Structural Subdomains Research Articles

Limited proteolysis of bovine alpha-lactalbumin: isolation and characterization of protein domains.

The partly folded states of alpha-lactalbumin (alpha-LA) exposed to acid solution at pH 2.0 (A-state) or at neutral pH upon EDTA-mediated removal of the single protein-bound calcium ion (apo form) have been probed by limited proteolysis experiments. These states are nowadays commonly considered to be molten globules and thus protein-folding intermediates. Pepsin was used for proteolysis at acid pH, while proteinase K and chymotrypsin at neutral pH. The expectations were that these proteolytic probes would detect sites and/or chain regions in the partly folded states of alpha-LA sufficiently dynamic, or even unfolded, capable of binding and adaptation to the specific stereochemistry of the protease's active site. A time-course analysis of the proteolytic events revealed that the fast, initial proteolytic cuts of the 123-residue chain of alpha-LA in its A-state or apo form by the three proteases occur at the same chain region 39-54, the actual site(s) of cleavage depending upon the protease employed. This region in native alpha-LA encompasses the beta-sheets of the protein. Subsequent cleavages occur mostly at chain regions 31-35 and 95-105. Four fragment species of alpha-LA have been isolated by reverse-phase high-performance liquid chromatography, and their conformational properties examined by circular dichroism and fluorescence emission spectroscopy. The single chain fragment 53-103, containing all the binding sites for calcium in native alpha-LA and cross-linked by two disulfide bridges, maintains in aqueous buffer and in the presence of calcium ions a folded structure characterized by the same content of alpha-helix of the corresponding chain segment in native alpha-LA. Evidence for some structure was also obtained for the two-chain species 1-40 and 104-123, as well as 1-31 and 105-123, both systems being covalently linked by two disulfide bonds. In contrast, the protein species given by fragment 1-34 connected to fragment 54-123 or 57-123 via four disulfide bridges adopts in solution a folded structure with the helical content expected for a native-like conformation. Of interest, the proteolytic fragment species herewith isolated correspond to the structural domains and subdomains of alpha-LA that can be identified by computational analysis of the three-dimensional structure of native alpha-LA (Siddiqui AS, Barton GI, 1995, Protein Sci 4:872-884). The fast, initial cleavages at the level of the beta-sheet region of native alpha-LA indicate that this region is highly mobile or even unfolded in the alpha-LA molten globule(s), while the rest of the protein chain maintains sufficient structure and rigidity to prevent extensive proteolysis. The subsequent cleavages at chain segment 95-105 indicate that also this region is somewhat mobile in the A-state or apo form of the protein. It is concluded that the overall domain topology of native alpha-LA is maintained in acid or at neutral pH upon calcium depletion. Moreover, the molecular properties of the partly folded states of alpha-LA deduced here from proteolysis experiments do correlate with those derived from previous NMR and other physicochemical measurements.

Functional analysis of microtubule-binding domain of bovine MAP4.

Bovine microtubule-associated protein 4 (MAP4) consists of an amino-terminal projection domain and a carboxyl-terminal microtubule-binding domain. The carboxyl-terminal domain of MAP4 is further divided into three subdomains: a region rich in proline and basic residues (Pro-rich region), a region containing four repeats of an assembly-promoting (AP) sequence, which consists of 22 amino acid residues (AP sequence region), and a hydrophobic tail region (Tail region). The subdomain structure of MAP4 microtubule binding domain is similar to those of other MAPs (MAP2 and tau). In order to study the function of each subdomain per se of bovine MAP4 microtubule-binding domain, we purified a series of truncated fragments of MAP4, expressed in Escherichia coil. Binding affinity of the PA4T fragment (containing the Pro-rich region, the AP sequence region and the Tail region) is only four times higher than that of the A4T fragment (containing the AP sequence region and the Tail region), while the microtubule nucleating activity of the PA4T fragment is far greater. We propose that the Pro-rich region promotes the nucleation of microtubule assembly. The A4 fragment (corresponding to the AP sequence region) stimulated the assembly of tubulin into coldstable amorphous aggregates. The AP sequence region of MAP4 failed to promote microtubule assembly. On the other hand, the fragment has an activity to stimulate microtubule elongation. The function of the MAP4 Tail region is not clear at present. The A4T fragment (containing the AP sequence region and the Tail region) promote both microtubule nucleation and elongation step, but the A4 fragment only promotes microtubule elongation, suggesting that the Tail region is indispensable for the nucleation step. However, the fragment containing only the Tail region could not bind to microtubule. Although MAP4 was considered to be long, thin and flexible molecule, never the Tail region may contribute to be the proper folding of MAP4, and/or may interact with other molecules. We concluded that both the Pro-rich region and the AP sequence region take part in the promotion of tubulin polymerization, and that the former is important for the lateral protofilament-protofilament interaction, and the latter is important for the longitudinal affinity between each tubulin dimer in a protofilament.

