Relationship between noise characteristics in protein expressions and regulatory structures of amino acid biosynthesis pathways.

Gene regulatory dynamics involves several stochastic chemical reactions. As a consequence, the copy number of given protein varies greatly among cells even in the case of isogeneic cells. Recently, the characteristics of noise in gene expressions were studied by using simple artificial gene networks. However, the noise characteristics in natural regulatory networks having complex interactions still remain unclear. In this study, we have focused on the noise in natural regulatory networks to understand the relationship between the characteristics of the noise and the structures of regulatory interactions. We targeted the expressions of genes related to amino acid biosynthesis (AAB) because of their well known regulatory structures. By measuring the noise of AAB genes in isogeneic Escherichia coli cells using flow cytometry, we found the noise amplitude in AAB genes to depend on the structure of the regulatory network. We categorised the regulatory networks with feedback regulation into two cases. In one case, the gene expression is negatively regulated by the final products of the AAB pathway known as feedback repression, whereas in another case, the gene expression is negatively regulated as a result of depletion of the substrate that is located upstream of the AAB pathway and activates the expression of the corresponding gene. Our data revealed that the noise amplitude of AAB genes in the former case is significantly smaller than the noise amplitude in the latter case. Furthermore, we found that the response time as a result of environmental changes is generally longer in the former case. This result provides a basis for understanding the role of natural regulatory networks better.

Molecular evolution in static and dynamical landscapes.

In an attempt to understand protein evolution, we address the issues ofhow much variety in the sequences is needed to prompt the evolution ofan enzyme from random polypeptides and how does cellular interactionaffect the dynamics of molecular evolution to allow genetic diversity inpopulation. The experimental evolution of phage-displayed randompolypeptides of about 140 amino acid residues panned with transition stateanalogue for an esterase reaction showed that even with a population sizeas small as ten, not only could significant varieties be found but also therandom polypeptides in each of the generation had great promise towardsdeveloping into functional proteins. Hence, it is evident that the enzymeevolution is prompted even within a small local area of the static landscapeof the sequence space. Considering that interaction among living cells is aninevitable event in natural evolution, its role was investigated through threeconsecutive rounds of random mutagenesis on the glutamine synthetasegene and chemostat culture of the transformed Escherichia colicellscontaining the mutated genes. The molecular phylogeny and populationdynamics show the coexistence of some mutants having different level ofglutamine synthetase at each generation. In addition, it was confirmed thatcellular interaction via the medium influences the stability of the coexistenceand bring forth fitness change to the coexisting members of the population,thereby, leading to a dynamical landscape. Based on experimental resultsreflecting the extent of interaction among members in population, here, Iproposed that protein evolution could change its mode from theoptimization on static landscape to diversification on dynamicallandscape.

Unique Colony Housing the Coexisting Escherichia coli and Dictyostelium discoideum.

Two well-characterized and phylogenetically diverse species, Escherichiacoli and Dictyostelium discoideum, were used as the modelorganisms. When the two species were mixed and allowed to grow onminimal agar plates at 22 (°)C, instead of the predator Dictyostelium exterminating E.coli, the two species remarkablyachieved a state of stable coexistence in about two weeks. In addition, theemerged colonies housing the coexisting species have a mucoidal naturethat is distinctive from its origin. The simplicity of the system and the shorttime span for the two species to develop the coexistence state, that isproven stable and reproducible on laboratory conditions, hence, providesa new model system for the study of symbiosis, particularly with referenceto the initial stages.

Plasticity of fitness and diversification process during an experimental molecular evolution.

A simplified experimental evolution encompassing the essence of natural one was designed in an attempt to understand the involved mechanism. In our system, molecular evolution was observed through three serial cycles of consecutive random mutagenesis of the glutamine synthetase gene and chemostat culture of the transformed Escherichia coli cells containing the mutated genes. Selection pressure was imposed solely on the glutamine synthetase gene when varieties of mutant genes compete in an unstructured environment of the chemostat. The molecular phylogeny and population dynamics were deduced from the nucleotide sequences of the genes isolated from each of the chemostat runs. An initial mutant population in each cycle, comprised of diversified closely-related genes, ended up with several varieties of mutants in a state of coexistence. Competition between two mutant genes in the final population of the first cycle ascertained that the observed coexisting state is not an incidental event and that cellular interaction via environmental nutrients is a possible mechanism of coexistence. In addition, the mutant gene once extinct in the previous passage was found to have the capacity to reinvade and constitute the gene pool of the later cycle of molecular evolution. These results, including the kinetic characteristics of the purified wild-type and mutant glutamine synthetases in the phylogenetic tree, revealed that the enzyme activity had diverged, rather than optimized, to a fittest value during the course of evolution. Here, we proposed that the plasticity of gene fitness in consequence of cellular interaction via the environment is an essential mechanism governing molecular evolution.

