Conservation and diversity of ancient hemoglobins in Bacteria

A group of single-domain proteins in Bacteria similar to thermoglobin, an oxygen-avid hemoglobin representative of the ancestral form, reveals the primordial structure, function, and evolvability of the family. Conserved residues at specific positions function to bind ligand or participate in hydrophobic packing of the protein core during protein folding. A potential hydrogen bond network consisting of a tyrosine and glutamine residue in the distal ligand-binding site of most hemoglobins suggests that the ancestral protein bound oxygen avidly. Two divergent hemoglobins with mutations at generally conserved positions contain non-canonical ligand-binding sites, illustrating plasticity of the fold. One binds heme in a manner similar to cytochromes and may represent an evolutionary link to the precursor of the hemoglobin fold. Conservation suggests specific biochemical properties of the ancestral protein; diversity suggests an evolvability of this group of hemoglobins tolerant of mutations that perturb conserved biochemical properties for adaptation to novel functions.     

The alphaD-globin gene originated via duplication of an embryonic alpha-like globin gene in the ancestor of tetrapod vertebrates

Gene duplication is thought to play an important role in the co-option of existing protein functions to new physiological pathways. The globin superfamily of genes provides an excellent example of the kind of physiological versatility that can be attained through the functional and regulatory divergence of duplicated genes that encode different subunit polypeptides of the tetrameric hemoglobin protein. In contrast to prevailing views about the evolutionary history of the alpha-globin gene family, here we present phylogenetic evidence that the alpha(A)- and alpha(D)-globin genes are not the product of a single, tandem duplication of an ancestral globin gene with adult function in the common ancestor of extant birds, reptiles, and mammals. Instead, our analysis reveals that the alpha(D)-globin gene of amniote vertebrates arose via duplication of an embryonic alpha-like globin gene that predated the radiation of tetrapods. The important evolutionary implication is that the distinct biochemical properties of alpha(D)-hemoglobin (HbD) are not exclusively derived characters that can be attributed to a post-duplication process of neofunctionalization. Rather, many of the distinct biochemical properties of HbD are retained ancestral characters that reflect the fact that the alpha(D)-globin gene arose via duplication of a gene that had a larval/embryonic function. These insights into the evolutionary origin of HbD illustrate how adaptive modifications of physiological pathways may result from the retention and opportunistic co-option of ancestral protein functions.     

A [2Fe-2S] protein from the hyperthermophilic bacterium Aquifex aeolicus.

Overexpression in Escherichia coli of the fdx4 gene from Aquifex aeolicus has allowed isolation and characterization of the first hyperthermophilic [2Fe-2S](Scys)(4) protein, a homodimer of M = 2 x 12.4 kDa with one [2Fe-2S] cluster per subunit. This protein is undamaged by heating to 100 degrees C for at least three hours. The primary structure, in particular the characteristic distribution of the four cysteine ligands of the metal site, and the spectroscopic properties of the A. aeolicus protein relate it to well characterized [2Fe-2S] proteins from Clostridium pasteurianum and Azotobacter vinelandii. These proteins are also homologous to subunits or domains of hydrogenases and NADH-ubiquinone oxidoreductase (Complex I) of respiratory chains. The A. aeolicus [2Fe-2S] protein is thus representative of a presumably novel protein fold involved in a variety of functions in very diverse cellular backgrounds.     

Expression, purification, and characterization of a new heterotetramer structure of leucyl-tRNA synthetase from Aquifex aeolicus in Escherichia coli

Aminoacyl-tRNA synthetases are key players in the interpretation of the genetic code. They constitute a textbook example of multi-domain proteins including insertion and terminal functional modules appended to one of the two class-specific active site domains. The non-catalytic domains usually have distinct roles in the aminoacylation reaction. Aquifex aeolicus leucyl-tRNA synthetase (LeuRS) is composed of a separated catalytic site and tRNA anticodon-binding site, which would represent one of the closest relics of the primordial aminoacyl-tRNA synthetase. Moreover, the essential catalytic site residues are split into the two different subunits. In all other class-I aminoacyl-tRNA synthetases, those two functional polypeptides are nowadays fused into a single protein chain. In this work, we report the isolation and the characterization, in Escherichia coli, of a novel oligomeric form (alphabeta)2 for A. aeolicus LeuRS, which is present in addition to the alphabeta heterodimer. A. aeolicus (alphabeta)2 LeuRS has been characterized by biochemical and biophysical methods. Native gel electrophoresis, mass spectrometry, analytical ultracentrifugation, and kinetic analysis confirmed that the (alphabeta)2 enzyme was a stable and active entity. By mass spectrometry we confirmed that the heterodimer alphabeta can bind one tRNALeu molecule whereas the heterotetramer (alphabeta)2 can bind two tRNALeu molecules. Active site titration and aminoacylation assays showed that two functional active sites are found per heterotetramer, suggesting that this molecular species might exist and be active in vivo. All those data suggest that the existence of the heterotetramer is certainly not an artifact of overexpression in E. coli.     

