Monitoring Genomic Sequences during SELEX Using High-Throughput Sequencing: Neutral SELEX

Background

SELEX is a well established in vitro selection tool to analyze the structure of ligand-binding nucleic acid sequences called aptamers. Genomic SELEX transforms SELEX into a tool to discover novel, genomically encoded RNA or DNA sequences binding a ligand of interest, called genomic aptamers. Concerns have been raised regarding requirements imposed on RNA sequences undergoing SELEX selection.

Methodology/Principal Findings

To evaluate SELEX and assess the extent of these effects, we designed and performed a Neutral SELEX experiment omitting the selection step, such that the sequences are under the sole selective pressure of SELEX''s amplification steps. Using high-throughput sequencing, we obtained thousands of full-length sequences from the initial genomic library and the pools after each of the 10 rounds of Neutral SELEX. We compared these to sequences obtained from a Genomic SELEX experiment deriving from the same initial library, but screening for RNAs binding with high affinity to the E. coli regulator protein Hfq. With each round of Neutral SELEX, sequences became less stable and changed in nucleotide content, but no sequences were enriched. In contrast, we detected substantial enrichment in the Hfq-selected set with enriched sequences having structural stability similar to the neutral sequences but with significantly different nucleotide selection.

Conclusions/Significance

Our data indicate that positive selection in SELEX acts independently of the neutral selective requirements imposed on the sequences. We conclude that Genomic SELEX, when combined with high-throughput sequencing of positively and neutrally selected pools, as well as the gnomic library, is a powerful method to identify genomic aptamers.     

Selection of genomic sequences that bind tightly to Ff gene 5 protein: primer-free genomic SELEX

Single-stranded DNA or RNA libraries used in SELEX experiments usually include primer-annealing sequences for PCR amplification. In genomic SELEX, these fixed sequences may form base pairs with the central genomic fragments and interfere with the binding of target molecules to the genomic sequences. In this study, a method has been developed to circumvent these artificial effects. Primer-annealing sequences are removed from the genomic library before selection with the target protein and are then regenerated to allow amplification of the selected genomic fragments. A key step in the regeneration of primer-annealing sequences is to employ thermal cycles of hybridization-extension, using the sequences from unselected pools as templates. The genomic library was derived from the bacteriophage fd, and the gene 5 protein (g5p) from the phage was used as a target protein. After four rounds of primer-free genomic SELEX, most cloned sequences overlapped at a segment within gene 6 of the viral genome. This sequence segment was pyrimidine-rich and contained no stable secondary structures. Compared with a neighboring genomic fragment, a representative sequence from the family of selected sequences had about 23-fold higher g5p-binding affinity. Results from primer-free genomic SELEX were compared with the results from two other genomic SELEX protocols.     

Genomic SELEX for Hfq-binding RNAs identifies genomic aptamers predominantly in antisense transcripts

An unexpectedly high number of regulatory RNAs have been recently discovered that fine-tune the function of genes at all levels of expression. We employed Genomic SELEX, a method to identify protein-binding RNAs encoded in the genome, to search for further regulatory RNAs in Escherichia coli. We used the global regulator protein Hfq as bait, because it can interact with a large number of RNAs, promoting their interaction. The enriched SELEX pool was subjected to deep sequencing, and 8865 sequences were mapped to the E. coli genome. These short sequences represent genomic Hfq-aptamers and are part of potential regulatory elements within RNA molecules. The motif 5′-AAYAAYAA-3′ was enriched in the selected RNAs and confers low-nanomolar affinity to Hfq. The motif was confirmed to bind Hfq by DMS footprinting. The Hfq aptamers are 4-fold more frequent on the antisense strand of protein coding genes than on the sense strand. They were enriched opposite to translation start sites or opposite to intervening sequences between ORFs in operons. These results expand the repertoire of Hfq targets and also suggest that Hfq might regulate the expression of a large number of genes via interaction with cis-antisense RNAs.     

