The appearance of new structures and functions in proteins during evolution

Sequence, Structure, and Functional Space of Drosophila De Novo Proteins

During de novo emergence, new protein coding genes emerge from previously nongenic sequences. The de novo proteins they encode are dissimilar in composition and predicted biochemical properties to conserved proteins. However, functional de novo proteins indeed exist. Both identification of functional de novo proteins and their structural characterization are experimentally laborious. To identify functional and structured de novo proteins in silico, we applied recently developed machine learning based tools and found that most de novo proteins are indeed different from conserved proteins both in their structure and sequence. However, some de novo proteins are predicted to adopt known protein folds, participate in cellular reactions, and to form biomolecular condensates. Apart from broadening our understanding of de novo protein evolution, our study also provides a large set of testable hypotheses for focused experimental studies on structure and function of de novo proteins in Drosophila.

Random, de novo, and conserved proteins: How structure and disorder predictors perform differently

Understanding the emergence and structural characteristics of de novo and random proteins is crucial for unraveling protein evolution and designing novel enzymes. However, experimental determination of their structures remains challenging. Recent advancements in protein structure prediction, particularly with AlphaFold2 (AF2), have expanded our knowledge of protein structures, but their applicability to de novo and random proteins is unclear. In this study, we investigate the structural predictions and confidence scores of AF2 and protein language model-based predictor ESMFold for de novo and conserved proteins from Drosophila and a dataset of comparable random proteins. We find that the structural predictions for de novo and random proteins differ significantly from conserved proteins. Interestingly, a positive correlation between disorder and confidence scores (pLDDT) is observed for de novo and random proteins, in contrast to the negative correlation observed for conserved proteins. Furthermore, the performance of structure predictors for de novo and random proteins is hampered by the lack of sequence identity. We also observe fluctuating median predicted disorder among different sequence length quartiles for random proteins, suggesting an influence of sequence length on disorder predictions. In conclusion, while structure predictors provide initial insights into the structural composition of de novo and random proteins, their accuracy and applicability to such proteins remain limited. Experimental determination of their structures is necessary for a comprehensive understanding. The positive correlation between disorder and pLDDT could imply a potential for conditional folding and transient binding interactions of de novo and random proteins.

Assessing structure and disorder prediction tools for de novo emerged proteins in the age of machine learning

Background: De novo protein coding genes emerge from scratch in the non-coding regions of the genome and have, per definition, no homology to other genes. Therefore, their encoded de novo proteins belong to the so-called "dark protein space". So far, only four de novo protein structures have been experimentally approximated. Low homology, presumed high disorder and limited structures result in low confidence structural predictions for de novo proteins in most cases. Here, we look at the most widely used structure and disorder predictors and assess their applicability for de novo emerged proteins. Since AlphaFold2 is based on the generation of multiple sequence alignments and was trained on solved structures of largely conserved and globular proteins, its performance on de novo proteins remains unknown. More recently, natural language models of proteins have been used for alignment-free structure predictions, potentially making them more suitable for de novo proteins than AlphaFold2. Methods: We applied different disorder predictors (IUPred3 short/long, flDPnn) and structure predictors, AlphaFold2 on the one hand and language-based models (Omegafold, ESMfold, RGN2) on the other hand, to four de novo proteins with experimental evidence on structure. We compared the resulting predictions between the different predictors as well as to the existing experimental evidence. Results: Results from IUPred, the most widely used disorder predictor, depend heavily on the choice of parameters and differ significantly from flDPnn which has been found to outperform most other predictors in a comparative assessment study recently. Similarly, different structure predictors yielded varying results and confidence scores for de novo proteins. Conclusions: We suggest that, while in some cases protein language model based approaches might be more accurate than AlphaFold2, the structure prediction of de novo emerged proteins remains a difficult task for any predictor, be it disorder or structure.

DNA methylation in stress and depression: from biomarker to therapeutics

Epigenetic mechanisms such as DNA methylation (DNAm) have been associated with stress responses and increased vulnerability to depression. Abnormal DNAm is observed in stressed animals and depressed individuals. Antidepressant treatment modulates DNAm levels and regulates gene expression in diverse tissues, including the brain and the blood. Therefore, DNAm could be a potential therapeutic target in depression. Here, we reviewed the current knowledge about the involvement of DNAm in the behavioural and molecular changes associated with stress exposure and depression. We also evaluated the possible use of DNAm changes as biomarkers of depression. Finally, we discussed current knowledge limitations and future perspectives.

Readthrough Errors Purge Deleterious Cryptic Sequences, Facilitating the Birth of Coding Sequences

De novo protein-coding innovations sometimes emerge from ancestrally noncoding DNA, despite the expectation that translating random sequences is overwhelmingly likely to be deleterious. The "preadapting selection" hypothesis claims that emergence is facilitated by prior, low-level translation of noncoding sequences via molecular errors. It predicts that selection on polypeptides translated only in error is strong enough to matter and is strongest when erroneous expression is high. To test this hypothesis, we examined noncoding sequences located downstream of stop codons (i.e., those potentially translated by readthrough errors) in Saccharomyces cerevisiae genes. We identified a class of "fragile" proteins under strong selection to reduce readthrough, which are unlikely substrates for co-option. Among the remainder, sequences showing evidence of readthrough translation, as assessed by ribosome profiling, encoded C-terminal extensions with higher intrinsic structural disorder, supporting the preadapting selection hypothesis. The cryptic sequences beyond the stop codon, rather than spillover effects from the regular C-termini, are primarily responsible for the higher disorder. Results are robust to controlling for the fact that stronger selection also reduces the length of C-terminal extensions. These findings indicate that selection acts on 3' UTRs in Saccharomyces cerevisiae to purge potentially deleterious variants of cryptic polypeptides, acting more strongly in genes that experience more readthrough errors.

