Five commercially-available antibodies react differentially with allelic forms of human HLA-DR beta chain.

Allelic variants of HLA-DRB1 have been associated with a variety of autoimmune and infectious diseases. Although the precise molecular mechanisms by which HLA-DRB1 alleles predispose to a particular disease are currently unclear, it has been shown that mRNA expression levels of HLA-DRB1 are dependent on the different alleles. We aimed to measure HLA-DR beta chain levels in peripheral blood mononuclear cells of individuals carrying HLA-DRB1*03:01/*04:01 and HLA-DRB1*03:01/*15:01 alleles by western blotting, using five commercially-available HLA-DRB antibodies. We observed highly heterogeneous binding of the tested antibodies to the different allelic forms of the HLA-DR beta chain. Overall, we show that current immunological research that employs available antibodies to detect HLA-DR beta chains is biased towards detection of specific variants of the protein; this may cause significant discrepancy in quantification of protein expression in a heterogeneous human population.

A rapid CRISPR competitive assay for in vitro and in vivo discovery of potential drug targets affecting the hematopoietic system.

CRISPR/Cas9 can be used as an experimental tool to inactivate genes in cells. However, a CRISPR-targeted cell population will not show a uniform genotype of the targeted gene. Instead, a mix of genotypes is generated - from wild type to different forms of insertions and deletions. Such mixed genotypes complicate analysis of the role of the targeted gene in the studied cell population. Here, we present a rapid and universal experimental approach to functionally analyze a CRISPR-targeted cell population that does not involve generating clonal lines. As a simple readout, we leverage the CRISPR-induced genetic heterogeneity and use sequencing to identify how different genotypes are enriched or depleted in relation to the studied cellular behavior or phenotype. The approach uses standard PCR, Sanger sequencing, and a simple sequence deconvoluting software, enabling laboratories without specific in-depth experience to perform these experiments. As proof of principle, we present examples studying various aspects related to hematopoietic cells (T cell development in vivo and activation in vitro, differentiation of macrophages and dendritic cells, as well as a leukemia-like phenotype induced by overexpressing a proto-oncogene). In conclusion, we present a rapid experimental approach to identify potential drug targets related to mature immune cells, as well as normal and malignant hematopoiesis.

In vitro and ex vivo functional characterization of human HLA-DRB1∗04 restricted T cell receptors.

Recent advances in single-cell sequencing technologies enable the generation of large-scale data sets of paired TCR sequences from patients with autoimmune disease. Methods to validate and characterize patient-derived TCR data are needed, as well as relevant model systems that can support the development of antigen-specific tolerance inducing drugs. We have generated a pipeline to allow streamlined generation of 'artificial' T cells in a robust and reasonably high throughput manner for in vitro and in vivo studies of antigen-specific and patient-derived immune responses. Hereby chimeric (mouse-human) TCR alpha and beta constructs are re-expressed in three different formats for further studies: (i) transiently in HEK cells for peptide-HLA tetramer validation experiments, (ii) stably in the TCR-negative 58 â€‹T cell line for functional readouts such as IL-2 production and NFAT-signaling, and lastly (iii) in human HLA-transgenic mice for studies of autoimmune disease and therapeutic interventions. As a proof of concept, we have used human HLA-DRB1∗04:01 restricted TCR sequences specific for a type I diabetes-associated GAD peptide, and an influenza-derived HA peptide. We show that the same chimeric TCR constructs can be used in each of the described assays facilitating sequential validation and prioritization steps leading to humanized animal models.

Biased TCR gene usage in citrullinated Tenascin C specific T-cells in rheumatoid arthritis.

We aimed to search for common features in the autoreactive T cell receptor (TCR) repertoire in patients with rheumatoid arthritis (RA), focusing on the newly identified candidate antigen citrullinated Tenascin C (cit-TNC). Mononuclear cells from peripheral blood or synovial fluid of eight RA-patients positive for the RA-associated HLA-DRB1*04:01 allele were in-vitro cultured with recently identified citrullinated peptides from Tenascin C. Antigen-specific T cells were isolated using peptide-HLA tetramer staining and subsequently single-cell sequenced for paired alpha/beta TCR analyses by bioinformatic tools. TCRs were re-expressed for further studies of antigen-specificity and T cell responses. Autoreactive T cell lines could be grown out from both peripheral blood and synovial fluid. We demonstrate the feasibility of retrieving true autoreactive TCR sequences by validating antigen-specificity in T cell lines with re-expressed TCRs. One of the Tenascin C peptides, cit-TNC22, gave the most robust T cell responses including biased TCR gene usage patterns. The shared TCR-beta chain signature among the cit-TNC22-specific TCRs was evident in blood and synovial fluid of different patients. The identification of common elements in the autoreactive TCR repertoire gives promise to the possibility of both immune monitoring of the autoimmune components in RA and of future antigen- or TCR-targeted specific intervention in subsets of patients.

