Mechanism and evolution of protein dimerization.

We have investigated the mechanism and the evolutionary pathway of protein dimerization through analysis of experimental structures of dimers. We propose that the evolution of dimers may have multiple pathways, including (1) formation of a functional dimer directly without going through an ancestor monomer, (2) formation of a stable monomer as an intermediate followed by mutations of its surface residues, and (3), a domain swapping mechanism, replacing one segment in a monomer by an equivalent segment from an identical chain in the dimer. Some of the dimers which are governed by a domain swapping mechanism may have evolved at an earlier stage of evolution via the second mechanism. Here, we follow the theory that the kinetic pathway reflects the evolutionary pathway. We analyze the structure-kinetics-evolution relationship for a collection of symmetric homodimers classified into three groups: (1) 14 dimers, which were referred to as domain swapping dimers in the literature; (2) nine 2-state dimers, which have no measurable intermediates in equilibrium denaturation; and (3), eight 3-state dimers, which have stable intermediates in equilibrium denaturation. The analysis consists of the following stages: (i) The dimer is divided into two structural units, which have twofold symmetry. Each unit contains a contiguous segment from one polypeptide chain of the dimer, and its complementary contiguous segment from the other chain. (ii) The division is repeated progressively, with different combinations of the two segments in each unit. (iii) The coefficient of compactness is calculated for the units in all divisions. The coefficients obtained for different cuttings of a dimer form a compactness profile. The profile probes the structural organization of the two chains in a dimer and the stability of the monomeric state. We describe the features of the compactness profiles in each of the three dimer groups. The profiles identify the swapping segments in domain swapping dimers, and can usually predict whether a dimer has domain swapping. The kinetics of dimerization indicates that some dimers which have been assigned in the literature as domain swapping cases, dimerize through the 2-state kinetics, rather than through swapping segments of performed monomers. The compactness profiles indicate a wide spectrum in the kinetics of dimerization: dimers having no intermediate stable monomers; dimers having an intermediate with a stable monomer structure; and dimers having an intermediate with a stable structure in part of the monomer. These correspond to the multiple evolutionary pathways for dimer formation. The evolutionary mechanisms proposed here for dimers are applicable to other oligomers as well.     

Nuclear protein modification and chromatin substructure. 2. Internucleosomal localization of poly(adenosine diphosphate-ribose) polymerase.

Definitive evidence for poly(ADP-Rib) polymerase activity is localized within internucleosomal "linker" regions of HeLa cell chromatin is presented. This evidence was based on the following criteria: the enzyme activity did not coincide with the position of core particles in a sucrose gradient but was displaced to that part of the gradient which is enriched in monomers with linker regions. This was not due to dimer contamination, since resedimentation did not affect the enzyme activity in relation to the monomer. A new method of assaying enzyme activity directly in polyacrylamide gels following the separation of monomers and dimers showed that only dimers and monomers with linker regions contained activity. When dimers were digested, the enzyme activity moved from the dimer to the monomer with linker.     

Self-association mode of a flavoenzyme D-amino acid oxidase from hog kidney. II. Stoichiometry of holoenzyme association and energetics of subunit association

