Fast, easy and efficient: site-specific insertion of transgenes into Enterobacterial chromosomes using Tn7 without need for selection of the insertion event
© McKenzie and Craig; licensee BioMed Central Ltd. 2006
Received: 24 January 2006
Accepted: 28 April 2006
Published: 28 April 2006
Inserting transgenes into bacterial chromosomes is generally quite involved, requiring a selection for cells carrying the insertion, usually for drug-resistance, or multiple cumbersome manipulations, or both. Several approaches use phage λ red recombination, which allows for the possibility of mutagenesis of the transgene during a PCR step.
We present a simple, rapid and highly efficient method for transgene insertion into the chromosome of Escherichia coli, Salmonella or Shigella at a benign chromosomal site using the site-specific recombination machinery of the transposon Tn7. This method requires very few manipulations. The transgene is cloned into a temperature-sensitive delivery plasmid and transformed into bacterial cells. Growth at the permissive temperature with induction of the recombination machinery leads to transgene insertion, and subsequent growth at the nonpermissive temperature cures the delivery plasmid. Transgene insertion is highly site-specific, generating insertions solely at the Tn7 attachment site and so efficient that it is not necessary to select for the insertion.
This method is more efficient and straightforward than other techniques for transgene insertion available for E. coli and related bacteria, making moving transgenes from plasmids to a chromosomal location a simple matter. The non-requirement for selection is particularly well suited for use in development of unmarked strains for environmental release, such as live-vector vaccine strains, and also for promoter-fusion studies, and experiments in which every bacterial cell must express a transgene construct.
Expression of transgenes in bacteria is routinely accomplished by expressing the gene from an extrachromosomal plasmid. However, chromosomal expression, with its single-copy nature and inherent stability, is preferable in many situations. For example, in experiments where physiological levels of protein are desired, including genetic complementation experiments, even low copy number plasmids can yield non-physiological levels of expression making single-copy expression a better choice . In addition, interpretation of experiments involving plasmids can often be confounded by the heterogeneity of the population arising from plasmid loss due to the imperfect segregation of most plasmids . Recent experiments in Salmonella have shown that merely having a plasmid present in a strain can have an effect on virulence . An additional difficulty with plasmids is that a selection is required to maintain them during growth. Typically this is provided by a drug-resistance (drugR) marker, however drugR markers are undesirable in some contexts such as when bacteria will be introduced into humans (e.g., for live vaccine purposes ). Expression of transgenes from the chromosome bypasses the sub-ideal features of plasmid-based expression. However, most systems for transgene insertion are somewhat cumbersome, involving multiple steps, and also require a drugR marker to allow selection for relatively rare integration events [4, 5]. Those that do not involve drug selection rely on other selection strategies that involve multiple steps . We describe a system that makes chromosomal transgene expression, with all its advantages, and without need of a selection, so simple to achieve that it can be routinely used.
We have developed a system for chromosomal transgene insertion that does not require a selection for transgene insertion based on the site-specific recombination system of the bacterial transposon, Tn7. Tn7 uses five proteins for transposition, TnsABCDE . TnsA and TnsB form the transposase that recognizes approximately 240 bp sequences at the ends of Tn7 and does the breakage and joining reactions of transposition, moving the transposon into the new site. TnsD and TnsE are alternative site selectors that determine where the transposon will insert, and TnsC links TnsA and TnsB to the site selectors, activating transposition through regulation of the ATP-bound state of TnsC. TnsE mediates transposition into conjugative plasmids, whereas TnsD promotes site-specific transposition into a chromosomal "safe haven" attachment site (attTn7) at the 3' end of the glmS gene. TnsD recognizes a site in the highly conserved glmS gene (essential in E. coli  and likely essential in nearly all bacteria), and inserts immediately downstream of glmS into attTn7, such that no gene is disrupted, and at no discernable fitness cost to the host . In cells expressing TnsABCD, transposition occurs exclusively into attTn7, and not into other locations . Tn7 appears to have a broad host range, having been demonstrated to function in many species of bacteria .