Glycosylation Tunes Neuroserpin Physiological and Pathological Properties.

Neuroserpin (NS) is a member of the serine protease inhibitors superfamily. Specific point mutations are responsible for its accumulation in the endoplasmic reticulum of neurons that leads to a pathological condition named familial encephalopathy with neuroserpin inclusion bodies (FENIB). Wild-type NS presents two N-glycosylation chains and does not form polymers in vivo, while non-glycosylated NS causes aberrant polymer accumulation in cell models. To date, all in vitro studies have been conducted on bacterially expressed NS, de facto neglecting the role of glycosylation in the biochemical properties of NS. Here, we report the expression and purification of human glycosylated NS (gNS) using a novel eukaryotic expression system, LEXSY. Our results confirm the correct N-glycosylation of wild-type gNS. The fold and stability of gNS are not altered compared to bacterially expressed NS, as demonstrated by the circular dichroism and intrinsic tryptophan fluorescence assays. Intriguingly, gNS displays a remarkably reduced polymerisation propensity compared to non-glycosylated NS, in keeping with what was previously observed for wild-type NS in vivo and in cell models. Thus, our results support the relevance of gNS as a new in vitro tool to study the molecular bases of FENIB.

Multi-HLA class II tetramer analyses of citrulline-reactive T cells and early treatment response in rheumatoid arthritis.

BACKGROUND: HLA class II tetramers can be used for ex vivo enumeration and phenotypic characterisation of antigen-specific CD4+ T cells. They are increasingly applied in settings like allergy, vaccination and autoimmune diseases. Rheumatoid arthritis (RA) is a chronic autoimmune disorder for which many autoantigens have been described. RESULTS: Using multi-parameter flow cytometry, we developed a multi-HLA class II tetramer approach to simultaneously study several antigen specificities in RA patient samples. We focused on previously described citrullinated HLA-DRB1*04:01-restricted T cell epitopes from α-enolase, fibrinogen-ß, vimentin as well as cartilage intermediate layer protein (CILP). First, we examined inter-assay variability and the sensitivity of the assay in peripheral blood from healthy donors (n = 7). Next, we confirmed the robustness and sensitivity in a cohort of RA patients with repeat blood draws (n = 14). We then applied our method in two different settings. We assessed lymphoid tissue from seropositive arthralgia (n = 5) and early RA patients (n = 5) and could demonstrate autoreactive T cells in individuals at risk of developing RA. Lastly, we studied peripheral blood from early RA patients (n = 10) and found that the group of patients achieving minimum disease activity (DAS28 < 2.6) at 6 months follow-up displayed a decrease in the frequency of citrulline-specific T cells. CONCLUSIONS: Our study demonstrates the development of a sensitive tetramer panel allowing simultaneous characterisation of antigen-specific T cells in ex vivo patient samples including RA 'at risk' subjects. This multi-tetramer approach can be useful for longitudinal immune-monitoring in any disease with known HLA-restriction element and several candidate antigens.

Memory T cells specific to citrullinated α-enolase are enriched in the rheumatic joint.