Universal and taxon-specific trends in protein sequences as a function of age

Extant protein-coding sequences span a huge range of ages, from those that emerged only recently to those present in the last universal common ancestor. Because evolution has had less time to act on young sequences, there might be 'phylostratigraphy' trends in any properties that evolve slowly with age. A long-term reduction in hydrophobicity and hydrophobic clustering was found in previous, taxonomically restricted studies. Here we perform integrated phylostratigraphy across 435 fully sequenced species, using sensitive HMM methods to detect protein domain homology. We find that the reduction in hydrophobic clustering is universal across lineages. However, only young animal domains have a tendency to have higher structural disorder. Among ancient domains, trends in amino acid composition reflect the order of recruitment into the genetic code, suggesting that the composition of the contemporary descendants of ancient sequences reflects amino acid availability during the earliest stages of life, when these sequences first emerged.

Foldability of a Natural De Novo Evolved Protein

The de novo evolution of protein-coding genes from noncoding DNA is emerging as a source of molecular innovation in biology. Studies of random sequence libraries, however, suggest that young de novo proteins will not fold into compact, specific structures typical of native globular proteins. Here we show that Bsc4, a functional, natural de novo protein encoded by a gene that evolved recently from noncoding DNA in the yeast S. cerevisiae, folds to a partially specific three-dimensional structure. Bsc4 forms soluble, compact oligomers with high β sheet content and a hydrophobic core, and undergoes cooperative, reversible denaturation. Bsc4 lacks a specific quaternary state, however, existing instead as a continuous distribution of oligomer sizes, and binds dyes indicative of amyloid oligomers or molten globules. The combination of native-like and non-native-like properties suggests a rudimentary fold that could potentially act as a functional intermediate in the emergence of new folded proteins de novo.

Computing the origin and evolution of the ribosome from its structure - Uncovering processes of macromolecular accretion benefiting synthetic biology

Accretion occurs pervasively in nature at widely different timeframes. The process also manifests in the evolution of macromolecules. Here we review recent computational and structural biology studies of evolutionary accretion that make use of the ideographic (historical, retrodictive) and nomothetic (universal, predictive) scientific frameworks. Computational studies uncover explicit timelines of accretion of structural parts in molecular repertoires and molecules. Phylogenetic trees of protein structural domains and proteomes and their molecular functions were built from a genomic census of millions of encoded proteins and associated terminal Gene Ontology terms. Trees reveal a ‘metabolic-first’ origin of proteins, the late development of translation, and a patchwork distribution of proteins in biological networks mediated by molecular recruitment. Similarly, the natural history of ancient RNA molecules inferred from trees of molecular substructures built from a census of molecular features shows patchwork-like accretion patterns. Ideographic analyses of ribosomal history uncover the early appearance of structures supporting mRNA decoding and tRNA translocation, the coevolution of ribosomal proteins and RNA, and a first evolutionary transition that brings ribosomal subunits together into a processive protein biosynthetic complex. Nomothetic structural biology studies of tertiary interactions and ancient insertions in rRNA complement these findings, once concentric layering assumptions are removed. Patterns of coaxial helical stacking reveal a frustrated dynamics of outward and inward ribosomal growth possibly mediated by structural grafting. The early rise of the ribosomal ‘turnstile’ suggests an evolutionary transition in natural biological computation. Results make explicit the need to understand processes of molecular growth and information transfer of macromolecules.

Ancestral Insertions and Expansions of rRNA do not Support an Origin of the Ribosome in Its Peptidyl Transferase Center

Phylogenetic reconstruction of ribosomal history suggests that the ribonucleoprotein complex originated in structures supporting RNA decoding and ribosomal mechanics. A recent study of accretion of ancestral expansion segments of rRNA, however, contends that the large subunit of the ribosome originated in its peptidyl transferase center (PTC). Here I re-analyze the rRNA insertion data that supports this claim. Analysis of a crucial three-way junction connecting the long-helical coaxial branch that supports the PTC to the L1 stalk and its translocation functions reveals an incorrect branch-to-trunk insertion assignment that is in conflict with the PTC-centered accretion model. Instead, the insertion supports the ancestral origin of translocation. Similarly, an insertion linking a terminal coaxial trunk that holds the L7–12 stalk and its GTPase center to a seven-way junction of the molecule again questions the early origin of the PTC. Unwarranted assumptions, dismissals of conflicting data, structural insertion ambiguities, and lack of phylogenetic information compromise the construction of an unequivocal insertion-based model of macromolecular accretion. Results prompt integration of phylogenetic and structure-based models to address RNA junction growth and evolutionary constraints acting on ribosomal structure.

Unusual hepatic mitochondrial arginase in an Indian air-breathing teleost, Heteropneustes fossilis: purification and characterization

A functional urea cycle with both cytosolic (ARG I) and mitochondrial (ARG II) arginase activity is present in the liver of an ureogenic air-breathing teleost, Heteropneustes fossilis. Antibodies against mammalian ARG II showed no cross-reactivity with the H. fossilis ARG II. ARG II was purified to homogeneity from H. fossilis liver. Purified ARG II showed a native molecular mass of 96 kDa. SDS-PAGE showed a major band at 48 kDa. The native enzyme, therefore, appears to be a homodimer. The pI value of the enzyme was 7.5. The purified enzyme showed maximum activity at pH 10.5 and 55 °C. The K(m) of purified ARG II for l-arginine was 5.25±1.12 mM. L-Ornithine and N(ω)-hydroxy-L-arginine showed mixed inhibition with K(i) values 2.16±0.08 and 0.02±0.004 mM respectively. Mn(+2) and Co(+2) were effective activators of arginase activity. Antibody raised against purified H. fossilis ARG II did not cross-react with fish ARG I, and mammalian ARG I and ARG II. Western blot with the antibodies against purified H. fossilis hepatic ARG II showed cross reactivity with a 96 kDa band on native PAGE and a 48 kDa band on SDS-PAGE. The molecular, immunological and kinetic properties suggest uniqueness of the hepatic mitochondrial ARG II in H. fossilis.