Effects of amino acid substitution on the physicochemical properties of artificial proteins with random sequences.

Physicochemical properties of four random proteins, each consisting of about 150 amino acid residues with different sequence identity, were compared to know the correlation between the physicochemical properties and its sequence. The results showed that the extent of the sequence alterations correlated well with the extent of differences in CD spectra, roughly with those in pH-solubility profiles and sedimentation velocity, and not with that in the binding of a hydrophobic fluorescent dye (ANS). Therefore, proteins with similar sequences can have different physicochemical properties, indicating that the extent of mutational effects varies in response to the sequence being altered. This warrants the evolution of a protein in a sequence-specific manner.

Synthesis of functional protein in liposome.

The liposome consisting of eggPC, cholesterol, and DSPE-PEG5000 with a molar ratio of 1.5:1:0.08 was used to entrap cell-free protein synthesis reaction mixture. The synthesis of a mutant green fluorescent protein in the liposome was confirmed by the fluorescence emitted from the liposome on flow cytometry analysis and fluorescence microscopy. The protein synthesized in the liposome is hence functional.

Glycoprotein IIb/IIIa receptor inhibitor attenuates platelet aggregation induced by thromboxane A2 during in vitro nonpulsatile ventricular assist circulation.

A recent development in antithrombotic research allows the inhibition of platelet aggregation via protection of the glycoprotein IIb/IIIa receptor on the platelet membrane. We hypothesized that a GP IIb/IIIa receptor inhibitor would inhibit thromboxane-induced platelet aggregation during circulation in our in vitro ventricular assist device (VAD) circuit and preserve long-term platelet function. Twenty-one in vitro nonpulsatile centrifugal VAD circuits were simulated for 4 days using 450 ml of fresh human whole blood with or without glycoprotein IIb/IIIa receptor inhibitor (tirofiban). Platelet aggregation and degranulation were measured in whole blood induced by ristocetin, collagen, ADP, and thromboxane A2 (TXA2). The tirofiban-treated group preserved the platelet count and tended to exert these beneficial effects by inhibiting pathologic platelet aggregation induced by TXA2, collagen, and ADP as well as degranulation. Tirofiban may be useful in preserving platelet number and function during clinical VAD use.

GroEL binds artificial proteins with random sequences.

Chaperonin GroEL from Escherichia coli binds to the non-native states of many unrelated proteins, and GroEL-recognizable structural features have been argued. As model substrate proteins of GroEL, we used seven artificial proteins (138 approximately 141 residues), each of which has a unique but randomly chosen amino acid sequence and no propensity to fold into a certain structure. Two of them were water-soluble, and the rest were soluble in 3 m urea. The soluble ones interacted with GroEL in a manner similar to that of a natural substrate; they stimulated the ATPase cycle of GroEL and GroEL/GroES and inhibited GroEL-assisted folding of other protein. All seven artificial proteins were able to bind to GroEL. The results suggest that the secondary structure as well as the specific sequence motif of the substrate proteins are not necessary to be recognized by GroEL.

Sympatric speciation: compliance with phenotype diversification from a single genotype.

A novel mechanism for sympatric speciation that takes into account complex bioprocesses within each individual organism is proposed. According to dynamical systems theory, organisms with identical genotypes can possess differentiated physiological states and may coexist 'symbiotically' through appropriate mutual interaction. With mutations, the phenotypically differentiated organisms gradually come to possess distinct genotypes while maintaining their symbiotic relationship. This symbiotic speciation is robust against sexual recombination, because offspring of mixed parentage with intermediate genotypes are less fit than their parents. This leads to sterility of the hybrid. Accordingly, a basis for mating preference also arises.

Construction and characterization of phage libraries displaying artificial proteins with random sequences.