MS2 coat proteins bound to yeast mRNAs block 5′ to 3′ degradation and trap mRNA decay products: implications for the localization of mRNAs by MS2-MCP system

Arrays of MS2 binding sites are placed into mRNAs and are commonly used to visualize the localization of mRNAs in vivo by the expression of an MS2-GFP fusion protein. In Saccharomyces cerevisiae, we observed that arrays of MS2 binding sites inhibit 5′ to 3′ degradation of the mRNA in yeast cells and lead to the accumulation of a 3′ mRNA fragment containing the MS2 binding sites. This accumulation can be dependent on the binding of the MS2 stem loops (MS2-SL) by MS2 coat proteins (MCPs). Since such decay fragments can still bind MCP-GFP, the localization of such mRNA fragments can complicate the interpretation of the localization of full-length mRNA in vivo.     

Isolation and structural analysis of ribosomal protein genes in Xenopus laevis. Homology between sequences present in the gene and in several different messenger RNAs

The ribosomal protein genes are present in two to four copies per haploid genome of Xenopus laevis. Using cloned complementary DNA probes, we have isolated, from a genomic library of X. laevis, several clones containing genes for two different ribosomal proteins (L1 and L14). These genes contain intervening sequences. In the case of the L1 gene, the exons are 100 to 200 base-pairs long and the introns, on average, 400 base-pairs. Along the genomic fragments, two different classes of repetitive DNA are present: highly and middle repetitive DNA. Both are evolutionarily unstable as shown by hybridization to Xenopus tropicalis DNA. Several introns of the gene coding for protein L1 contain middle repetitive sequences. Hybridization and hybrid-released translation experiments have shown that sequences inside the two genes hybridize to several poly(A) messenger RNAs. Some of the products encoded by these mRNA have electrophoretic properties of ribosomal proteins.     

Mutually-induced Conformational Switching of RNA and Coat Protein Underpins Efficient Assembly of a Viral Capsid

Single-stranded RNA viruses package their genomes into capsids enclosing fixed volumes. We assayed the ability of bacteriophage MS2 coat protein to package large, defined fragments of its genomic, single-stranded RNA. We show that the efficiency of packaging into a T = 3 capsid in vitro is inversely proportional to RNA length, implying that there is a free-energy barrier to be overcome during assembly. All the RNAs examined have greater solution persistence lengths than the internal diameter of the capsid into which they become packaged, suggesting that protein-mediated RNA compaction must occur during assembly. Binding ethidium bromide to one of these RNA fragments, which would be expected to reduce its flexibility, severely inhibited packaging, consistent with this idea. Cryo-EM structures of the capsids assembled in these experiments with the sub-genomic RNAs show a layer of RNA density beneath the coat protein shell but lack density for the inner RNA shell seen in the wild-type virion. The inner layer is restored when full-length virion RNA is used in the assembly reaction, implying that it becomes ordered only when the capsid is filled, presumably because of the effects of steric and/or electrostatic repulsions. The cryo-EM results explain the length dependence of packaging. In addition, they show that for the sub-genomic fragments the strongest ordered RNA density occurs below the coat protein dimers forming the icosahedral 5-fold axes of the capsid. There is little such density beneath the proteins at the 2-fold axes, consistent with our model in which coat protein dimers binding to RNA stem-loops located at sites throughout the genome leads to switching of their preferred conformations, thus regulating the placement of the quasi-conformers needed to build the T = 3 capsid. The data are consistent with mutual chaperoning of both RNA and coat protein conformations, partially explaining the ability of such viruses to assemble so rapidly and accurately.     