Metabolic basis for the self-referential genetic code

An investigation of the biosynthesis pathways producing glycine and serine was necessary to clarify an apparent inconsistency between the self-referential model (SRM) for the formation of the genetic code and the model of coevolution of encodings and of amino acid biosynthesis routes. According to the SRM proposal, glycine was the first amino acid encoded, followed by serine. The coevolution model does not state precisely which the first encodings were, only presenting a list of about ten early assignments including the derivation of glycine from serine-this being derived from the glycolysis intermediate glycerate, which reverses the order proposed by the self-referential model. Our search identified the glycine-serine pathway of syntheses based on one-carbon sources, involving activities of the glycine decarboxylase complex and its associated serine hydroxymethyltransferase, which is consistent with the order proposed by the self-referential model and supports its rationale for the origin of the genetic code: protein synthesis was developed inside an early metabolic system, serving the function of a sink of amino acids; the first peptides were glycine-rich and fit for the function of building the early ribonucleoproteins; glycine consumption in proteins drove the fixation of the glycine-serine pathway.

The action of key factors in protein evolution at high temporal resolution

Background

Protein evolution is particularly shaped by the conservation of the amino acids' physico-chemical properties and the structure of the genetic code. While conservation is the result of negative selection against proteins with reduced functionality, the codon sequences determine the stochastic aspect of amino acid exchanges. Thus far, it is known that the genetic code is the dominant factor if little time has elapsed since the divergence of one gene into two, but physico-chemical forces gain importance at greater evolutionary distances. Further details, however, on how the influence of these factors varies with time are unknown to date.

Methodology/Principal Findings

Here, we derive each 10,000 divergence specific substitution matrices for orthologues and paralogues from the Pfam collection of multiple protein alignments and quantify the action of three physico-chemical forces and of the structure of the genetic code at high resolution using correlation analysis. For closely related proteins, the codon sequence similarity is the most influential factor controlling protein evolution, but its influence decreases rapidly as divergence grows. From a protein sequence divergence of about 20 percent on the maintenance of the hydrophobic character of an amino acid is the most influential factor. All factors lose importance from about 40 percent divergence on. This suggests that the original protein structure often does no longer represent a constraint to the protein sequence. The proteins then become free to adopt new functions. We furthermore show that the constraints exerted by both physico-chemical forces and by the genetic code are quite comparable for orthologues and paralogues, however somewhat weaker for paralogues than for orthologues in weakly or moderately diverged proteins.

Conclusion/Significance

Our analysis substantiates earlier findings that protein evolution is mainly governed by the structure of the genetic code in the early phase after divergence and by the conservation of physico-chemical properties at the later phase. We determine the level of sequence divergence from which on the conservation of the hydrophobic character is gaining importance over the genetic code to be 20 percent. The evolution of orthologues and paralogues is shaped by evolutionary forces in quite comparable ways.

The neoselectionist theory of genome evolution

The vertebrate genome is a mosaic of GC-poor and GC-rich isochores, megabase-sized DNA regions of fairly homogeneous base composition that differ in relative amount, gene density, gene expression, replication timing, and recombination frequency. At the emergence of warm-blooded vertebrates, the gene-rich, moderately GC-rich isochores of the cold-blooded ancestors underwent a GC increase. This increase was similar in mammals and birds and was maintained during the evolution of mammalian and avian orders. Neither the GC increase nor its conservation can be accounted for by the random fixation of neutral or nearly neutral single-nucleotide changes (i.e., the vast majority of nucleotide substitutions) or by a biased gene conversion process occurring at random genome locations. Both phenomena can be explained, however, by the neoselectionist theory of genome evolution that is presented here. This theory fully accepts Ohta's nearly neutral view of point mutations but proposes in addition (i) that the AT-biased mutational input present in vertebrates pushes some DNA regions below a certain GC threshold; (ii) that these lower GC levels cause regional changes in chromatin structure that lead to deleterious effects on replication and transcription; and (iii) that the carriers of these changes undergo negative (purifying) selection, the final result being a compositional conservation of the original isochore pattern in the surviving population. Negative selection may also largely explain the GC increase accompanying the emergence of warm-blooded vertebrates. In conclusion, the neoselectionist theory not only provides a solution to the neutralist/selectionist debate but also introduces an epigenomic component in genome evolution.

An analysis of the Sargasso Sea resource and the consequences for database composition

Background

The environmental sequencing of the Sargasso Sea has introduced a huge new resource of genomic information. Unlike the protein sequences held in the current searchable databases, the Sargasso Sea sequences originate from a single marine environment and have been sequenced from species that are not easily obtainable by laboratory cultivation. The resource also contains very many fragments of whole protein sequences, a side effect of the shotgun sequencing method.

These sequences form a significant addendum to the current searchable databases but also present us with some intrinsic difficulties. While it is important to know whether it is possible to assign function to these sequences with the current methods and whether they will increase our capacity to explore sequence space, it is also interesting to know how current bioinformatics techniques will deal with the new sequences in the resource.

Results

The Sargasso Sea sequences seem to introduce a bias that decreases the potential of current methods to propose structure and function for new proteins. In particular the high proportion of sequence fragments in the resource seems to result in poor quality multiple alignments.

Conclusion

These observations suggest that the new sequences should be used with care, especially if the information is to be used in large scale analyses. On a positive note, the results may just spark improvements in computational and experimental methods to take into account the fragments generated by environmental sequencing techniques.

The structure of the protein universe and genome evolution

Despite the practically unlimited number of possible protein sequences, the number of basic shapes in which proteins fold seems not only to be finite, but also to be relatively small, with probably no more than 10,000 folds in existence. Moreover, the distribution of proteins among these folds is highly non-homogeneous -- some folds and superfamilies are extremely abundant, but most are rare. Protein folds and families encoded in diverse genomes show similar size distributions with notable mathematical properties, which also extend to the number of connections between domains in multidomain proteins. All these distributions follow asymptotic power laws, such as have been identified in a wide variety of biological and physical systems, and which are typically associated with scale-free networks. These findings suggest that genome evolution is driven by extremely general mechanisms based on the preferential attachment principle.