Our overall strategy is as follows: tnsABCD and a transgene flanked by recognition sites for the tranposase are delivered into cells on a temperature-sensitive plasmid that can later be cured. We use transformation (or conjugation) to establish a culture of plasmid-bearing cells, induce expression of the TnsABCD proteins to promote transposition into the attTn7 site, and then cure the plasmid. Using this strategy the majority of cells treated acquire a transgene insertion at attTn7, as detailed below.
Frequency of transgene insertion at E. coli attTn7 in the absence of selection. For each experiment shown three independent electrotransformations of pGRG25, which carries the multiple cloning site as the transgene, into strain MG1655 were performed. One single colony from each transformation was streaked out once at 32°C on LB with ampicillin, and then a culture of each colony was grown overnight in LB at 32°C. Appropriate dilutions were then plated on LB at 42°C. Ten colonies from each culture were examined by PCR to determine if they carried an insertion at the attachment site. The values shown are the average frequency from 3 cultures with 10 colonies examined per culture (i.e., a total of 30 colonies examined per experiment). The final value shows a mean average of all 6 experiments and the standard error of the mean (SEM).
Frequency of colonies carrying Tn7 at attachment site
Average ± SEM
0.79 ± 0.05
Frequency of transgene insertion into sites other than attTn7. For each experiment shown, three independent electrotransformations of pGRG19 (carrying a ChlR transgene) into MG1655 and MG1655 attTn7::Tn7kan were done. The presence of the Tn7 at attTn7 prevents insertion events there and allows measurement of the frequency at which Tn7 can transpose to other sites as discussed in the text . One single colony from each transformation was streaked out once at 32°C on LB with ampicillin. From this, a culture was grown overnight in LB at 32°C, and appropriate dilutions were plated at 42°C on LB and LB with chloramphenicol. Values shown are the frequency of ChlR
Frequency of ChlR cells
Average ± SEM
7.0 × 10-1
7.2 × 10-1
(7.9 ± 0.8) × 10-1
9.4 × 10-1
6.8 × 10-5
4.6 × 10-5
(5.5 ± 0.7) × 10-5
5.2 × 10-5
Although the above experiment confirms that insertion is specific to attTn7, we wanted to further ensure that this system was not promoting insertions at other sites, given the extraordinary frequency of transposon insertion observed. To detect rare non-attTn7 insertion events, we made use of a property of Tn7 known as transposition immunity whereby Tn7 blocks transposition into a site already occupied by a Tn7 . We compared the ability of our system to deliver a ChlR transgene into a pair of strains, one with Tn7 already in the attachment site (to measure non-specific transposition frequencies) and one without Tn7 (to measure integration frequencies at attTn7). Selecting for the ChlR transgene, we found that the frequency of insertion (ChlR cells) was nearly 104 fold lower in the strain with a blocked attachment site (Table 2). Thus insertion at secondary sites can occur but the frequency is so much lower that it is extremely unlikely to recover a secondary non-site-specific insertion (approximately 1/104 isolated colonies will carry an insertion at a site other than attTn7).
Discussion & conclusion
This system for creating chromosomal transgenes is fast, easy and efficient. Transposition occurs in 79% of cells treated (Table 1), and occurs very specifically into the Tn7 attachment site. This is the highest Tn7 transposition seen in any in vivo system , with frequencies of closer to 1% being more typical. In our system, transposition occurs at such high frequencies that selection for transgene insertion is not required. This is not the first system to employ Tn7 for transgene integration. However, previously described systems utilize multiple plasmids and are not efficient enough to allow detection of insertion events in the absence of a selection [14–16]. Non-specific transposition occurs at a frequency of 6.8 × 10-5 in our system (Table 2), indicating that there is little worry of non-specific or secondary transposition events. We have demonstrated the utility of this system in ten members of the E. coli reference collection of environmental isolates (data not shown)  and others have successfully inserted transgenes into Enteropathogenic and Enterohemorrhagic E. coli isolates (R. Keller, J.W. Yoon and J. Kaper, personal communication). We have also demonstrated the utility of this system in another Enterobacterial species, Salmonella typhimurium (data not shown) indicating the broad utility of our approach. Our transgene integration method can be used in any bacterium that supports replication of the pSC101 origin (i.e., members of the E. coli, Salmonella, or Shigella families).
