Novel restriction factor RNA-associated early-stage anti-viral factor (REAF) inhibits human and simian immunodeficiency viruses
© Marno et al.; licensee BioMed Central Ltd. 2014
Received: 27 March 2013
Accepted: 28 November 2013
Published: 10 January 2014
The discovery of novel anti-viral restriction factors illuminates unknown aspects of innate sensing and immunity. We identified RNA-associated Early-stage Anti-viral Factor (REAF) using a whole genome siRNA screen for restriction factors to human immunodeficiency virus (HIV) that act in the early phase of viral replication.
We observed more than 50 fold rescue of HIV-1 infection, using a focus forming unit (FFU) assay, following knockdown of REAF by specific siRNA. Quantitative PCR was used to show that REAF knockdown results in an increase of early and late reverse transcripts which impacts the level of integration. REAF thus appears to act at an early stage of the viral life cycle during reverse transcription. Conversely when REAF is over-expressed in target cells less infected cells are detectable and fewer reverse transcripts are produced. Human REAF can also inhibit HIV-2 and simian immunodeficiency virus (SIV) infection. REAF associates with viral nucleic acids and may act to prevent reverse transcription.
This report firmly places REAF alongside APOBECs and SAMHD1 as a potent inhibitor of HIV replication acting early in the replication cycle, just after cell entry. We propose that REAF is part of an anti-viral surveillance system destroying incoming retroviruses. This novel mechanism could apply to invasion of cells by any intracellular pathogen.
The extent of the cellular armoury against viral infection is becoming increasingly appreciated. In particular human immunodeficiency virus (HIV) must overcome many cellular obstacles on its replication pathway to the nucleus. Once HIV enters the cytoplasm its genomic RNA is reverse transcribed by the virally encoded reverse transcriptase (RT), resulting in hybrid RNA:DNA intermediates. The RNase H activity of RT degrades the RNA from these hybrids resulting in single stranded (ss)DNA from which the second DNA strand is synthesised. Immediately upon initiation the process of reverse transcription is susceptible to members of the apolipoprotein B mRNA-editing, enzyme-catalytic, polypeptide-like (APOBEC) family by inducing deoxycytidine to deoxyuridine mutations in the nascent DNA . In primary macrophages, dendritic cells and resting CD4+ T cells SAMHD1 degrades dNTPs which are required for efficient reverse transcription [2–4]. Two other factors p21 and PAF1 act at an early stage however their mechanism of action is not yet understood [5, 6]. Once reverse transcription is complete the pre-integration complex (PIC) containing the double stranded (ds)DNA is then formed and integrated into the genome of the host cell. This process is inhibited by the TRIM28 (KAP1)/SETDB1 complex [6, 7]. Once the provirus is integrated the late phase of the replication cycle begins with the production of viral proteins . Finally, a plasma membrane located restriction factor tetherin/BST-2/CD317, prevents viruses from leaving the cell at the late budding stage of the life cycle .
Using a whole genome siRNA screen , we identified RPRD2 (here called RNA-associated Early-stage Anti-viral Factor; REAF) as a potential restriction factor. REAF is a protein of previously unknown function. Our data suggest that it acts at an early post-entry stage of HIV-1 replication, during or following the initiation of reverse transcription.
Characterising the action of a novel restriction factor reveals molecular details of HIV replication and innate host immunity and more importantly may lead to novel therapeutics for HIV and possibly other viral infections.
REAF inhibits HIV and SIV infection and occurs at an early post-entry stage of replication
To quantify the ability of REAF to rescue infection, HeLa-CD4 cells were challenged with dual tropic wild type HIV-189.6 following siRNA knockdown. HIV-1 infection was strongly rescued following REAF silencing (>50 fold); from 6.9 × 102 to 3.5 × 104 focus forming units (FFU)/ml (Fold change cf. CB control; Figure 1B, Raw FFU data; Additional file 2). To determine if this anti-viral effect was restricted to HIV-1 we also tested for activity against HIV-2CBL-23 and various strains of simian immunodeficiency virus (SIV); African Green Monkey (TYO-1, SIVAGM), Macaque (32H, SIVMAC) and Sooty Mangabey (B670, SIVSM). We observed significant rescue of all viruses between 15–126 fold (Fold change cf. CB control; Figure 1B, Raw FFU data; Additional file 2).