The self-association pattern of D-amino acid oxidase holoenzyme in 0.1 M sodium pyrophosphate, pH 8.3, at 25 degrees C was examined by the low-angle laser light-scattering method. As to the results of nonlinear least-squares analysis of the apparent weight-average molecular weight (Mwapp) versus protein concentration (c) data, the following three models fitted equally well the data over the concentration range of 0.03-11.4 mg/ml: 1) the model of isodesmic indefinite self-association of the monomer where the dimerization constant differs from the isodesmic association constant, 2) the model which involves the dimerization of the monomer and isodesmic indefinite self-association of the dimer, and 3) the model which involves the trimerization of the monomer and isodesmic indefinite self-association of the trimer. In a more limited concentration range (0.3-11.4 mg/ml), a model of isodesmic indefinite self-association of the stable dimer where the dimer does not dissociate into the monomers cannot be excluded from the above three models. Measurements with the concentration range lowered to 0.03 mg/ml enabled us to exclude unequivocally the model involving such a stable dimer and to extrapolate the Mwapp data to the Mr of the monomer at infinite dilution as in the case of the apoenzyme. The observed sedimentation boundary profiles were qualitatively consistent with the idealized boundary profiles calculated with the model which involves the dimerization of the monomer and isodesmic indefinite self-association of the dimer, so this model is the most probable of the models examined. These results provide the first evidence that the association mode of the holoenzyme is different from that of the apoenzyme, i.e. isodesmic indefinite self-association of the monomer (Tojo, H., Horiike, K., Shiga, K., Nishina, Y., Watari, H., and Yamano, T. (1985) J. Biol. Chem. 260, 12607-12614). The overall linkage scheme, between binding of coenzyme FAD and subunit association, was considered, and the overall free energy change in each process in the scheme was calculated. The total stabilization energies of the intersubunit interaction in the holoenzyme relative to the apoenzyme were found to be -2.2 kcal/mol at the dimerization step and -0.5 kcal/mol at the step of the addition of the dimer to any 2i-mer (i = 1,2, ...).     

On the binding of tRNA to Escherichia coli RNA polymerase.

The fixation of tRNA to Escherichia coli RNA polymerase has been investigated. Bound and free tRNA have been separated and quantified after filtration through cellulose nitrate filters, centrifugation or sucrose gradients or electrophoresis in polyacrylamide gels. We detect no differences between the fixation of E. coli fMet-tRNAfMet, Met-tRNAmMet or uncharged unfractionated tRNA to RNA polymerase. Tight complexes, with a long residence time, are formed between core enzyme and tRNA with a dissociation constant of less than 1 nM. Complexes exist between tRNA and both monomer and dimer forms of the core enzyme. In the monomer complex, one tRNA is bound per alpha 2 beta beta' unit, whereas in the dimer complex only 0.5 tRNA molecule is fixed per alpha 2 beta beta' unit. In contrast to the core enzyme, very little tRNA fixes tightly to the holoenzyme at salt concentrations greater than 80 mM. At lower salt concentrations tRNA fixation results in a loss of sigma subunit from the holo enzyme to the resulting core enzyme where it binds tightly. DNA fixation reduces the binding of tRNA to RNA polymerase and tRNA fixation reduces the binding of DNA. However, binding of DNA to polymerase is not competitive with binding of tRNA, and ternary complexes between RNA polymerase, DNA and tRNA are shown to exist. Our results are discussed in relation to other studies concerning the effects of tRNA upon RNA polymerase.     

The higher order repeat structure of chromatin is built up of globular particles containing eight nucleosomes

Mild nuclease digestion of rat liver chromatin generates particles with sedimentation coefficients of about 33S, 60S, and 90S (in 50 mM NaCl). The kinetics of appearance and disappearance of these particles with progressive digestion suggest that they are produced by cleavage from a higher order repeat structure, the 33S particle representing the monomer. At an intermediate stage of digestion, about 75 % of the nuclear chromatin can be recovered as monomers to trimers of this higher order structure. Sedimentation profiles indicate that monomer particles containing 7–8 nucleosomes occur at the highest frequency. The DNA fragments in monomers have a size corresponding to hepta- and octanucleosomes, and those in dimers have a size corresponding to chains of sixteen nucleosomes. The higher order repeat structure is only stable between 30 and 200 mM NaCl; the particles unfold below 30 and above 200 mM NaCl. When examined by electron microscopy, monomers and dimers appear as compact globular structures. Relaxation by lowering the salt concentration results in the appearance of polynucleosomes with a chain length of eight beads in the monomer and sixteen in the dimer particle. These results indicate that the unit particle of the higher order repeat structure of rat liver chromatin contains eight nucleosomes.     

