Cost-effective large-scale expression of proteins for NMR studies.

We present protocols for high-level expression of isotope-labelled proteins in E. coli in cost-effective ways. This includes production of large amounts of unlabeled proteins and 13C-methyl methionine labeling in rich media, where yields of up to a gram of soluble protein per liter of culture are reached. Procedures for uniform isotope labeling of 2H, 13C and 15N using auto-induction or isopropyl-ß-D-1-thiogalactopyranoside-induction are described, with primary focus on minimal isotope consumption and high reproducibility of protein expression. These protocols are based on high cell-density fermentation, but the key procedures are easily transferred to shake flask cultures.

Enabling low cost biopharmaceuticals: high level interferon alpha-2b production in Trichoderma reesei.

BACKGROUND: The filamentous fungus Trichoderma reesei has tremendous capability to secrete over 100 g/L of proteins and therefore it would make an excellent host system for production of high levels of therapeutic proteins at low cost. We have developed T. reesei strains suitable for production of therapeutic proteins by reducing the secreted protease activity. Protease activity has been the major hindrance to achieving high production levels. We have constructed a series of interferon alpha-2b (IFNα-2b) production strains with 9 protease deletions to gain knowledge for further strain development. RESULTS: We have identified two protease deletions that dramatically improved the production levels. Deletion of the subtilisin protease slp7 and the metalloprotease amp2 has enabled production levels of IFNα-2b up to 2.1 and 2.4 g/L, respectively. With addition of soybean trypsin protease inhibitor the level of production improved to 4.5 g/L, with an additional 1.8 g/L still bound to the secretion carrier protein. CONCLUSIONS: High levels of IFNα-2b were produced using T. reesei strains with reduced protease secretion. Further strain development can be done to improve the production system by reducing protease activity and improving carrier protein cleavage.

Opportunities and Challenges in Developing a Cryptosporidium Controlled Human Infection Model for Testing Antiparasitic Agents.

Cryptosporidiosis is a leading cause of moderate-to-severe diarrhea in low- and middle-income countries, responsible for high mortality in children younger than two years of age, and it is also strongly associated with childhood malnutrition and growth stunting. There is no vaccine for cryptosporidiosis and existing therapeutic options are suboptimal to prevent morbidity and mortality in young children. Recently, novel therapeutic agents have been discovered through high-throughput phenotypic and target-based screening strategies, repurposing malaria hits, etc., and these agents have a promising preclinical in vitro and in vivo anti-Cryptosporidium efficacy. One key step in bringing safe and effective new therapies to young vulnerable children is the establishment of some prospect of direct benefit before initiating pediatric clinical studies. A Cryptosporidium controlled human infection model (CHIM) in healthy adult volunteers can be a robust clinical proof of concept model for evaluating novel therapeutics. CHIM could potentially accelerate the development path to pediatric studies by establishing the safety of a proposed pediatric dosing regimen and documenting preliminary efficacy in adults. We present, here, perspectives regarding the opportunities and perceived challenges with the Cryptosporidium human challenge model.

Automated seamless DNA co-transformation cloning with direct expression vectors applying positive or negative insert selection.

BACKGROUND: Molecular DNA cloning is crucial to many experiments and with the trend to higher throughput of modern approaches automated techniques are urgently required. We have established an automated, fast and flexible low-cost expression cloning approach requiring only vector and insert amplification by PCR and co-transformation of the products. RESULTS: Our vectors apply positive selection for the insert or negative selection against empty vector molecules and drive strong expression of target proteins in E.coli cells. Variable tags are available both in N-terminal or C-terminal position. A newly developed beta-lactamase (DeltaW290) selection cassette contains a segment inside the beta-lactamase open reading frame encoding a stretch of hydrophilic amino acids that result in a T7 promoter when back-translated. This position of the promoter permits positive selection and attenuated expression of fusion proteins with C-terminal tags. We have tested eight vectors by inserting six target sequences of variable length, provenience and function. The target proteins were cloned, expressed and detected using an automated Tecan Freedom Evo II liquid handling work station. Only two colonies had to be picked to score with 85% correct inserts while 80% of those were positive in expression tests. CONCLUSIONS: Our results establish co-transformation and positive/negative selection cloning in conjunction with the provided vectors and selection cassettes as an automatable alternative to commercialized high-throughput cloning systems like Gateway or ligase-independent cloning (LIC) .