This transgene integration system accommodates large transgenes, as we have seen similar frequencies of insertion with inserts as large as 3.2 kb (data not shown). Others have used pGRG25 to insert the 7.1 kb luxCDABE operon into Extraintestinal pathogenic E. coli and observed similar insertion frequencies (P. Germon, personal communication). It is likely to be possible to insert even larger transgenes, as wildtype Tn7 carries 14 kb between the left and right ends.
Because the Tn7 attachment site is highly conserved amongst diverse bacterial species  this system may be useful in many different bacteria. Possible limitations of the system with regard to whether it will work in other organisms are: 1) the ability to deliver DNA via transformation or conjugation, 2) the ability of the plasmid to be replicated sufficiently to support transposase expression, and 3) the activity of the araBAD promoter driving transposase expression. Restriction of foreign DNA can have a dramatic negative effect on transformation efficiency, however in cases where restriction may be a problem, for example, with environmental isolates, the Ocr type 1 restriction inhibitor  (Epicentre Biotechnologies) can greatly improve transformation efficiency. Systems utilizing broad host range plasmids have been described that use Tn7 transposition functions to insert transgenes into the chromosome (e.g. ) in numerous different bacteria. Although these other systems work in diverse bacteria and are very useful tools for that reason, they have more components and are several orders of magnitude less efficient requiring drugR markers to detect transgene insertions. The drugR marker can be removed after insertion with additional steps, but clearly the system we describe here for single-step insertion without selection is simpler and easier in the organisms where we have demonstrated functionality (E.coli/Salmonella/Shigella).
Expanding the utility of our system to include bacteria outside the Enterobacteriaceae might be possible by incorporating a broad host range replicon combined with a negative selection for loss of the plasmid. Indeed, Tn7 seems the ideal molecular machine for delivering transgenes given that the attTn7 sequence appears to be present in most genomes examined (it occurs twice in some, but this can be dealt with as in ). Tn7 has been demonstrated to move in a broad variety of Gram negative bacteria  and the presence of Tn7-like sequences in Bacillus cereus and Clostridium thermocellum suggests that it would transpose in Gram positives as well ( and GenBank accession #NZ_AABG04000054).
A commonly used method for transgene insertion available prior to this work  is based on phage λ red recombination. It usually requires PCR amplification of the substrate, transformation of the red gam expression system and transformation of the recombination substrate, followed by curing and PCR verification. Removal of the necessary drugR marker requires an additional transformation and curing. It works at frequencies at or below 10-6, meaning that a selection is required. The Tn7-based method described here inserts genes at such high frequencies that selection is not required. It requires very few steps: cloning the transgene, transformation into the strain of interest and growth at 2 temperatures, followed by PCR verification of insertion. Clearly our system is an improved method with fewer steps and greatly improved efficiency.
We anticipate this system to be useful for studies requiring single-copy gene expression, for example genetic studies, including complementation experiments. In addition, we envision that this system will be useful for development of bacterial strains for environmental release (that may not have antibiotic-resistance markers), including live-vector vaccine strains . Expression of heterologous antigens in vaccine strains is currently largely plasmid-based, one reason being that plasmid copy number facilitates expression of sufficient antigen to engender an immune response. However, recent work has demonstrated that immunologically relevant levels of antigen can be expressed from the chromosome . The construction of stable and selection-free chromosomal transgenes using this Tn7-based system should support the development of chromosomal antigen expression technology. It is worth noting that because Tn7 recognizes an attTn7 site in the human genome  it may be possible to create vectors for use in human cells to deliver genes at high-frequency into an innocuous site in the human genome for gene therapy.