Organization and expression of a-tubulin genes in Drosophila melanogaster: One member of the a-tubulin multigene family is transcribed in both oogenesis and later embryonic development

Genomic clones containing α-tubulin sequences were isolated from a library of Drosophila melanogaster DNA and identified by a hybridization-selection and in vitro-translation procedure. The in vitro translation products were identical to the two electrophoretic variants of α-tubulin present in Drosophila embryos. They co-assembled with an embryonic tubulin fraction to form microtubules in vitro and generated the same partial proteolytic fragments as Drosophila α-tubulins. Hybridization in situ to polytene chromosomes revealed that the α-tubulin sequences constitute a multigene family localized on the right arm of chromosome 3 at sites 84 B3–6, 84 D4–8 and 85 E6–10. Clones hybridizing to these sites corresponded to the three major α-tubulin sequences in genomic DNA. The α-tubulin sequences at 84 B3–6 were present twice per haploid genome, embedded in a large duplicated DNA segment. The sequences of the three genomic α-tubulin genes showed considerable divergence, making it possible to conclude that both of the α-tubulin variants present in embryos are encoded by the genes at 84 B3–6. Furthermore, the abundance of this α-tubulin messenger RNA changes with the requirements for microtubule synthesis during embryo development. The genes at 84 B3–6 encoded both the stored maternal mRNA of the oocyte and the major mRNA transcribed during embryonic development.     

Primer-Free Aptamer Selection Using A Random DNA Library

Aptamers are highly structured oligonucleotides (DNA or RNA) that can bind to targets with affinities comparable to antibodies ¹. They are identified through an in vitro selection process called Systematic Evolution of Ligands by EXponential enrichment (SELEX) to recognize a wide variety of targets, from small molecules to proteins and other macromolecules ^2-4. Aptamers have properties that are well suited for in vivo diagnostic and/or therapeutic applications: Besides good specificity and affinity, they are easily synthesized, survive more rigorous processing conditions, they are poorly immunogenic, and their relatively small size can result in facile penetration of tissues.Aptamers that are identified through the standard SELEX process usually comprise ~80 nucleotides (nt), since they are typically selected from nucleic acid libraries with ~40 nt long randomized regions plus fixed primer sites of ~20 nt on each side. The fixed primer sequences thus can comprise nearly ~50% of the library sequences, and therefore may positively or negatively compromise identification of aptamers in the selection process ³, although bioinformatics approaches suggest that the fixed sequences do not contribute significantly to aptamer structure after selection ⁵. To address these potential problems, primer sequences have been blocked by complementary oligonucleotides or switched to different sequences midway during the rounds of SELEX ⁶, or they have been trimmed to 6-9 nt ^{7, 8}. Wen and Gray ⁹ designed a primer-free genomic SELEX method, in which the primer sequences were completely removed from the library before selection and were then regenerated to allow amplification of the selected genomic fragments. However, to employ the technique, a unique genomic library has to be constructed, which possesses limited diversity, and regeneration after rounds of selection relies on a linear reamplification step. Alternatively, efforts to circumvent problems caused by fixed primer sequences using high efficiency partitioning are met with problems regarding PCR amplification ¹⁰.We have developed a primer-free (PF) selection method that significantly simplifies SELEX procedures and effectively eliminates primer-interference problems ^{11, 12}. The protocols work in a straightforward manner. The central random region of the library is purified without extraneous flanking sequences and is bound to a suitable target (for example to a purified protein or complex mixtures such as cell lines). Then the bound sequences are obtained, reunited with flanking sequences, and re-amplified to generate selected sub-libraries. As an example, here we selected aptamers to S100B, a protein marker for melanoma. Binding assays showed Kd s in the 10^-7 - 10^-8 M range after a few rounds of selection, and we demonstrate that the aptamers function effectively in a sandwich binding format.Download video file.^{(128M, mp4)}     

Libraries for genomic SELEX.

An increasing number of proteins are being identified that regulate gene expression by binding specific nucleic acidsin vivo. A method termed genomic SELEX facilitates the rapid identification of networks of protein-nucleic acid interactions by identifying within the genomic sequences of an organism the highest affinity sites for any protein of the organism. As with its progenitor, SELEX of random-sequence nucleic acids, genomic SELEX involves iterative binding, partitioning, and amplification of nucleic acids. The two methods differ in that the variable region of the nucleic acid library for genomic SELEX is derived from the genome of an organism. We have used a quick and simple method to construct Escherichia coli, Saccharomyces cerevisiae, and human genomic DNA PCR libraries that can be transcribed with T7 RNA polymerase. We present evidence that the libraries contain overlapping inserts starting at most of the positions within the genome, making these libraries suitable for genomic SELEX.     