There are multiple stages in the virus life cycle vulnerable to attack by cellular restriction factors. To gain insight into the action of REAF anti-viral activity we determined the point in the viral life cycle which is impeded. Cells were challenged with HIV-189.6 after REAF was down modulated. Using real time quantitative PCR (qPCR) the initiation and completion of negative strand strong-stop DNA (−sssDNA; early) and full length (late) RT products were measured at various time points after infection (Figure 1C). Knockdown of REAF enhanced both early and late transcript levels within 1–2 hours of viral challenge, indicating that the reverse transcription process is targeted. An Alu-PCR assay subsequently found an increase in the amount of detectable integrated provirus (Figure 1D). This is probably a result of more available full length viral cDNA transcripts and correlates with a block by REAF during the reverse transcription process.
Over-expression of REAF confers restriction which is viral route of entry dependent
We previously reported that the Lv2 post entry restriction to HIV can be dependent on the envelope mediated route of entry [10–13]. To determine whether the anti-viral activity of REAF is route of entry dependent, HIV-189.6Δenv/VSV-G pseudotyped virus was used to challenge HeLa-CD4 cells following REAF knockdown by siRNA. The 89.6Δenv construct contains the full length 89.6 genome, with a stop codon in the Env ORF, so these pseudotyped virions still produce p24 following integration and thus infection was measured by p24 immunostaining as above. In contrast to HIV-189.6, no rescue of infected foci was observed after challenge with HIV-189.6Δenv/VSV-G pseudotypes (Figure 2D). VSV-G triggers entry by clathrin-mediated endocytosis through a ubiquitously expressed glycolipid. These results suggest that REAF-mediated restriction of HIV-189.6 is circumvented by a VSV-G mediated route of entry.
REAF interacts with viral nucleic acids
To investigate this further, HEK 293 T cells were transfected with vectors expressing either REAF-EGFP or pEGFP-C3 along with the HIV-189.6 molecular clone DNA. All samples were lysed and incubated with either rabbit IgG (control) or α-REAF antibody conjugated protein A/G beads before immunoprecipitation (IP). Specific isolation of REAF protein by IP was confirmed by Western (Figure 3B). Following extensive washing, total RNA was isolated from the bead-bound complexes and reverse transcription (RT)-qPCR, using viral specific primers (RU5 and late), was performed. All results were normalised to input and the negligible level of qPCR products detected in IgG IP samples. Following this analysis, in cells only expressing endogenous REAF (transfected with pEGFP-C3), a 3.8 and 5.5 fold enrichment of viral RU5 and late transcripts respectively was measured compared with samples that were not transfected with virus (first half of Figure 3C). A further enrichment in both RU5 (16 fold) and late (32 fold) viral cDNA was seen following over-expression of REAF-EGFP (second half of Figure 3C), indicating a specific association of REAF protein and viral nucleic acids. To confirm the RT-qPCR results, a standard PCR was performed on the isolated cDNA and terminated at a non-saturating number of amplification cycles. These PCR products were run on a 4% agarose gel and visually confirm the reported qPCR results (Additional file 5).
The low amounts of recovered viral nucleic acids in this assay require initial transfection of viral DNA as previously published . It must be noted therefore, that the viral RNA is thus not being delivered to the cell as in normal infection and reverse transcription is not taking place. Nevertheless our results suggest an association of REAF with viral nucleic acids. Since the restriction acts in the early phase of replication we hypothesise that REAF interacts with RNA:DNA intermediates or dsDNA products of HIV reverse transcription, either directly or indirectly targeting them for degradation.
REAF is downregulated in response to HIV-1 infection and restriction requires reverse transcription
To further characterise this virus-REAF interplay we investigated whether reverse transcription itself plays a role. A similar time course of infection was performed on HeLa-CD4 cells that had been pre-treated with AZT (Figure 4C). In contrast to untreated cells, REAF was not degraded in the presence of AZT, suggesting that initiation of reverse transcription is required to trigger degradation of REAF. We hypothesise that REAF interacts either with the products of reverse transcription (RNA:DNA intermediates or dsDNA) directly or indirectly to restrict replication and is degraded in the proteasome.
