Effective inhibition of HCMV UL49 gene expression and viral replication by oligonucleotide external guide sequences and RNase P
© Zhang et al; licensee BioMed Central Ltd. 2010
Received: 22 March 2010
Accepted: 18 May 2010
Published: 18 May 2010
Human cytomegalovirus (HCMV) is a ubiquitous herpesvirus that typically causes asymptomatic infections in healthy individuals but may lead to serious complications in newborns and immunodeficient individuals. The emergence of drug-resistant strains of HCMV has posed a need for the development of new drugs and treatment strategies. Antisense molecules are promising gene-targeting agents for specific regulation of gene expression. External guide sequences (EGSs) are oligonucleotides that consist of a sequence complementary to a target mRNA and recruit intracellular RNase P for specific degradation of the target RNA. The UL49-deletion BAC of HCMV was significantly defective in growth in human foreskin fibroblasts. Therefore, UL49 gene may serve as a potential target for novel drug development to combat HCMV infection. In this study, DNA-based EGS molecules were synthesized to target the UL49 mRNA of human cytomegalovirus (HCMV).
By cleavage activity assessing in vitro, the EGS aimed to the cleavage site 324 nt downstream from the translational initiation codon of UL49 mRNA (i.e. EGS324) was confirmed be efficient to direct human RNase P to cleave the target mRNA sequence. When EGS324 was exogenously administered into HCMV-infected human foreskin fibroblasts (HFFs), a significant reduction of ~76% in the mRNA and ~80% in the protein expression of UL49 gene, comparing with the cells transfected with control EGSs. Furthermore, a reduction of about 330-fold in HCMV growth were observed in HCMV-infected HFFs treated with the EGS.
These results indicated that UL49 gene was essential for replication of HCMV. Moreover, our study provides evidence that exogenous administration of a DNA-based EGS can be used as a potential therapeutic approach for inhibiting gene expression and replication of a human virus.
Human cytomegalovirus is a ubiquitous herpesvirus that typically causes asymptomatic infections in healthy individuals but may lead to serious complications in newborns and immunodeficient individuals. For example, this virus accounts for one of the most common opportunistic infections in AIDS patients (i.e., CMV retinitis). Moreover, HCMV infection is the leading viral cause of birth defects in newborns and a major cause of morbidity and mortality in bone marrow and solid organ transplant recipients . The emergence of drug-resistant strains of HCMV has posed a need for the development of new drugs and treatment strategies .
RNA-based EGSs have been expressed endogenously as transgenes in both bacteria and mammalian cells , and were effective in inhibiting the gene expression of herpes simplex and influenza virus and in abolishing the replication of influenza virus in human cells [13, 14]. In vitro studies have also shown that DNA-based EGSs, as well as EGS molecules with modified nucleotides, can direct M1 RNA or human RNase P to cleave a mRNA sequence, although their targeting efficiencies are lower than those of unmodified RNA-based EGSs . However, little is known about whether DNA-based EGSs are functionally active in cultured cells. Also, whether DNA-based EGSs can be exogenously administered into human cells to inhibit gene expression and growth of human viruses is yet to be determined.
The UL49-deletion BAC of HCMV was significantly defective in growth in human foreskin fibroblasts . Furthermore, the expression of HCMV UL49 protein has been shown in our mock infected human foreskin fibroblasts (unpublished data, doctoral dissertation). Therefore, UL49 gene may serve as a potential target for novel drug development to combat HCMV infection. In this study, we provide direct evidence that exogenous administration of chemically synthesized DNA-based EGS is highly effective in inhibiting HCMV gene expression and growth in human cell culture and, furthermore, demonstrate the feasibility of using DNA-based EGSs for the studies and treatment of infections caused by human viruses including HCMV.
