Coevolution of DNA uptake sequences and bacterial proteomes.

Dramatic examples of repeated sequences occur in the genomes of some naturally competent bacteria, which contain hundreds or thousands of copies of short motifs called DNA uptake signal sequences. Here, we analyze the evolutionary interactions between coding-region uptake sequences and the proteomes of Haemophilus influenzae, Actinobacillus pleuropneumoniae, and Neisseria meningitidis. In all three genomes, uptake sequence accumulation in coding sequences has approximately doubled the frequencies of those tripeptides specified by each species' uptake sequence. The presence of uptake sequences in particular reading frames correlated most strongly with the use of preferred codons at degenerately coded positions, but the density of uptake sequences correlated only poorly with protein functional category. Genes lacking homologs in related genomes also lacked uptake sequences, strengthening the evidence that uptake sequences do not drive lateral gene transfer between distant relatives but instead accumulate after genes have been transferred. Comparison of the uptake sequence-encoded peptides of H. influenzae and N. meningitidis proteins with their homologs from related bacteria without uptake sequences indicated that uptake sequences were also preferentially located in poorly conserved genes and at poorly conserved amino acids. With few exceptions, amino acids at positions encoded by uptake sequences were as well conserved as other amino acids, suggesting that extant uptake sequences impose little or no constraint on coding for protein function. However, this state is likely to be achieved at a substantial cost because of the selective deaths required to eliminate maladaptive mutations that improve uptake sequences.

Target recognition by calmodulin: dissecting the kinetics and affinity of interaction using short peptide sequences.

The interaction between calmodulin (CaM) and peptide M13, its target binding sequence from skeletal muscle myosin light chain kinase, involves predominantly two sets of interactions, between the N-terminal target residues and the C-domain of calmodulin, and between the C-terminal target residues and the N-domain of calmodulin (Ikura M et al., 1992, Science 256:632-638). Using short synthetic peptides based on the two halves of the target sequence, the interactions with calmodulin and its separate C-domain have been studied by fluorescence and CD spectroscopy, calcium binding, and kinetic techniques. Peptide WF10 (residues 1-10 of M13) binds to CaM with Kd approximately 1 microM; peptide FW10 (residues 9-18 of M13, with Phe-17-->Trp substitution) binds to CaM with Kd approximately 100 microM. The effect of peptide WF10 on calcium binding to calmodulin produces a biphasic saturation curve, with marked enhancement of affinity for the binding of two calcium ions to the C-domain, forming a stable half-saturated complex, Ca2-CaM-peptide, and confirming the functional importance of the interaction of this sequence with the C-domain. Stopped-flow studies show that the EGTA-induced dissociation of WF10 from Ca4-CaM proceeds by a reversible relaxation mechanism from a kinetic intermediate state, also involving half-saturation of CaM, and the same mechanism is evident for the full target peptide. Interaction of the N-terminal target residues with the C-domain is energetically the most important component, but interaction of calmodulin with the whole target sequence is necessary to induce the full cooperative interaction of the two contiguous elements of the target sequence with both N- and C-domains of calmodulin. Thus, the interaction of calmodulin with the M13 sequence can be dissected on both a structural and kinetic basis into partial reactions involving intermediates comprising distinct regions of the target sequence. We propose a general mechanism for the calcium regulation of calmodulin-dependent enzyme activation, involving an intermediate complex formed by interaction of the calmodulin C-domain and the corresponding part of the target sequence. This intermediate species can function to regulate the overall calcium sensitivity of activation and to determine the affinity of the calmodulin target interaction.

Role of the N-terminal region of the skeletal muscle myosin light chain kinase target sequence in its interaction with calmodulin.

The binding of calmodulin (CaM) to four synthetic peptide analogues of the skeletal muscle myosin light chain kinase (sk-MLCK) target sequence has been studied using 1H-NMR. The 18-residue peptide WFF is anchored to CaM via the interaction of the Trp 4 side chain with the C-domain and the Phe 17 side chain with the N-domain of the protein. A peptide corresponding to the first 10 residues (WF10) does not provide the second anchoring residue and is not long enough to span both domains of CaM. 1H-NMR spectroscopy indicates that the WF10 peptide interacts specifically with the C-domain of CaM, and the chemical shifts of the bound Trp side chain are very similar in the CaM:WF10 and CaM:WFF complexes. Binding of the C-domain of CaM to the strongly basic region around Trp 4 of this MLCK sequence may be an important step in target recognition. Comparison of 1H-NMR spectra of CaM bound to WFF, a Trp 4-->Phe analogue (FFF), or a Trp 4-->Phe/Phe 17-->Trp analogue (FFW) suggests that all three peptides bind to CaM in the same orientation, i.e., with the peptide side chain in position 4 interacting with the C-domain and the side chain in position 17 interacting with the N-domain. This indicates that a Trp residue in position 4 is not an absolute requirement for binding this target sequence and that interchanging the Trp 4 and Phe 17 residues does not reverse the orientation of the bound peptide, in confirmation of the deduction from previous indirect studies using circular dichroism (Findlay WA, Martin SR, Beckingham K, Bayley PM, 1995, Biochemistry 34:2087-2094). Molecular modeling/energy minimization studies indicate that only minor local changes in the protein structure are required to accommodate binding of the bulkier Trp 17 side chain of the FFW peptide to the N-domain of CaM.

