The Drosophila methyl-DNA binding protein MBD2/3 interacts with the NuRD complex via p55 and MI-2
© Marhold et al; licensee BioMed Central Ltd. 2004
Received: 02 July 2004
Accepted: 29 October 2004
Published: 29 October 2004
Methyl-DNA binding proteins help to translate epigenetic information encoded by DNA methylation into covalent histone modifications. MBD2/3 is the only candidate gene in the Drosophila genome with extended homologies to mammalian MBD2 and MBD3 proteins, which represent a co-repressor and an integral component of the N ucleosome R emodelling and D eacetylase (NuRD) complex, respectively. An association of Drosophila MBD2/3 with the Drosophila NuRD complex has been suggested previously. We have now analyzed the molecular interactions between MBD2/3 and the NuRD complex in greater detail.
The two MBD2/3 isoforms precisely cofractionated with NuRD proteins during gel filtration of extracts derived from early and late embryos. In addition, we demonstrate that MBD2/3 forms multimers, and engages in specific interactions with the p55 and MI-2 subunits of the Drosophila NuRD complex.
Our data provide novel insights into the association between Drosophila MBD2/3 and NuRD proteins. Additionally, this work provides a first analysis of the architecture of the Drosophila NuRD complex.
Methyl-DNA binding proteins are connecting DNA methylation to transcriptional silencing [1–4]. Up to now, six methyl-DNA binding proteins could be identified in vertebrates . MeCP2, MBD2 and MBD3 can be found in large chromatin complexes containing histone deacetylase activity [1, 6, 4, 3] whereas MBD4 is involved in DNA mismatch-repair . MBD1 has been shown to repress transcription in cell culture  and recruits the histone H3-K9 methyltransferase SETDB1 to the chromatin assembly factor CAF-1 during S phase . MBD2, which can bind methylated DNA , is a transcriptional repressor recruiting a Nu cleosome R emodelling and D eacetylase complex (NuRD) to methylated CpG dinucleotides [6, 3], whereas mammalian MBD3, which is not able to bind methylated DNA  is an integral component of NuRD . Kaiso, a transcriptional repressor protein, can bind directly to CpG methylated DNA even though it lacks a conserved methyl-DNA binding domain . Kaiso is a component of a subpopulation of MeCP1 complexes that lack MBD2 .
The Drosophila gene MBD2/3 encodes a protein, which shares high homology to mammalian MBD2 and MBD3 [12, 4]. Due to differential splicing, Drosophila MBD2/3 is expressed in two isoforms, the smaller one is lacking part of the putative methyl-DNA binding domain [12–15]. The large isoform is expressed during early development, whereas the small isoform can only be detected during late embryogenesis [14, 15]. In insect cells expressing only the small MBD2/3 isoform, this protein was found to be associated with components of the Drosophila NuRD complex [12–14]. Moreover, this protein could repress transcription effectively in transfected Drosophila cells [13, 14].
The NuRD complexes of vertebrates are a heterogeneous group of complexes containing both histone deacetylase and nucleosome remodelling activities . NuRD complexes comprise at least seven proteins. The ATP- dependent nucleosome-remodelling activity is mediated by MI-2, which contains a SWI2/SNF2-type helicase/ATPase domain, two chromodomains and two PHD fingers . The related MTA1, MTA2 and MTA3 proteins have been found in various complex preparations [17, 4, 3, 18]. MTA1 was originally identified as being overexpressed in metastatic carcinomas . The histone deacetylases HDAC1 and HDAC2 and the two histone binding proteins RbAp46 and RbAp48 form the histone deacetylase core of NuRD complexes. Finally, as mentioned above, mammalian MBD3 is an integral component of at least some NuRD complexes .
Strikingly, the Drosophila genome contains clear homologues for all verterbrate NuRD proteins. Recombinant Drosophila MI-2 was shown to have ATPase and nucleosome mobilization properties . In Drosophila the HDAC gene Rpd3 is important for segmentation of the embryo . Drosophila p55, a WD-40 protein, is homologous to the histone deacetylase-associated proteins RbAp46 and RbAp48 . Finally, Drosophila MTA-like displays extensive homology to the vertebrate MTA proteins .