A hyperthermophilic plant-type [2Fe-2S] ferredoxin from Aquifex aeolicus is stabilized by a disulfide bond

A [2Fe-2S] ferredoxin (Fd1) from the hyperthermophilic bacterium Aquifex aeolicus has been obtained by heterologous expression of the encoding gene in Escherichia coli. Sequence comparisons show that this protein belongs to the extended family of plant- and mammalian-type [2Fe-2S] ferredoxins but also indicate that it is not closely similar to either the plant-type or mammalian-type subfamilies. Instead, it appears to bear some similarity to novel members of this family, in particular the Isc-type ferredoxins involved in the assembly of iron-sulfur clusters in vivo. The two redox levels of the [2Fe-2S](2+/+) metal site of A. aeolicus ferredoxin have been studied by UV-visible, resonance Raman, EPR, variable temperature magnetic circular dichroism, and M?ssbauer spectroscopies. A full-spin Hamiltonian analysis is given for the M?ssbauer spectra. In aggregate, the spectroscopic data reveal differences with both the plant-type and mammalian-type ferredoxins, in keeping with the sequence comparisons. The midpoint potential of the [2Fe-2S](2+/+) couple, at -375 mV versus the normal hydrogen electrode, is more negative than those of mammalian-type ferredoxins and at the upper end of the range covered by plant-type ferredoxins. A. aeolicus ferredoxin contains two cysteines in addition to the four that are committed as ligands of the [2Fe-2S] cluster. These two residues have been shown by chemical modification and site-directed mutagenesis to form a disulfide bridge in the native protein. While that cystine unit plays a significant role in the exceptional thermostability of A. aeolicus ferredoxin (T(m) = 121 degrees C at pH 7 versus T(m) = 113 degrees C in a molecular variant where the disulfide bridge has been removed), it does not bear on the properties of the [2Fe-2S](2+/+) chromophore. This observation is consistent with the large distance (ca. 20 A) that is predicted to separate the iron-sulfur chromophore from the disulfide bridge.     

A novel carbamoyl-phosphate synthetase from Aquifex aeolicus

Aquifex aeolicus, an extreme hyperthermophile, has neither a full-length carbamoyl-phosphate synthetase (CPSase) resembling the enzyme found in all mesophilic organisms nor a carbamate kinase-like CPSase such as those present in several hyperthermophilic archaea. However, the genome has open reading frames encoding putative proteins that are homologous to the major CPSase domains. The glutaminase, CPS.A, and CPS.B homologs from A. aeolicus were cloned, overexpressed in Escherichia coli, and purified to homogeneity. The isolated proteins could catalyze several partial reactions but not the overall synthesis of carbamoyl phosphate. However, a stable 124-kDa complex could be reconstituted from stoichiometric amounts of CPS.A and CPS.B proteins that synthesized carbamoyl phosphate from ATP, bicarbonate, and ammonia. The inclusion of the glutaminase subunit resulted in the formation of a 171-kDa complex that could utilize glutamine as the nitrogen-donating substrate, although the catalytic efficiency was significantly compromised. Molecular modeling, using E. coli CPSase as a template, showed that the enzyme has a similar structural organization and interdomain interfaces and that all of the residues known to be essential for function are conserved and properly positioned. A steady state kinetic study at 78 degrees C indicated that although the substrate affinity was similar for bicarbonate, ammonia, and glutamine, the K(m) for ATP was appreciably higher than that of any known CPSase. The A. aeolicus complex, with a split gene encoding the major synthetase domains and relatively inefficient coupling of amidotransferase and synthetase functions, may be more closely related to the ancestral precursor of contemporary mesophilic CPSases.     