Plasmid pCW4 , containing tnsABCD, was digested with Pac I and partially digested with Drd I, to liberate a 6636 bp Pac I-Drd I fragment. The overhanging ends were made blunt with Klenow fragment and ligated to Sma I linearized pBAD18 . The plasmid containing tnsABCD in the correct orientation relative to the PBAD promoter was named pGRG2. pGRG2 and pMAK700 , a pSC101 plasmid with a temperature-sensitive replicon, were digested with Nae I, releasing a 9541 bp tnsABCD containing fragment and a 2416 bp pSC101 origin-containing fragment, respectively. These two fragments were ligated together to create pGRG3, which contains the bla gene encoding β-lactamase, the araC gene and the PBAD-tnsABCD genes, and the temperature-sensitive pSC101 origin of replication. KanR and ChlR mini-Tn7 s were inserted into pGRG3 using red-gam recombination as follows: The primers ATCATGGCAATTCTGGAAGAAATAGCGCTTTCAGCCtgtgggcggacaaaatagttgggaactggga and CATGAGCAGATCCTCTACGCCGGACGCATCGTGGCCtgtgggcggacaataaagtcttaaactgaa were used to amplify the KanR mini-Tn7 from pGPS1.1 and the ChlR mini-Tn7 from pGPS2.1 (both from New England Biolabs). The lower case letters are complementary to Tn7 and the upper case letters are homologous to sequences in pGRG3. The resulting PCR products were transformed into DH5α carrying the plasmid pTP223 , which expresses the phage λ red recombinase products, and KanR or ChlR recombinants were selected. The resulting KanR and ChlR mini-Tn7 containing plasmids were named pGRG8 and pGRG6. Both Tn7 inserts in the plasmids were sequenced to ensure that no mutations were inserted by PCR. Any mutations were corrected to the wildtype sequence by subcloning from samples that did not contain those mutations. pGRG17 was created by digesting pGRG8 with Dra III and Eag I to remove the kanamycin resistance gene. This was replaced with the annealed oligonucleotides GGCCGTGGCGCGCCTCCTAGGTGCTCGAGTGGCGGCCGCTATTGAGGGATCTGATTAATTAAAAC and TTAATTAATCAGATCCCTCAATAGCGGCCGCCACTCGAGCACCTAGGAGGCGCGCCAC to create a multiple cloning site with the following unique restriction sites: Avr II, Not I, Pac I, and Xho I.
To provide conjugation as an alternative way to introduce these plasmids into bacterial strains, we cloned the oriT site from the RP4 plasmid sequences in the strain SM10 . oriT was amplified by PCR using primers GGCGCCGGCCAGCCTCGCAGAGCA and GGCGCCGGGCAGGATAGGTGAAGT, and cloned into pCR2.1-topo using the Topo TA cloning kit (Invitrogen). The sequence and transfer activity was confirmed, then the oriT sequence was moved by moving the 113 bp Sfo I fragment into the Nae I site (resulting in the plasmid in parentheses) of pGRG6 (pGRG19), pGRG8 (pGRG20), and pGRG17 (pGRG25). pGRG36 is identical to pGRG25, except that it contains a Sma I site in the multiple cloning site. Both pGRG25 and pGRG36 sequences can be downloaded at GenBank using NCBI accession # DQ460223. The ability of the delivery plasmids to be transferred by conjugation was confirmed using pGRG25-containing SM10 as donor in a standard conjugation assay at 32°C .
To facilitate the cloning of large fragments into the delivery vehicle we cloned the RfC.1 element from the Gateway cloning system (Invitrogen) into the Sma I site of pGRG36, yielding pGRG37. This allows in vitro movement of transgenes cloned in Gateway cloning vectors into pGRG37, bypassing the need for standard cloning techniques .
Plasmid and strain handling
Plasmids and strains were prepared as described  with the exception that transposition plasmids were grown in LB medium with ampicillin and 0.1% glucose, to repress transposition, and plasmids were prepared using Qiagen very low-copy plasmid protocols.
Transformation and transposition protocol
E. coli electro-competent cells were prepared normally, as described in , with the exception that they were grown in Luria-Bertani (LB) medium with 0.1% glucose to repress expression of the transposase during recovery. Transformants were selected on LB + 100 μg/mL ampicillin at 32°C and streaked out once, then grown overnight in LB at 32°C, with no drugs present. Dilutions of this culture were plated on LB + 30 μg/mL chloramphenicol (pGRG19) or on LB alone (pGRG25). Colonies were streaked at least once at 42°C to ensure plasmid loss. Transposition was verified by absence of the ampicillin resistance marker and by PCR using the primers 5'GATGCTGGTGGCGAAGCTGT and 5'GATGACGGTTTGTCACATGGA, which flank the attTn7 site.