Three phage libraries, PL1, PL2, and PL3, displaying artificial proteins with random sequences were constructed. The artificial proteins, which are model of ancestral proteins, are derivatives of the 25 kinds of random proteins with about 140 amino acid residues produced via random mutagenesis and combinatorial recombination. The random proteins were displayed on the surface of filamentous bacteriophage as fusion protein with the pIII coat protein at an estimated average number on the phage particles in PL1, PL2, and PL3 of 0.32, 0.32, and 0.08, respectively. Each library was shown to express 10(5) to 10(6) kinds of random proteins. With the phage libraries displaying long random peptides, we now have an effective selection system to observe in vitro evolution of new functional proteins from artificial proteins with random sequences.

Evolution of genetic codes through isologous diversification of cellular states.

Evolution of genetic codes is studied as change in the choice of enzymes that are used to synthesize amino acids from the genetic information of nucleic acids. We propose the following theory: the differentiation of physiological states of a cell allows for a choice of enzymes, and this choice is later fixed genetically through evolution. To demonstrate this theory, a dynamical systems model consisting of the concentrations of metabolites, enzymes, amino acyl tRNA synthetase, and tRNA - amino acid complexes in a cell is introduced and studied numerically. It is shown that the biochemical states of cells are differentiated by cell-cell interactions, and each differentiated type starts to use a different synthetase. Through the mutation of genes, this difference in the genetic code is amplified and stabilized. The relevance of this theory to the evolution of non-universal genetic code in mitochondria is suggested. The present theory is based on our recent theory of isologous symbiotic speciation, which is briefly reviewed. According to the theory, phenotypes of organisms are first differentiated into distinct types through the interaction and developmental dynamics, even though they have identical genotypes; later, with mutation in the genotype, the genotype also differentiates into discrete types, while maintaining the "symbiotic" relationship between the types. Relevance of the theory to natural as well as artificial evolution is discussed.

Isologous diversification for robust development of cell society.

Isologous diversification, proposed for cell differentiation, is shown to be stable against molecular and other external fluctuations, where amplification of noise-induced slight difference between cells leads to a noise-tolerant society with differentiated cell types. It is a general consequence of interacting cells with biochemical networks and cell divisions, as is confirmed by several model simulations. According to the theory, differentiation proceeds first by loss of synchrony of intracellular oscillations as the number of cells increases. Then the chemical composition of the cells is differentiated. The differentiated compositions become inherited by the next generation, and lead to determined cell types. As a result of successive occurrence of the cell differentiation, the cell society will be composed of different cell types. The whole developmental process is robust not only against molecular fluctuations but also against the removal of a cluster of cells. This robustness is a remarkable feature of isologous diversification, in contrast to the conventional threshold-type mechanism for development. As a testable consequence of the theory, we also discuss interaction-dependent tumor formation and negative correlation between growth speed and chemical diversity.

Gradual development of protein-like global structures through functional selection.

This work focuses on streamlining the exploration of all possible sequences in an attempt to find polypeptides capable of folding into unique structures. Using a computer simulation, we have demonstrated the efficacy of constraining an 'active site' toward the correct configuration, in this case a particular conformation of a four-residue sequence, to bring about protein-like structure from a significant fraction of random sequences. The successive selections for a correct local configuration lead also to the gradual development of overall folding ability, helicity and compactness within 200 generations. The selection thus imposed alleviates an exhaustive search in sequence space.

Evolutionary molecular engineering by random elongation mutagenesis.

We describe a new method of random mutagenesis that employs the addition of peptide tails with random sequences to the C-terminal of enzyme molecules. A mutant population of catalase I from Bacillus stearothermophilus prepared by this method has a diversity in thermostability and enzyme activity equal to that obtained after random point mutagenesis. When a triple mutant of catalase I (I108T/D130N/1222T)-the thermostability of which is much lower than that of the wild type-was subjected to random elongation mutagenesis, we generated a mutant population containing only mutants with higher thermostability than the triple mutant. Some had an even higher stability than the wild-type enzyme, whose thermostability is considered to be optimized. These results indicate that peptide addition expands the protein sequence space resulting in a new fitness landscape. The enzyme can then move along the routes of the new landscape until it reaches a new optimum. The combination of random elongation mutagenesis with random point mutagenesis should be a useful approach to the in vitro evolution of proteins with new properties.

Nonadditivity of mutational effects on the properties of catalase I and its application to efficient directed evolution.