Materials and methods
Construction of the RNA substrate and EGSs
The DNA sequence that encodes HCMV UL49 gene was constructed by PCR using AD169 genomic DNA as a template and oligonucleotides 49F: 5'-AGTTGGATCCATGGCCAGTCGTCGTCTC-3' and 49R: 5'-ATCTAAGCTTGTAGACATGGGGCAGGCCGT-3' as 5' and 3' primers, respectively. The PCR fragments were digested with BamH I and Hind III, and then inserted into vector pGEM3z (Invitrogen) to form a recombinant plasmid, UL49-pGEM3z. The RNA substrates were synthesized by bacteriophage T7 RNA polymerase using the Hind III-linearized UL49-pGEM3z plasmid as transcriptional templates and labeled with 32P-UTP. The restriction endonucleases, LA Taq polymerase, T4 DNA ligase and T7 RNA polymerase were purchased from Takara (Dalian, China). All oligonucleotides used as PCR primers were purchased from Invitrogen (Guangzhou, China). EGS oligonucleotides used in this report including EGS324(5' CGACACACTTGGGGTTTCCCCACGCAGGTTCGAATCCTGCGCCATTCCCA3'), EGS1539(5'GTCGCGGTTGGGGTTTCCCCACGCAGGTTCGAATCCTGCCTCGCCCCCA3'), C-EGS324(5' CGACACACTTGGGGTTTCCCCACGCAGGAAGGAATCCTGCGCCATTCCCA3'), C-EGS1539(5'GTCGCGGTTGGGGTTTCCCCACGCAGGAAGGAATCCTGCCTCGCCCCCA3'), and TK-EGS(5'TACGTCGGTGCGGTCTCCGCGCGCAGGTTCAAATCCTGCCGCAGACACCA3') were chemically synthesized directly by using a DNA synthesizer.
Purification of human RNase P
RNase P was purified from HeLa cells according to the method as described  with the following modifications. The DEAE-Sepharose Fast-flow column was equilibrated and washed with buffer A containing 10 mM Tris-HCl, pH 7.5, 2.5 mM MgCl2, 5 mM KCl, 2 mM DTT, and 1 mM Pefabloc (proteinase inhibitor). The enzyme was eluted with a linear gradient of 100~500 mM KCl in the same buffer. The active fractions were pooled and concentrated to a final volume of approximately 5 ml. The active fractions were pooled and dialyzed against a buffer containing 10 mM Tris-HCl, pH 7.5, 2.5 mM MgCl2, 5 mM KCl, 2 mM DTT, 1 mM Pefabloc and 50% glycerol. The samples were then stored at -80°C before use. The specific activity of the RNase P was determined as the cleavage of 0.02 pmol yeast tRNAser precursor per min per μl.
In vitro cleavage studies
The EGSs (50 nM) and 32P-labeled UL49 mRNA (50 nM) were incubated with human RNase P (5 units). The cleavage reactions were carried out at 37°C for 30 min in buffer A (50 mM Tris-HCl, pH7.5, 100 mM NH4Cl, 10 mM MgCl2). To disrupt aggregates, which might form during storage, EGS and substrate together were heated to 90°C for 1 min and then cooled to room temperature before the addition of other components. Substrate and cleavage products were separated under denaturing conditions on a 15% acrylamide gel containing 8 M urea. The amount of radioactivity per band was quantified using a Typhoon 9200 phosphorimager. The procedures to measure the equilibrium dissociation constants (K d ) of the EGS-UL49 mRNA complexes were carried out as described . In brief, various concentrations of EGSs were preincubated in buffer B (50 mM Tris, pH 7.5, 100 mM NH4Cl, 20 mM MgCl2, 3% glycerol, 0.1% xylene cyanol, 0.1% bromphenol blue) at 37°C for 10 min before mixing with an equal volume of different concentrations of 32P-labeled substrate RNA preheated under identical conditions. The samples were incubated for 15 min to allow binding, loaded on a 5% polyacrylamide gel, and run at 10 watts. The electrophoresis running buffer contained 100 mM Tris-Hepes, pH 7.5, and 10 mM MgCl2. Each band was quantitated with a Typhoon 9200 phosphorimager. The value of Kd was then extrapolated from a graph plotting percentage of product bound versus EGS concentration. The values were the average of three experiments.
EGS internalization and viral infection
Lipofectamine 2000 reagent (GIBCO) was diluted in 100 μl of Opti-MEM medium with the EGS to give a final concentration of 10 ug/ml lipid-100 nM EGS. The transfection experiments were carried out by using 100 nM EGS. The EGS-lipid mixtures were prepared according to the manufacturer's recommendation (GIBCO) and incubated with cells for 7 hrs. HCMV (strain AD169) was propagated in human foreskin fibroblasts in DMEM supplemented with 10% FBS (GIBCO). Cells (1 × 106) were treated with liposome complexes in the absence and presence of EGS, followed by HCMV infection or mock infection at a MOI of 1 in an inoculum of 1.5 ml DMEM supplemented with 1% fetal calf serum (FCS). Total cellular RNAs and proteins were isolated from cells as described 72 hrs postinfection .