The authors would like to thank J. Kaper, S.R. Kushner and S. Maloy for plasmids and strains. G.J.M. would like to thank L.S. Frost, R. Keller, M.-J. Lombardo and J.F. Petrosino for helpful discussions, P. Germon and J. Kaper for sharing unpublished data and B.C. Bazinet for graphical assistance preparing figures. G.J.M. holds an Alberta Heritage Foundation for Medical Research postdoctoral fellowship. N.L.C. is an Investigator of the Howard Hughes Medical Institute.
- Grozdanov L, Zähringer U, Blum-Oehler G, Brade L, Henne A, Knirel YA, Schombel U, Schulze J, Sonnenborn U, Gottschalk G, Hacker J, Rietschel ET, Dobrindt U: A single nucleotide exchange in the wzy gene is responsible for the semirough O6 lipopolysaccharide phenotype and serum sensitivity of Escherichia coli strain Nissle 1917. J Bacteriol. 2002, 184 (21): 5912-25. 10.1128/JB.184.21.5912-5925.2002.PubMed CentralView ArticlePubMedGoogle Scholar
- Bumann D, Hueck C, Aebischer T, Meyer TF: Recombinant live Salmonella spp. for human vaccination against heterologous pathogens. FEMS Immunol Med Microbiol. 2000, 27 (4): 357-364.View ArticlePubMedGoogle Scholar
- Knodler LA, Bestor A, Ma C, Hansen-Wester I, Hensel M, Vallance BA, Steele-Mortimer O: Cloning vectors and fluorescent proteins can significantly inhibit Salmonella enterica virulence in both epithelial cells and macrophages: implications for bacterial pathogenesis studies. Infect Immun. 2005, 73 (10): 7027-7031. 10.1128/IAI.73.10.7027-7031.2005.PubMed CentralView ArticlePubMedGoogle Scholar
- Datsenko KA, Wanner BL: One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci USA. 2000, 97 (12): 6640-6645. 10.1073/pnas.120163297.PubMed CentralView ArticlePubMedGoogle Scholar
- Boyd D, Weiss DS, Chen JC, Beckwith J: Towards single-copy gene expression systems making gene cloning physiologically relevant: lambda InCh, a simple Escherichia coli plasmid-chromosome shuttle system. J Bacteriol. 2000, 182 (3): 842-847. 10.1128/JB.182.3.842-847.2000.PubMed CentralView ArticlePubMedGoogle Scholar
- Gumbiner-Russo LM, Lombardo M-J, Ponder RP, Rosenberg SM: The TGV transgenic vectors for single copy gene expression in the E. coli chromosome. Gene. 2001, 273 (1): 97-104. 10.1016/S0378-1119(01)00565-0.View ArticlePubMedGoogle Scholar
- Peters JE, Craig NL: Tn7: smarter than we thought. Nat Rev Mol Cell Biol. 2001, 2 (11): 806-814. 10.1038/35099006.View ArticlePubMedGoogle Scholar
- Milewski S: Glucosamine-6-phosphate synthase – the multi-facets enzyme. Biochim Biophys Acta. 2002, 1597 (2): 173-192.View ArticlePubMedGoogle Scholar
- Craig NL: Transposon Tn7. Mobile DNA. Edited by: Berg DE, Howe MM. 1989, Washington, D. C.: American Society for Microbiology, 211-225.Google Scholar
- Guzman LM, Belin D, Carson MJ, Beckwith J: Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol. 1995, 177 (14): 4121-4130.PubMed CentralPubMedGoogle Scholar
- Hamilton CM, Aldea M, Washburn BK, Babitzke P, Kushner SR: New method for generating deletions and gene replacements in Escherichia coli. J Bacteriol. 1989, 171 (9): 4617-4622.PubMed CentralPubMedGoogle Scholar
- Sharpe PL, Craig NL: Host proteins can stimulate Tn7 transposition: a novel role for the ribosomal protein L29 and the acyl carrier protein. EMBO J. 1998, 17 (19): 5822-5831. 10.1093/emboj/17.19.5822.PubMed CentralView ArticlePubMedGoogle Scholar