Our results demonstrate that REAF is a new potent anti-HIV/SIV restriction factor. The timing of both the restriction to viral replication and the virus’ potential counterattack supports our model that REAF acts upon a viral nucleic acid target at an early post-entry stage of infection, before the viral genome reaches the nucleus.
The phenotype of a restriction factor is most readily observed for REAF following its knockdown in HeLa-CD4 cells by siRNA. These experiments consistently show >50 fold rescue of viral titre for HIV-1 and even greater for HIV-2. Interestingly, REAF is also able to act against other non-human retroviruses, suggesting a common viral target. This restriction is confirmed by over-expression of REAF-EGFP, resulting in a decrease in the amount of both FFU/ml and HIV-1 DNA amplified by qPCR compared with cells transfected with empty vector (pEGFP-C3) and challenged with virus. REAF is similar to already identified restriction factors TRIM5α, APOBECs and SAMHD1 because it acts at an early post-entry stage of infection. An increase in the earliest detectable form of RT products (−sssDNA) is seen within the first 1–2 hours following viral challenge in cells down modulated for REAF, and this difference grows steadily with time. The accumulation of full length HIV-1 DNA transcripts results in higher levels of integration, which will impact the ability of the virus to spread more quickly through multiple rounds of infection. Of course we cannot rule out that REAF may additionally block a step after reverse transcription. The apparent ability of REAF to bind DNA, along with our results showing higher rescue of infected cells relative to RT products in REAF knockdown cells, would be consistent with this hypothesis.
The viral target itself probably involves nucleic acid recognition: indeed we demonstrate that REAF associates with viral nucleic acids. However we cannot be sure whether the interaction is direct or indirect. The interaction with nucleic acids suggests that the surveillance mechanism may have implications for the invasion of cells by any RNA virus.
A characteristic of restriction factors of HIV-1 is that viruses evolve to mitigate their effects. Here we observed that REAF is destroyed after viral infection. As early as 1 hour after viral challenge, REAF protein levels are depleted in HeLa-CD4 cells, yet REAF mRNA levels remain constant throughout infection. Depletion of REAF is prevented in the presence of proteasomal inhibitor suggesting that REAF is specifically targeted for degradation through cellular proteasomal pathways. The early time frame for this virally induced degradation coincides with the point in the viral life cycle restricted by REAF. It is possible that HIV-1 targets cellular REAF precisely to prevent inhibition of its critical initiation of reverse transcription. The hypothesis that REAF targets viral replication after the initiation of reverse transcription is supported by the abrogation of REAF protein down modulation in the presence of the AZT. However we cannot exclude that REAF is being destroyed by a cellular programme activated after it recognises its target as part of the restriction process.
In recent years, several cellular factors with the ability to suppress HIV-1 replication early in the viral life cycle have been described. It would therefore not be surprising if the virus has evolved to counteract this onslaught. Studies so far have shown that APOBEC and tetherin, as well as the recently discovered SAMHD1, are antagonised by viral proteins HIV-1 Vif, Vpu and HIV-2 Vpx respectively [9, 15, 17]. The potential of a viral antagonist for REAF has not yet been determined. The destruction of REAF is similar to that of SAMHD1 which is targeted for cellular proteasomal degradation by HIV-2 Vpx .
The mechanism of REAF anti-viral activity shares similarities with known restriction factors APOBEC, tetherin and SAMHD1 and is an important new addition to the cellular armoury against viral infection. It is vital to understand the interaction of HIV and innate cellular restriction factors as they represent a relatively untapped source of therapeutic potential. Small molecule pharmaceuticals capable of regulating a pathway of interest to enhance the interaction between a viral protein and its cellular antagonist may be a valuable approach.
Plasmids and virus production
The REAF-EGFP expression plasmid was generated by PCR amplifying the open reading frame from HeLa-CD4 cDNA. The vector used was pEGFP-C3 (Clontech). The infectious molecular clone for HIV-189.6 was obtained from the Centre for AIDS Research (NIBSC, UK). Plasmid construct HIV-189.6Δenv was generated from the HIV-189.6 molecular clone using overlap extension PCR. Clones were confirmed by plasmid sequencing (Source BioScience). Primers for all constructs are available upon request. Virus stocks were prepared from infectious full-length or pseudotyped HIV clones by polyethylenimine (Polysciences) or Lipofectamine 2000 (Invitrogen) transfection of HEK 293 T cells. HIV-2CBL-23 and SIV stocks were grown in C8166 cells.