Nonuniform size distribution of nascent peptides. The effect of messenger RNA structure upon the rate of translation.

The size distribution of bacteriophage MS2 coat protein nascent chains purified from MS2-infected Escherichia coli has been determined. Accumulations of nascent chains of discrete sizes were observed, providing evidence that the rate of chain elongation during coat protein biosynthesis is not uniform. A correlation of the size of nascent peptides which accumulate during MS2 coat protein biosynthesis and the position on the MS2 coat protein mRNA of the ribosome carrying those lengths of nascent peptides may be made. This correlation leads to the hypothesis that regions of mRNA secondary structure impede the movement of ribosomes during chain elongation and thus serve as the origin of the accumulation of nascent chains of discrete sizes.     

SELEX experiments: new prospects, applications and data analysis in inferring regulatory pathways

Systematic Evolution of Ligands by EXponential enrichment (SELEX) is an experimental procedure that allows extraction, from an initially random pool of oligonucleotides, of the oligomers with a desired binding affinity for a given molecular target. The procedure can be used to infer the strongest binders for a given DNA or RNA binding protein, and the highest affinity binding sequences isolated through SELEX can have numerous research, diagnostic and therapeutic applications. Recently, important new modifications of the SELEX protocol have been proposed. In particular, a modification of the standard SELEX procedure allows generating a dataset from which protein-DNA interaction parameters can be determined with unprecedented accuracy. Another variant of SELEX allows investigating interactions of a protein with nucleic-acid fragments derived from the entire genome of an organism. We review here different SELEX-based methods, with particular emphasis on the experimental design and on the applications aimed at inferring protein-DNA interactions. In addition to the experimental issues, we also review relevant methods of data analysis, as well as theoretical modeling of SELEX.     

Viral Genomic Single-Stranded RNA Directs the Pathway Toward a T = 3 Capsid

The molecular mechanisms controlling genome packaging by single-stranded RNA viruses are still largely unknown. It is necessary in most cases for the protein to adopt different conformations at different positions on the capsid lattice in order to form a viral capsid from multiple copies of a single protein. We showed previously that such quasi-equivalent conformers of RNA bacteriophage MS2 coat protein dimers (CP₂) can be switched by sequence-specific interaction with a short RNA stem-loop (TR) that occurs only once in the wild-type phage genome. In principle, multiple switching events are required to generate the phage T = 3 capsid. We have therefore investigated the sequence dependency of this event using two RNA aptamer sequences selected to bind the phage coat protein and an analogous packaging signal from phage Qβ known to be discriminated against by MS2 coat protein both in vivo and in vitro. All three non-cognate stem-loops support T = 3 shell formation, but none shows the kinetic-trapping effect seen when TR is mixed with equimolar CP₂. We show that this reflects the fact that they are poor ligands compared with TR, failing to saturate the coat protein under the assay conditions, ensuring that sufficient amounts of both types of dimer required for efficient assembly are present in these reactions. Increasing the non-cognate RNA concentration restores the kinetic trap, confirming this interpretation. We have also assessed the effects of extending the TR stem-loop at the 5′ or 3′ end with short genomic sequences. These longer RNAs all show evidence of the kinetic trap, reflecting the fact that they all contain the TR sequence and are more efficient at promoting capsid formation than TR. Mass spectrometry has shown that at least two pathways toward the T = 3 shell occur in TR-induced assembly reactions: one via formation of a 3-fold axis and another that creates an extended 5-fold complex. The longer genomic RNAs suppress the 5-fold pathway, presumably as a consequence of steric clashes between multiply bound RNAs. Reversing the orientation of the extension sequences with respect to the TR stem-loop produces RNAs that are poor assembly initiators. The data support the idea that RNA-induced protein conformer switching occurs throughout assembly of the T = 3 shell and show that both positional and sequence-specific effects outside the TR stem-loop can have significant impacts on the precise assembly pathway followed.     