Recovery of native structure by calcium binding site mutants of calmodulin upon binding of sk-MLCK target peptides.

The calcium-dependent binding of two synthetic 18-residue peptides derived from the calmodulin binding region of skeletal myosin light chain kinase to wild-type Drosophila melanogaster calmodulin and four calcium binding site calmodulin mutants has been investigated using optical spectroscopy. The WFF peptide (with W4 and F17) and the FFW peptide (with F4 and W17) both bind to wild-type calmodulin with 1:1 stoichiometry and Kd values of < or = 0.2 and 1.6 nM, respectively. Near-UV CD spectra of the protein-peptide complexes suggest that both peptides bind in the same orientation, with the side chain of residue 4 interacting with the C-domain of calmodulin and that of residue 17 with the N-domain [as in the structure of the calmodulin-M13 peptide complex determined by Ikura et al. [Ikura, M., Clore, G. M., Gronenborn, A. M., Zhu, G., Klee, C. B., & Bax, A. (1992) Science 256, 632-638]]. Both peptides have lower affinities for all the mutant calmodulins than for the wild-type protein. Fluorescence measurements suggest that mutation of calcium binding site 2 in the N-domain does not affect the interaction of the W4 side chain of the WFF peptide with the C-domain of calmodulin. However, the E67Q (B2Q) but not the E67K (B2K) mutation (site 2, N-domain) alters the interaction of W17 of the FFW peptide with the protein. In contrast, the E140K (B4K) mutation has a much greater effect than the E140Q (B4Q) mutation (site 4, C-domain) on the interaction of calmodulin with both peptides.(ABSTRACT TRUNCATED AT 250 WORDS)

Solution structure of the TR1C fragment of skeletal muscle troponin-C.

Residues 12-87 (TR1C fragment) of turkey skeletal muscle troponin-C comprises two helix-loop-helix calcium-binding motifs which are the regulatory calcium-binding sites in the N-terminal domain of the protein. We have used the combined distance geometry-simulated annealing protocol DGII (Havel, T. F. (1991) Prog. Biophys. Mol. Biol. 56, 43-78) to determine the structure of this 76-residue polypeptide in solution from 475 1H NMR-derived distance restraints. The nuclear Overhauser enhancement-derived distance constraints used in the DGII protocol were supplemented by introducing generic hydrogen bond distance restraints for slowly exchanging amide hydrogens in regular secondary structure elements, by restricting the available phi angle space to -180 degrees to 0 degrees for all residues except glycines, and by tailoring the distance boundaries used for quantitating the nuclear Overhauser enhancement intensities to correspond to characteristic distances found in helices. This improved the geometry of the four helices in the resulting structures. The relative positions of helices A and B which flank calcium-binding loop 1, helix D which follows calcium-binding loop 2, and the beta-sheet between the two calcium-binding loops were well defined and had an overall root-mean-square deviation for 20 converged structures of 1.4 +/- 0.2 A for backbone atoms. The structure and relative orientations of these regions are very similar to these of the corresponding regions of the protein in the crystal structure of intact turkey skeletal troponin C (Herzberg, O., and James, M. N. G. (1988) Nature 313, 653-659). The structure of helix C was well defined, but its relative position to the other helices was not defined. It occupied a range of positions in the set of 20 DGII structures, the average of which was quite similar to the orientation of helix C in the x-ray structure. The overall structure of the apo regulatory domain of troponin-C is therefore not affected by the loss of the N-helix, or the low pH conditions used for the x-ray structure, but may be more flexible in regions known to be involved in contacts with other skeletal muscle regulatory proteins.

1H-NMR resonance assignments, secondary structure, and global fold of the TR1C fragment of turkey skeletal troponin C in the calcium-free state.