Self-organizing neural networks bridge the biomolecular resolution gap

Topology-representing neural networks are employed to generate pseudo-atomic structures of large-scale protein assemblies by combining high-resolution data with volumetric data at lower resolution. As an application example, actin monomers and structural subdomains are located in a three-dimensional (3D) image reconstruction from electron micrographs. To test the reliability of the method, the resolution of the atomic model of an actin polymer is lowered to a level typically encountered in electron microscopic reconstructions. The atomic model is restored with a precision nine times the nominal resolution of the corresponding low-resolution density. The presented self-organizing computing method may be used as an information-processing tool for the synthesis of structural data from a variety of biophysical sources.

Stable rearrangements of the beta3-beta4 hairpin loop of HIV-1 reverse transcriptase in plasma viruses from patients receiving combination therapy.

To study the genetic rearrangements of HIV-1 reverse transcriptase (RT) in circulating viruses from patients under combination therapy, and to determine the impact of these changes on the virological response to treatment. Blood samples were extracted from total RNA and amplified by RT-PCR. The HIV-1 RT and protease genes were sequenced by fluorescent dye terminator cycle sequencing. Specific rearrangements in the RT coding region (between amino acids 66 and 71) were documented in nine patients. This region, which corresponds to a loop between the beta3 and beta4 strands of the fingers subdomain of RT, is involved in the interaction between the enzyme and the template primer. In vitro data with recombinant enzymes have shown the importance of this domain in the processive polymerization of HIV-1 RT. The rearrangements (eight deletions/insertions and one deletion with conservation of the reading frame) did not affect the overall secondary structure of the fingers subdomain, as assessed by the Garnier Osguthorpe Robson prediction method. The changes were generally stable over a follow-up of 10-12 months. With the exception of two cases, most of the patients of this study did not respond efficiently to antiretroviral therapy as assessed by measurements of plasma viraemia. Correspondingly, the RT and protease genes sequenced from these patients displayed numerous resistance-associated mutations. Functional and stable rearrangements in the beta3-beta4 hairpin of HIV-1 RT can be found in circulating viruses from patients under combination therapy. These rearrangements may affect the virological response to antiretroviral therapy by increasing the processivity of RT, an enzymatic parameter that reflects the fidelity of the polymerization process.

Structural dissection of the DNA-binding domain of the yeast transcriptional activator GAL4 reveals an alpha-helical region responsible for dimerization.

Limited proteolysis of the DNA-binding domain (residues 1-147) of the yeast transcriptional activator GAL4 has been used to define more precisely the subdomain structure required for DNA binding and dimerization. Two regions of the protein were found to be resistant to proteolysis: the cysteine-rich, zinc-binding region (residues 6-43) and a hydrophobic sequence between residues 52 and 97. Carboxy-terminal deletion fragments of the DNA-binding domain were generated and assayed by DNase 1 footprinting. This showed that the affinity of DNA binding depends on the sequence between residues 65 and 94. Structural comparisons by UV circular dichroism (CD) were made and the difference CD spectra indicate that strong alpha-helical content is found specifically in the region between residues 65 and 94, which previous studies have shown to enable dimerization and in this study the formation of a stable protein-DNA complex.

RNA folding causes secondary structure rearrangement.

The secondary structure of the P5abc subdomain (a 56-nt RNA) of the Tetrahymena thermophila group I intron ribozyme has been determined by NMR. Its base pairing in aqueous solution in the absence of magnesium ions is significantly different from the RNA in a crystal but is consistent with thermodynamic predictions. On addition of magnesium ions, the RNA folds into a tertiary structure with greatly changed base pairing consistent with the crystal structure: three Watson-Crick base pairs, three G.U base pairs, and an extra-stable tetraloop are lost. The common assumption that RNA folds by first forming secondary structure and then forming tertiary interactions from the unpaired bases is not always correct.

