- Top of page
- Materials and methods
- Supporting Information
Decapping scavenger (DcpS) assists in precluding inhibition of cap-binding proteins by hydrolyzing cap species remaining after mRNA 3′5′ degradation. Its significance was reported in splicing, translation initiation and microRNA turnover. Here we examine the structure and binding mode of DcpS from Caenorhabditis elegans (CeDcpS) using a large collection of chemically modified methylenebis(phosphonate), imidodiphosphate and phosphorothioate cap analogs. We determine that CeDcpS is a homodimer and propose high accuracy structural models of apo- and m7GpppG-bound forms. The analysis of CeDcpS regioselectivity uncovers that the only site of hydrolysis is located between the β and γ phosphates. Structure–affinity relationship studies of cap analogs for CeDcpS reveal molecular determinants for efficient cap binding: a strong dependence on the type of substituents in the phosphate chain, and reduced binding affinity for either methylated hydroxyl groups of m7Guo or an extended triphosphate chain. Docking analysis of cap analogs in the CeDcpS active site explains how both phosphate chain mobility and the orientation in the cap-binding pocket depend on the number of phosphate groups, the substituent type and the presence of the second nucleoside. Finally, the comparison of CeDcpS with its well known human homolog provides general insights into DcpS–cap interactions.
- Top of page
- Materials and methods
- Supporting Information
The regulation of gene expression depends on mutual interactions between mRNA and cap-binding proteins, including the cap-binding complex, eukaryotic translation initiation factor 4E (eIF4E), decapping protein 2 (Dcp2) and decapping scavenger (DcpS), which are involved in mRNA maturation, transport, translation and turnover. After entering the cytoplasm, mRNAs can be translated, stored for later translation in stress granules or P bodies, or degraded [1-3]. The translation and turnover of mRNAs are modulated by the mRNA 5′ cap, which is a modified nucleoside (N7-methylguanosine, m7Guo) linked to the first transcribed nucleoside (N) of the mRNA chain through a 5′,5′-triphosphate bridge, m7Gp(γ)p(β)p(α)N. The cap structure serves as the assembly site for translation initiation factors and degradation complexes [4, 5]. Eukaryotic mRNAs are degraded through two major decay pathways, both initiated by poly(A) tail shortening . In the 5′3′ decay pathway, Dcp2 cleaves a capped mRNA to generate 7-methylguanosine diphosphate (m7GDP) and an mRNA chain containing a phosphate group at its 5′ end . In the 3′5′ decay pathway, mRNA is degraded by the exosome, releasing a cap-containing dinucleotide or a short capped oligonucleotide, which is subsequently hydrolyzed to 7-methylguanosine monophosphate (m7GMP) by the DcpS enzyme [7, 8].
DcpSs are members of the HIT family of pyrophosphatases containing an evolutionarily conserved histidine triad (HIT) motif which is crucial for DcpS-mediated hydrolysis of the 5′,5′-triphosphate bridge within the cap . The general mechanism of catalysis proposed for all HIT proteins involves the central histidine residue of the HIT motif [H277 for human DcpS (HsDcpS)] in the formation of a covalent nucleotidyl phosphohistidyl intermediate, which functions as the nucleophilic agent for the γ phosphate group of dinucleoside polyphosphate substrates . The crystal structures of HsDcpS–cap complexes demonstrate that the active site residues can accommodate m7GpppG or m7GpppA dinucleotides as substrates (catalytically inactive H277N HsDcpS mutant was used ), or m7GDP mononucleotide , which is a non-hydrolyzable, competitive DcpS inhibitor . The formation of DcpS–cap complexes for other species has not been presented until now.
The DcpS hydrolytic activity shows a strong preference of this enzyme for dinucleotide substrates [8, 14]. However, cap-containing oligonucleotides up to four and 10 nucleotides can also be hydrolyzed by Caenorhabditis elegans DcpS (CeDcpS) and HsDcpS, respectively [8, 14]. Such short capped species are able to inhibit cap-dependent processes of mRNA metabolism [15-17]. Consequently, DcpS inhibition may serve as a treatment strategy for diseases where the decrease of DcpS activity reduces pathology, such as spinal muscular atrophy . It has been suggested that, in several types of cancer with overexpression of eIF4E [19, 20], cap analogs might be used as therapeutic agents targeting eIF4E. Particularly useful for this purpose should be analogs resistant to DcpS, because of their potentially increased cellular stability [21, 22]. Recently, it was discovered that DcpS contributes to the control of microRNA (miRNA) turnover in C. elegans by stimulating the XRN1-mediated miRNA degradation .