Evolution of DNA binding motifs and operators

A gene regulatory protein with helix-turn-helix (HTH) DNA-binding motif, GalS contains a functional operator within the DNA sequences encoding the HTH region (Nature 369 (1994) 314). We searched for operator-like sequences within the DNA sequences encoding the DNA binding motifs of other regulatory proteins. Five such proteins, DeoR, CytR, LRP, LuxR and PurR, were found to have actual operator or operator-like sequences in the DNA sequences encoding the DNA-binding motif. Except DeoR, all of them including GalS, are known to be auto-regulated. Auto-regulation in case of DeoR has not been investigated. Seven other proteins containing a HTH motif, do not have operator-like sequences in the DNA sequences encoding the HTH motif; none of them, except MerR, are known to be auto-regulated. The DNA binding proteins may have evolved from a common ancestor containing a DNA binding site within its gene segment that encodes the DNA-binding motif to facilitate auto-regulation. We have discussed current evidence for monophyletic or polyphyletic origin of such sequences.

A unifold, mesofold, and superfold model of protein fold use

As more and more protein structures are determined, there is increasing interest in the question of how many different folds have been used in biology. The history of the rate of discovery of new folds and the distribution of sequence families among known folds provide a means of estimating the underlying distribution of fold use. Previous models exploiting these data have led to rather different conclusions on the total number of folds. We present a new model, based on the notion that the folds used in biology fall naturally into three classes: unifolds, that is, folds found only in a single narrow sequence family; mesofolds, found in an intermediate number of families; and the previously noted superfolds, found in many protein families. We show that this model fits the available data well and has predicted the development of SCOP over the past 2 years. The principle implications of the model are as follows: (1) The vast majority of folds will be found in only a single sequence family; (2) the total number of folds is at least 10,000; and (3) 80% of sequence families have one of about 400 folds, most of which are already known.

Multifunctional enzymes and evolution of biosynthetic pathways: retro-evolution by jumps

A likely scenario of evolution of biosynthetic pathways is believed to have occurred by retro-evolution through recruitment of existing enzymes rather than generation of de novo classes. It had been proposed that such retro-evolution occurred in steps as a response to depletion of an essential metabolite and availability of another related substance in the environment. In this article, I argue that because of instability of many such extant intermediates, it is unlikely that retro-evolution had occurred in steps. I further propose that such evolution in many cases has taken place by jumps, i.e., by recruitment of a multifunctional enzyme capable of catalyzing several steps at a time, albeit inefficiently. I further speculate that in some cases one primordial multienzyme may have catalyzed the whole sequence of reaction of a biosynthetic pathway, i.e., the pathway may have evolved by a single leap. Gene duplications and further evolution to more efficient enzymes led to extant pathways. Such a mechanism predicts that some or all enzymes of a pathway must have descended from a common ancestor. Sequence and structural homologies among extant enzymes of a biosynthetic pathway have been examined.

ATP binding site of P2X channel proteins: structural similarities with class II aminoacyl-tRNA synthetases

The extracellular loop of P2X channel proteins contains a sequence stretch (positions 170-330) that exhibits similarities with the catalytic domains of class II aminoacyl-tRNA synthetases as shown by secondary structure predictions and sequence alignments. The arrangement of several conserved cysteines (positions 110-170) shows similarities with metal binding regions of metallothioneins and zinc finger motifs. Thus, for the extracellular part of P2X channel proteins a metal binding domain and an antiparallel six-stranded beta-pleated sheet containing the ATP binding site are very probable. The putative channel forming H5 part (positions 320-340) shows similarities with the enzyme motif 1 responsible for aggregation of subunits to the holoenzyme.

Sequence fossils, triplet expansion, and reconstruction of earliest codons

mRNA sequences are known to carry a hidden periodical pattern (GCU)n, which may be considered a remnant of sequence organization of mRNA early in its evolution, dominated by codons for alanine and their point mutation derivatives. A similar pattern is characteristic of the master (consensus) tRNA sequence derived in 1981 by Eigen and Winkler-Oswatitsch. The master tRNA sequence is thought to represent one of the earliest mRNA. From analysis of literature and from our own calculations presented in this work, the (GCU)n pattern appears to be the most expandable in the norm and in disease. The speculation is put forward that (GCU)n and polyalanine have been key players at the beginning of the triplet code, and the first codons, apart from the GCU triplet, were point change derivatives of the generic triplet GCU, coding for amino acids present in the early prebiotic-biotic environment. The set of the earliest amino acids is derived on the basis of structural simplicity, presence in imitated prebiotic conditions and involvement with class II aminoacyl-tRNA synthetases. The set consists of six amino acids: Ala, Asp, Gly, Pro, Ser and Thr. All these amino acids are, indeed, encoded by the GCU triplet and its derivatives, as predicted. Thus, the pairs GCN (Ala), GAU (Asp), GGU (Gly), CCU (Pro), UCU (Ser) and ACU (Thr) can be viewed as an early triplet code.

Population statistics of protein structures: lessons from structural classifications

Structural classifications aid the interpretation of proteins by describing degrees of structural and evolutionary relatedness. They have also recently revealed strikingly skewed distributions at all levels; for example, a small number of folds are far more common than others, and just a few superfamilies are known to have diverged widely. The classifications also provide an indication of the total number of superfamilies in nature.