Construct Validity of Scores on a Developmental Assessment with Mathematical Patfierns Tasks

Three analytic approaches were employed within the framework of classical test theory to examine construct validity of a mathematics assessment made up of 16 ordered constructed-response tasks. The tasks focused on problem solving by identifying and explaining mathematical patterns. Descriptive analyses across four age groups suggested a developmental structure of individual tasks and subdomains that was generally consistent with the intended design of the test. Acceptable fit was obtained using structural equation modeling procedures for a five-factor, first-order structure with directional paths specified (RMSEA = .068) and for two second-order factors (RMSEA = .068). Convergent validity coefficients with standardized tests in mathematics computation, concepts of numbers, and applications indicated a modest overlap in constructs across the two assessments (r = .47) with Stanford Achievement Test (SAT) total mathematics scores. Findings are discussed with reference to intended uses of the assessment and its scores in the context of current mathematics education reforms.

Molecular Cloning and Characterization of a Novel Nuclear Protein Kinase in Mice

We cloned cDNAs which encode a mouse liver nuclear protein with an apparent molecular mass of 51 kDa, using sequences derived from a purified protein as the basis for designing specific primers. The deduced amino acid sequences revealed that the 51-kDa protein contains characteristic subdomain structures of a protein kinase. The bacterially expressed recombinant 51-kDa protein catalyzed phosphorylation of general substrates such as casein and was autophosphorylated at serine residue(s). This 51-kDa protein kinase, designated 51PK, is 40% identical to the 34-kDa protein kinase encoded by the vaccinia virus B1 gene and 25% identical to the casein kinase I isoforms, including yeast HRR25. The 51PK mRNA was expressed as two splice variants and the 51PK protein was exclusively localized in nuclei. Northern hybridization showed that 51PK mRNA was expressed in various tissues, with highest levels in testis, spleen, lung, and liver. These results, therefore, indicate that 51PK is a nuclear serine/threonine kinase and a novel distinct member of the protein kinase superfamily.

Two structural subdomains of barstar detected by rapid mixing NMR measurement of amide hydrogen exchange

Equilibrium amide hydrogen exchange studies of barstar have been carried out at pH 6.7, 32 degrees C using one- and two-dimensional nuclear magnetic resonance. An unusually large fraction of the backbone amide hydrogens of barstar exchange too fast to be measured, and the exchange rates of only fifteen slow-exchanging amide sites including indole amides of two tryptophans could be measured in the presence of 0 to 1.8 M guanidine hydrochloride (GdnHCl). Measurement of exchange occurring in tens of seconds in the unfolding transition region was possible by the use of a fast stopped-flow mixing method. The observed exchange rates have been simulated in the EX2 limit according to a two-process model that incorporates two exchange-competent states: a transiently unfolded state (U*) in which many amide hydrogens are completely accessible to solvent-exchange, and a near-native locally unfolded state (N*), in which only one or a few amide hydrogens are completely accessible to solvent-exchange. The two-process model appears to account for the observed exchange behavior over the entire range of GdnHCl concentrations studied. For several measurable slow-exchanging amide hydrogens, the free energies of production of exchange-competent states from the exchange-incompetent native state are significantly higher than the free-energy of production of the equilibrium unfolded state from the native state, when the latter is determined from circular dichroism- or fluorescence-monitored equilibrium unfolding curves. The result implies that U*, which forms transiently in the strongly native-like conditions used for the hydrogen exchange studies, is higher in energy than the equilibrium-unfolded state. The higher energy of this transiently unfolded exchange-competent state can be attributed to either proline isomerization or to the presence of residual structure. On the basis of the free energies of production of exchange-competent states, the measured amide sites of barstar appear to define two structural subdomains--a three-helix unit and a two-beta-strand unit in the core of the protein.

Two structural subdomains of barstar detected by rapid mixing NMR measurement of amide hydrogen exchange

Two structural subdomains of barstar detected by rapid mixing NMR measurement of amide hydrogen exchange

Solution structure of the Mu end DNA-binding ibeta subdomain of phage Mu transposase: modular DNA recognition by two tethered domains.