Monomeric and Dimeric CXCL8 Are Both Essential for In Vivo Neutrophil Recruitment

Rapid mobilization of neutrophils from vasculature to the site of bacterial/viral infections and tissue injury is a critical step in successful resolution of inflammation. The chemokine CXCL8 plays a central role in recruiting neutrophils. A characteristic feature of CXCL8 is its ability to reversibly exist as both monomers and dimers, but whether both forms exist in vivo, and if so, the relevance of each form for in vivo function is not known. In this study, using a ‘trapped’ non-associating monomer and a non-dissociating dimer, we show that (i) wild type (WT) CXCL8 exists as both monomers and dimers, (ii) the in vivo recruitment profiles of the monomer, dimer, and WT are distinctly different, and (iii) the dimer is essential for initial robust recruitment and the WT is most active for sustained recruitment. Using a microfluidic device, we also observe that recruitment is not only dependent on the total amount of CXCL8 but also on the steepness of the gradient, and the gradients created by different CXCL8 variants elicit different neutrophil migratory responses. CXCL8 mediates its function by binding to CXCR2 receptor on neutrophils and glycosaminoglycans (GAGs) on endothelial cells. On the basis of our data, we propose that dynamic equilibrium between CXCL8 monomers and dimers and their differential binding to CXCR2 and GAGs mediates and regulates in vivo neutrophil recruitment. Our finding that both CXCL8 monomer and dimer are functional in vivo is novel, and indicates that the CXCL8 monomer-dimer equilibrium and neutrophil recruitment are intimately linked in health and disease.     

Characterization of mitogen-activated protein kinase (MAPK) dimers

Phosphorylated ERK2 has an increased capacity to form homodimers relative to unphosphorylated ERK2. We have characterized the nature of the ERK2 dimer and have mutated residues in the crystal dimer interface to examine the impact of dimerization on ERK2 activity. Analysis of the mutants by gel filtration indicates that at least five residues must be mutated simultaneously to produce an ERK2 mutant that is predominantly monomeric. Mutants, whether monomers or dimers, have specific protein kinase activities under fixed assay conditions that are roughly equivalent to wild-type ERK2. The ratio of dimers to monomers is increased as the salt concentration increases, consistent with a strong hydrophobic contribution to the energy of dimer formation. ERK2 dimerization also requires divalent cations. Sedimentation analysis indicates that the related c-Jun N-terminal kinase SAPKalphaI/JNK2 also forms dimers, but dimerization displays no dependence on phosphorylation; the unphosphorylated and phosphorylated forms of the kinase behave similarly, with low micromolar dimer dissociation constants.     

The flavin dimers I. The application of absorption in anti-stokes excitation region to investigate the flavin dimer formation

The dimer formation process of the flavin in aqueous solution has been studied. The difference absorption spectra with the change of concentration in Stokes and anti-Stokes excitation region of the flavomononucleotide and riboflavin were measured. The highest temperature in which the dimers still appear is discussed. It is suggested that this temperature T_d can be treated as one of the empirical parameters which describe the dimer formation process of the dyes in solutions. The aqueous solution of flavins with the concentration c?5·10^?5 M at room temperature can be treated as a flavin monomers solution. For higher concentrations the flavin monomers and dimers exist in a solution at room temperature.     

Aggregation equilibria of Escherichia coli RNA polymerase: evidence for anion-linked conformational transitions in the protomers of core and holoenzyme

The aggregation equilibria of Escherichia coli RNA polymerase core and holoenzyme have been studied by velocity sedimentation as a function of [NaCl] both in the presence and in the absence of MgCl2. Effects of other anions (F- and I-), pH, and temperature have also been examined. Diffusion coefficients obtained by quasi-elastic light scattering (QLS) at high and low salt concentrations were used in conjunction with sedimentation coefficients under these conditions to obtain molecular weights of the protomer and aggregates of the core enzyme. At low salt concentration, core aggregates to a tetramer in the absence of MgCl2 and to an octamer in the presence of MgCl2. Some ambiguity exists in the interpretation of the sedimentation and QLS data for holoenzyme. The sedimentation results are consistent with the formation of dimers at low salt, both in the presence and in the absence of MgCl2. In all cases, equilibrium constants were calculated assuming a simple monomer--j-mer stoichiometry. These equilibrium constants are extremely sensitive functions of the concentration and type of monovalent anion. In Cl-, aggregation of both core and holoenzyme begins abruptly when the salt concentration is reduced below approximately 0.2 M (at a protein concentration of approximately 0.30 mg/mL); for core, substitution of I- for Cl- suppresses aggregation while F- enhances aggregation at a fixed anion concentration. No specific effect of monovalent cations (Na+, NH4+) is observed; Mg2+ has no effect on holoenzyme dimerization and has little effect on the salt range of core aggregation, though the stoichiometries of the core aggregates in the presence and absence of Mg2+ differ. Anion effects on these equilibria were modeled by assuming that a class of anion-binding sites on the protomer is not present in the aggregate, so that anion release accompanies aggregation. Analytical expressions for several models of the effect of anions on the aggregation equilibria were derived by using the method of binding polynomials. The salt dependence of the aggregation equilibria in the absence of Mg2+ appears inconsistent with a model in which the anion-binding sites on the protomer are independent (noncooperative), but it is well described by a model in which anion binding to the protomers occurs in a completely cooperative manner. The molecular basis of this apparent cooperative effect of anions on the aggregation equilibria is proposed to be an allosteric effect of anions on conformational equilibria of the protomers of core polymerase and the holoenzyme. Implications of such a salt-dependent conformational transition for the DNA-binding interactions of the enzyme are considered.     