The TR1C fragment of turkey skeletal muscle TnC (residues 12-87) comprises the two regulatory calcium binding sites of the protein. Complete assignments of the 1H-NMR resonances of the backbone and amino acid side chains of this domain in the absence of metal ions have been obtained using 2D 1H-NMR techniques. Sequential (i,i+1) and short-range (i,i+3) NOE connectivities define two helix-loop-helix calcium binding motifs, and long-range NOE connectivities indicate a short two-stranded beta-sheet formed between the two calcium binding loops. The two calcium binding sites are different in secondary structure. In terms of helix length, site II conforms to a standard "EF-hand" motif with the first helix ending one residue before the first calcium ligand and the second helix starting one residue after the beta-sheet. In site I, the first helix ends three residues before the first calcium ligand, and the second helix starts three residues after the beta-sheet. A number of long-range NOE connectivities between the helices define their relative orientation and indicate formation of a hydrophobic core between helices A, B, and D. The secondary structure and global fold of the TR1C fragment in solution in the calcium-free state are therefore very similar to those of the corresponding region in the crystal structure of turkey skeletal TnC [Herzberg, O., & James, M.N.G. (1988) J. Mol. Biol. 203, 761-779].

An improved procedure for the purification of formiminotransferase-cyclodeaminase from pig liver. Kinetics of the transferase activity with tetrahydropteroylpolyglutamates.

Formiminotransferase-cyclodeaminase is stabilized and activated approx. 40% in the presence of low concentrations (equal or less than 0.2%) of Triton X-100, possibly because the average hydrophobicity (1.10 kcal per residue) and the frequency of large non-polar side-chains (0.34) of this protein are both somewhat higher than average. This stabilization enabled us to develop a new purification procedure for the enzyme using chromatography on Matrex Gel Orange A and heparin-Sepharose columns in the presence of Triton X-100. This procedure is easier, much more reproducible, and gives slightly higher yield than the previous method described by Drury, et al. Further investigations of the role of tetrahydropteroylpolyglutamates with formiminotransferase-cyclodeaminase reveal that the use of polyglutamylated folate substrates does not change the mechanism of the transferase reaction, but decreases the K(m) for formininoglutamate, the second substrate, more than 10-fold, bringing it closer to the expected physiological concentration.

Renaturation of formiminotransferase-cyclodeaminase from guanidine hydrochloride. Quaternary structure requirements for the activities and polyglutamate specificity.

Formiminotransferase-cyclodeaminase denatured in 6 M guanidine hydrochloride (Gdn.HCl) refolds and reassembles to the native octameric structure upon dilution into buffer. Both enzymic activities are recovered to greater than 90%, and the renatured enzyme "channels" the formiminotetrahydropteroylpentaglutamate intermediate. Under conditions where the two activities are recovered simultaneously, the rate-limiting step in reactivation is first order with respect to protein, with k = 1.9 X 10(-5) s-1 at 22 degrees C and delta E approximately equal to 15 kcal mol-1. In the presence of 1.5 M urea, renaturation is arrested at the level of dimers having only transferase activity. Subsequent dialysis to remove the urea leads to recovery of deaminase activity and formation of octamer. Kinetic studies with mono- and pentaglutamate derivatives of the folate substrates demonstrated that native and renatured enzyme as well as deaminase-active dimers [Findlay, W. A., & MacKenzie, R. E (1987) Biochemistry 26, 1948-1954] have much higher affinity for polyglutamate substrates, while the transferase-active dimers do not. These results indicate that the transferase activity is associated with one type of subunit-subunit interaction in the native tetramer of dimers and that the polyglutamate binding site and the deaminase activity are associated with the other interface. A dimeric transferase-active fragment generated by limited proteolysis of the native enzyme can also be renatured from 6 M Gdn.HCl, confirming that it is an independently folding domain capable of reforming one type of subunit interaction.

Dissociation of the octameric bifunctional enzyme formiminotransferase-cyclodeaminase in urea. Isolation of two monofunctional dimers.

Partial denaturation of the circular octameric bifunctional enzyme formiminotransferase-cyclodeaminase in increasing urea concentrations leads to sequential dissociation via dimers to inactive monomers. In potassium phosphate buffer, dissociation to dimers in 3 M urea coincides with loss of both activities and a major decrease in intensity of intrinsic tryptophan fluorescence. In the presence of folic acid, these dimers retain the deaminase activity, but with folylpolyglutamates, both activities are protected and the native octameric structure is retained. The protection profiles with polyglutamates are cooperative with a Hill coefficient greater than 2, suggesting that binding of more than one folylpolyglutamate per octamer is required to stabilize the native structure. In triethanolamine hydrochloride buffer, transferase-active dimers that retain the intrinsic tryptophan fluorescence can be obtained in 1 M urea and stabilized at higher urea concentration by the addition of glutamate. Deaminase-active dimers are obtained by the protection of folate in 3 M urea. Proteolysis of the two kinds of dimers by chymotrypsin leads to very different fragmentation patterns, indicating that they are structurally different. We propose that the two dimers retain different subunit-subunit interfaces, one of which is required for each activity. This suggests that the native octameric structure is required for expression of both activities and therefore for "channeling" of intermediates.