Preparation of protein lysates
Prior to analysis on SDS-PAGE gels, cell pellets were lysed in ice cold lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1% NP-40, 0.25% sodium deoxycholate) containing protease inhibitor cocktail (Roche Applied Sciences) and 5 mM of sodium fluoride, β-glycerophosphate and sodium orthovanadate.
Western blot and immunoprecipitations
SDS-PAGE separated proteins, immobilised on nitrocellulose membrane (Novex) were detected with the primary rabbit polyclonal antibody against REAF (RbpAb-RPRD2), rabbit pAb-GAPDH, rabbit pAb-PABP or rat pAb-tubulin (Abcam) followed by corresponding horseradish peroxidise-conjugated goat α-rabbit/rat antibody (Abcam). Proteins were visualised using a chemiluminescence kit (ECL/ECL Prime, GE Healthcare).
Oligo (dT) immunoprecipitation
Prior to oligo (dT) addition, control samples were treated with DNase (10U) or increasing amounts of an RNaseA/H cocktail (250-2500U RNaseA, 20U RNaseH) and incubated at 37°C for 15 min. DNase (DNA-free™ Kit; Invitrogen) was physically removed from the appropriate sample according to manufacturer’s instructions. Ribonucleoprotein properties of REAF were then investigated by mixing lysates prepared as above with oligo (dT) magnetic beads (Invitrogen). After 15 min at 4°C the oligo (dT)-protein complexes were washed extensively with lysis buffer and TE. The complex was eluted into 1× PAGE loading buffer and analysed by Western blot.
Viral RNA IP and RT-qPCR analysis
Immunoprecipitation of viral RNA was performed as previously reported with slight modifications . Briefly, HEK 293 T cells were transfected with either pEGFP-C3 or REAF-EGFP and HIV-189.6 molecular clone DNA using Lipofectamine 2000 (Invitrogen). Cells were washed 2× with ice cold PBS then 2× with ice cold swelling buffer (25 mM HEPES, 1.5 mM MgCl2, 85 mM KCl, pH 8.0) and lysed in ice cold lysis buffer (25 mM HEPES, 1.5 mM MgCl2, 85 mM KCl, 0.02% NP-40, 1% Triton X-100, pH 8.0) containing RNaseOUT (500U/ml, Invitrogen), protease inhibitor cocktail (Roche Applied Sciences) and 10 mM of sodium fluoride, sodium pyrophosphate, β-glycerophosphate, and sodium orthovanadate. After centrifugation, the lysate was incubated with 2U of DNase (Ambion) for 5 min at 37°C and pre-cleared with Protein A/G-coupled rabbit IgG beads for 30 min at 4°C. One tenth of the volume was kept for total cellular RNA input, one tenth for Western blotting input and the rest mixed with Protein A/G-coupled REAF Ab beads and incubated for 1 hr at 4°C. The bead bound complexes were washed 2× with lysis buffer, 2× with wash buffer (25 mM HEPES, 1.5 mM MgCl2, 85 mM KCl, 0.25% Triton X-100, pH 8.0) containing 0.1U/ml RNaseOUT, 2× in wash buffer supplemented with 0.4 M NaCl followed by 2 washes with wash buffer supplemented with 0.1U/ml RNaseOUT and 0.1% v/v SDS. An aliquot of the complex was kept for Western blotting while the rest was incubated with DNase for 10 min followed by two phenol:chloroform:isoamyl alcohol (50:49:1) and one chloroform extraction. After precipitation with ethanol and glycogen carrier (Ambion), the RNA pellet was resuspended in DEPC-treated water and reverse transcribed using Superscript III First Strand Synthesis System (Invitrogen) according to manufacturer’s instructions. The resulting cDNA was used in qPCR analysis with primers for RU5 and late HIV-1 DNA as previously described . Primers are available upon request. This qPCR was repeated in a standard PCR machine for a non-saturating number of amplification cycles. These PCR products were resolved on a 4% agarose gel for visual confirmation.
siRNA transfection and infection with replication competent virus
HeLa-CD4 cells were seeded at 2.5 × 104 cells/well in 24-well plates. siRNA transfection (30nM) was performed using HiPerfect (QIAGEN) according to the manufacturer’s instructions. 72 hr after siRNA transfection, cells were challenged with virus (MOI 0.2) for up to 5 hr. Infection was assessed up to 48 hr by intracellular p24 staining or qPCR analysis. When required, reverse transcriptase inhibitor AZT (100 μM) and proteasomal inhibitor MG132 (10 μM) were added 1 hr before infection.