ACPA-positive rheumatoid arthritis (RA) is associated with distinct HLA-DR alleles and immune responses to many citrullinated self-antigens. Herein we investigated the T cell epitope confined within α-enolase326-340 in the context of HLA-DRB1*04:01 and assessed the corresponding CD4+ T cells in both the circulation and in the rheumatic joint. Comparative crystallographic analyses were performed for the native and citrullinated α-enolase326-340 peptides in complex with HLA-DRB1*04:01. HLA-tetramers assembled with either the native or citrullinated peptide were used for ex vivo and in vitro assessment of α-enolase-specific T cells in peripheral blood, synovial fluid and synovial tissue by flow cytometry. The native and modified peptides take a completely conserved structural conformation within the peptide-binding cleft of HLA-DRB1*04:01. The citrulline residue-327 was located N-terminally, protruding towards TCRs. The frequencies of T cells recognizing native eno326-340 were similar in synovial fluid and peripheral blood, while in contrast, the frequency of T cells recognizing cit-eno326-340 was significantly elevated in synovial fluid compared to peripheral blood (3.6-fold, p = 0.0150). Additionally, citrulline-specific T cells with a memory phenotype were also significantly increased (1.6-fold, p = 0.0052) in synovial fluid compared to peripheral blood. The native T cell epitope confined within α-enolase326-340 does not appear to lead to complete negative selection of cognate CD4+ T cells. In RA patient samples, only T cells recognizing the citrullinated version of α-enolase326-340 were found at elevated frequencies implicating that neo-antigen formation is critical for breach of tolerance.

Corrigendum: Functional and Structural Characterization of a Novel HLA-DRB1*04:01-Restricted α-Enolase T Cell Epitope in Rheumatoid Arthritis.

[This corrects the article on p. 494 in vol. 7, PMID: 27895642.].

Functional and Structural Characterization of a Novel HLA-DRB1*04:01-Restricted α-Enolase T Cell Epitope in Rheumatoid Arthritis.

Antibodies to citrullinated proteins, common in rheumatoid arthritis (RA) patients, are strongly associated to a specific set of HLA-DR alleles including HLA-DRB1*04:01, *04:04, and *01:01. Here, we first demonstrate that autoantibody levels toward the dominant citrullinated B cell epitope from α-enolase are significantly elevated in HLA-DRB1*04:01-positive RA patients. Furthermore, we identified α-enolase-derived T cell epitopes and demonstrated that native and citrullinated versions of several peptides bind with different affinities to HLA-DRB1*04:01, *04:04, and *01:01. The citrulline residues in the eight identified peptides are distributed throughout the entire length of the presented epitopes and more specifically, localized at peptide positions p-2, p2, p4, p6, p7, p10, and p11. Importantly, in contrast to its native version peptide 26 (TSKGLFRAAVPSGAS), the HLA-DRB1*04:01-restricted citrullinated peptide Cit26 (TSKGLFCitAAVPSGAS) elicited significant functional T cell responses in primary cells from RA patients. Comparative analysis of the crystal structures of HLA-DRB1*04:01 in complex with peptide 26 or Cit26 demonstrated that the posttranslational modification did not alter the conformation of the peptide. And since citrullination is the only structural difference between the two complexes, this indicates that the neo-antigen Cit26 is recognized by T cells with high specificity to the citrulline residue.

A hexameric peptide barrel as building block of amyloid-ß protofibrils.

Oligomeric and protofibrillar aggregates formed by the amyloid-ß peptide (Aß) are believed to be involved in the pathology of Alzheimer's disease. Central to Alzheimer pathology is also the fact that the longer Aß42 peptide is more prone to aggregation than the more prevalent Aß40 . Detailed structural studies of Aß oligomers and protofibrils have been impeded by aggregate heterogeneity and instability. We previously engineered a variant of Aß that forms stable protofibrils and here we use solid-state NMR spectroscopy and molecular modeling to derive a structural model of these. NMR data are consistent with packing of residues 16 to 42 of Aß protomers into hexameric barrel-like oligomers within the protofibril. The core of the oligomers consists of all residues of the central and C-terminal hydrophobic regions of Aß, and hairpin loops extend from the core. The model accounts for why Aß42 forms oligomers and protofibrils more easily than Aß40 .

High-affinity binding to staphylococcal protein A by an engineered dimeric Affibody molecule.

Affibody molecules are engineered binding proteins, in which the three-helix bundle motif of the Z domain derived from protein A is used as a scaffold for sequence variation. We used phage display to select Affibody binders to staphylococcal protein A itself. The best binder, called ZpA963, binds with similar affinity and kinetics to the five homologous E, D, A, B and C domains of protein A, and to a five-domain protein A construct with an average dissociation constant, K(D), of ~20 nM. The structure of ZpA963 in complex with the Z domain shows that it interacts with a surface on Z that is identical in the five protein A domains, which explains the multi-domain affinity. This property allows for high-affinity binding by dimeric Affibody molecules that simultaneously engage two protein A domains in a complex. We studied two ZpA963 dimers in which the subunits were linked by a C-terminal disulfide in a symmetric dimer or head-to-tail in a fusion protein, respectively. The dimers both bind protein A with high affinity, very slow off-rates and with saturation-dependent kinetics that can be understood in terms of dimer binding to multiple sites. The head-to-tail (ZpA963)2htt dimer binds with an off-rate of k(off) ≤ 5 × 10(-6) s(-1) and an estimated K(D) ≤ 16 pM. The results illustrate how dimers of selected monomer binding proteins can provide an efficient route for engineering of high-affinity binders to targets that contain multiple homologous domains or repeated structural units.