Neutral and nonneutral mutations: the creative mix--evolution of complexity in gene interaction systems

Random drift, while indifferent to the functionality of the molecular features on which it acts, may nevertheless affect evolving molecular mechanisms. It can lead to functional novelty in either gene structure or regulation. In particular, a nearly neutral (in the sense of Ohta), somewhat deleterious mutation can result in a loss of efficiency in gene regulation, and this loss is expected at times to be compensated by a selected event of a particular type: the use of an additional regulatory factor. An accumulation of additional regulatory factors, implying a combination of events of drift and selection, can permit regulatory systems to achieve an increase in both specificity and complexity as mere byproducts of a particular repair process. Nearly neutral mutations thus may, at times, constitute a required pathway for increases in gene interaction complexity. The process seems to point to an inbuilt drive-built into the gene interaction system itself-toward the evolution of higher organisms. This is a matter worthy of experimental exploration, since the general foundations for the evolution of "higher" from "lower" organisms seems so far to have largely eluded analysis.

MIPS: a database for protein sequences, homology data and yeast genome information

The MIPS group (Martinsried Institute for Protein Sequences) at the Max-Planck-Institute for Biochemistry, Martinsried near Munich, Germany, collects, processes and distributes protein sequence data within the framework of the tripartite association of the PIR-International Protein Sequence Database (,). MIPS contributes nearly 50% of the data input to the PIR-International Protein Sequence Database. The database is distributed on CD-ROM together with PATCHX, an exhaustive supplement of unique, unverified protein sequences from external sources compiled by MIPS. Through its WWW server (http://www.mips.biochem.mpg.de/ ) MIPS permits internet access to sequence databases, homology data and to yeast genome information. (i) Sequence similarity results from the FASTA program () are stored in the FASTA database for all proteins from PIR-International and PATCHX. The database is dynamically maintained and permits instant access to FASTA results. (ii) Starting with FASTA database queries, proteins have been classified into families and superfamilies (PROT-FAM). (iii) The HPT (hashed position tree) data structure () developed at MIPS is a new approach for rapid sequence and pattern searching. (iv) MIPS provides access to the sequence and annotation of the complete yeast genome (), the functional classification of yeast genes (FunCat) and its graphical display, the 'Genome Browser' (). A CD-ROM based on the JAVA programming language providing dynamic interactive access to the yeast genome and the related protein sequences has been compiled and is available on request.

The coevolution of gene family trees

Gene duplication mutants arise spontaneously at a high rate in bacteria, bacteriophages, insects and mammalian cells, and are generally viable. Thus, the rate-limiting step in the evolutionary process of gene duplication and divergence was probably not gene duplication per se. Rather, it is likely that only a small fraction of all duplicated genes were retained, and were able to diverge into new specificities. Furthermore, gene duplications and functionally related gene families often show similarities in divergence dates, functional specificities, and phylogenetic tree topologies. These correlations suggest that the family trees of functionally related gene families co-evolved because functionally complementary gene duplication and divergence events tended to be retained by natural selection.

Cornering and catching the common protein fold

Protein coding regions (genomic DNA) and cDNA sequences can be screened rapidly to help identify them. If a protein's three-dimensional fold can also be identified, then modelling can provide an estimate of the protein's structure, which can assist both in drug design and in probing structure-function relationships. Methods linking one-dimensional information, represented by the sequence, to three-dimensional information of the protein fold are key to this progression of events. The major difficulty lies with the identification of related proteins whose sequences have diverged beyond recognition, while their folds remain similar enough to confirm their descent from a common ancestor.

Segmented structure of protein sequences and early evolution of genome by combinatorial fusion of DNA elements

A theory of an early stage of genome evolution by combinatorial fusion of circular DNA units is suggested, based on protein sequence "fossil" evidence. The evidence includes preference of protein sequence lengths for certain sizes--multiples of 123 aa for eukaryotes and multiples of 152 aa for prokaryotes. At the DNA level these sizes correspond to 350-450 base pairs--the known optimal range for DNA ring closure. The methionine residues repeatedly appear along the sequences with the same period of about 120 aa (in eukaryotes), presumably marking the sites of insertion of the early genes--rings of protein-coding DNA. No torsional constraint in this DNA results in very sharp estimate of the helical periodicity of the early DNA, indistinguishable from the experimental mean value for extant DNA. According to the combinatorial fusion theory, based on the above evidence, in the pregenomic, prerecombinational stage the genes and the noncoding sequences existed in form of autonomously replicating DNA rings of close to standard size, randomly segregating between dividing cells, like modern plasmids do. In the recombinational early genomic stage the rings started to fuse, forming larger DNA molecules consisting of several unit genes connected in various combinations and forming long protein-coding sequences (combinatorial fusion). This process, which involved, perhaps, noncoding sequences as well, eventually resulted in the formation of large genomes. The dispersed circular DNA--or, rather, evolutionarily advanced derivatives thereof--may still exist in the form of various mobile DNA elements.

Periodic recurrence of methionines: fossil of gene fusion?

As we have recently shown, approximately 20% of proteins are made of uniform size units of approximately 123 aa for eukaryotes and approximately 152 aa for prokaryotes. Such regularity may reflect certain past events in protein evolution by fusion (molecular recombination) of a spectrum of standard-size protein-coding DNA segments--the early genes. Consequently, methionines, as start residues, would mark those locations in proteins that correspond to the DNA recombination sites--the borders between the fused genes. This positional preference of the methionines may still survive as a fossil of the early protein sequence organization. In this study we address the question how methionines are distributed in modern protein sequences. This analysis of eukaryotic sequences shows that methionine residues do preferentially appear at the positions corresponding to the multiples of the unit size, as predicted.

Analysis of peptides from known proteins: clusterization in sequence space

A combinatorial sequence space (CSS) model was introduced to represent sequences as a set of overlapping k-tuples of some fixed length which correspond to points in the CSS. The aim was to analyze clusterization of protein sequences in the CSS and to test various hypotheses about the possible evolutionary basis of this clusterization. The authors developed an easy-to-use technique which can reveal and analyze such a clusterization in a multidimensional CSS. Application of the technique led to an unexpectedly high clusterization of points in the CSS corresponding to k-tuples from known proteins. The clusterization could not be inferred from nonuniform amino acid frequencies or be explained by the influence of homologous data. None of the tested possible evolutionary and structural factors could explain the clusterization observed either. It looked as if certain protein sequence variations occurred and were fixed in the early course of evolution. Subsequent evolution (predominantly neutral) allowed only a limited number of changes and permitted new variants which led to preservation of certain k-tuples during the course of evolution. This was consistent with the theory of exon shuffling and protein block structure evolution. Possible applications of sequence space features found were also discussed.

