Search

Finding structural similarities between proteins often helps reveal shared functionality, which otherwise might not be detected by native sequence information alone. Such similarity is usually detected and quantified by protein structure alignment. Determining the optimal alignment between two protein structures, however, remains a hard problem. An alternative approach is to approximate each three-dimensional protein structure using a sequence of motifs derived from a structural alphabet. Using this approach, structure comparison is performed by comparing the corresponding motif sequences or structural sequences. In this article, we measure the performance of such alphabets in the context of the protein structure classification problem. We consider both local and global structural sequences. Each letter of a local structural sequence corresponds to the best matching fragment to the corresponding local segment of the protein structure. The global structural sequence is designed to generate the best possible complete chain that matches the full protein structure. We use an alphabet of 20 letters, corresponding to a library of 20 motifs or protein fragments having four residues. We show that the global structural sequences approximate well the native structures of proteins, with an average coordinate root mean square of 0.69 A over 2225 test proteins. The approximation is best for all alpha-proteins, while relatively poorer for all beta-proteins. We then test the performance of four different sequence representations of proteins (their native sequence, the sequence of their secondary-structure elements, and the local and global structural sequences based on our fragment library) with different classifiers in their ability to classify proteins that belong to five distinct folds of CATH. Without surprise, the primary sequence alone performs poorly as a structure classifier. We show that addition of either secondary-structure information or local information from the structural sequence considerably improves the classification accuracy. The two fragment-based sequences perform better than the secondary-structure sequence but not well enough at this stage to be a viable alternative to more computationally intensive methods based on protein structure alignment.

During translational initiation in prokaryotes, the 3' end of the 16S rRNA binds to a region just upstream of the initiation codon. The relationship between this Shine-Dalgarno (SD) region and the binding of ribosomes to translation start-points has been well studied, but a unified mathematical connection between the SD, the initiation codon and the spacing between them has been lacking. Using information theory, we constructed a model that treats these three components uniformly by assigning to the SD and the initiation region (IR) conservations in bits of information, and by assigning to the spacing an uncertainty, also in bits. To build the model, we first aligned the SD region by maximizing the information content there. The ease of this process confirmed the existence of the SD pattern within a set of 4122 reviewed and revised Escherichia coli gene starts. This large data set allowed us to show graphically, by sequence logos, that the spacing between the SD and the initiation region affects both the SD site conservation and its pattern. We used the aligned SD, the spacing, and the initiation region to model ribosome binding and to identify gene starts that do not conform to the ribosome binding site model. A total of 569 experimentally proven starts are more conserved (have higher information content) than the full set of revised starts, which probably reflects an experimental bias against the detection of gene products that have inefficient ribosome binding sites. Models were refined cyclically by removing non-conforming weak sites. After this procedure, models derived from either the original or the revised gene start annotation were similar. Therefore, this information theory-based technique provides a method for easily constructing biologically sensible ribosome binding site models. Such models should be useful for refining gene-start predictions of any sequenced bacterial genome.

The information required to generate a protein structure is contained in its amino acid sequence, but how three-dimensional information is mapped onto a linear sequence is still incompletely understood. Multiple structure alignments of similar protein structures have been used to investigate conserved sequence features but contradictory results have been obtained, due, in large part, to the absence of subjective criteria to be used in the construction of sequence profiles and in the quantitative comparison of alignment results. Here, we report a new procedure for multiple structure alignment and use it to construct structure-based sequence profiles for similar proteins. The definition of "similar" is based on the structural alignment procedure and on the protein structural distance (PSD) described in paper I of this series, which offers an objective measure for protein structure relationships. Our approach is tested in two well-studied groups of proteins; serine proteases and Ig-like proteins. It is demonstrated that the quality of a sequence profile generated by a multiple structure alignment is quite sensitive to the PSD used as a threshold for the inclusion of proteins in the alignment. Specifically, if the proteins included in the aligned set are too distant in structure from one another, there will be a dilution of information and patterns that are relevant to a subset of the proteins are likely to be lost. In order to understand better how the same three-dimensional information can be encoded in seemingly unrelated sequences, structure-based sequence profiles are constructed for subsets of proteins belonging to nine superfolds. We identify patterns of relatively conserved residues in each subset of proteins. It is demonstrated that the most conserved residues are generally located in the regions where tertiary interactions occur and that are relatively conserved in structure. Nevertheless, the conservation patterns are relatively weak in all cases studied, indicating that structure-determining factors that do not require a particular sequential arrangement of amino acids, such as secondary structure propensities and hydrophobic interactions, are important in encoding protein fold information. In general, we find that similar structures can fold without having a set of highly conserved residue clusters or a well-conserved sequence profile; indeed, in some cases there is no apparent conservation pattern common to structures with the same fold. Thus, when a group of proteins exhibits a common and well-defined sequence pattern, it is more likely that these sequences have a close evolutionary relationship rather than the similarities having arisen from the structural requirements of a given fold.