Caenorhabditis elegans is one of the best-studied models of eukaryotic organisms, widely used to research metazoan genetics and biology , whose genome was the first from multicellular organisms to be completely sequenced . However, prior to the current research little was known about its DcpS [13, 14, 26], including the enzyme's structure and affinity for native substrates or their derivatives. The recognition of DcpS significance in miRNA turnover in C. elegans  highlights that CeDcpS is worthy of continued basic biophysical, biochemical and computational studies.
We previously examined the substrate specificity of CeDcpS towards a series of dinucleotide cap analogs modified within the m7Guo or the second nucleoside . Modifications of nucleosides did not alter cap degradation, and from the Michaelis constants (KM) and maximum velocities (vmax) we established the impact of the functional groups of the nucleosides on substrate specificity and hydrolysis rate . We also demonstrated that a DcpS substrate, regardless of human or nematode origin (C. elegans and Ascaris suum), could be the m7Guo with at least three phosphate groups . Interestingly, a diphosphate chain was sufficient for strong cap binding, although not for hydrolysis .
In order to find other molecular determinants for efficient cap binding by DcpS, we analyzed the structure–affinity relationship for the interaction of CeDcpS with methylenebis(phosphonate), imidodiphosphate and phosphorothioate cap analogs, bearing CH2, NH or S modification respectively within the phosphate bridge (Fig. 1). These analogs differ not only in the size or electronegativity of a substituent but also in the modification site: CH2 and NH groups are bridging substitutions (the oxygen atom from the α/β or β/γ position is replaced), while the S atom represents a non-bridging substitution (the oxygen atom from the γ position is replaced; Fig. 1). Among 15 dinucleotide cap analogs modified in the phosphate bridge, we selected those resistant to degradation by CeDcpS (Table 1), by means of HPLC. Using the non-hydrolyzable compounds in the subsequent time-synchronized fluorescence titration measurements, we were able to determine the equilibrium association constants (KAS) and Gibbs free energies of binding (ΔG0) for the wild-type enzyme, not affecting its native binding affinity.
Table 1. Specificity and binding affinity of CeDcpS towards cap analogs, determined in 50 mm Tris/HCl buffer containing 200 mm KCl, 0.5 mm EDTA and 1 mm dithiothreitol (pH 7.6), at 20 °C. Reaction products were identified via HPLC, while the equilibrium association constants (KAS) and Gibbs free energies of binding (ΔG0) were obtained via fluorescence titration. Enzyme concentration was 0.2 μm, cap analog concentration was 15 μm in the HPLC analysis and from 0.1 to 15 μm in titration measurements
|Type||No.||Cap analog||Hydrolysis products||KAS (μm−1)||ΔG0 (kcal·mol−1)|
|No substituent in the phosphate chain||1||m7GMPa||Resistant||8.00 ± 0.65||−9.25 ± 0.05|
|2||m7GDPa||Resistant||97.04 ± 12.52||−10.71 ± 0.08|
|3||m7GTPa||m7GMP + pp||–||–|
|4||m7GpppGa,b||m7GMP + GDP||–||–|
|5||m7GppppG||m7GMP + GTP||–||–|
|6||m27,2′-OGpppGb||m27,2′-OGMP + GDP||–||–|
|7||m27,3′-OGpppGb||m27,3′-OGMP + GDP||–||–|
|8||bn7GpppGb||bn7GMP + GDP||–||–|
|Methylenebis(phosphonates)||9||m7GpCH2ppG||Resistant||8.57 ± 0.43||−9.29 ± 0.03|
|10||m7GpCH2pppG||Resistant||2.74 ± 0.23||−8.63 ± 0.05|
|11||m27,3′-OGpCH2ppG||Resistant||2.10 ± 0.82||−8.48 ± 0.23|
|12||bn7GpCH2ppG||Resistant||2.13 ± 0.17||−8.48 ± 0.05|
|13||m7GppCH2pG||m7GMP + pCH2pG||–||–|
|14||m27,2′-OGppCH2pG||m27,2′-OGMP + pCH2pG||–||–|
|15||m27,3′-OGppCH2pG||Resistant||0.33 ± 0.12||−7.40 ± 0.21|
|Imidodiphosphates||16||m7GpNHppG||Resistant||65.59 ± 8.78||−10.48 ± 0.08|