A robust metallo-oxidase from the hyperthermophilic bacterium Aquifex aeolicus

The gene, Aquifex aeolicus AAC07157.1, encoding a multicopper oxidase (McoA) and localized in the genome as part of a putative copper-resistance determinant, has been cloned, over-expressed in Escherichia coli and the recombinant enzyme purified to homogeneity. The purified enzyme shows spectroscopic and biochemical characteristics typical of the well-characterized multicopper oxidase family of enzymes. McoA presents higher specificity (k(cat)/K(m)) for cuprous and ferrous ions than for aromatic substrates and is therefore designated as a metallo-oxidase. Addition of copper is required for maximal catalytic efficiency. A comparative model structure of McoA has been constructed and a striking structural feature is the presence of a methionine-rich region (residues 321-363), reminiscent of those found in copper homeostasis proteins. The kinetic properties of a mutant enzyme, McoADeltaP321-V363, deleted in the methionine-rich region, provide evidence for the key role of this region in the modulation of the catalytic mechanism. McoA has an optimal temperature of 75 degrees C and presents remarkable heat stability at 80 and 90 degrees C, with activity lasting for up to 9 and 5 h, respectively. McoA probably contributes to copper and iron homeostasis in A. aeolicus.     

Structural and biochemical characterization of a novel Mn2+-dependent phosphodiesterase encoded by the yfcE gene

Escherichia coli YfcE belongs to a conserved protein family within the calcineurin-like phosphoesterase superfamily (Pfam00149) that is widely distributed in bacteria and archaea. Superfamily members are metallophosphatases that include monoesterases and diesterases involved in a variety of cellular functions. YfcE exhibited catalytic activity against bis-p-nitrophenyl phosphate, a general substrate for phosphodiesterases, and had an absolute requirement for Mn2+. However, no activity was observed with phosphodiesters and over 50 naturally occurring phosphomonoesters. The crystal structure of the YfcE phosphodiesterase has been determined to 2.25 A resolution. YfcE has a beta-sandwich architecture similar to metallophosphatases of common ancestral origin. Unlike its more complex homologs that have added structural elements for regulation and substrate recognition, the relatively small 184-amino-acid protein has retained its ancestral simplicity. The tetrameric protein carries two zinc ions per active site from the E. coli extract that reflect the conserved di-Mn2+ active site geometry. A cocrystallized sulfate inhibitor mimics the binding of phosphate moeities in known ligand/phosphatase complexes. Thus, YfcE has a similar active site and biochemical mechanism as well-characterized superfamily members, while the YfcE phosphodiester-containing substrate is unique.     

A present-day aminoacyl-tRNA synthetase with ancestral editing properties

Leucyl-, isoleucyl-, and valyl-tRNA synthetases form a subgroup of related aminoacyl-tRNA synthetases that attach similar amino acids to their cognate tRNAs. To prevent amino acid misincorporation during translation, these enzymes also hydrolyze mischarged tRNAs through a post-transfer editing mechanism. Here we show that LeuRS from the deep-branching bacterium Aquifex aeolicus edits the complete set of aminoacylated tRNAs generated by the three enzymes: Ile-tRNA(Ile), Val-tRNA(Ile), Val-tRNA(Val), Thr-tRNA(Val), and Ile-tRNA(Leu). This unusual enlarged editing property was studied in a model of a primitive editing system containing a composite minihelix carrying the triple leucine, isoleucine, and valine identity mimicking the primitive tRNA precursor. We found that the freestanding LeuRS editing domain can edit this precursor in contrast to IleRS and ValRS editing domains. These results suggest that A. aeolicus LeuRS carries editing properties that seem more primitive than those of IleRS and ValRS. They suggest that the A. aeolicus editing domain has preserved the ambiguous editing property from the ancestral common editing domain or, alternatively, that this plasticity results from a specific metabolic adaptation.     

The PUF family of RNA-binding proteins: does evolutionarily conserved structure equal conserved function?

Drosophila Pumilio (Pum) protein is a founder member of a novel family of RNA-binding proteins, known as the PUF family. The PUF proteins constitute an evolutionarily highly conserved family of proteins present from yeast to humans and plants, and are characterized by a highly conserved C-terminal RNA-binding domain, composed of eight tandem repeats. The conserved biochemical features and genetic function of PUF family members have emerged from studies of model organisms. PUF proteins bind to related sequence motifs in the 3' untranslated region (3'UTR) of specific target mRNAs and repress their translation. Frequently, PUF proteins function asymmetrically to create protein gradients, thus causing asymmetric cell division and regulating cell fate specification. Thus, it was recently proposed that the primordial role of PUF proteins is to sustain mitotic proliferation of stem cells. Here we review the evolution, conserved genetic and biochemical properties of PUF family of proteins, and discuss protein interactions, upstream regulators and downstream targets of PUF proteins. We also suggest that a conserved mechanism of PUF function extends to the newly described mammalian members of the PUF family (human PUM1 and PUM2, and mouse Pum1 and Pum2), that show extensive homology to Drosophila Pum, and could have an important role in cell development, fate specification and differentiation.     