Amyloid-ß protofibrils: size, morphology and synaptotoxicity of an engineered mimic.

Structural and biochemical studies of the aggregation of the amyloid-ß peptide (Aß) are important to understand the mechanisms of Alzheimer's disease, but research is complicated by aggregate inhomogeneity and instability. We previously engineered a hairpin form of Aß called Aßcc, which forms stable protofibrils that do not convert into amyloid fibrils. Here we provide a detailed characterization of Aß42cc protofibrils. Like wild type Aß they appear as smooth rod-like particles with a diameter of 3.1 (±0.2) nm and typical lengths in the range 60 to 220 nm when observed by atomic force microscopy. Non-perturbing analytical ultracentrifugation and nanoparticle tracking analyses are consistent with such rod-like protofibrils. Aß42cc protofibrils bind the ANS dye indicating that they, like other toxic protein aggregates, expose hydrophobic surface. Assays with the OC/A11 pair of oligomer specific antibodies put Aß42cc protofibrils into the same class of species as fibrillar oligomers of wild type Aß. Aß42cc protofibrils may be used to extract binding proteins in biological fluids and apolipoprotein E is readily detected as a binder in human serum. Finally, Aß42cc protofibrils act to attenuate spontaneous synaptic activity in mouse hippocampal neurons. The experiments indicate considerable structural and chemical similarities between protofibrils formed by Aß42cc and aggregates of wild type Aß42. We suggest that Aß42cc protofibrils may be used in research and applications that require stable preparations of protofibrillar Aß.

Structural basis of L-phosphoserine binding to Bacillus alcalophilus phosphoserine aminotransferase.

Phosphoserine aminotransferase is a vitamin B6-dependent enzyme that catalyzes the reversible conversion of 3-phosphohydroxypyruvate to L-phosphoserine using glutamate as an amine donor. In an effort to gain insight into the substrate-recognition mechanism of the enzyme, crystal structures of Bacillus alcalophilus phosphoserine aminotransferase in the presence or absence of L-phosphoserine were determined to resolutions of 1.5 and 1.6 Å, respectively. Local conformational changes induced upon substrate binding were identified. However, in contrast to other aminotransferases, no domain or subunit movements were observed. Two Arg residues (Arg42 and Arg328) and two His residues (His41 and His327) were found to form a tight binding site for the phosphate group of L-phosphoserine. Comparison with Escherichia coli phosphoserine aminotransferase in complex with the substrate analogue α-methylglutamate revealed more extensive structural changes in the case of L-phosphoserine binding. Based on the structural analysis, the flexibility of Arg328 is proposed to be critical for substrate recognition.

Allosteric mechanism controls traffic in the chaperone/usher pathway.

Many virulence organelles of Gram-negative bacterial pathogens are assembled via the chaperone/usher pathway. The chaperone transports organelle subunits across the periplasm to the outer membrane usher, where they are released and incorporated into growing fibers. Here, we elucidate the mechanism of the usher-targeting step in assembly of the Yersinia pestis F1 capsule at the atomic level. The usher interacts almost exclusively with the chaperone in the chaperone:subunit complex. In free chaperone, a pair of conserved proline residues at the beginning of the subunit-binding loop form a "proline lock" that occludes the usher-binding surface and blocks usher binding. Binding of the subunit to the chaperone rotates the proline lock away from the usher-binding surface, allowing the chaperone-subunit complex to bind to the usher. We show that the proline lock exists in other chaperone/usher systems and represents a general allosteric mechanism for selective targeting of chaperone:subunit complexes to the usher and for release and recycling of the free chaperone.

Structural basis for high-affinity HER2 receptor binding by an engineered protein.