In situ immunostaining for p24 antigen
Infected cells were fixed with cold (−20°C) methanol:acetone (1:1), washed with PBS then immunostained for p24 using mouse anti-HIV-1 p24 monoclonal antibodies EVA365 and 366 (NIBSC, UK) (1:50) or anti-SIV p27 monoclonal antibodies (1:500) (to detect SIV infected cells), as previously described . Infected cells were blue (regarded as foci of infection (FFU/ml)) and quantitated by light microscopy.
First round Alu-PCR
DNA was extracted at various time points after infection with the QIAamp DNA Blood Mini Kit (QIAGEN). Integrated HIV-1 DNA was measured by nested PCR, as previously described .
qPCR for HIV-1 DNA
The isolated DNA was subjected to qPCR to determine the number of early (negative strand strong stop DNA, -sssDNA), RU5 and late (gag) transcripts, normalised for cell number by genomic GAPDH as previously described .
cDNA synthesis and mRNA analysis
Total HeLa-CD4 RNA was extracted using an RNeasy Plant Mini Kit (QIAGEN) and cDNA was synthesised with Superscript III First Strand Synthesis System (Invitrogen), according to manufacturer’s instructions. The cDNA produced was subjected to qPCR as previously described . Primer sequences used to determine REAF mRNA levels are available upon request.
The monoclonal antibodies to p24 (EVA365 and 366) were provided by the EU Programme EVA Centre for AIDS Reagents, NIBSC, UK (AVIP Contract Number LSHP-CT-2004-503487).
- Bishop KN, Holmes RK, Sheehy AM, Malim MH: APOBEC-mediated editing of viral RNA. Science. 2004, 305: 645-10.1126/science.1100658.View ArticlePubMedGoogle Scholar
- Descours B, Cribier A, Chable-Bessia C, Ayinde D, Rice G, Crow Y, Yatim A, Schwartz O, Laguette N, Benkirane M: SAMHD1 restricts HIV-1 reverse transcription in quiescent CD4(+) T-cells. Retrovirology. 2012, 9: 87-10.1186/1742-4690-9-87.PubMed CentralView ArticlePubMedGoogle Scholar
- Lahouassa H, Daddacha W, Hofmann H, Ayinde D, Logue EC, Dragin L, Bloch N, Maudet C, Bertrand M, Gramberg T, et al: SAMHD1 restricts the replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside triphosphates. Nat Immunol. 2012, 13: 223-228. 10.1038/ni.2236.PubMed CentralView ArticlePubMedGoogle Scholar
- St Gelais C, Wu L: SAMHD1: a new insight into HIV-1 restriction in myeloid cells. Retrovirology. 2011, 8: 55-10.1186/1742-4690-8-55.PubMed CentralView ArticlePubMedGoogle Scholar
- Bergamaschi A, David A, Le Rouzic E, Nisole S, Barre-Sinoussi F, Pancino G: The CDK inhibitor p21Cip1/WAF1 is induced by FcgammaR activation and restricts the replication of human immunodeficiency virus type 1 and related primate lentiviruses in human macrophages. J Virol. 2009, 83: 12253-12265. 10.1128/JVI.01395-09.PubMed CentralView ArticlePubMedGoogle Scholar
- Liu L, Oliveira NM, Cheney KM, Pade C, Dreja H, Bergin AM, Borgdorff V, Beach DH, Bishop CL, Dittmar MT, McKnight A: A whole genome screen for HIV restriction factors. Retrovirology. 2011, 8: 94-10.1186/1742-4690-8-94.PubMed CentralView ArticlePubMedGoogle Scholar
- Allouch A, Di Primio C, Alpi E, Lusic M, Arosio D, Giacca M, Cereseto A: The TRIM family protein KAP1 inhibits HIV-1 integration. Cell Host Microbe. 2011, 9: 484-495. 10.1016/j.chom.2011.05.004.View ArticlePubMedGoogle Scholar