The dimer of the beta subunit of Escherichia coli DNA polymerase III holoenzyme is dissociated into monomers upon binding magnesium(II)

The beta subunit of Escherichia coli DNA polymerase III holoenzyme binds Mg2+. Reacting beta with fluoresceinmaleimide (FM) resulted in one label per beta monomer with full retention of activity. Titration of FM-beta with Mg2+ resulted in a saturable 11% fluorescence enhancement. Analysis indicated that there was one noncooperative magnesium binding site per beta monomer with a dissociation constant of 1.7 mM. Saturable fluorescence enhancement was also observed when titration was with Ca2+ or spermidine(3+) but not with the monovalent cations Na+ and K+. The Mg2+-induced fluorescence enhancement was specific for FM-beta and was not observed with FM-glutathione, dimethoxystilbenemaleimide-beta, or pyrenylmaleimide-beta. Gel filtration studies indicated that the beta dimer-monomer dissociation occurred at physiologically significant beta concentrations and that the presence of 10 mM Mg2+ shifted the dimer-monomer equilibrium to favor monomers. Both the gel-filtered dimers and the gel-filtered monomers were active in the replication assay. These and other results suggested that the fluorescence increase which accompanies beta dissociation is due to a relief from homoquenching of FM when the beta dimer dissociates into monomers.     

Multicopy plasmid instability: the dimer catastrophe hypothesis

Multimer formation reduces plasmid copy number and is an established cause of segregational instability. Nevertheless, it is difficult to rationalize observations that low levels of dimers can cause severe instability, if we assume they are distributed evenly in cell populations. We report here that dimer distribution is in fact heterogeneous in recombination-proficient strains. Most cells in the population contain only monomers; dimers are confined to a small sub-population from which plasmid-free daughters arise at high frequency. In a rec⁺ culture where 4% of pBR322 molecules are dimers, more than half are in dimer-only cells. We show that this situation is inevitable because dimers replicate at twice the rate of monomers. Runaway multimerization is avoided because dimer-containing cells grow more slowly than their monomer-containing counterparts. A computer simulation is used to show how dimers proliferate after formation by homologous recombination. The equilibrium concentration of dimers is proportional to the inter-plasmid recombination rate and is essentially independent of the rate at which homologous recombination converts dimers to monomers.     

The aggregation state of bovine heart cytochrome c oxidase and its kinetics in monomeric and dimeric form