The human epidermal growth factor receptor 2 (HER2) is specifically overexpressed in tumors of several cancers, including an aggressive form of breast cancer. It is therefore a target for both cancer diagnostics and therapy. The 58 amino acid residue Zher2 affibody molecule was previously engineered as a high-affinity binder of HER2. Here we determined the structure of Zher2 in solution and the crystal structure of Zher2 in complex with the HER2 extracellular domain. Zher2 binds to a conformational epitope on HER2 that is distant from those recognized by the therapeutic antibodies trastuzumab and pertuzumab. Its small size and lack of interference may provide Zher2 with advantages for diagnostic use or even for delivery of therapeutic agents to HER2-expressing tumors when trastuzumab or pertuzumab are already employed. Biophysical characterization shows that Zher2 is thermodynamically stable in the folded state yet undergoing conformational interconversion on a submillisecond time scale. The data suggest that it is the HER2-binding conformation that is formed transiently prior to binding. Still, binding is very strong with a dissociation constant K(D) = 22 pM, and perfect conformational homogeneity is therefore not necessarily required in engineered binding proteins. A comparison of the original Z domain scaffold to free and bound Zher2 structures reveals how high-affinity binding has evolved during selection and affinity maturation and suggests how a compromise between binding surface optimization and stability and dynamics of the unbound state has been reached.

Stabilization of neurotoxic Alzheimer amyloid-beta oligomers by protein engineering.

Soluble oligomeric aggregates of the amyloid-beta peptide (Abeta) have been implicated in the pathogenesis of Alzheimer's disease (AD). Although the conformation adopted by Abeta within these aggregates is not known, a beta-hairpin conformation is known to be accessible to monomeric Abeta. Here we show that this beta-hairpin is a building block of toxic Abeta oligomers by engineering a double-cysteine mutant (called Abetacc) in which the beta-hairpin is stabilized by an intramolecular disulfide bond. Abeta(40)cc and Abeta(42)cc both spontaneously form stable oligomeric species with distinct molecular weights and secondary-structure content, but both are unable to convert into amyloid fibrils. Biochemical and biophysical experiments and assays with conformation-specific antibodies used to detect Abeta aggregates in vivo indicate that the wild-type oligomer structure is preserved and stabilized in Abetacc oligomers. Stable oligomers are expected to become highly toxic and, accordingly, we find that beta-sheet-containing Abeta(42)cc oligomers or protofibrillar species formed by these oligomers are 50 times more potent inducers of neuronal apoptosis than amyloid fibrils or samples of monomeric wild-type Abeta(42), in which toxic aggregates are only transiently formed. The possibility of obtaining completely stable and physiologically relevant neurotoxic Abeta oligomer preparations will facilitate studies of their structure and role in the pathogenesis of AD. For example, here we show how kinetic partitioning into different aggregation pathways can explain why Abeta(42) is more toxic than the shorter Abeta(40), and why certain inherited mutations are linked to protofibril formation and early-onset AD.

Conserved hydrophobic clusters on the surface of the Caf1A usher C-terminal domain are important for F1 antigen assembly.

The outer membrane usher protein Caf1A of the plague pathogen Yersinia pestis is responsible for the assembly of a major surface antigen, the F1 capsule. The F1 capsule is mainly formed by thin linear polymers of Caf1 (capsular antigen fraction 1) protein subunits. The Caf1A usher promotes polymerization of subunits and secretion of growing polymers to the cell surface. The usher monomer (811 aa, 90.5 kDa) consists of a large transmembrane ß-barrel that forms a secretion channel and three soluble domains. The periplasmic N-terminal domain binds chaperone-subunit complexes supplying new subunits for the growing fiber. The middle domain, which is structurally similar to Caf1 and other fimbrial subunits, serves as a plug that regulates the permeability of the usher. Here we describe the identification, characterization, and crystal structure of the Caf1A usher C-terminal domain (Caf1A(C)). Caf1A(C) is shown to be a periplasmic domain with a seven-stranded ß-barrel fold. Analysis of C-terminal truncation mutants of Caf1A demonstrated that the presence of Caf1A(C) is crucial for the function of the usher in vivo, but that it is not required for the initial binding of chaperone-subunit complexes to the usher. Two clusters of conserved hydrophobic residues on the surface of Caf1A(C) were found to be essential for the efficient assembly of surface polymers. These clusters are conserved between the FGL family and the FGS family of chaperone-usher systems.