Molecular pathways to parallel evolution: I. Gene nexuses and their morphological correlates

Aspects of the regulatory interactions among genes are probably as old as most genes are themselves. Correspondingly, similar predispositions to changes in such interactions must have existed for long evolutionary periods. Features of the structure and the evolution of the system of gene regulation furnish the background necessary for a molecular understanding of parallel evolution. Patently "unrelated" organs, such as the fat body of a fly and the liver of a mammal, can exhibit fractional homology, a fraction expected to become subject to quantitation. This also seems to hold for different organs in the same organism, such as wings and legs of a fly. In informational macromolecules, on the other hand, homology is indeed all or none. In the quite different case of organs, analogy is expected usually to represent attenuated homology. Many instances of putative convergence are likely to turn out to be predominantly parallel evolution, presumably including the case of the vertebrate and cephalopod eyes. Homology in morphological features reflects a similarity in networks of active genes. Similar nexuses of active genes can be established in cells of different embryological origins. Thus, parallel development can be considered a counterpart to parallel evolution. Specific macromolecular interactions leading to the regulation of the c-fos gene are given as an example of a "controller node" defined as a regulatory unit. Quantitative changes in gene control are distinguished from relational changes, and frequent parallelism in quantitative changes is noted in Drosophila enzymes. Evolutionary reversions in quantitative gene expression are also expected. The evolution of relational patterns is attributed to several distinct mechanisms, notably the shuffling of protein domains. The growth of such patterns may in part be brought about by a particular process of compensation for "controller gene diseases," a process that would spontaneously tend to lead to increased regulatory and organismal complexity. Despite the inferred increase in gene interaction complexity, whose course over evolutionary time is unknown, the number of homology groups for the functional and structural protein units designated as domains has probably remained rather constant, even as, in some of its branches, evolution moved toward "higher" organisms. In connection with this process, the question is raised of parallel evolution within the purview of activating and repressing master switches and in regard to the number of levels into which the hierarchies of genic master switches will eventually be resolved.

How long did it take for life to begin and evolve to cyanobacteria?

There is convincing paleontological evidence showing that stromatolite-building phototactic prokaryotes were already in existence 3.5 x 10(9) years ago. Late accretion impacts may have killed off life on our planet as late as 3.8 x 10(9) years ago. This leaves only 300 million years to go from the prebiotic soup to the RNA world and to cyanobacteria. However, 300 million years should be more than sufficient time. All known prebiotic reactions take place in geologically rapid time scales, and very slow prebiotic reactions are not feasible because the intermediate compounds would have been destroyed due to the passage of the entire ocean through deep-sea vents every 10(7) years or in even less time. Therefore, it is likely that self-replicating systems capable of undergoing Darwinian evolution emerged in a period shorter than the destruction rates of its components (<5 million years). The time for evolution from the first DNA/protein organisms to cyanobacteria is usually thought to be very long. However, the similarities of many enzymatic reactions, together with the analysis of the available sequence data, suggest that a significant number of the components involved in basic biological processes are the result of ancient gene duplication events. Assuming that the rate of gene duplication of ancient prokaryotes was comparable to today's present values, the development of a filamentous cyanobacterial-like genome would require approximately 7 x 10(6) years--or perhaps much less. Thus, in spite of the many uncertainties involved in the estimates of time for life to arise and evolve to cyanobacteria, we see no compelling reason to assume that this process, from the beginning of the primitive soup to cyanobacteria, took more than 10 million years.

On the recombinational origin of protein-sequence-subunit structure

Since 1929 the concept that proteins are built from subunits of certain standard size (Svedberg 1929) has been revisited several times, each time with a new demonstration that, indeed, there are certain preferred protein sizes. According to recent estimates the overrepresented sizes are close to multiples of 125 amino acid (aa) residues for eukaryotes and 150 residues for prokaryotes. To explain these preferences, a hypothesis is suggested, and quantitatively developed, on the recombinational nature of this regularity. The protein-coding sequences are assumed to evolve at some early stage via recombinational events--insertions of DNA circles of a certain optimal size. The contour lengths of the protein-coding DNA circles had to be simultaneously divisible by three and, to minimize torsional constraint, by the DNA helical repeat. With these two conditions satisfied, the calculated contour lengths of the DNA circles, 250-500 base pairs (bp), turn out to correspond well to known optimal DNA circularization sizes and to the predicted range of the protein sequence subunit sizes: 80-170 aa residues, which covers experimentally observed values. The subunit size is found to be strongly influenced by the helical repeat of DNA. The sizes 125 and 150 aa are derived when the corresponding helical repeats of DNA are set within fractions of promilles from the 10.54 bp/turn value. This fits to the experimentally estimated mean for natural mixed DNA sequences, 10.53-10.57 bp/turn.(ABSTRACT TRUNCATED AT 250 WORDS)

On the early emergence of reverse transcription: theoretical basis and experimental evidence

