotnikov, unpublished data), and 2H1R. Organisms represented are as follows. Eukaryotes: Arabidopsis thaliana (at), Dictyostelium discoideum (dd), Leishmania brazilensis (lb), Giardia lamblia (gl), Plasmodium vivax (pv), Homo sapiens (hs), Saccharomyces cerevisiae (sc), Drosophila melanogaster (dm), and Caenorhabditis elegans (ce). Archaea: Methanopyrus kandleri (mk), Methanosaeta thermophila (mth), Haloquadratum walsbyi (hw), Methanoculleus marisnigri (mma), Methanocaldococcus jannaschii (mj), Pyrococcus horikoshii (ph), Methanosphaera stadtmanae (ms), Picrophilus torridus (pt), Archaeoglobus fulgidus (af), Aeropyrum pernix (ap), Sulfolobus solfataricus (ss), Pyrobaculum aerophilum (pa), and Cenarchaeum symbiosum (cs). Bacteria: Synechococcus elongatus (se), Bacillus subtilis (bs), Mycobacterium tuberculosis (mtu), Thermus thermophilus (tt), Bacteroides fragilis (bf), Chlamydia trachomatis (ct), Borrelia burgdorferi (bb), and Escherichia coli (ec). Accession numbers for each sequence are found in Additional file 4. (PDF 280 KB)
Additional file 2: Sequence alignment of mtTFB, mtTFB1, and mtTFB2 proteins. The structure-based sequence alignment was perfomed using the program Expresso . The structure 1I4W was used for the alignment. Organisms represented are Caenorhabditis elegans (ce), Homo sapiens (hs), Drosophila melanogaster (dm), Anopheles gambiae (ag), Apis mellifera (am), Xenopus laevis (xl), Takifugu rubripes (tr), Ciona intestinalis (ci), Rattus norvegicus (rn), Pan troglodytes (pt), Mus musculus (mmu), Bos taurus (bt), Tetraodon nigroviridis (tn), Saccharomyces cerevisiae (sc), Schizosaccharomyces pombe (sp), Kluyveromyces lactis (kl), Eremothecium gossypii (eg), Candida albicans (ca), Dictyostelium discoideum (dd), Trypanosoma brucei (tb), and Leishmania major (lm). Accession numbers for each sequence are found in Additional file 4. (PDF 355 KB)
Additional file 3: Sequence alignment of Erm enzymes. Erm enzymes were identified using the Nomenclature Center for MLS Genes, maintained by Dr. Marilyn C. Roberst . One member of each class was chosen, with two exceptions. ErmI was not used because a corresponding sequence could not be found. Erm32 was not used because this enzyme methylates G748 rather than A2058 . The structure based sequence alignment was performed with Expresso . Structures used for the alignment were 1QAM and 1YUB. Organisms represented are Staphylococcus aureus, Enterococcus faecalis, Bacillus subtilis, Bacillus licheniformis, Saccharopolyspora erythraea, Bacteroides fragilis, Lysinibacillus sphaericus, Streptomyces thermotolerans, Streptomyces fradiae, Streptomyces coelicolor, Clostridium perfringens, Aeromicrobium erythreum, Lactobacillus reuteri, Streptomyces lincolnensis, Streptomyces viridochromogenes, Micromonospora griseorubida, Corynebacterium jeikeium, Streptomyces ambofaciens, Streptomyces venezuelae, Staphylococcus sciuri, Bacillus clausii, Bacteroides coprosuis, Micrococcus luteus, Mycobacterium tuberculosis, Mycobacterium smegmatis, Mycobacterium fortuitum, Mycobacterium mageritense, and Mycobacterium abscessus, Accession numbers are found in Additional file 4. (PDF 230 KB)
Additional file 4: Sequences used in protein alignments. Sequences were compiled from NCBI and Ensembl; organisms and accession numbers are indicated. (PDF 88 KB)
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.
About this article
Cite this article
O'Farrell, H.C., Xu, Z., Culver, G.M. et al. Sequence and structural evolution of the KsgA/Dim1 methyltransferase family. BMC Res Notes 1, 108 (2008). https://doi.org/10.1186/1756-0500-1-108