The reconstruction of a fragment of primordial genetic information from modern proteins: PGI-FI (AHAP).

In a continuation of earlier studies the connection between a reconstructed ancestral histone IV peptide and various sequences from non-histone proteins was investigated. This peptide, AHAP, was found to be related to partial peptides from the human encephalitogenic protein, immunoglobulin L-chains, fibrinopeptide A, glyceraldehyde-3-phosphate-dehydrogenase, TMV coat protein and several other proteins and protein families; for comparison, nodal sequences were employed wherever possible. The widespread occurrence of genetic information expressed in AHAP caused us to rename the peptide PGI-FI (primordial genetic information-fragment 1) since this peptide is apparently part of a very primitive and ancient genetic information, and may be called "protogene".     

Evolution of the albumin: alpha-fetoprotein ancestral gene from the amplification of a 27 nucleotide sequence

The genes for alpha-fetoprotein and albumin arose by duplication of an ancestral gene that contained three genetic domains. These domains were generated by the triplication of a primordial genetic domain composed of five exons or subdomains. That the primordial domain itself arose by amplification of a simpler sequence is suggested by nucleotide sequence homologies among the subdomains of the mouse alpha-fetoprotein gene. A detailed analysis of these homologies reveals that each of the five subdomain families contains remnants of a 27-base-long repeat from which the entire alpha-fetoprotein coding sequence has been assembled. A consensus sequence for the 27 nucleotide repeat is derived, and the positions of the repeats within each subdomain are described. A model is proposed for the evolution of the primordial domain by the amplification and divergence of the 27 base-pair sequence, along with the condensation of the repeats into subdomains separated by intervening sequences. It is postulated that the role of intervening sequences may be to limit sequence amplification in genes such as alpha-fetoprotein and albumin whose protein products cannot tolerate size variation.     

tRNA-rRNA sequence homologies: a model for the origin of a common ancestral molecule, and prospects for its reconstruction

A model is proposed for the early evolution of the coding mechanism. A primordial RNA embodies the functions of today's nucleic acids in a single molecule. The molecule is generated by successive rounds of self-priming and -templating. After proximity is assured by enclosure in a cell, the functions can be partitioned among more efficient specialized molecules. The prediction of sequence homologies in later forms prompted a search for matches between t- and r-RNAs. These are described. Their distributions offer clues to their origins. The existance of overlapping homologies indicates an approach to the reconstruction of an ancestral molecule.     

RusA proteins from the extreme thermophile Aquifex aeolicus and lactococcal phage r1t resolve Holliday junctions

The RusA protein of Escherichia coli is a DNA structure-specific endonuclease that resolves Holliday junction intermediates formed during DNA replication, recombination and repair by introducing symmetrically paired incisions 5' to CC dinucleotides. It is encoded by the defective prophage DLP12, which raises the possibility that it may be of bacteriophage origin. We show that rusA-like sequences are indeed often associated with prophage sequences in the genomes of several bacterial species. They are also found in many bacteriophages, including Lactococcus lactis phage r1t. However, rusA is also present in the chromosome of the hyperthermophilic bacterium Aquifex aeolicus. In this case, there is no obvious association of rusA with prophage-like sequences. Given the ancient lineage of Aquifex aeolicus, this observation provides the first indication that RusA may be of bacterial origin. The RusA proteins of A. aeolicus and bacteriophage r1t were purified and shown to resolve Holliday junctions. The r1t enzyme also promotes DNA repair in strains lacking the RuvABC resolvase. Both enzymes cleave junctions in a sequence-dependent manner, but the A. aeolicus RusA shows a different sequence preference (3' to TG) from the E. coli protein (5' to CC), and the r1t RusA has relaxed sequence dependence, requiring only a single cytosine.     

tRNA-rRNA sequence homologies: A model for the origin of a common ancestral molecule,and prospects for its reconstruction

A model is proposed for the early evolution of the coding mechanism. A primordial RNA embodies the functions of today's nucleic acids in a single molecule. The molecule is generated by successive rounds of self-priming and-templating. After proximity is assured by enclosure in a cell, the functions can be partitioned among more efficient specializel molecules. The prediction of sequence homologies in later forms prompted a search for matches between t- and r-RNAs. These are described. Their distributions offer clues to their origins. The existance of overlapping homologies indicates an approach to the reconstruction of an ancestral molecule.     