Effect of pH on the structure and stability of Bacillus circulans ssp. alkalophilus phosphoserine aminotransferase: thermodynamic and crystallographic studies.

pH is one of the key parameters that affect the stability and function of proteins. We have studied the effect of pH on the pyridoxal-5'-phosphate-dependent enzyme phosphoserine aminotransferase produced by the facultative alkaliphile Bacillus circulans ssp. alkalophilus using thermodynamic and crystallographic analysis. Enzymatic activity assay showed that the enzyme has maximum activity at pH 9.0 and relative activity less than 10% at pH 7.0. Differential scanning calorimetry and circular dichroism experiments revealed variations in the stability and denaturation profiles of the enzyme at different pHs. Most importantly, release of pyridoxal-5'-phosphate and protein thermal denaturation were found to occur simultaneously at pH 6.0 in contrast to pH 8.5 where denaturation preceded cofactor's release by approximately 3 degrees C. To correlate the observed differences in thermal denaturation with structural features, the crystal structure of phosphoserine aminotransferase was determined at 1.2 and 1.5 A resolution at two different pHs (8.5 and 4.6, respectively). Analysis of the two structures revealed changes in the vicinity of the active site and in surface residues. A conformational change in a loop involved in substrate binding at the entrance of the active site has been identified upon pH change. Moreover, the number of intramolecular ion pairs was found reduced in the pH 4.6 structure. Taken together, the presented kinetics, thermal denaturation, and crystallographic data demonstrate a potential role of the active site in unfolding and suggest that subtle but structurally significant conformational rearrangements are involved in the stability and integrity of phosphoserine aminotransferase in response to pH changes.

Strain relief at the active site of phosphoserine aminotransferase induced by radiation damage.

The X-ray susceptibility of the lysine-pyridoxal-5'-phosphate Schiff base in Bacillus alcalophilus phosphoserine aminotransferase has been investigated using crystallographic data collected at 100 K to 1.3 A resolution, complemented by on-line spectroscopic studies. X-rays induce deprotonation of the internal aldimine, changes in the Schiff base conformation, displacement of the cofactor molecule, and disruption of the Schiff base linkage between pyridoxal-5'-phosphate and the Lys residue. Analysis of the "undamaged" structure reveals a significant chemical strain on the internal aldimine bond that leads to a pronounced geometrical distortion of the cofactor. However, upon crystal exposure to the X-rays, the strain and distortion are relaxed and eventually diminished when the total absorbed dose has exceeded 4.7 x 10(6) Ggamma. Our data provide new insights into the enzymatic activation of pyridoxal-5'-phosphate and suggest that special care should be taken while using macromolecular crystallography to study details in strained active sites.

Enzyme adaptation to alkaline pH: atomic resolution (1.08 A) structure of phosphoserine aminotransferase from Bacillus alcalophilus.

The crystal structure of the vitamin B(6)-dependent enzyme phosphoserine aminotransferase from the obligatory alkaliphile Bacillus alcalophilus has been determined at 1.08 A resolution. The model was refined to an R-factor of 11.7% (R(free) = 13.9%). The enzyme displays a narrow pH optimum of enzymatic activity at pH 9.0. The final structure was compared to the previously reported structure of the mesophilic phosphoserine aminotransferase from Escherichia coli and to that of phosphoserine aminotransferase from a facultative alkaliphile, Bacillus circulans subsp. alkalophilus. All three enzymes are homodimers with each monomer comprising a two-domain architecture. Despite the high structural similarity, the alkaliphilic representatives possess a set of distinctive structural features. Two residues directly interacting with pyridoxal-5'-phosphate are replaced, and an additional hydrogen bond to the O3' atom of the cofactor is present in alkaliphilic phosphoserine aminotransferases. The number of hydrogen bonds and hydrophobic interactions at the dimer interface is increased. Hydrophobic interactions between the two domains in the monomers are enhanced. Moreover, the number of negatively charged amino acid residues increases on the solvent-accessible molecular surface and fewer hydrophobic residues are exposed to the solvent. Further, the total amount of ion pairs and ion networks is significantly reduced in the Bacillus enzymes, while the total number of hydrogen bonds is increased. The mesophilic enzyme from Escherichia coli contains two additional beta-strands in a surface loop with a third beta-strand being shorter in the structure. The identified structural features are proposed to be possible factors implicated in the alkaline adaptation of phosphoserine aminotransferase.