- Goff SP: Host factors exploited by retroviruses. Nat Rev Microbiol. 2007, 5: 253-263. 10.1038/nrmicro1541.View ArticlePubMedGoogle Scholar
- Neil SJ, Zang T, Bieniasz PD: Tetherin inhibits retrovirus release and is antagonized by HIV-1 Vpu. Nature. 2008, 451: 425-430. 10.1038/nature06553.View ArticlePubMedGoogle Scholar
- Harrison IP, McKnight A: Cellular entry via an actin and clathrin-dependent route is required for Lv2 restriction of HIV-2. Virology. 2011, 415: 47-55. 10.1016/j.virol.2011.04.001.View ArticlePubMedGoogle Scholar
- Marchant D, Neil SJ, Aubin K, Schmitz C, McKnight A: An envelope-determined, pH-independent endocytic route of viral entry determines the susceptibility of human immunodeficiency virus type 1 (HIV-1) and HIV-2 to Lv2 restriction. J Virol. 2005, 79: 9410-9418. 10.1128/JVI.79.15.9410-9418.2005.PubMed CentralView ArticlePubMedGoogle Scholar
- Schmitz C, Marchant D, Neil SJ, Aubin K, Reuter S, Dittmar MT, McKnight A: Lv2, a novel postentry restriction, is mediated by both capsid and envelope. J Virol. 2004, 78: 2006-2016. 10.1128/JVI.78.4.2006-2016.2004.PubMed CentralView ArticlePubMedGoogle Scholar
- Oliveira NM, Trikha R, McKnight A: A novel envelope mediated post entry restriction of murine leukaemia virus in human cells is Ref1/TRIM5alpha independent. Retrovirology. 2010, 7: 81-10.1186/1742-4690-7-81.PubMed CentralView ArticlePubMedGoogle Scholar
- Burdick R, Smith JL, Chaipan C, Friew Y, Chen J, Venkatachari NJ, Delviks-Frankenberry KA, Hu WS, Pathak VK: P body-associated protein Mov10 inhibits HIV-1 replication at multiple stages. J Virol. 2010, 84: 10241-10253. 10.1128/JVI.00585-10.PubMed CentralView ArticlePubMedGoogle Scholar
- Laguette N, Sobhian B, Casartelli N, Ringeard M, Chable-Bessia C, Segeral E, Yatim A, Emiliani S, Schwartz O, Benkirane M: SAMHD1 is the dendritic- and myeloid-cell-specific HIV-1 restriction factor counteracted by Vpx. Nature. 2011, 474: 654-657. 10.1038/nature10117.PubMed CentralView ArticlePubMedGoogle Scholar
- Mehle A, Strack B, Ancuta P, Zhang C, McPike M, Gabuzda D: Vif overcomes the innate antiviral activity of APOBEC3G by promoting its degradation in the ubiquitin-proteasome pathway. J Biol Chem. 2004, 279: 7792-7798.View ArticlePubMedGoogle Scholar
- Sheehy AM, Gaddis NC, Choi JD, Malim MH: Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature. 2002, 418: 646-650. 10.1038/nature00939.View ArticlePubMedGoogle Scholar
- Salahuddin SZ, Markham PD, Wong-Staal F, Franchini G, Kalyanaraman VS, Gallo RC: Restricted expression of human T-cell leukemia–lymphoma virus (HTLV) in transformed human umbilical cord blood lymphocytes. Virology. 1983, 129: 51-64. 10.1016/0042-6822(83)90395-1.View ArticlePubMedGoogle Scholar
- Cheney KM, McKnight A: Interferon-alpha mediates restriction of human immunodeficiency virus type-1 replication in primary human macrophages at an early stage of replication. PLoS One. 2010, 5: e13521-10.1371/journal.pone.0013521.PubMed CentralView ArticlePubMedGoogle Scholar
- McKnight A, Clapham PR, Weiss RA: HIV-2 and SIV infection of nonprimate cell lines expressing human CD4: restrictions to replication at distinct stages. Virology. 1994, 201: 8-18. 10.1006/viro.1994.1260.View ArticlePubMedGoogle 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.