Efficient elimination of nonstoichiometric enzyme inhibitors from HTS hit lists.

High-throughput screening often identifies not only specific, stoichiometrically binding inhibitors but also undesired compounds that unspecifically interfere with the targeted activity by nonstoichiometrically binding, unfolding, and/or inactivating proteins. In this study, the effect of such unwanted inhibitors on several different enzyme targets was assessed based on screening results for over a million compounds. In particular, the shift in potency on variation of enzyme concentration was used as a means to identify nonstoichiometric inhibitors among the screening hits. These potency shifts depended on both compound structure and target enzyme. The approach was confirmed by statistical analysis of thousands of dose-response curves, which showed that the potency of competitive and therefore clearly stoichiometric inhibitors was not affected by increasing enzyme concentration. Light-scattering measurements of thermal protein unfolding further verified that compounds that stabilize protein structure by stoichiometric binding show the same potency irrespective of enzyme concentration. In summary, measuring inhibitor IC(50) values at different enzyme concentrations is a simple, cost-effective, and reliable method to identify and eliminate compounds that inhibit a specific target enzyme via nonstoichiometric mechanisms.

Seamless Insert-Plasmid Assembly at High Efficiency and Low Cost.

Seamless cloning methods, such as co-transformation cloning, sequence- and ligation-independent cloning (SLIC) or the Gibson assembly, are essential tools for the precise construction of plasmids. The efficiency of co-transformation cloning is however low and the Gibson assembly reagents are expensive. With the aim to improve the robustness of seamless cloning experiments while keeping costs low, we examined the importance of complementary single-stranded DNA ends for co-transformation cloning and the influence of single-stranded gaps in circular plasmids on SLIC cloning efficiency. Most importantly, our data show that single-stranded gaps in double-stranded plasmids, which occur in typical SLIC protocols, can drastically decrease the efficiency at which the DNA transforms competent E. coli bacteria. Accordingly, filling-in of single-stranded gaps using DNA polymerase resulted in increased transformation efficiency. Ligation of the remaining nicks did not lead to a further increase in transformation efficiency. These findings demonstrate that highly efficient insert-plasmid assembly can be achieved by using only T5 exonuclease and Phusion DNA polymerase, without Taq DNA ligase from the original Gibson protocol, which significantly reduces the cost of the reactions. We successfully used this modified Gibson assembly protocol with two short insert-plasmid overlap regions, each counting only 15 nucleotides.

Structure-based design, synthesis, and memapsin 2 (BACE) inhibitory activity of carbocyclic and heterocyclic peptidomimetics.

Molecular modeling based on the X-ray crystal structure of the Tang-Ghosh heptapeptide inhibitor 1 (OM99-2) of BACE led to the design and synthesis of a series of constrained P(1)' analogues. A cyclopentane ring was incorporated in 1 spanning the P(1)' Ala methyl group and the adjacent methylene carbon atom of the chain. Progressive truncation at the P(2)'-P(4)' sites led to a potent truncated analogue 5 with good selectivity over Cathepsin D. Using the same backbone replacement concept, a series of cyclopentane, cyclopentanone, tetrahydrofuran, pyrrolidine, and pyrrolidinone analogues were synthesized with considerable variation at the P and P' sites. The cyclopentanone and 2-pyrrolidinone analogues 45 and 57 showed low nM BACE inhibition. X-ray cocrystal structures of two analogues 5 and 45 revealed excellent convergence with the original inhibitor 1 structure while providing new insights into other interactions which could be exploited for future modifications.

Enabling Low Cost Biopharmaceuticals: A Systematic Approach to Delete Proteases from a Well-Known Protein Production Host Trichoderma reesei.