The monomeric and dimeric forms of bovine cytochrome c oxidase (EC 1.9.3.1) were obtained from gel filtration chromatography on Ultrogel AcA 34 and analyzed. Both species contained all 12-13 subunits described for this enzyme. In the dimer 320 molecules [3H]dodecyl-beta-D-maltoside were bound per heme aa3 and in the monomer 360 molecules per heme aa3. The monomers contained 10 mol of tightly bound phospholipid/mol heme aa3 and the dimers 14. Sedimentation coefficients of 15.5-18 S for the dimer and 9.6 S for the monomer were calculated from sucrose density centrifugation analysis and analytical centrifugation. By the laser beam light-scattering technique a Stokes radius of 70 A for the dimeric detergent-lipid-protein complex was measured. From those parameters and the densitometric determined partial specific volumes of the detergent and the enzyme, the molecular weights of 400,000 for the protein moiety of the dimer and 170,000-200,000 for the monomer were calculated. Under very low ionic strength conditions the monomer/dimer equilibrium was found to be dependent on the protein concentration. At low enzyme concentrations (10(-9) M) monomers were predominant, whereas at concentrations above 5 X 10(-6) M the amounts of dimers and higher aggregates were more represented. The cytochrome c oxidase activity, measured spectrophotometrically and analyzed by Eadie-Hofstee plot, was biphasic as a function of cytochrome c concentration for the dimeric enzyme. Pure monomers gave monophasic kinetics. The data, fitting with a homotropic negative cooperative mechanism for the dimer of cytochrome c oxidase, are discussed and compared with other described mechanisms.     

Photochemical redox transformations of dimers of NAD+ and N'-methylnicotinamide

Electrochemically generated dimers of N′-methylnicotinamide and NAD⁺ both undergo photooxidation on irradiation at 254 nm in aqueous medium to yield the respective parent monomers. For the NAD dimer the quantum yield for photodissociation was about 0.01, whether irradiated at 254 nm or at wavelengths to the red of 320 nm. Irradiation at the latter wavelengths, where the NAD⁺ monomer itself does not absorb and is not photosensitive, led to quantitative regeneration of coenzyme activity in the alcohol dehydrogenase system. The photodissociation reaction exhibited no oxygen effect. The photochemically generated dimer of N′-methylnicotinamide was also photooxidized to the parent monomer by irradiation at 254 nm at pH 9.5. The foregoing process of electrochemical (or photochemical) reduction and photochemical oxidation, comprising a closed cycle of electron and proton transport, is similar to that previously observed for a number of pyrimidine analogs. Furthermore, the NAD dimer is a substrate of snake venom nucleotide pyrophosphatase and is hydrolyzed to release NMN dimer which, on irradiation at 254 nm, also undergoes photooxidation to the parent NMN⁺ monomer. A mechanism for the photooxidation reaction is formulated and relevant biological implications of the foregoing are presented.     

Crystal structure of a C-terminal deletion mutant of human protein kinase CK2 catalytic subunit

Protein kinase CK2 (formerly called: casein kinase 2) is a heterotetrameric enzyme composed of two separate catalytic chains (CK2alpha) and a stable dimer of two non-catalytic subunits (CK2beta). CK2alpha is a highly conserved member of the superfamily of eukaryotic protein kinases. The crystal structure of a C-terminal deletion mutant of human CK2alpha was solved and refined to 2.5A resolution. In the crystal the CK2alpha mutant exists as a monomer in agreement with the organization of the subunits in the CK2 holoenzyme. The refined structure shows the helix alphaC and the activation segment, two main regions of conformational plasticity and regulatory importance in eukaryotic protein kinases, in active conformations stabilized by extensive contacts to the N-terminal segment. This arrangement is in accordance with the constitutive activity of the enzyme. By structural superimposition of human CK2alpha in isolated form and embedded in the human CK2 holoenzyme the loop connecting the strands beta4 and beta5 and the ATP-binding loop were identified as elements of structural variability. This structural comparison suggests that the ATP-binding loop may be the key region by which the non-catalytic CK2beta dimer modulates the activity of CK2alpha. The beta4/beta5 loop was found in a closed conformation in contrast to the open conformation observed for the CK2alpha subunits of the CK2 holoenzyme. CK2alpha monomers with this closed beta4/beta5 loop conformation are unable to bind CK2beta dimers in the common way for sterical reasons, suggesting a mechanism to protect CK2alpha from integration into CK2 holoenzyme complexes. This observation is consistent with the growing evidence that CK2alpha monomers and CK2beta dimers can exist in vivo independently from the CK2 holoenzyme and may possess physiological roles of their own.     