The strong conservation between vertebrate NuRD complexes and the Drosophila NuRD complex implies a conserved function during the animal development. In previous studies cell lines were analysed that express only the small isoform of MBD2/3, lacking part of the putative methyl-DNA binding domain and a Drosophila specific domain. In order to analyse isoform-specific differences of MBD2/3 and their ability to bind NuRD proteins, we now extend our analysis to both isoforms in Drosophila embryos.
MBD2/3 is associated with Drosophila NuRD proteins
Interactions between MBD2/3 and NuRD homologues
We then performed the MBD2/3-NuRD interaction assay using these conditions and observed strong interactions only between the long and small MBD2/3 isoforms (Fig. 2B, MBD2/3li and MBD2/3si, respectively) and p55. To confirm the association between MBD2/3 and p55 in Drosophila embryos we immunoprecipitated nuclear extracts with MBD2/3-specific antibodies and with myc-specific control antibodies. Precipitates were analysed by Western blot for the presence of p55 and GAGA factor. This revealed a specific association of MBD2/3 with the Drosophila NuRD homologue p55, but not with the unrelated GAGA factor (Fig. 2C).
A weaker interaction could also be detected between the small MBD2/3 isoform and MI-2 (Fig. 2B). Neither RPD3 nor MTA-like binding exceeded the background level defined by our pilot experiment with SV40 large T protein (Fig. 2B). Our results thus identified p55 and MI-2 as the direct interaction partners of MBD2/3 in the NuRD complex. In addition, our results from the GST-pulldown assay indicate that Drosophila MBD2/3 forms dimers or multimers, similar to mammalian MBD2 and MBD3 . To discriminate between the latter two possibilities we expressed a recombinant long MBD2/3 isoform and RPD3 in insect cells using a baculovirus expression system. Extracts from infected cells were subjected to Superdex 200 gel filtration and fractions were analysed by Western blot. The 58 kDa RPD3 protein eluted in fractions corresponding to molecular weights ranging from 66 kDa to 158 kDa, which suggested that RPD3 exists as monomers or dimers in solution (Fig. 2D). In contrast, the 36 kDa MBD2/3 isoform showed a strikingly different elution profile (Fig. 2D). The long MBD2/3 isoform eluted in a broad peak ranging from more than 158 kDa up to high molecular weight fractions (>400 kDa). No MBD2/3 was detected in fractions corresponding to the expected size of MBD2/3 monomers or dimers. This result is consistent with our earlier observation that MBD2/3 forms distinct aggregates in embryonic nuclei  and suggests that the protein efficiently oligomerises to form high-molecular weight complexes in solution.
Next we sought to delineate the domains that mediate the association between the MBD2/3 isoforms and their interacting proteins. To this end, we generated several MBD2/3 deletion mutants and tested their interaction with other proteins in a yeast two-hybrid assay. In a first series of experiments we tested the MBD2/3 mutants for their ability to interact with p55. This identified the region between residues 178 and 305 of the MBD2/3 long isoform as the p55 interaction domain (Fig. 3B). We also delineated the domain that mediates the interaction with MI-2. Our experiments revealed a strongly reduced interaction between a MBD2/3 small isoform derivative lacking the coiled-coil domain and MI-2 (Fig. 3C). It has been shown previously that vertebrate MBD2 and MBD3 form homo- and heterodimers via their N-terminal MBD domain and their C-terminal coiled-coil like sequences . Our findings might reflect a direct interaction between the coiled-coil domain and MI-2 or a requirement for efficient MBD2/3 multimerization for MI-2 interaction. In conclusion, our results identify distinct regions in the MBD2/3 protein that mediate protein-protein interactions with other NuRD proteins.
It has been previously suggested that MBD2/3 is associated with the Drosophila NuRD complex . This study determined that the small isoform MBD2/3 coelutes with some putative Drosophila NuRD subunits during fractionation of extracts derived from a Drosophila cell line. We have now extended this analysis to show that both isoforms of MBD2/3 coelute with NuRD homologues during fractionation of embryonic extracts. This data provides further evidence for a direct interaction between MBD2/3 and the NuRD complex.