- DeBoy RT, Craig NL: Tn7 transposition as a probe of cis interactions between widely separated (190 kilobases apart) DNA sites in the Escherichia coli chromosome. J Bacteriol. 1996, 178 (21): 6184-6191.PubMed CentralPubMedGoogle Scholar
- Choi KH, Gaynor JB, White KG, Lopez C, Bosio CM, Karkhoff-Schweizer RR, Schweizer HP: A Tn7 -based broad-range bacterial cloning and expression system. Nat Methods. 2005, 2 (6): 443-448. 10.1038/nmeth765.View ArticlePubMedGoogle Scholar
- Koch B, Jensen LE, Nybroe O: A panel of Tn7-based vectors for insertion of the gfp marker gene or for delivery of cloned DNA into Gram-negative bacteria at a neutral chromosomal site. J Microbiol Methods. 2001, 45 (3): 187-195. 10.1016/S0167-7012(01)00246-9.View ArticlePubMedGoogle Scholar
- Choi YJ, Bourque D, Morel L, Groleau D, Míguez CB: Multicopy integration and expression of heterologous genes in Methylobacterium extorquens ATCC 55366. Appl Environ Microbiol. 2006, 72 (1): 753-759. 10.1128/AEM.72.1.753-759.2006.PubMed CentralView ArticlePubMedGoogle Scholar
- Ochman H, Selander RK: Standard reference strains of Escherichia coli from natural populations. J Bacteriol. 1984, 157 (2): 690-693.PubMed CentralPubMedGoogle Scholar
- Kuduvalli PN, Mitra R, Craig NL: Site-specific Tn7 transposition into the human genome. Nucleic Acids Res. 2005, 33 (3): 857-863. 10.1093/nar/gki227.PubMed CentralView ArticlePubMedGoogle Scholar
- Walkinshaw MD, Taylor P, Sturrock SS, Atanasiu C, Berge T, Henderson RM, Edwardson JM, Dryden DT: Structure of Ocr from bacteriophage T7, a protein that mimics B-form DNA. Mol Cell. 2002, 9 (1): 187-194. 10.1016/S1097-2765(02)00435-5.View ArticlePubMedGoogle Scholar
- Rasko DA, Ravel J, Økstad OA, Helgason E, Cer RZ, Jiang L, Shores KA, Fouts DE, Tourasse NJ, Angiuoli SV, Kolonay J, Nelson WC, Kolstø AB, Fraser CM, Read TD: The genome sequence of Bacillus cereus ATCC 10987 reveals metabolic adaptations and a large plasmid related to Bacillus anthracis pXO1. Nucleic Acids Res. 2004, 32 (3): 977-988. 10.1093/nar/gkh258.PubMed CentralView ArticlePubMedGoogle Scholar
- Stratford R, McKelvie ND, Hughes NJ, Aldred E, Wiseman C, Curtis J, Bellaby T, Bentley M, Hindle Z, Brennan FR, Chatfield SN, Dougan G, Khan SA: Optimization of Salmonella enterica serovar typhi ΔaroC ΔssaV derivatives as vehicles for delivering heterologous antigens by chromosomal integration and in vivo inducible promoters. Infect Immun. 2005, 73 (1): 362-368. 10.1128/IAI.73.1.362-368.2005.PubMed CentralView ArticlePubMedGoogle Scholar
- Waddell CS, Craig NL: Tn7 transposition: two transposition pathways directed by five Tn7-encoded genes. Genes Dev. 1988, 2 (2): 137-149.View ArticlePubMedGoogle Scholar
- Poteete AR, Fenton AC: λ red-dependent growth and recombination of phage P22. Virology. 1984, 134 (1): 161-167. 10.1016/0042-6822(84)90281-2.View ArticlePubMedGoogle Scholar
- Simon R, Priefer U, Pühler A: A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Bio/Technology. 1983, 1 (9): 784-791. 10.1038/nbt1183-784.View ArticleGoogle Scholar
- Cold Spring Harbor Laboratory Press, Miller JH: A short course in bacterial genetics. 1992, Cold Spring Harbor Laboratory PressGoogle Scholar
- Hartley JL, Temple GF, Brasch MA: DNA cloning using in vitro site-specific recombination. Genome Res. 2000, 10 (11): 1788-1795. 10.1101/gr.143000.PubMed CentralView ArticlePubMedGoogle Scholar
- Sambrook J, Fritsch EF, Maniatis T: Molecular cloning: a laboratory manual. Cold Spring Harbor. 1989, NY: Cold Spring Harbor Press, 2Google 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.