Fluorescence Quantitative RT-PCR
Fluorescence Quantitative RT-PCR (FQ-RT-PCR) was performed using standard protocols on an Applied Biosystem's 7300 Sequence Detection System. Briefly, total RNAs were extracted using Trizol, then 5 ml of a 1/100 dilution of cDNA in water was added into 12.5 ml of the 2 × SYBR green PCR master mix (Takara), with 800 nM of each primer in a total volume of 25 ml. All reactions were run in triplicate using Applied Biosystem's 7300 Sequence Detection System. UL49 gene sense primer: 5'-CGTTCTTGCGTCCTTCATCT-3'; anti-sense primer: 5'-CACAAAGTAGGGCTTGGTCAT-3'. As internal standard, a fragment of human endogenous β actin was amplified simultaneously in each PCR. β actin sense primer: 5'-TCGTCCACCGCAAATGCTTCTAG-3', anti-sense primer: 5'-ACTGCTGTCACCTTCACCGTTCC-3'. Quantitative RT-PCR conditions: 95°C for 10 s, 1 cycle, followed by 95°C for 5 s, 60°C for 20 s, 40 cycles for amplifying UL49 and β actin.
For Western analyses, the denatured, solubilized polypeptides were separated on 12% (vol/vol) SDS polyacrylamide gels cross-linked with N, N"-methylenebisacylamide, and then transferred onto PVDF membranes (Amersham, Freiburg, Germany). The membranes were incubated in the primary antibody solution (including antibody against human β actin, ul49, IE1/IE2, pp28, or gB), then were incubated in a goat anti-mouse HRP-conjugated antibody solution. After washing with PBS containing 0.05% Tween 20, the membranes were subsequently stained with a chemiluminescent substrate with the aid of a Western chemiluminescent substrate kit (Pierce, Rockford) and quantitated with a Typhoon 9200 phosphorimager . The polyclonal antibodies against HCMV UL49 protein were kindly provided by Doctor Feng Zhu (BOAOSEN, Beijing). The monoclonal antibodies, which react with HCMV proteins gB, pp28 and IE1/IE2, respectively, were purchased from SANTA CRUS. The monoclonal antibodies against human β actin was purchased from SIGMA.
Assay of viral growth in EGS-treated cells
To determine the level of the inhibition of viral growth, 5 × 105 human foreskin fibroblasts were first incubated with liposome complexes in the absence and presence of EGSs, and then mock-infected or infected with HCMV AD169 at a MOI of 1. The cells and medium were harvested at 1, 2, 3, 4, 5, 6, and 7 days postinfection. The viral stocks were prepared by adding an equal volume of 10% skim milk, followed by sonication. The titers of the viral stocks were determined by infecting 2 × 105 foreskin fibroblasts in 6-well plates and counting the number of plaques 7~10 days postinfection. The values obtained were the average from triplicate experiments.
Design of EGSs
To achieve optimal cleavage, it is critical to choose the target regions of the UL49 mRNA. According to general principles of designing EGSs , two positions, 324 nt and 1539nt downstream from the translational initiation codon, were chosen as the cleavage sites for EGSs. The flanking sequence of these cleavage sites exhibited several sequence features that need to be present in order to interact with an EGS and RNase P to achieve efficient cleavage. These features include that ➀ the nucleotides 3' and 5' adjacent to the site of cleavage are a guanosine and a pyrimidine, respectively, and ➁ a U is 8 nt downstream from this cleavage site. The interactions of these sequence elements with the EGS facilitate the formation of the mRNA-EGS complex into a ptRNA-like structure while those with RNase P are critical for recognition and cleavage by the enzyme . On the basis of the target sites, two EGSs (EGS324 and EGS1539) were constructed respectively, which resemble a part of the tRNASer structure, containing a T-loop, a stem, and a variable region but not the anticodon region (Fig. 1C and 1E). The anticodon domain has been shown to be dispensable for EGS activity. Binding of EGSs to the target mRNA results in a helix structure of 7 bp that is equivalent to the D stem of a tRNA, which usually has 4 bp. Our choice of a D stem-equivalent structure of 7 bp instead of 4 bp is to increase the targeting sequence specificity of the EGS (Fig. 1C and 1E). The two control (C-EGS324 and C-EGS1539) were derived respectively from EGS324 and EGS1539 by introducing base substitution mutations in three positions of the T-loop (Fig. 1D and 1F). The nucleotides in these three positions are highly conserved among tRNA molecules and are important for the recognition of tRNA molecules by RNase P .