It is generally accepted that many different protein sequences have similar folded structures, and that there is a relatively high probability that a new sequence possesses a previously observed fold. An indirect consequence of this is that protein design should define the sequence space accessible to a given structure, rather than providing a single optimized sequence. We have recently developed a new approach for protein sequence design, which optimizes the complete sequence of a protein based on the knowledge of its backbone structure, its amino acid composition and a physical energy function including van der Waals interactions, electrostatics, and environment free energy. The specificity of the designed sequence for its template backbone is imposed by keeping the amino acid composition fixed. Here, we show that our procedure converges in sequence space, albeit not to the native sequence of the protein. We observe that while polar residues are well conserved in our designed sequences, non-polar amino acids at the surface of a protein are often replaced by polar residues. The designed sequences provide a multiple alignment of sequences that all adopt the same three-dimensional fold. This alignment is used to derive a profile matrix for chicken triose phosphate isomerase, TIM. The matrix is found to recognize significantly the native sequence for TIM, as well as closely related sequences. Possible application of this approach to protein fold recognition is discussed.

We describe a completely automated approach to identifying local sequence motifs that transcend protein family boundaries. Cluster analysis is used to identify recurring patterns of variation at single positions and in short segments of contiguous positions in multiple sequence alignments for a non-redundant set of protein families. Parallel experiments on simulated data sets constructed with the overall residue frequencies of proteins but not the inter-residue correlations show that naturally occurring protein sequences are significantly more clustered than the corresponding random sequences for window lengths ranging from one to 13 contiguous positions. The patterns of variation at single positions are not in general surprising: chemically similar amino acids tend to be grouped together. More interesting patterns emerge as the window length increases. The patterns of variation for longer window lengths are in part recognizable patterns of hydrophobic and hydrophilic residues, and in part less obvious combinations. A particularly interesting class of patterns features highly conserved glycine residues. The patterns provide a means to abstract the information contained in multiple sequence alignments and may be useful for comparison of distantly related sequences or sequence families and for protein structure prediction.

The replication initiator protein RepA of plasmid P1 can bind to 14 sites on the plasmid. These sites are variously used to autoregulate RepA synthesis and for initiation and control of DNA replication. Analysis of information (degree of conservation) at the sites revealed three sequence patches of high conservation. By saturation mutagenesis, the conservation at the outer two patches was found to contribute to RepA binding more critically. The guanine bases that are likely to contact RepA through the major groove were identified by methylation interference and methylation protection experiments. These bases mapped to the outer two patches and were separated by one turn of the helix. Therefore, they belong to major grooves on the same face of DNA. All backbone contacts of the protein, determined by hydroxyl radical footprinting, also mapped to the same face. We conclude from this that RepA binds to its site on one face of the DNA. Information analysis of binding sites for several prokaryotic repressors and activators, where the nature of DNA—protein contacts are known, revealed a correlation between the positions of high conservation and the positions of major grooves that faced the protein. The middle patch of high conservation in the RepA binding sites is an exception since in this region a minor groove is likely to face the protein. The simplest model for minor groove contacts suggests that in B-form DNA a T · A base-pair cannot easily be distinguished from an A · T pair by inspection of the minor groove. Yet in the RepA site, a T → A mutation in the middle patch significantly affects binding. Therefore, the simplest models for both minor and major groove contacts are unlikely. It is possible that the patch determines the proper conformation of the site and thereby contributes to recognition indirectly.

Calculation of dot-matrices is a widespread tool in the search for sequence similarities. When sequences are distant, even this approach may fail to point out common regions. If several plots calculated for all members of a sequence set consistently displayed a similarity between them, this would increase its credibility. We present an algorithm to delineate dot-plot agreement. A novel procedure based on matrix multiplication is developed to identify common patterns and reliably aligned regions in a set of distantly related sequences. The algorithm finds motifs independent of input sequence lengths and reduces the dependence on gap penalties. When sequences share greater similarity, the same approach converts to a multiple sequence alignment procedure.

Nucleic acid polymers selected from random sequence space constitute an enormous array of catalytic, diagnostic and therapeutic molecules. Despite the fact that proteins are robust polymers with far greater chemical and physical diversity, success in unlocking protein sequence space remains elusive. We have devised a combinatorial strategy for accessing nucleic acid sequence space corresponding to proteins comprising selected amino acid alphabets. Using the SynthOMIC approach (synthesis of ORFs by multimerizing in-frame codons), representative libraries comprising four amino acid alphabets were fused in-frame to the lambda repressor DNA-binding domain to provide an in vivo selection for self-interacting proteins that re-constitute lambda repressor function. The frequency of self-interactors as a function of amino acid composition ranged over five orders of magnitude, from approximately 6% of clones in a library comprising the amino acid residues LARE to approximately 0.6 in 10(6) in the MASH library. Sequence motifs were evident by inspection in many cases, and individual clones from each library presented substantial sequence identity with translated proteins by BLAST analysis. We posit that the SynthOMIC approach represents a powerful strategy for creating combinatorial libraries of open reading frames that distils protein sequence space on the basis of three inherent properties: it supports the use of selected amino acid alphabets, eliminates redundant sequences and locally constrains amino acids.