The filamentous fungus Trichoderma reesei has tremendous capability to secrete proteins. Therefore, it would be an excellent host for producing high levels of therapeutic proteins at low cost. Developing a filamentous fungus to produce sensitive therapeutic proteins requires that protease secretion is drastically reduced. We have identified 13 major secreted proteases that are related to degradation of therapeutic antibodies, interferon alpha 2b, and insulin like growth factor. The major proteases observed were aspartic, glutamic, subtilisin-like, and trypsin-like proteases. The seven most problematic proteases were sequentially removed from a strain to develop it for producing therapeutic proteins. After this the protease activity in the supernatant was dramatically reduced down to 4% of the original level based upon a casein substrate. When antibody was incubated in the six protease deletion strain supernatant, the heavy chain remained fully intact and no degradation products were observed. Interferon alpha 2b and insulin like growth factor were less stable in the same supernatant, but full length proteins remained when incubated overnight, in contrast to the original strain. As additional benefits, the multiple protease deletions have led to faster strain growth and higher levels of total protein in the culture supernatant.

The x-ray crystal structure of the first RNA recognition motif and site-directed mutagenesis suggest a possible HuR redox sensing mechanism.

Hu-antigen R (HuR) is a ubiquitous RNA-binding protein that comprises three RNA recognition motifs (RRMs). The first two tandem RRMs are known to bind to AU-rich elements (AREs) in the 3'-untranslated region of many mRNAs. The third RRM is connected to the second RRM through a basic hinge region that contains a localization signal termed HuR nucleocytoplasmic shuttling. Binding of HuR to the ARE in the 3'-untranslated region of mRNA leads to nuclear export, stabilization, and/or translational de-repression of the mRNA, resulting in upregulation of the encoded protein. Among the various ARE binding proteins known to date, HuR is still the only known ubiquitous antagonist of posttranscriptional gene silencing by AREs. Given the wide repertoire of known and suspected targets of HuR, it is considered to be a central node in the ARE pathway. Here, the x-ray crystal structure of the first RRM of HuR (amino acids 18-99) at 2.0 A resolution is presented. The overall fold consists of two alpha-helices and a four-stranded beta-sheet, with a beta1-alpha1-beta2-beta3-alpha2-beta4 topology and a beta-hairpin between alpha2 and beta4. The asymmetric unit consists of four chains. The large crystal contact interfaces observed between chains A/B and C/D contain hydrophobic residues located at the alpha-helix side of the fold, opposite to the RNA-binding interface. This hydrophobic region structurally resembles the protein-protein interaction site of RRM domains of other proteins. Because the nature of the assumed HuR homodimerization is mechanistically not well understood to date, we used site-directed mutagenesis, analytical size-exclusion chromatography and multiangle light scattering to investigate HuR interactions via the RRM hydrophobic region. Our data indicate that in vitro, HuR RRM1 and RRM1,2 homodimerization involves a disulfide bond at cysteine 13. This homodimerization mode may have a functional significance in redox modulation of HuR activity in response to oxidative stress. Because HuR is involved in many diseases (e.g., cancer, cachexia, and inflammatory bowel disease), the presented structure may provide a basis for rational drug design.

Positive-selection vector for direct protein expression.

We describe the development of a novel positive-selection vector, RHP-AmpS, that is suitable for seamless cloning and high-level protein expression in Escherichia coli. In this vector, beta-lactamase (Bla) was rendered nonfunctional by replacing the codon for the C-terminal amino acid of the beta-lactamase gene (bla) with a stop codon. Insertion of a target gene in the correct orientation (tail to tail) results in the reconstruction of the C-terminal codon (W290) of bla. This restores the function of the gene and allows the selection of positive recombinants on agar plates containing ampicillin. To allow a high level of protein expression, this selection cassette was inserted into the T7 polymerase-based expression cassette of the Novagen pET28a expression vector. To our knowledge, this is the first example of true positive-selection cloning and direct, high-level expression from a single vector.

An improved method for fast, robust, and seamless integration of DNA fragments into multiple plasmids.