Using several independent assays, we have demonstrated that MBD2/3 engages in homotypic interactions to form multimers. This effect is consistent with the formation of foci in embryonic nuclei  and also reminiscent of the interactions described for vertebrate MBD2 and MBD3 .
In addition, our data provides new insights into the association between MBD2/3 and NuRD. For example, we have shown that p55 appears to be the primary interaction partner of MBD2/3. We also observed a strong interaction between the small isoform MBD2/3 and MI-2, but not between MBD2/3 small isoform and MTA-like or RPD3. Additionally, we found interactions between p55 and all NuRD proteins, as well as a p55 homotypic interaction. The last finding is consistent with the fact that in the vertebrate NuRD complex the two p55 homologues RbAp46 and RbAp48 were identified as integral components . We note that the vertebrate NuRD complex also contains the two RPD3 homologues HDAC1 and HDAC2 ; unexpectedly, we were not able to detect any homotypic interaction of RPD3 in the yeast two-hybrid assay. One explanation could be that the dose of the two co-expressed RPD3 bait and prey proteins could interfere with the reporter gene activity due to their inherent histone deacetylation activity. However, the gel filtration assay revealed that the 58 kDa RPD3 protein eluted in fractions corresponding to molecular weights ranging from 66 kDa to 158 kDa, which suggested that RPD3 can interact homotypically (Fig. 3C).
It is possible that more complex interactions are involved in the assembly of the NuRD complex but they might not have been detectable under the stringent conditions of our assays. The interaction between MBD2/3 and MI-2 detected in our assays could also contribute to the specific association between MBD2/3 and the NuRD complex.
The interaction between MBD2/3 and p55 could promote the assembly of specialized chromatin structures in the fly. p55 is a WD-40 repeat protein that is involved in many aspects of chromatin organization . For example, p55 has been shown to be a component of the Drosophila CAF-1 complex that promotes nucleosome assembly . In addition, p55 is also contained in the NURF chromatin remodelling complex  and in the E(Z) complex that regulates homeotic gene expression [25, 26]. A mutant Drosophila allele for p55 is not available, but results obtained from Arabidopsis mutants with decreased levels of a p55 homologue indicate that the protein plays an important role in stabilizing epigenetic chromatin structures .
The goal of this study was to identify the interacting partners of the Drosophila methyl-DNA binding protein MBD2/3 within the Drosophila NuRD complex. We identified p55 and MI-2 as the primary interacting partners. We also found homo- and heterotypic interactions of the MBD2/3 isoforms, similar to vertebrate MBD2 and MBD3. Additionally, yeast two-hybrid assays revealed that p55 is able to specifically interact with all other NuRD proteins and can form homotypic interactions. Our data provides for the first time information about the architecture of the Drosophila NuRD complex. This allows us to develop a structural model of the NuRD complex.
Constructs for the yeast two-hybrid assay were generated by PCR amplification from cDNA clones  using specific primers (Tab. 1) with an attached restriction endonuclease target site for cloning. The PCR products were subsequently cloned into pGBKT7 or pGADT7 (Clontech) using standard procedures. All constructs were sequenced and in vitro translated to confirm the expression of corresponding proteins. Additionally, the bait constructs were expressed in yeast and expression of the fusion proteins was confirmed by Western analysis.
The following antibodies have been described previously: rabbit anti-MI-2 , rabbit anti-MTA1 [19, 4], rabbit anti-RPD3 , rabbit anti-p55 , rabbit anti-MBD2/3  and rabbit anti-GAGA .
Immunoprecipitations were caried out in 300 mM KCl supplemented with 0.2 % NP-40. Rabbit anti-MBD2/3 or mouse anti-myc (Clontech) antibodies were added to 75 μl of nuclear extracts prepared as mentioned below and incubated for 12 h at 4°C. Protein G beads (Amersham) were blocked in 3 mg/ml BSA for 20 min at room temperature, washed three times with 300 mM KCl, 0.2 % NP-40 and added to the samples. Incubation was carried out for an additional 1 h at 4°C. The beads were collected by centrifugation and washed four times in 1 ml 300 mM KCl, 0.2 % NP-40. The beads were resuspended in loading buffer for SDS-PAGE and vigorously vortexed for 15 sec. The immunoprecipitates were separated by SDS-PAGE, without futher boiling, transferred to a PVDF membrane and probed with antibodies against RPD3, p55 and GAGA factor, respectively.