In vitro studies of targeting activities of the EGSs
Inhibition of HCMV UL49 expression in EGS-treated cells
Inhibition of UL49 expression is not expected to affect the expression of other viral genes
Inhibiting rates of the mRNA and protein expression of different viral genes in cells.
76% ± 3%
80% ± 3%
Inhibition of HCMV growth in the EGS-treated cells
Compared to other nucleic acid-based gene interference approaches, the EGS technology with the use of endogenous human RNase P exhibits several unique and attractive features as a gene targeting tool. First, the mechanism of the EGS technology for degradation of a specific mRNA is different from other RNA- or DNA-based gene-targeting approaches. It uses the endogenous RNase P, which is one of the most ubiquitous, abundant, stable and efficient enzymes in all type of cells. This essential enzyme is highly expressed (5 × 104 copies per cell) and is responsible for the processing of all tRNA precursors that account for approximately 2% of total cellular RNA. The action of RNase P with the EGS will result in irreversible cleavage of the target mRNA in a highly efficient catalytic fashion. Second, the sequence specificity of the EGS technology is governed by two different types of interactions between the EGS and the target mRNA: the base-pairing interactions in which the sequence of 12 nucleotides in the EGS hybridizes with the target mRNA, and the interactions between the target mRNA and the other part of the EGS sequence (equivalent to the T-stem and T-loop, and variable regions of a tRNA) which are required for folding of the RNase P-recognizable tertiary structure. Thus, the EGS-based technology is highly specific and does not generate nonspecific "irrelevant cleavage" that is observed in RNase H-mediated cleavage induced by conventional antisense phosphorothioate molecules .
Our study used exogenous administration of DNA-based EGSs for antiviral applications and demonstrated that RNase P-mediated targeting directed by DNA-based EGS was highly effective in inhibiting HCMV gene expression and viral growth in cultured cells. We showed that these EGS molecules directed human RNase P to cleave the UL49 mRNA sequence efficiently in vitro. Moreover, we also showed that these EGSs were readily delivered in cultured cells. A reduction of about 76~80% in the UL49 expression was achieved with a functional EGS, EGS324, whereas a reduction of less than 10% was observed in cells that were treated with C-EGS324 or TK-EGS. C-EGS324 bound to UL49 mRNA substrate in vitro as well as EGS324 but contained nucleotide mutations that disrupted RNase P recognition. These results suggested that the overall observed inhibition with EGS324 was primarily due to targeted cleavage by RNase P as opposed to the antisense effect or other nonspecific effects of the EGSs. Moreover, the antiviral effect of the EGS treatment (inhibition of viral growth) appeared to be due to the reduction of the UL49 expression. Only the expression of UL49 mRNA and protein was found to be reduced in cells treated with EGS324. We found no significant change in the expression of IE2, UL44, pp28, and gB (Table 1). The observed level of inhibition of UL49 expression was consistent with the extent of the reduction of viral growth.
HCMV is a member of the human herpesvirus family, which includes seven other different viruses such as HSV and Epstein-Barr virus. UL49 is well conserved and necessary for the growth of this virus . Therefore, UL49 is considered as an ideal target for antiherpes therapy. To develop EGSs as a conventional drug that can be used as an exogenous agent for intracellular delivery in antiviral therapy, EGSs containing modified oligonucleotides are required to increase their stability in vivo[26, 27]. It has been recently reported that chemically synthesized RNA-based EGSs with modified nucleotides can be administered exogenously into human cells and inhibit cellular gene expression . Our study showed that exogenous administration of DNA-based EGSs is highly efficient in inhibiting gene expression and viral growth. Further understanding of how the functional groups in the nucleotides of an EGS interact with human RNase P and the mRNA substrate will lead to the construction of highly active and stable EGSs with either different bases or modifications at these nucleotide positions. Moreover, engineering different designs of EGSs  for increasing their targeting activity, as well as developing new means for improving their delivery, are needed to increase the efficacy of the EGSs in vivo. These studies will facilitate the development of the EGS-based technology for gene-targeting applications in both basic research and clinical therapy of HCMV infections.