A 2kb DNA region of the broad-host-range plasmid pCU1 carries all of the information essential for the stable maintenance of the plasmid and to express the same host-range specificity. It was predicted that the protein required to initiate replication from at least one of the three origins of the plasmid is encoded by the longest open-reading frame (ORF239) of the three overlapping in-frame ORFs located within the 2 kb region. The product of ORF239 has been named RepA. The initiator protein was overexpressed, purified and used for in vitro binding studies. Gel mobility shift experiments were performed to localize RepA binding sites. The DNA sequence protected by the bound RepA molecule(s) was determined by DNase I footprinting and 19 of a 20 bp long sequence that is part of the protected sequence were located in two clusters flanking the repA gene. A plasmid created by linking a 310 bp fragment (nucleotides 238 to 547) of the 2 kb region to the antibiotic resistance genes carried by the omega fragment, can be maintained stably if the RepA protein is supplied in trans. We conclude that this 310 bp DNA fragment, which consists of a short G+C and a long A+T rich region and the cluster of five RepA binding sites, carries a functional origin of the plasmid-protein dependent replication. The position of this origin indicates that it is oriB, one of the three origins previously identified by electron microscopy. The second cluster of RepA binding sites is downstream of the repA gene and consists of 14 sites that are in inverted orientation compared with the binding sites located in the oriB region. They are part of the region that was shown formerly to be involved in controlling the copy number of the plasmid. In contrast to oriB, binding of RepA to neither the oriS nor oriV region was detected.

A highly repetitive long interspersed sequence from rat DNA has been isolated and partly characterized. This sequence comprises at least a 1300 base-pair and a 2400 base-pair EcoRI fragment and probably additional elements. The 2400 base-pair segment has been analyzed in detail. It appears to be part of the chromosomal DNA in rat cells. The 2400 base-pair repeat is likely to be distributed over several regions in the rat genome. The 2400 base-pair segment has been cloned, mapped for restriction sites, and part of its nucleotide sequence has been determined. The 2400 base-pair sequence is a member of a typical highly repetitive long interspersed sequence with high copy number and restriction site polymorphism. There are sequence homologies to mouse and human DNA. A striking homology has been detected to the flanking sequences of a repetitive mouse DNA sequence that has been described to be located adjacent to one of the kappa-immunoglobulin variable genes. Elements in the 2400 base-pair rat repeat are transcribed in cells from most rat organs and from several continuous rat cell lines. This RNA from rat cell lines was found polyadenylated or not polyadenylated. The nucleotide sequence of parts of the 2400 base-pair DNA segment revealed open reading frames for polypeptide sequences. Such open reading frames have been detected in two different segments of the 2400 base-pair DNA repeat. Open reading frames exist in the two complementary strands in the same DNA segment. The hypothetical polypeptide whose sequence has been determined in toto has a length of 190 amino acid residues and is enriched in hydrophobic amino acids, reminiscent of the amino acid composition in membrane proteins. Hence, it is conceivable that the 2400 base-pair repeat sequence from rat DNA, at least in part, encodes messenger RNAs that might be translated into functional proteins.

To evaluate the importance of the surrounding nucleotide sequence in the selection of a splice site for mRNA, we have carried out computer studies of eukaryotic protein genes whose entire nucleotide sequences were available. A splice site-like sequence that has a significant homology to the consensus splice junction sequences is frequently found within an intron and exon. It is found that the higher the homology of a candidate donor site sequence to the nine-nucleotide consensus sequence, the higher is its probability of being a donor site. For most of the donors, the stability of presumed base-pairing with U1-RNA is higher than that of donor-like sequences, if any, in the adjacent exon and intron. However, homology of a candidate acceptor sequence to the 15-nucleotide consensus is a poor criterion of an acceptor site. The presence of a sequence that could serve as a branch-point 18 to 37 nucleotides before an acceptor does not seem to be critical in distinguishing it from an acceptor-like sequence. For genes of human, rat, mouse and chicken, respectively, nucleotide frequencies around splice junctions of many genes have been calculated. They seem to be different at some positions around a donor site from species to species. The acceptors for these vertebrates have longer pyrimidine-rich regions than the previous consensus sequence. The newly derived nucleotide frequencies were used as the standard to calculate the weighted homology score of a candidate splice site sequence in a gene of the four species. This weighted homology score of the 40 to 60-nucleotide intron-exon sequence is a much better criterion of an acceptor. These results suggest that the most important signal in the selection of a splice resides in the surrounding nucleotide sequence. It is also suggested that the surrounding nucleotide sequence alone is not generally sufficient for the selection.