The phage Mu transposase (MuA) binds to the ends of the Mu genome during the assembly of higher order nucleoprotein complexes. We investigate the structure and function of the MuA end-binding domain (Ibetagamma). The three-dimensional solution structure of the Ibeta subdomain (residues 77-174) has been determined using multidimensional NMR spectroscopy. It comprises five alpha-helices, including a helix-turn-helix (HTH) DNA-binding motif formed by helices 3 and 4, and can be subdivided into two interacting structural elements. The structure has an elongated disc-like appearance from which protrudes the recognition helix of the HTH motif. The topology of helices 2-4 is very similar to that of helices 1-3 of the previously determined solution structure of the MuA Igamma subdomain and to that of the homeodomain family of HTH DNA-binding proteins. We show that each of the two subdomains binds to one half of the 22 bp recognition sequence, Ibeta to the more conserved Mu end distal half (beta subsite) and Igamma to the Mu end proximal half (gamma subsite) of the consensus Mu end-binding site. The complete Ibetagamma domain binds the recognition sequence with a 100- to 1000-fold higher affinity than the two subdomains independently, indicating a cooperative effect. Our results show that the Mu end DNA-binding domain of MuA has a modular organization, with each module acting on a specific part of the 22 bp binding site. Based on the present binding data and the structures of the Ibeta and Igamma subdomains, a model for the interaction of the complete Ibetagamma domain with DNA is proposed.

Structure-function analysis of the triphosphatase component of vaccinia virus mRNA capping enzyme.

The N-terminal 60 kDa (amino acids 1 to 545) of the D1 subunit of vaccinia virus mRNA capping enzyme is an autonomous bifunctional domain with triphosphatase and guanylyltransferase activities. We previously described two alanine cluster mutations, R77 to A (R77A)-K79A and E192A-E194A, which selectively inactivated the triphosphatase component. Here, we characterize the activities of 11 single alanine mutants-E37A, E39A, Q60A, E61A, T67A, T69A, K75A, R77A, K79A, E192A, and E194A-and a quadruple mutant in which four residues (R77, K79, E192, and E194) were replaced by alanine. We report that Glu-37, Glu-39, Arg-77, Glu-192, and Glu-194 are essential for gamma-phosphate cleavage. The five essential residues are conserved in the capping enzymes of Shope fibroma virus, molluscum contagiosum virus, and African swine fever virus. Probing the structure of D1(1-545) by limited V8 proteolysis suggested a bipartite subdomain structure. The essential residue Glu-192 is the principal site of V8 cleavage. Secondary cleavage by V8 occurs at the essential residue Glu-39. The triphosphatase-defective quadruple mutant transferred GMP to the triphosphate end of poly(A) to form a tetraphosphate cap structure, GppppA. We report that GppppA-capped RNA is a poor substrate for cap methylation by the vaccinia virus and Saccharomyces cerevisiae RNA (guanine-7) methyltransferases. The transcription termination factor activity of the D1-D12 capping enzyme heterodimer was not affected by mutations that abrogated ATPase activity. Thus, the capping enzyme is not responsible for the requirement for ATP hydrolysis during transcription termination.

Atomic structure of a thermostable subdomain of HIV-1 gp41.

Infection by HIV-1 involves the fusion of viral and cellular membranes with subsequent transfer of viral genetic material into the cell. The HIV-1 envelope glycoprotein that mediates fusion consists of the surface subunit gp120 and the transmembrane subunit gp41. gp120 directs virion attachment to the cell-surface receptors, and gp41 then promotes viral-cell membrane fusion. A soluble, alpha-helical, trimeric complex within gp41 composed of N-terminal and C-terminal extraviral segments has been proposed to represent the core of the fusion-active conformation of the HIV-1 envelope. A thermostable subdomain denoted N34(L6)C28 can be formed by the N-34 and C-28 peptides connected by a flexible linker in place of the disulfide-bonded loop region. Three-dimensional structure of N34(L6)C28 reveals that three molecules fold into a six-stranded helical bundle. Three N-terminal helices within the bundle form a central, parallel, trimeric coiled coil, whereas three C-terminal helices pack in the reverse direction into three hydrophobic grooves on the surface of the N-terminal trimer. This thermostable subdomain displays the salient features of the core structure of the isolated gp41 subunit and thus provides a possible target for therapeutics designed selectively to block HIV-1 entry.

Solution structure of the Iγ subdomain of the Mu end DNA-binding domain of phage Mu transposase

The MuA transposase of phase Mu is a large modular protein that plays a central role in transposition. We show that the Mu end DNA-binding domain, Iβγ, which is responsible for binding the DNA attachment sites at each end of the Mu genome, comprises two subdomains, Iβ and Iγ, that are structurally autonomous and do not interact with each other in the absence of DNA. The solution structure of the Iγ subdomain has been determined by multidimensional NMR spectroscopy. The structure of Iγ comprises a four helix bundle and, despite the absence of any significant sequence identity, the topology of the first three helices is very similar to that of the homeodomain family of helix-turn-helix DNA-binding proteins. The helix-turn-helix motif of Iγ, however, differs from that of the homeodomains in so far as the loop is longer and the second helix is shorter, reminiscent of that in the POU-specific domain.