We thank Doctor Feng Zhu for anti-UL49 antibody and Professor Fengyong Liu (University of California, Berkeley) for technical assistance. This work was supported by grants from the National Natural Science Foundation of China (Grant No. 90608024 & 303707762); the Key Science & Technology Project of Guangzhou (Grant No. 2006J1-C0111); the Planned Science and Technology Project of Guangdong Province (Grant No. 2006B35502002); the Project supported by the Natural Science Foundation of Guangdong Province (Grant No. 9451022401003429); and the China Postdoctoral Science Foundation Funded Project (Grant No. 20080430845). This work was also supported by 211 grant of MOE.
- Mocarski ES, Courcelle CT: Cytomegaloviruses and their replication. In Fields virology. Edited by: Knipe DM, Howley PM. Philadelphia PA: Lippincott-William & Wilkins; 2001:2629-2673.
- Baldanti F, Lurain N, Gerna G: Clinical and biologic aspects of human cytomegalovirus resistance to antiviral drugs. Human Immunology 2004, 65: 403-409. 10.1016/j.humimm.2004.02.007PubMedView Article
- Stein CA, Cheng Y: Antisense oligonucleotides as therapeutic agents--is the bullet really magical? Science 1993, 261: 1004-1012. 10.1126/science.8351515PubMedView Article
- Margraf S, Bittoova M, Vogel JU, Kotchekov R, Doerr HW, Cinatl JJ: Antisense oligonucleotide ISIS 2922 targets IE-expression and prevents HCMV-IE-induced suppression of TSP-1 and TSP-2 expression. Nucleotides Nucleic Acids 2001, 20: 1425-1428. 10.1081/NCN-100002569View Article
- Zhang G, Raghavan B, Kotur M, Cheatham J, Sedmak D, Cook C, Waldman J, Trgovcich J: Antisense transcription in the human cytomegalovirus transcriptome. J Virol 2007, 81: 11267-11281. 10.1128/JVI.00007-07PubMedPubMed CentralView Article
- Forster AC, Altman S: External guide sequences for an RNA enzyme. Science 1990, 249: 783-786. 10.1126/science.1697102PubMedView Article
- Werner M, Rosa E, Nordstrom JL, Goldberg AR, George ST: Short oligonucleotides as external guide sequences for site-specific cleavage of RNA molecules with human RNase P. RNA 1990, 4: 847-855. 10.1017/S1355838298980323View Article
- Zhang WJ, Li HJ, Li YQ, He HQ, Tang DS, Zhang X, Zhou TH: Construction of an effective M1GS ribozyme targeting HCMV UL97 mRNA segment in vitro . Yi Chuan Xue Bao 2005, 32: 1205-1212.PubMed
- Su YZ, Li HJ, Li YQ, Chen HJ, Tang DS, Zhang X, Jiang H, Zhou TH: In vitro construction of effective M1GS ribozymes targeting HCMV UL54 RNA segments. Acta Biochim Biophys Sin 2005, 37: 210-214.PubMedView Article
- Frank DN, Pace NR: Ribonuclease P: unity and diversity in a tRNA processing ribozyme. Annu Rev Biochem 1998, 67: 153-180. 10.1146/annurev.biochem.67.1.153PubMedView Article
- Yuan Y, Hwang ES, Altman S: Targeted cleavage of mRNA by human RNase P. Proc Natl Acad Sci USA 1992, 89: 8006-8010. 10.1073/pnas.89.17.8006PubMedPubMed CentralView Article
- Guerrier-Takada C, Li Y, Altman S: Artificial regulation of gene expression in Escherichia coli by RNase P. Proc Natl Acad Sci USA 1995, 92: 11115-11119. 10.1073/pnas.92.24.11115PubMedPubMed CentralView Article
- Kawa D, Wang J, Yuan Y, Liu F: Inhibition of viral gene expression by human ribonuclease P. RNA 1998, 4: 1397-1406. 10.1017/S1355838298980918PubMedPubMed CentralView Article
- Plehn-Dujowich D, Altman S: Effective inhibition of influenza virus production in cultured cells by external guide sequences and ribonuclease P. Proc Natl Acad Sci USA 1998, 95: 7327-7332. 10.1073/pnas.95.13.7327PubMedPubMed CentralView Article