Catalase I of Bacillus stearothermophilus has high catalatic and low peroxidatic activities. The mutant from the first random mutant population, D130N, which has higher peroxidatic and lower catalatic activities than those exhibited by the wild-type enzyme, was subjected to second random mutagenesis in observance of the change in reaction specificity. From the second mutant population, the mutant I108T/D130N/I222T was selected and examined. The reaction specificity of the purified enzymes revealed that catalase I being originally 98% catalase and 2% peroxidase was brought to 58% specificity to peroxidase after two-step adaptive walks. From the statistical analysis of the two random mutant populations, the average degree of nonadditivity of the mutational effects was estimated to be 0.13 irrespective of the properties of the enzyme. It was demonstrated that the distribution pattern of a property of the second mutant population can be predicted well from the data of the first mutant population by taking into consideration the degree of nonadditivity. The strategy for an efficient adaptive walk in directed evolution of enzymes through the prediction of appropriate mutation rate and effective sample size for further mutation and selection was presented and discussed.

Characterization of random-sequence proteins displayed on the surface of Escherichia coli RNase HI.

In a previous study, random-sequence proteins of 120-130 amino acid residues were inserted into the surface loop region of the enzyme, Escherichia coli RNase HI [Doi et al. (1997) FEBS Lett. 402, 177-1801. Here we established that the RNase H activity of the insertion mutants is correlated with their secondary structure contents evaluated by circular dichroism measurement at 222 nm. The random-sequence insert of a mutant enzyme possessing relatively high RNase H activity was detached from the RNase HI scaffold, and its characterization indicated that the random-sequence protein maintains its secondary structure after separation from the scaffold. Thus, the structural features of random-sequence proteins were suggested to be monitored by measuring the activity of the scaffold enzyme into which these proteins have been inserted.

Characterization of soluble artificial proteins with random sequences.

The structural and catalytic properties of two soluble random proteins, RP3-42 and RP3-45, of 141 amino acid residues were investigated. Although no marked secondary structure was detected by CD spectrum, sedimentation equilibrium and small-angle X-ray scattering studies showed that they form an oligomeric structure and are as compact as the molten globule. The random proteins have low but distinct esterase activity; the values of the second-order rate constant for the hydrolysis of p-nitrophenol were 0.78 and 1.39 M(-1) s(-1) for RP3-42 and RP3-45, respectively. The differences in the properties of the random and the native proteins are discussed from the evolutionary point of view.

Properties of artificial proteins with random sequences.

A library of artificial proteins of 141 amino acid residues, of which 95 are random and which include 20 kinds of amino acids, was prepared. As the properties of the artificial random proteins are free from the evolutionary constraint, they can be used as a standard to discriminate the specialized properties of natural proteins. Out of the 25 identified random proteins, 5 are soluble in the cell lysate, indicating that about 20% of the random proteins expressed in Escherichia coli are expected to be soluble. Therefore, as natural soluble or insoluble proteins can arise from the line of soluble or insoluble ancestry, respectively, solubility does not seem a specialized property of natural proteins. The soluble random proteins RP3-42 and RP3-45 were purified and their properties were investigated.

Evolution of the folding ability of proteins through functional selection.

An evolutionary process is simulated with a simple spin-glass-like model of proteins to examine the origin of folding ability. At each generation, sequences are randomly mutated and subjected to a simulation of the folding process based on the model. According to the frequency of local configurations at the active sites, sequences are selected and passed to the next generation. After a few hundred generations, a sequence capable of folding globally into a native conformation emerges. Moreover, the selected sequence has a distinct energy minimum and an anisotropic funnel on the energy surface, which are the imperative features for fast folding of proteins. The proposed model reveals that the functional selection on the local configurations leads a sequence to fold globally into a conformation at a faster rate.

General equation of steady-state enzyme kinetics using net rate constants and its applicaiton to the kinetic analysis of catalase reaction.

The steady-state velocity equation is derived for the general reaction scheme containing n kinds of enzyme species connected by a network of reversible reaction steps. The general equation is represented by the net rate constants as well as the true rate constants of the individual reaction steps. Using a general equation, equations for simpler schemes can easily be derived. Furthermore, the velocity equation expressed by net rate constants is useful for understanding the dynamic state of any reaction be it simple or complex. The generality of the presented equation is tested in a complex reaction involving Bacillus stearothermophilus catalase I. In addition, the applicability of the general equation in analysing the reaction specificity and dynamic state of the reaction is shown.