The regions on the adenovirus type 2 genome which encode the three segments of the common tripartite leader of late adenovirus messenger RNA have been identified and sequenced. Each segment of the tripartite leader is specified by a contiguous sequence of DNA. A comparison between the established sequences and the sequence of the 5′ non-coding region of the hexon mRNA (Akusjarvi & Pettersson, 1979) suggests that the first leader segment (map position 16.3) is 42 nucleotides long, including the m 7 G, the second (map position 19.6) 71 nucleotides long and the third segment (map position 26.5) 89 nucleotides long. These estimates are, however, in all cases ambiguous due to the presence of short tandemly repeated sequences at the borders of each individual intervening sequence. These repeats makes it impossible to pin-point the exact position where the splicing event has taken place. A comparison of the six junction regions where leaders are connected to intervening sequences shows that they have similar although not identical sequences. All junctions contain the trinucleotide GGT at the 5′ side and the dinucleotide A-G at the 3′ side of the intervening sequence. Furthermore the established junction sequences exhibit a striking similarity to the corresponding regions of eukaryotic genes, in particular to the sequences which flank the large intervening sequence in the gene for mouse β maj -globin (Konkel et al. , 1978). Two regions in the promoter for late adenovirus transcription at map position 16.3 share homology with the corresponding region of the gene for β maj -globin (Konkel et al. , 1978). These two regions include sequences which precede and overlap the initiation site for late adenovirus transcription and the sequence T-A-T-A-A-A-A which is present 24 nucleotides leftwards of the cap site. This heptanucleotide resembles a sequence found in prokaryotic promoters (Pribnow, 1975).

The nucleotide sequence of the region coding for the F protein of bacteriophage φX174 has now been completed, in conjunction with extended peptide sequence analysis of the gene F protein. The protein is 426 amino acids in length, with a molecular weight, calculated from the sequence, of 48,340.

The nucleotide sequence of the principal bacteriophage φX174 ribosome-protected DNA fragment has been determined. Details of the approaches and methods used for direct DNA sequence analysis are described, The resulting sequence allows prediction of the first nine amino acids at the amino terminus of a protein which has been identified as the product of φX174 gene G . The sequence has several features of biological relevance including two termination codons to the left of the initiator triplet.

Application of three different methods to obtain nucleotide sequences from X174 DNA has yielded a continuous sequence of 281 nucleotides and partial sequences totalling another 263 nucleotides from adjoining regions. Simultaneous work on amino acid sequences of X174 coat proteins showed that these nucleotide sequences are part of the F gene, coding for the largest coat protein of the bacteriophage. A total of 136 codons has been identified, of which 56% have thymine as the base in the third position of the coding triplet.

The nucleotide sequences of eight randomly selected, cloned, repetitive sequence elements were determined. No homologies exist among the eight sequences that are sufficient to promote cross-reaction between them under standard conditions of measurement. Thus, each sequence is representative of a different repeat sequence family. Statistically significant but short (8 to 41 nucleotides) internal direct and inverse repetitions occupy a minor fraction of the sequence length in five of the eight repeat sequences. None contains internal reverse repeats sufficiently long to permit inclusion in the “foldback” DNA fraction. The general lack of internal sequence homology means that the sequence complexity of the eight clones is approximately equal to the length of the cloned inserts. Nucleotide sequences of three different members of one particular short interspersed repeat sequence family are also reported. Comparison of these sequences reveals that both the number and order of internal sequence subelements differ among family members. The results show that both fine-scale rearrangement and sequence divergence have occurred during the evolution of this repeat family.

A band of 300 nucleotide long duplex DNA is released by treating renatured repeated human DNA with the single strand-specific endonuclease S 1 . Since many of the interspersed repeated sequences in human DNA are 300 nucleotides long, this band should be enriched in such repeats. We have determined the nucleotide sequences of 15 clones constructed from these 300 nucleotide S 1 -resistant repeats. Ten of these cloned sequences are members of the Alu family of interspersed repeats. These ten sequences share a recognizable consensus sequence from which individual clones have an average divergence of 12.8%. The 300 nucleotide Alu family consensus sequence has a dimeric structure and was evidently formed from a head to tail duplication of an ancestral monomeric sequence. Three of the remaining clones are variations on a simple pentanucleotide sequence previously reported for human satellite III DNA. Two of the 15 clones have distinct and complex sequences and may represent other families of interspersed repeated sequences.