- Ma M, Benimetskaya L, Lebedeva I, Dignam J, Takle G, Stein CA: Intracellular mRNA cleavage induced through activation of RNase P by nuclease-resistant external guide sequences. Nat Biotechnol 2000, 18: 58-61. 10.1038/81113PubMedView Article
- Dunn W, Chou C, Li H, Hai R, Patterson D, Stolc V, Zhu H: Functional profiling of a human cytomegalovirus genome. Proc Natl Acad Sci USA 2003, 100: 14223-14228. 10.1073/pnas.2334032100PubMedPubMed CentralView Article
- Zhu J, Trang P, Kim K, Zhou T, Deng H, Liu F: Effective inhibition of Rta expression and lytic replication of Kaposi's sarcoma-associated herpesvirus by human RNase P. Proc Natl Acad Sci USA 2004, 101: 9073-9078. 10.1073/pnas.0403164101PubMedPubMed CentralView Article
- Yang Y, Li H, Zhou T, Kim K, Liu F: Engineered external guide sequences are highly effective in inducing RNase P for inhibition of gene expression and replication of human cytomegalovirus. Nucleic Acids Res 2006, 34: 575-583. 10.1093/nar/gkj431PubMedPubMed CentralView Article
- Trang P, Lee M, Nepomuceno E, Kim J, Zhu H, Liu F: Effective inhibition of human cytomegalovirus gene expression and replication by a ribozyme derived from the catalytic RNA subunit of RNase P from Escherichia coli. Proc Natl Acad Sci USA 2000, 97: 5812-5817. 10.1073/pnas.100101797PubMedPubMed CentralView Article
- Zeng Z, Li H, Li Y, Cui Y, Zhou Q, Zou Y, Yang G, Zhou T: Effective inhibition of human cytomegalovirus gene expression by DNA-based external guide sequences. Acta Biochim Biophys Sin 2009, 41: 389-398. 10.1093/abbs/gmp024PubMedView Article
- Liu F, Altman S: Requirements for cleavage by a modified RNase P of a small model substrate. Nucleic Acids Res 1996, 24: 2690-2096. 10.1093/nar/24.14.2690PubMedPubMed CentralView Article
- Yuan Y, Altman S: Selection of guide sequences that direct efficient cleavage of mRNA by human ribonuclease P. Science 1994, 263: 1269-1273. 10.1126/science.8122108PubMedView Article
- Altman S, Kirsebom LA: Ribonuclease P. In The RNA world. Edited by: Gesteland RF. New York: Cold Spring Harbor Laboratory Press; 1999:351-380.
- Gopalan V, Vioque A, Altman S: RNase P: variations and uses. J Biol Chem 2002, 277: 6759-6762. 10.1074/jbc.R100067200PubMedView Article
- Trang P, Kim K, Liu F: Developing RNase P ribozymes for gene-targeting and antiviral therapy. Cell Microbiol 2004, 6: 499-508. 10.1111/j.1462-5822.2004.00398.xPubMedView Article
- Verma S, Eckstein F: Modified oligonucleotides: synthesis and strategy for users. Annu Rev Biochem 1998, 67: 99-134. 10.1146/annurev.biochem.67.1.99PubMedView Article
- Kim K, Liu F: Inhibition of gene expression in human cells using RNase P-derived ribozymes and external guide sequences. Biochim Biophys Acta 2007, 1769: 603-612.PubMedPubMed CentralView Article
- Li H, Trang P, Kim K, Zhou T, Liu F: Effective inhibition of human cytomegalovirus gene expression and growth by intracellular expression of external guide sequence RNA. RNA 2006, 12: 63-72. 10.1261/rna.2184706PubMedPubMed CentralView Article
- Yang YH, Li H, Zhou T, Kim K, Liu F: Engineered external guide sequences are highly effective in inducing RNase P for inhibition of gene expression and replication of human cytomegalovirus. Nucleic Acids Res 2006, 34: 575-583. 10.1093/nar/gkj431PubMedPubMed CentralView Article
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.