Preparation of protein extracts and gel filtration
For baculovirus expression, the MBD2/3 cDNA  was subcloned into the pVL1392 baculovirus transfer vector. Transfer vector and linearised baculovirus DNA were cotransfected into Sf9 cells using the Bac'n'Blue transfection kit (Invitrogen) and recombinant virus was amplified according to the manufacturer's instructions. Whole cell extracts of infected Sf9 cells were generated by resuspending cell pellets in lysis buffer (20 mM Hepes pH 7.6, 200 mM KCl, 0.1 % NP40), incubation on ice for 10 min, three freeze/thaw cycles and sonication. Nuclear extracts from Drosophila embryos were prepared as described previously . Extracts were cleared by centrifugation and passaged through a 0.2 μm filter. 200 μl of cleared extract was applied to Superdex 200 HR 10/30 or Superose 6 HR 10/30 gel filtration columns (Amersham Pharmacia) and resolved in 20 mM Hepes pH 7.6 and 300 mM KCl on an Äkta Purifier system (Amersham Pharmacia) according to the manufacturer's instructions.
35S-methionine-labelled proteins were generated by in vitro trancription/translation of pGADT7_T, pGADT7_Mi-2, pGADT7_MTA-like, pGADT7_Rpd3, pGADT7_p55, pGADT7_MBD2/3 long isoform and pGADT7_MBD2/3 small isoform using the TNT coupled reticulocyte lysate system (Promega) according to the manufacturer's protocol. GST-MBD2/3 long isoform and GST-MBD2/3 small isoform fusionproteins were obtained by cloning the coding region of the two isoforms in pGEX4T1 (Amersham) and subsequent expressing of the constructs in BL-21 bacteria according to the manufacturer's protocol. GST-MBD2a was obtained from Hidetoshi Fujita . GST-pull downs were performed under the following conditions: As incubation and washing buffer we used 20 mM HEPES, pH 7.8, 300 mM NaCl, 0.1 % desoxycholate, 0.1 % IGEPAL, 10 % glycerol. The radioactively labelled proteins were incubated in incubation buffer for 3 h at room temperature with GST fusion proteins coupled to Sepharose 4B (Amersham) or with GST alone coupled to Sepharose 4B. Beads were then washed five times for 10 min at room temperature. After the last washing step, beads were boiled in SDS loading buffer for 10 min and loaded onto standard SDS-polyacrylamide gels. After separation by SDS-PAGE, proteins were blotted onto a PVDF membrane, which was stained with Ponceau S, dried and exposed on X-ray films.
Yeast two-hybrid assays
The Matchmaker Two-Hybrid system 3 (Clontech) was used for all experiments, according to the manufacturer's instructions. The AH109 strain was used as host and transformed with various constructs (see above) using standard procedures. Transformants were selected on SD/-Ade/-His/-Leu/-Trp plates.
We would like to thank Bodo Brueckner and Frank Lyko for critical revision of the paper and helpful guidance. We would also like to thank Jürg Müller and Jessica Tyler for antibodies, and Angelika Mitterweger and Michael Korenjak for help with protein extract preparation.
- Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe AP: Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet. 1998, 19: 187-91. 10.1038/561View ArticlePubMed
- Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, Bird A: Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature. 1998, 393: 386-9. 10.1038/30764View ArticlePubMed
- Zhang Y, Ng HH, Erdjument-Bromage H, Tempst P, Bird A, Reinberg D: Analysis of the NuRD subunits reveals a histone deacetylase core complex and a connection with DNA methylation. Genes Dev. 1999, 13: 1924-35. 10.1101/gad.13.18.2388PubMed CentralView ArticlePubMed
- Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP: Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nat Genet. 1999, 23: 62-6.PubMed
- Ballestar E, Wolffe AP: Methyl-CpG-binding proteins. Targeting specific gene repression. Eur J Biochem. 2001, 268: 1-6. 10.1046/j.1432-1327.2001.01869.xView ArticlePubMed
- Ng HH, Zhang Y, Hendrich B, Johnson CA, Turner BM, Erdjument-Bromage H, Tempst P, Reinberg D, Bird A: MBD2 is a transcriptional repressor belonging to the MeCP1 histone deacetylase complex. Nat Genet. 1999, 23: 58-61.PubMed
- Hendrich B, Hardeland U, Ng HH, Jiricny J, Bird A: The thymine glycosylase MBD4 can bind to the product of deamination at methylated CpG sites. Nature. 1999, 401: 301-4. 10.1038/45843View ArticlePubMed
- Fujita N, Takebayashi S, Okumura K, Kudo S, Chiba T, Saya H, Nakao M: Methylation-mediated transcriptional silencing in euchromatin by methyl-CpG binding protein MBD1 isoforms. Mol Cell Biol. 1999, 19: 6415-26.PubMed CentralView ArticlePubMed
- Sarraf SA, Stancheva I: Methyl-CpG Binding Protein MBD1 Couples Histone H3 Methylation at Lysine 9 by SETDB1 to DNA Replication and Chromatin Assembly. Mol Cell. 2004, 15: 595-605. 10.1016/j.molcel.2004.06.043View ArticlePubMed
- Hendrich B, Bird A: Identification and characterization of a family of mammalian methyl-CpG binding proteins. Mol Cell Biol. 1998, 18: 6538-47.PubMed CentralView ArticlePubMed
- Prokhortchouk A, Hendrich B, Jorgensen H, Ruzov A, Wilm M, Georgiev G, Bird A, Prokhortchouk E: The p120 catenin partner Kaiso is a DNA methylation-dependent transcriptional repressor. Genes Dev. 2001, 15: 1613-8. 10.1101/gad.198501PubMed CentralView ArticlePubMed
- Tweedie S, Ng HH, Barlow AL, Turner BM, Hendrich B, Bird A: Vestiges of a DNA methylation system in Drosophila melanogaster?. Nat Genet. 1999, 23: 389-90. 10.1038/70490View ArticlePubMed
- Roder K, Hung MS, Lee TL, Lin TY, Xiao H, Isobe KI, Juang JL, Shen CJ: Transcriptional repression by Drosophila methyl-CpG-binding proteins. Mol Cell Biol. 2000, 20: 7401-9. 10.1128/MCB.20.19.7401-7409.2000PubMed CentralView ArticlePubMed
- Ballestar E, Pile LA, Wassarman DA, Wolffe AP, Wade PA: A Drosophila MBD family member is a transcriptional corepressor associated with specific genes. Eur J Biochem. 2001, 268: 5397-406. 10.1046/j.0014-2956.2001.02480.xView ArticlePubMed
- Marhold J, Zbylut M, Lankenau DH, Li M, Gerlich D, Ballestar E, Mechler BM, Lyko F: Stage-specific chromosomal association of Drosophila dMBD2/3 during genome activation. Chromosoma. 2002, 111: 13-21. 10.1007/s00412-002-0188-2View ArticlePubMed
- Zhang Y, LeRoy G, Seelig HP, Lane WS, Reinberg D: The dermatomyositis-specific autoantigen Mi2 is a component of a complex containing histone deacetylase and nucleosome remodeling activities. Cell. 1998, 95: 279-89. 10.1016/S0092-8674(00)81758-4View ArticlePubMed
- Fujita N, Jaye DL, Kajita M, Geigerman C, Moreno CS, Wade PA: MTA3, a Mi-2/NuRD complex subunit, regulates an invasive growth pathway in breast cancer. Cell. 2003, 113: 207-19. 10.1016/S0092-8674(03)00234-4View ArticlePubMed