We describe an improved, universal method for the seamless integration of DNA fragments into plasmids at any desired position. The protocol allows in vitro joining of insert and linearized plasmid at terminal homology regions using the BD In-Fusion cloning system. According to the standard BD In-Fusion protocol, vectors are linearized by restriction enzyme digestion. Linearization of plasmids by polymerase chain reaction (PCR), instead of restriction enzyme digestion, extends the usefulness of the method by rendering it independent of restriction endonuclease recognition sites and by allowing seamless insertion of DNA fragments at any position, without introduction of unwanted nucleotides flanking the site of insertion. The combination of PCR linearization of plasmids and BD In-Fusion technology has shown to be very useful for the insertion of genes into the expression regions of multiple plasmids for the heterologous expression of proteins in Escherichia coli. Hands-on time is minimal and there is no need for preparative gel electrophoresis. The protocol is very simple and only involves PCR and liquid handling steps. The method should therefore theoretically have a good potential for automation.

A Rubredoxin based system for screening of protein expression conditions and on-line monitoring of the purification process.

Rubredoxin (Rub) from Thermotoga maritima, a 6.1-kDa red protein containing an Fe(III)-cysteine(4) center, was evaluated for its usefulness as a colored fusion tag for expression of recombinant proteins in E. coli. Here, we describe the Rub features relevant to accelerating screening for optimal high yield soluble expression conditions and automating the ensuing purification process. Spectroscopic properties and the yield of Rub fused to a typical target protein were compared to analogous GFP and Flavodoxin constructs, showing Rub absorption to be sufficient for structural genomics purposes while being produced at much higher soluble levels than GFP constructs. Based entirely on Rub absorption at 380 nm, both generic and affinity purification of crude cell lysate were performed: thus guided anion exchange purification of a Rub fusion construct as well as automated Ni-NTA purification resulted in pure protein. Rub is stable over a wide range of pH, temperature, and buffer environments, enabling robust purification protocols. Across a variety of fusion constructs, including N- and C-terminal Rub, quantitation via the Rub signal was shown to reliably correlate with analytical HPLC data obtained at 220 nm. We propose the "RubyTag" as an alternative to conventional protein fusion tags, as it combines a specific absorption signal with convenient biochemical and biological properties. Further, it allows direct on-line readout on conventional chromatography systems, holding promise for automated multi-step chromatography.

1.3 A X-ray structure of an antibody Fv fragment used for induced membrane-protein crystallization.

The antibody Fv fragment 7E2 has previously been employed in the induced crystallization of the integral membrane protein cytochrome c oxidase from Paracoccus denitrificans. The 1.3 A X-ray structure of the uncomplexed antibody fragment reveals conserved water networks on the surfaces of the framework regions. A novel consensus motif for water coordination, XX(S/T), is found along the edges of the beta-sandwich, where a water molecule forms hydrogen bonds to the carbonyl O atom of a residue at position N and the OG hydroxyl groups of conserved serines or threonines at position N + 2. Multiple conformations were found in the hydrophobic core for residues IleL21, LeuL33 and the disulfide bridges. An internal water molecule that is compatible with only one of the three packing states of the V(L) core suggests local 'breathing' of the variable domain. TrpH47, a conserved key residue of the V(H)/V(L) interface, is crucially involved in the formation of the antigen-binding site by adopting a novel conformation that specifically stabilizes the non-canonical CDR-L3 loop. Finally, a comparison with 7E2-cytochrome c oxidase complexes demonstrates that binding of this membrane-bound antigen proceeds without major conformational changes of the 7E2 antibody fragment.

Structure and allosteric regulation of the alpha X beta 2 integrin I domain.

The integrin alpha X beta 2 (CD11c/CD18, p150,95) binds ligands through the I domain of the alpha X subunit. Ligands include the complement factor fragment iC3b, a key component in the innate immune defense, which, together with the expression of alpha X beta 2 on dendritic cells and on other leukocytes, suggests a role in the immune response. We now report the structure of the alpha X I domain resolved at 1.65 A by x-ray crystallography. To analyze structural requirements for ligand binding we made a mutation in the alpha X I domain C-terminal helix, which increased the affinity for iC3b approximately 200-fold to 2.4 microM compared with the wild-type domain affinity of approximately 400 microM. Gel permeation chromatography supported a conformational change between the wild-type and mutated domains. Conservation of allosteric regulation in the alpha X I domain points to the functional importance of this phenomenon.