|17||m7GpNHpppG||Resistant||6.44 ± 0.61||−9.13 ± 0.06|
|18||m27,2′-OGpNHppG||Resistant||0.74 ± 0.24||−7.87 ± 0.19|
|19||m27,2′-OGpNHpppG||Resistant||1.02 ± 0.34||−8.05 ± 0.19|
|20||m7GppNHpG||m7GMP + pNHpG||–||–|
|21||m27,2′-OGppNHpG||m27,2′-OGMP + pNHpG||–||–|
|Phosphorothioates||22||m7GpSpp D2||Resistant||19.53 ± 1.63||−9.77 ± 0.05|
|23||m7GpSppG D2c||Resistant||68.65 ± 5.63||−10.51 ± 0.05|
|24||m27,2′-OGpSppG D2||Resistant||2.43 ± 0.77||−8.56 ± 0.18|
Figure 1. Structures of cap analogs with the specific types of modifications indicated by van der Waals spheres. Atoms are colored as follows: C, grey; O, red; N, blue; P, orange; S, yellow-green.
Download figure to PowerPoint
Remarkably, neither the structure nor the subunit composition of CeDcpS has been presented until now. Applying gel filtration chromatography, we show that CeDcpS is a homodimer and propose the homology models of its structure based on the resolved structures of HsDcpS dimer [11, 12]. To identify particular residues important for the molecular recognition and binding of a cap, we carried out docking studies of representative experimentally examined cap analogs within the CeDcpS active site. These synergistic experimental and computational approaches enabled us to compare structural features and binding affinities between CeDcpS and its well characterized human homolog [11-14, 18, 21, 27-31]. The identification of features common to CeDcpS and HsDcpS could probably be further exploited in the design and validation of DcpS inhibitors that might be therapeutically relevant, thereby offering the ability to perform a large-scale genetic analysis of the effects of DcpS inhibition in C. elegans. On the other hand, by finding differences between human and nematode DcpS enzymes we could propose potential strategies for designing potent anti-parasitic agents.
- Top of page
- Materials and methods
- Supporting Information
We thank Magdalena Wypijewska, Anna Nowicka and Katarzyna Wnek (Division of Biophysics, Institute of Experimental Physics, Faculty of Physics, University of Warsaw) for help with HPLC experiments and Anna Szakiel (Institute of Biochemistry, Faculty of Biology, University of Warsaw) for the reference compounds used in gel filtration studies. We acknowledge time-synchronized titration training and exceptionally helpful discussions with Joanna Zuberek (Division of Biophysics, Institute of Experimental Physics, Faculty of Physics, University of Warsaw) and Anna Niedzwiecka (Institute of Physics, Polish Academy of Sciences). Thanks to Frederic Picard-Jean and Moheshwarnath Issur (Department of Biochemistry, University of Sherbrooke) for sharing knowledge about molecular docking and their encouragement. We also acknowledge helpful discussion on the experimental work with Professor Richard E. Davis (Department of Biochemistry and Molecular Genetics, University of Colorado, School of Medicine, Aurora) and on computational procedures with Dr Andrei-Jose Petrescu (IBAR).
This work was supported by grants from the Polish Ministry of Science and Higher Education (N 301 096339 to E.D.), the National Science Centre, Poland (UMO-2012/05/E/ST5/03893 to J.J.), the National Center of Research and Development (02/EuroNanoMed/2011 to E.D.) and the project of the Institute of Experimental Physics, Faculty of Physics, University of Warsaw (BST 163500/BF to A.W.). M.D.S and A.L.M. were supported by the Romanian Academy (project 3 of IBAR). A.L.M. acknowledges a postdoctoral program from the European Social Fund (POSDRU/89/1.5/S/60746). M.D.S. acknowledges a grant (ID3-0342-181/2011) within the Exploratory Research Projects funded by the Romanian Ministry of Research and Education.