- Humphrey GW, Wang Y, Russanova VR, Hirai T, Qin J, Nakatani Y, Howard BH: Stable histone deacetylase complexes distinguished by the presence of SANT domain proteins CoREST/kiaa0071 and Mta-L1. J Biol Chem. 2001, 276: 6817-24. 10.1074/jbc.M007372200View ArticlePubMed
- Toh Y, Pencil SD, Nicolson GL: A novel candidate metastasis-associated gene, mta1, differentially expressed in highly metastatic mammary adenocarcinoma cell lines. cDNA cloning, expression, and protein analyses. J Biol Chem. 1994, 269: 22958-63.PubMed
- Brehm A, Langst G, Kehle J, Clapier CR, Imhof A, Eberharter A, Muller J, Becker PB: dMi-2 and ISWI chromatin remodelling factors have distinct nucleosome binding and mobilization properties. Embo J. 2000, 19: 4332-41. 10.1093/emboj/19.16.4332PubMed CentralView ArticlePubMed
- Mannervik M, Levine M: The Rpd3 histone deacetylase is required for segmentation of the Drosophila embryo. Proc Natl Acad Sci U S A. 1999, 96: 6797-801. 10.1073/pnas.96.12.6797PubMed CentralView ArticlePubMed
- Tyler JK, Bulger M, Kamakaka RT, Kobayashi R, Kadonaga JT: The p55 subunit of Drosophila chromatin assembly factor 1 is homologous to a histone deacetylase-associated protein. Mol Cell Biol. 1996, 16: 6149-59.PubMed CentralView ArticlePubMed
- Bowen NJ, Fujita N, Kajita M, Wade PA: Mi-2/NuRD: multiple complexes for many purposes. Biochim Biophys Acta. 2004, 1677: 52-7.View ArticlePubMed
- Feng Q, Zhang Y: The MeCP1 complex represses transcription through preferential binding, remodeling, and deacetylating methylated nucleosomes. Genes Dev. 2001, 15: 827-32.PubMed CentralPubMed
- Tie F, Furuyama T, Prasad-Sinha J, Jane E, Harte PJ: The Drosophila Polycomb Group proteins ESC and E(Z) are present in a complex containing the histone-binding protein p55 and the histone deacetylase RPD3. Development. 2001, 128: 275-86.PubMed
- Muller J, Hart CM, Francis NJ, Vargas ML, Sengupta A, Wild B, Miller EL, O'Connor MB, Kingston RE, Simon JA: Histone methyltransferase activity of a Drosophila Polycomb group repressor complex. Cell. 2002, 111: 197-208. 10.1016/S0092-8674(02)00976-5View ArticlePubMed
- Li B, Fields S: Identification of mutations in p53 that affect its binding to SV40 large T antigen by using the yeast two-hybrid system. Faseb J. 1993, 7: 957-63.PubMed
- Tatematsu KI, Yamazaki T, Ishikawa F: MBD2-MBD3 complex binds to hemi-methylated DNA and forms a complex containing DNMT1 at the replication foci in late S phase. Genes Cells. 2000, 5: 677-88. 10.1046/j.1365-2443.2000.00359.xView ArticlePubMed
- Henikoff S: Versatile assembler. Nature. 2003, 423: 814-5. 817. 10.1038/423814aView ArticlePubMed
- Martinez-Balbas MA, Tsukiyama T, Gdula D, Wu C: Drosophila NURF-55, a WD repeat protein involved in histone metabolism. Proc Natl Acad Sci U S A. 1998, 95: 132-7. 10.1073/pnas.95.1.132PubMed CentralView ArticlePubMed
- Hennig L, Taranto P, Walser M, Schonrock N, Gruissem W: Arabidopsis MSI1 is required for epigenetic maintenance of reproductive development. Development. 2003, 130: 2555-65. 10.1242/dev.00470View ArticlePubMed
- Rubin GM, Hong L, Brokstein P, Evans-Holm M, Frise E, Stapleton M, Harvey DA: A Drosophila complementary DNA resource. Science. 2000, 287: 2222-4. 10.1126/science.287.5461.2222View ArticlePubMed
- Sandaltzopoulos R, Mitchelmore C, Bonte E, Wall G, Becker PB: Dual regulation of the Drosophila hsp26 promoter in vitro. Nucleic Acids Res. 1995, 23: 2479-87.PubMed CentralView ArticlePubMed
- Fujita H, Fujii R, Aratani S, Amano T, Fukamizu A, Nakajima T: Antithetic effects of MBD2a on gene regulation. Mol Cell Biol. 2003, 23: 2645-57. 10.1128/MCB.23.8.2645-2657.2003PubMed CentralView ArticlePubMed
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.