Alpha-fetoprotein as a biologic response modifier: relevance to domain and subdomain structure.

In the present review, the structure of alpha-fetoprotein (AFP) is discussed in consideration of AFP membership and position in the albuminoid supergene family in relation to other gene family members. Ontogenetic AFP gene expression is then discussed in view of AFP mRNA presence in various tissues at different times during development. The multiple molecular forms of AFP is also presented in relation to published reports of AFP binding proteins and cell surface receptors. The review proceeds on to present AFP as a potential model of a modular/cassette protein based on sequence comparison with cleaved fragments of prohormones and biological response modifiers. Such cleaved fragments could potentially serve as peptide messengers for vascular, neuroendocrine, and digestive biological activities. Following a discussion on fibrin binding and serine proteases, AFP-cytoskeletal, extracellular matrix, and cellular adhesion interactions are considered. AFP as a carrier/transport protein based on structural relationships is further elucidated by examination of the various ligands bound to AFP and its hormone interaction. Since AFP binds heavy metals, the question is posed of whether AFP could function as an antioxidant. An analysis of transcription factors, tumor suppressors, and homeodomain proteins follows, which is interfaced with the concept of programmed cell death in light of amino acid sequence matches detected on the AFP molecule. Emphasis was naturally placed upon the homeodomain protein sequence stretches since AFP is a fetal, phase-specific protein found throughout embryogenesis, histogenesis, and organogenesis. In keeping with histogenesis, a discussion of AFP and eye lens protein development is presented. Finally, AFP sequence analysis presented in light of members of the immunoglobulin superfamily, autoimmune disorders, and various disease states culminates the review. A closing discussion then summarizes regions of presumptive matched protein identities on each of AFP's three domains.

Base Miscoding and Strand Misalignment Errors by Mutator Klenow Polymerases with Amino Acid Substitutions at Tyrosine 766 in the O Helix of the Fingers Subdomain

A mutant derivative of Klenow fragment DNA polymerase containing serine substituted for tyrosine at residue 766 has been shown by kinetic analysis to have an increased misinsertion rate relative to wild-type Klenow fragment, but a decreased rate of extension from the resulting mispairs (Carroll, S. S., Cowart, M., and Benkovic, S. J. (1991) Biochemistry 30, 804-813). In the present study we use an M13mp2-based fidelity assay to study the error specificity of this mutator polymerase. Despite its compromised ability to extend mispairs, the Y766S polymerase and a Y766A mutant both have elevated base substitution error rates. The magnitude of the mutator effect is mispair-specific, from no effect for some mispairs to rates elevated by 60-fold for misincorporation of TMP opposite template G. The results with the Y766S mutant are remarkably consistent with the earlier kinetic analysis of misinsertion, demonstrating that either approach can be used to identify and characterize mutator polymerases. Both the Y766S and Y766A mutant polymerases are also frameshift mutators, having elevated rates for two-base deletions and a 276-base deletion between a direct repeat sequence. However, neither mutant polymerase has an increased error rate for single-base frameshifts in repetitive sequences. This error specificity suggests that the deletions generated by the mutator polymerases are initiated by misinsertion rather than by strand slippage. When considered with recent structure-function studies of other polymerases, the data indicate that the nucleotide misinsertion and strand-slippage mechanisms for polymerization infidelity are differentially affected by changes in distinct structural elements of DNA polymerases that share similar subdomain structures.

Molecular dissection of a LIM domain.

LIM domains are novel sequence elements that are found in more than 60 gene products, many of which function as key regulators of developmental pathways. The LIM domain, characterized by the cysteine-rich consensus CX2CX16-23HX2CX2CX2CX16-21 CX2-3(C/H/ D), is a specific mental-binding structure that consists of two distinct zinc-binding subdomains. We and others have recently demonstrated that the LIM domain mediates protein-protein interactions. However, the sequences that define the protein-binding specificity of the LIM domain had not yet been identified. Because structural studies have revealed that the C-terminal zinc-binding module of a LIM domain displays a tertiary fold compatible with nucleic acid binding, it was of interest to determine whether the specific protein-binding activity of a LIM domain could be ascribed to one of its two zinc-binding subdomains. To address this question, we have analyzed the protein-binding capacity of a model LIM peptide, called zLIM1, that is derived from the cytoskeletal protein zyxin. These studies demonstrate that the protein-binding function of zLIM1 can be mapped to sequences contained within its N-terminal zinc-binding module. The C-terminal zinc-binding module of zLIM1 may thus remain accessible to additional interactive partners. Our results raise the possibility that the two structural subdomains of a LIM domain are capable of performing distinct biochemical functions.