Perfect hemihedral twinning in crystals of the γ-subunit of translation initiation factor 2 from <Emphasis Type="Italic">Sulfolobus solfataricus</Emphasis>: Cause and effect

The crystal structure of the γ-subunit of translation initiation factor 2 from the archaeon Sulfolobus solfataricus (SsoIF2γ) has been solved based on perfectly hemihedral twinned data. The protein was cocrystallized with the 10-fold molar excess of GTP analog (GDPCP) over protein. However, no nucleotide was found in the structure, and the model demonstrated the apo form of the protein. Two slightly different molecules in the asymmetric unit of the crystal are related by the non-crystallographic 2-fold axis and form a tightly associated dimer. This dimer is stabilized by an intermolecular hydrophobic core and hydrogen bonds. Lack of GDPCP in the nucleotide-binding pocket of the γ-subunit and significant excess of dimers over monomers in the crystallization solution suggest that these dimers are the building blocks of the crystal. Contrary to SsoIF2γ monomers, these dimers are able to crystallize in two oppositely oriented slightly different crystal domains, thus forming a twinned crystal. Comparison of crystallization conditions for the twinned and untwinned crystals of apo SsoIF2γ showed that stabilization of the dimers in the solution may be caused by higher sodium salt concentration. Since amino acid residues involved in intermolecular contacts in the dimer are responsible for binding of the γand α-subunits within SsoIF2, increase in sodium salt concentration may prevent functioning of SsoIF2 in the cell.     

Letter: Cytoplasmic DNA polymerase: polymeric forms and their conversion into an active monomer resembling nuclear DNA polymerase

DNA polymerase isolated from the cytoplasm is an aggregate enzyme which can exist in forms of dimers, trimers and tetramers of a monomer of 54,000 molecular weight. Trimers and tetramers could be quantitatively converted into dimers and monomers by sedimentation in sucrose gradients at high ionic strength (0.45 m). Molecular properties of the monomer resembled nuclear DNA polymerase: sedimentation coefficient (3.5 S), template specificity and resistance to inhibition by N-ethylmaleimide and heparin.     

Reactivation and refolding of reassociated dimers of rabbit muscle creatine kinase

Creatine kinase (ATP:creatine N-phosphotransferase, EC 2.7.3.2) is a good model for studying dissociation and reassociation during unfolding and refolding. This study compares self-reassociated CK dimers and CK dimers that contain hybrid dimers under proper conditions. Creatine kinase forms a monomer when denatured in 6 M urea for 1 h which will very quickly form a dimer when the denaturant is diluted under suitable conditions. After modification by DTNB, CK was denatured in 6 M urea to form a modified CK monomer. Dimerization of this modified subunit of CK occurred upon dilution into a suitable buffer containing DTT. Therefore, three different types of reassociated CK dimers including a hybrid dimer can be made from two different CK monomers in the proper conditions. The CK monomers are from a urea-denatured monomer of DTNB-modified CK and from an unmodified urea dissociated monomer. Equal enzyme concentration ratios of these two monomers were mixed in the presence of urea, then diluted into the proper buffer to form the three types of reassociated CK dimers including the hybrid dimer. Reassociated CK dimers including all three different types recover about 75% activity following a two-phase course (k ₁ = 4.88 × 10^–3 s^–1, k ₂ = 0.68 × 10^–3 s^–1). Intrinsic fluorescence spectra of the three different CK monomers which were dissociated in 6 M urea, dissociated in 6 M urea after DTNB modification, and a mixture of the first two dissociated enzymes were studied in the presence of the denaturant urea. The three monomers had different fluorescence intensities and emission maxima. The intrinsic fluorescence maximum intensity changes of the reassociated CK dimers were also studied. The refolding processes also follow biphasic kinetics (k ₁ = 3.28 × 10^–3 s^–1, k ₂ = 0.11 × 10^–3 s ^–1) after dilution in the proper solutions. Tsou's method [Tsou (1988), Adv. Enzymol. Rel. Areas Mol. Biol. 61, 381–436] was also used to measure the kinetic reactivation rate constants for the different three types of reassociated CK dimers, with different kinetic reactivation rate constants observed for each type. CK dissociation and reassociation schemes are suggested based on the results.