Reverse transcriptase (RT) was first discovered as an essential catalyst in the biological cycle of retroviruses. However, in the past years evidence has accumulated showing that RTs are involved in a surprisingly large number of RNA-mediated transpositional events that include both viral and nonviral genetic entities. Although it is probable that some RT-bearing genetic elements like the different types of AIDS viruses and the mammalian LINE family have arisen in recent geological times, the possibility that reverse transcription first took place in the early Archean is supported by (1) the hypothesis that RNA preceded DNA as cellular genetic material; (2) the existence of homologous regions of the subunit tau of the E. coli DNA polymerase III with the simian immunodeficiency virus RT, the hepatitis B virus RT, and the beta' subunit of the E. coli RNA polymerase (McHenry et al. 1988); (3) the presence of several conserved motifs, including a 14-amino-acid segment that consists of an Asp-Asp pair flanked by hydrophobic amino acids, which are found in all RTs and in most cellular and viral RNA polymerases. However, whether extant RTs descend from the primitive polymerase involved in the RNA-to-DNA transition remains unproven. Substrate specificity of the AMV and HIV-1 RTs can be modified in the presence of Mn2+, a cation which allows them to add ribonucleotides to an oligo (dG) primer in a template-dependent reaction. This change in specificity is comparable to that observed under similar conditions in other nucleic acid polymerases. This experimentally induced change in RT substrate specificity may explain previous observations on the misincorporation of ribonucleotides by the Maloney murine sarcoma virus RT in the minus and plus DNA of this retrovirus (Chen and Temin 1980). Our results also suggest that HIV-infected macrophages and T-cell cells may contain mixed polynucleotides containing both ribo- and deoxyribonucleotides. The evolutionary significance of these changes in substrate specificities of nucleic acid polymerases is also discussed.

Intermolecular homologies of human interferon-alpha

Human interferon-alpha 2 (IFN) was analyzed by homology search computer program with the use of protein primary structures data bases. Results indicate that four domains with heightened ability to form homology pairs with different proteins exist in the IFN molecule. These domains occupy regions 35-56, 72-85, 97-110 and 124-136, mainly between the alpha-helical cylinders on the tertiary structure models. Additionally, results show in IFN structure the presence of amino-acid motifs that create the opportunity for this cytokine to influence directly the processes of DNA functioning in cell nuclei.

Phylogenetic continuum indicates "galaxies" in the protein universe: preliminary results on the natural group structures of proteins

The markedly nonuniform, even systematic distribution of sequences in the protein "universe" has been analyzed by methods of protein taxonomy. Mapping of the natural hierarchical system of proteins has revealed some dense cores, i.e., well-defined clusterings of proteins that seem to be natural structural groupings, possibly seeds for a future protein taxonomy. The aim was not to force proteins into more or less man-made categories by discriminant analysis, but to find structurally similar groups, possibly of common evolutionary origin. Single-valued distance measures between pairs of superfamilies from the Protein Identification Resource were defined by two chi 2-like methods on tripeptide frequencies and the variable-length subsequence identity method derived from dot-matrix comparisons. Distance matrices were processed by several methods of cluster analysis to detect phylogenetic continuum between highly divergent proteins. Only well-defined clusters characterized by relatively unique structural, intracellular environmental, organismal, and functional attribute states were selected as major protein groups, including subsets of viral and Escherichia coli proteins, hormones, inhibitors, plant, ribosomal, serum and structural proteins, amino acid synthases, and clusters dominated by certain oxidoreductases and apolar and DNA-associated enzymes. The limited repertoire of functional patterns due to small genome size, the high rate of recombination, specific features of the bacterial membranes, or of the virus cycle canalize certain proteins of viruses and Gram-negative bacteria, respectively, to organismal groups.

Revisiting junk DNA

The distribution of functions within genomes of higher organisms relative to processes that lead to the spread of mutations in populations is examined in its general outlines. A number of points are enumerated that collectively put in question the concept of junk DNA: the plausible compatibility of DNA function with rapid substitution rates; the likelihood of superimposed functions along much of eukaryotic DNA; the potential for a merely conditional functionality in sequence repeats; the apparent adoption of macromolecular waste as a strategy for maintaining a function without selective grooming of individual sequence repeats that carry out the function; the likely requirement that any DNA sequence must be "polite" vis-'a-vis (compatible with) functional sequences in its genomic environment; the existence in germ-cell lineages of selective constraints that are not apparent in populations of individuals; and the fact that DNA techtonics - the appearance and disappearance of genomic DNA - are not incompatible with function. It is pointed out that the inverse correlation between functional constraints and rates of substitution cannot be claimed to be pillar of the neutral theory, because it is also predicted from a selectionist viewpoint. The dispensability of functional structures is brought into relation with the concept of reproductive sufficiency the survivability of genotypes in the absence of fitter alleles.

Exons--original building blocks of proteins?

In a recent paper, Walter Gilbert's group has estimated the number of original exons from which all extant proteins might have been constructed. The approach used is subjected to a critical analysis here. It is shown that there are flawed assumptions about both the mechanism and generality of exon-shuffling and in the sequence comparison procedures employed, the latter failing to distinguish chance similarity from similarity due to common ancestry. These methodological errors lead to the omission of many known cases of exon-shuffling and the inclusion of others which may not be genuine. In consequence, the analysis from the Gilbert group cannot give a reliable estimate of those modules that actually participated in exon-shuffling and provides no information on the number of protein archetypes that did not participate in these processes.

Extreme differences in charge changes during protein evolution

The maintenance of a proper distribution of charged amino acid residues might be expected to be an important factor in protein evolution. We therefore compared the inferred changes in charge during the evolution of 43 protein families with the changes expected on the basis of random base substitutions. It was found that certain proteins, like the eye lens crystallins and most histones, display an extreme avoidance of changes in charge. Other proteins, like phospholipase A2 and ferredoxin, apparently have sustained more charged replacements than expected, suggesting a positive selection for changes in charge. Depending on function and structure of a protein, charged residues apparently can be important targets for selective forces in protein evolution. It appears that actual biased codon usage tends to decrease the proportion of charged amino acid replacements. The influence of nonrandomness of mutations is more equivocal. Genes that use the mitochondrial instead of the universal code lower the probability that charge changes will occur in the encoded proteins.