Structure of the single-stranded-DNA-binding domain of replication protein A bound to DNA.

The single-stranded-DNA-binding proteins (SSBs) are essential for DNA function in prokaryotic and eukaryotic cells, mitochondria, phages and viruses. The structures of four SSBs have been solved, but the molecular details of the interaction of SSBs with DNA remain speculative. We report here the crystal structure at 2.4 A resolution of the single-stranded-DNA-binding domain of human replication protein A (RPA) bound to DNA. Replication protein A is a heterotrimeric SSB that is highly conserved in eukaryotes. The largest subunit, RPA70, binds to single-stranded (ss)DNA and mediates interactions with many cellular and viral proteins. The DNA-binding domain, which lies in the middle of RPA70, comprises two structurally homologous subdomains oriented in tandem. The ssDNA lies in a channel that extends from one subdomain to the other. The structure of each RPA70 subdomain is similar to those of the bacteriophage SSBs, indicating that the mechanism of ssDNA-binding is conserved.

Chloroplastic aspartate aminotransferase from Arabidopsis thaliana: an examination of the relationship between the structure of the gene and the spatial structure of the protein.

A clone encoding a plastid isoenzyme of aspartate amino-transferase (AAT5) was isolated from an Arabidopsis genomic library and its complete sequence determined. The gene for AAT5 (asp5) contains an open reading frame of 2447 bp comprising 11 exons separated by introns ranging in length from 74 to 207 bp. The upstream regulatory region contains a putative TATA box and multiple copies of two sequence motifs, CTCTT and AAAGAT, previously associated with nodule-specific gene activity in legumes. The deduced primary amino acid sequence of the protein product of asp5 was used to generate a three-dimensional structure of the AAT5 protein by using the computer program Sybyl: Biopolymer Composer and known AAT structures on the protein databases. Both the mature protein and its precursor protein containing a putative N-terminal transit peptide were modelled. The resulting structure of the precursor protein indicated that the transit peptide might also inhibit dimerization of the protein until after its translocation across the chloroplast membrane. The derived structure of the mature protein was then analysed in terms of its component elements of secondary structure, and the positions on the polypeptide back-bone corresponding to intron insertion sites were determined. It is observed that the introns tend to map to regions between structural subdomains of the protein and also map to sites on the surface of the molecule. The asp5 gene in Arabidopsis is thus consistent with Gilbert's exon-shuffling theory of gene evolution [Gilbert (1985) Science 228, 823-824]. A high degree of conservation of intron insertion sites between AAT genes from different plants and animals is observed, particularly within the part of the gene encoding a large beta-sheet structure that forms the structural and functional core of the protein. This beta-sheet structure is thus believed to compromise an ancient and very highly conserved moiety of the molecule.

The first step in sugar transport: crystal structure of the amino terminal domain of enzyme I of the E. coli PEP: sugar phosphotransferase system and a model of the phosphotransfer complex with HPr

Background: The bacterial phosphoenolpyruvate (PEP): sugar phosphotransferase system (PTS) transports exogenous hexose sugars through the membrane and tightly couples transport with phosphoryl transfer from PEP to the sugar via several phosphoprotein intermediates. The phosphate group is first transferred to enzyme I, second to the histidine-containing phosphocarrier protein HPr, and then to one of a number of sugar-specific enzymes II. The structures of several HPrs and enzymes IIA are known. Here we report the structure of the N-terminal half of enzyme I from Escherichia coli (EIN). Results The crystal structure of EIN (MW ∼30 kDa) has been determined and refined at 2.5 å resolution. It has two distinct structural subdomains; one contains four α helices arranged as two hairpins in a claw-like conformation. The other consists of a β sandwich containing a three-stranded antiparallel β sheet and a four-stranded parallel β sheet, together with three short α helices. Plausible models of complexes between EIN and HPr can be made without assuming major structural changes in either protein. Conclusion The α/ β subdomain of EIN is topologically similar to the phospho-histidine domain of the enzyme pyruvate phosphate dikinase, which is phosphorylated by PEP on a histidyl residue but does not interact with HPr. It is therefore likely that features of this subdomain are important in the autophosphorylation of enzyme I. The helical subdomain of EIN is not found in pyruvate phosphate dikinase; this subdomain is therefore more likely to be involved in phosphoryl transfer to HPr.