General Stress Protein CTC from Bacillus subtilis Specifically Binds to Ribosomal 5S RNA

Two recombinant proteins of the CTC family were prepared: the general stress protein CTC from Bacillus subtilis and its homolog from Aquifex aeolicus. The general stress protein CTC from B. subtilis forms a specific complex with 5S rRNA and its stable fragment of 60 nucleotides, which contains internal loop E. The ribosomal protein TL5 from Thermus thermophilus, which binds with high affinity to 5S rRNA in the loop E region, was also shown to replace the CTC protein from B. subtilis in its complexes with 5S rRNA and its fragment. The findings suggest that the protein CTC from B. subtilis binds to the same site on 5S rRNA as the protein TL5. The protein CTC from A. aeolicus, which is 50 amino acid residues shorter from the N-terminus than the proteins TL5 from T. thermophilus and CTC from B. subtilis, does not interact with 5S rRNA.     

Cloning and characterization of thermostable endoglucanase (Cel8Y) from the hyperthermophilic Aquifex aeolicus VF5

Aquifex aeolicus is the hyperthermophilic bacterium known, with growth-temperature maxima near 95 degrees C. The cel8Y gene, encoding a thermostable endoglucanase (Cel8Y) from Aquifex aeolicus VF5, was cloned into a vector for expression and expressed in Escherichia coli XL1-Blue. A clone of 1.7 kb fragment containing endoglucanase activity, designated pKYCY100, was sequenced and found to contain an ORF of 978 bp encoding a protein of 325 amino acid residues, with a calculated molecular mass of 38,831 Da. This endoglucanase was designated cel8Y gene. The endoglucanase has an 18-amino-acid signal peptide but not cellulose-binding domain. The endoglucanase of A. aeolicus VF5 had significant amino acid sequence similarities with endoglucanases from glycosyl hydrolase family 8. The predicted amino acid sequence of the Cel8Y protein was similar to that of CMCase of Cellulomonas uda, BcsC of Escherichia coli, CelY of Erwinia chrysanthemi, and CMCase of Acetobacter xylinum. The molecular mass of Cel8Y was calculated to be 36,750 Da, which is consistent with the value obtained from result of CMC-SDS-PAGE of the purified enzyme. Cel8Y was thermostable, exhibiting maximal activity at 80 degrees C and pH optima of 7.0 and with half-lives of 2 h at 100 degrees C, 4 h at 90 degrees C.     

IFITM/Mil/fragilis family proteins IFITM1 and IFITM3 play distinct roles in mouse primordial germ cell homing and repulsion

The family of interferon-induced transmembrane protein (Ifitm/mil/fragilis) genes encodes cell surface proteins that may modulate cell adhesion and influence cell differentiation. Mouse Ifitm1 and -3, which are expressed in primordial germ cells (PGCs), are implicated to have roles in germ cell development, but the specific functions have been unclear. Our results show that Ifitm1 activity is required for PGC transit from the mesoderm into the endoderm, and that it appears to act via a repulsive mechanism, such that PGCs avoid Ifitm1-expressing tissues. In contrast, Ifitm3, which is expressed in migratory PGCs, is sufficient to confer autonomous PGC-like homing properties to somatic cells. These guidance activities are mediated by the N-terminal extracellular domain of the specific IFITM, which cannot be substituted by that of another family member. Complex homo- and/or heterotypic intercellular interactions among various IFITMs in PGCs and neighboring cells may underpin coordinated germ cell guidance in mice.     

Differential Effects of Replacing Escherichia coli Ribosomal Protein L27 with Its Homologue from Aquifex aeolicus

The rpmA gene, which encodes 50S ribosomal subunit protein L27, was cloned from the extreme thermophile Aquifex aeolicus, and the protein was overexpressed and purified. Comparison of the A. aeolicus protein with its homologue from Escherichia coli by circular dichroism analysis and proton nuclear magnetic resonance spectroscopy showed that it readily adopts some structure in solution that is very stable, whereas the E. coli protein is unstructured under the same conditions. A mutant of E. coli that lacks L27 was found earlier to be impaired in the assembly and function of the 50S subunit; both defects could be corrected by expression of E. coli L27 from an extrachromosomal copy of the rpmA gene. When A. aeolicus L27 was expressed in the same mutant, an increase in the growth rate occurred and the "foreign" L27 protein was incorporated into E. coli ribosomes. However, the presence of A. aeolicus L27 did not promote 50S subunit assembly. Thus, while the A. aeolicus protein can apparently replace its E. coli homologue functionally in completed ribosomes, it does not assist in the assembly of E. coli ribosomes that otherwise lack L27. Possible explanations for this paradoxical behavior are discussed.