Compositional patterns in vertebrate genomes: conservation and change in evolution

The evolution of vertebrate genomes can be investigated by analyzing their regional compositional patterns, namely the compositional distributions of large DNA fragments (in the 30-100-kb size range), of coding sequences, and of their different codon positions. This approach has shown the existence of two evolutionary modes. In the conservative mode, compositional patterns are maintained over long times (many million years), in spite of the accumulation of enormous numbers of base substitutions. In the transitional, or shifting, mode, compositional patterns change into new ones over much shorter times. The conservation of compositional patterns, which has been investigated in mammalian genomes, appears to be due in part to some measure of compositional conservation in the base substitution process, and in part to negative selection acting at regional (isochore) levels in the genome and eliminating deviations from a narrow range of values, presumably corresponding to optimal functional properties. On the other hand, shifts of compositional patterns, such as those that occurred between cold-blooded and warm-blooded vertebrates, appear to be due essentially to both negative and positive selection again operating at the isochore level, largely under the influence of changes in environmental conditions, and possibly taking advantage of mutational biases in the replication/repair enzymes and/or in the enzyme make-up of nucleotide precursor pools. Other events (like translocations and changes in chromosomal structure) also play a role in the transitional mode of genome evolution. The present findings (1) indicate that isochores, which correspond to the DNA segments of individual or contiguous chromatin domains, represent selection units in the vertebrate genome; and (2) shed new light on the selectionist-neutralist controversy.

Evolution of chorion structural genes and regulatory mechanisms in two wild silkmoths: a preliminary analysis

We report a preliminary analysis of structural and regulatory evolution of the A and B chorion gene families in two wild silkmoths, Antheraea pernyi and Antheraea polyphemus. Homospecific and heterospecific dot hybridizations were performed between previously characterized A. polyphemus complementary DNA clones and total or stage-specific follicular mRNAs from the two species. The hybridization patterns indicated substantial interspecies changes in the abundance of corresponding mRNA sequences (heteroposic evolution) without substantial changes in their developmental specificities (heterochronic evolution). In addition, the proteins encoded in the two species by corresponding mRNAs were determined by hybrid-selected translation followed by electrophoretic analysis. The results suggested that the proteins evolve in size, presumably through internal deletions and duplications.

Ribosomal RNA and the major lines of evolution: a perspective

Does the "universal tree" based on small-subunit ribosomal RNA sequences show the phylogenetic relationship of all modern organisms? The answer is "yes" only if all these rRNAs are orthologous. Herein I argue that the major rRNA lineages (e.g. eubacterial, one or more archaebacterial and eukaryotic nucleocytoplasmic) probably arose from a divergent population of rRNAs in the progenote, antedating the universal common ancestral organism. Thus the major lineages of rRNA are probably not orthologous, but paralogous. The extrapolated date for the origin of the common ancestral small-subunit rRNA (3.6-4.7 x 10(9) years ago) is consistent with major rRNA lineages being paralogous. This perspective on the early evolution of genes and organisms rationalizes the presence of unexpected ribosomal characters in microsporidia, and bears on xenogenous and endogenous theories of the origin of the organelles in eukaryotes.

On the molecular evolutionary clock

The conceptual framework surrounding the origin of the molecular evolutionary clock and circumstances of this origin are described. In regard to the quest for the best available molecular clocks, a return to protein clocks is conditionally recommended. On the basis of recent data and certain considerations, it is pointed out that the realm of neutrality in evolution is probably less extensive than is now commonly thought, in the three distinct senses of the term neutrality--neutrality as nonfunctionality of mutations, neutrality as equifunctionality of mutations, and neutrality as a mode of fixation of mutations. The possibility is raised that complex sets of interacting components forming a system that is bounded with respect to its environment may quite generally display an intrinsic trend to a quasi-clockwise evolutionary behavior.

Polite DNA: functional density and functional compatibility in genomes

Certain as yet poorly defined functions of DNA appear to involve collectively domain-sized sequences. It is proposed that most sequence segments within a domain may be either functionally superfluous or instrumental, depending on how many related sequences are present in the domain. When redundant and functionally dispensable, such DNA segments presumably still have to conform to compositional or sequence-motif patterns that characterize the domain. In its relations with neighboring sequences, such DNA is required to be "polite." Polite DNA is DNA that, without being crucially involved in function, is subject to constraints of conformity and, through its base composition, respects a function for which it is not required. This concept is developed by contrasting the distribution of specific and general functions over DNA with this distribution as found in proteins and by distinguishing functional compatibility from pivotal functionality. The sequence constraints to which heterochromatin as well as, apparently, long interspersed repetitive sequences are known to be subject seem to imply that DNA, even when it does not carry out a pivotal function, is indeed, at the very least, required to be polite.

A measure of the similarity of sets of sequences not requiring sequence alignment

Determination of first- and second-order Markov chain homogeneity of sets of nuclear eukaryotic DNA sequences, both coding and noncoding, finds similarities imperceptible to the standard Needleman-Wunsch base matching or dot-matrix algorithms. These measures of the similarities of the distributions of adjacent pairs or triplets are in agreement with accepted evolutionary-tree topologies. Hierarchical clustering of the distributions of doublets of 30 miscellaneous coding sequences gives clusters in reasonable agreement with accepted biological classifications. In addition to similarity by homology, there is also observed similarity of disparate genes in the same organism--for example, all three disparate yeast genes (two enzymes and actin) form a well-distinguished cluster.

Markov chain analysis finds a significant influence of neighboring bases on the occurrence of a base in eucaryotic nuclear DNA sequences both protein-coding and noncoding

Sixty-four eucaryotic nuclear DNA sequences, half of them coding and half noncoding, have been examined as expressions of first-, second-, or third-order Markov chains. Standard statistical tests found that most of the sequences required at least second-order Markov chains for their representation, and some required chains of third order. For all 64 sequences the observed one-step second-order transition count matrices were effective in predicting the two-step transition count matrices, and 56 of 64 were effective in predicting the three-step transition count matrices. The departure from random expectation of the observed first- and second-order transition count matrices meant that a considerable sample of eucaryotic nuclear DNA sequences, both protein coding and noncoding, have significant local structure over subsequences of three to five contiguous bases, and that this structure occurs throughout the total length of the sequence. These results suggested that present DNA sequences may have arisen from the duplication, concatenation, and gradual modification of very early short sequences.