Cistron Specific Translation Control Protein in <Emphasis Type="Italic">Escherichia coli</Emphasis>

GENE expression may be controlled during translation by ribosomal selection of mRNAs or even individual cistrons. Escherichia coli initiation factors associated with ribosomes affect the binding of ribosomes to mRNA^1,2; initiation factor IF3, for instance, influences the specificity of mRNA-ribosome interaction^3,4. IF3 activity has been separated into several fractions which show various specificities for different mRNA cistrons^4–9. An important problem is the possibility of intracellular changes in IF3 activity^10–12. From uninfected E. coli, we have now isolated a protein which changes the specificity of IF3 toward different mRNAs; we call this interference factor i. Pure factor i binds to IF3 and specifically affects the translation of T4 and MS2 RNA. Whereas the initiation of translation of the MS2 coat protein cistron is inhibited by factor i, the synthetase cistron—when available—is more rapidly initiated in the presence of factor i. The overall translation of T4 mRNA appears unchanged by factor i, but certain cistrons are stimulated at the expense of others. Interfering factors such as factor i could be important in controlling translation in E. coli.     

The wacA gene of Dictyostelium discoideum is a developmentally regulated member of the MIP family

We isolated a cDNA from Dictyostelium discoideum that encodes a 30 kDa protein with significant similarity to members of the major intrinsic protein (MIP) family of membrane transporters. The most closely related protein in the public data bases is an aquaporin from Cicadella viridis which shows 34% identity. The cDNA was used to isolate and characterize genomic fragments carrying the Dictyostelium gene which we named wacA. Genomic probes were used to recognize wacA mRNA isolated at various stages of development. The results showed that the gene is developmentally regulated such that the mRNA first appears at 12 h of development and is retained throughout the remainder of development. In situ hybridization of whole mounts prepared at 15 h of development showed that wacA mRNA accumulates exclusively in prespore cells and is absent from prestalk cells. Although wacA expression is prespore specific, disruption of the gene by homologous recombination did not result in observable alterations in the formation of spores or their resistance to osmotic challenges.     

A polymerase chain reaction (PCR) application for free‐living nematodes (Rhabditida) discrimination

Synthetic primers were designed and constructed based on Caenorhabditis elegans genomic sequences, targeting HSP70 related sequences. Genomic DNA derived from several free‐living bacterial feeding nematode species was subjected to polymerase chain reaction (PCR), attempting discrimination among them. The amplified DNA fragments exhibited a distinct and reproducible pattern that characterizes different nematode species and populations.     

Crystal Structure of the Bacteriophage Qβ Coat Protein in Complex with the RNA Operator of the Replicase Gene

The coat proteins of single-stranded RNA bacteriophages specifically recognize and bind to a hairpin structure in their genome at the beginning of the replicase gene. The interaction serves to repress the synthesis of the replicase enzyme late in infection and contributes to the specific encapsidation of phage RNA. While this mechanism is conserved throughout the Leviviridae family, the coat protein and operator sequences from different phages show remarkable variation, serving as prime examples for the co-evolution of protein and RNA structure. To better understand the protein–RNA interactions in this virus family, we have determined the three-dimensional structure of the coat protein from bacteriophage Qβ bound to its cognate translational operator. The RNA binding mode of Qβ coat protein shares several features with that of the widely studied phage MS2, but only one nucleotide base in the hairpin loop makes sequence-specific contacts with the protein. Unlike in other RNA phages, the Qβ coat protein does not utilize an adenine-recognition pocket for binding a bulged adenine base in the hairpin stem but instead uses a stacking interaction with a tyrosine side chain to accommodate the base. The extended loop between β strands E and F of Qβ coat protein makes contacts with the lower part of the RNA stem, explaining the greater length dependence of the RNA helix for optimal binding to the protein. Consequently, the complex structure allows the proposal of a mechanism by which the Qβ coat protein recognizes and discriminates in favor of its cognate RNA.