Structure-activity studies of human tumour necrosis factors

The mechanism by which tumour necrosis factors (TNF and lymphotoxin,also called TNFß and TNFß respectively) exerttheir cytotoxic activity on many malignant cells, remains largelyunknown. Furthermore, the broad array of differentiation (geneinduction) and mitogenic activities towards many primary cellsis still a subject of intensive investigation. TNF is an importantmediator in inflammation, immune responses and infection-relatedphenomena and these activities contribute to the severe toxidtyseen when TNF is used as an anticancer agent. The first stepin the mechanism of action is the specific binding of the ligandto its receptors and dissection of the molecular mechanism involvedin this interaction is the subject of this review. The reasonsfor the interest in this aspect are obvious: first, the developmentof strong antagonistic TNF analogues can be useful in dampeningthe potentially lethal or debilitating effects of an overproductionof the cytokine (as in septic shock or rheumatoid arthritis).Secondly, since two distinct TNF receptors exist, constructionof TNF muteins that distinguish between both types may leadto derivatives of this plekrtropic agent with a more restrictedbiological activity pattern. Ideally, one would like to developa TNF mutant that has retained its cytotoxic action on tumourcells without inducing the deleterious systemic toxteity. Suchan optimized TNF molecule could become a potent anticancer agent     

Aspartic acid 50 and tyrosine 108 are essential for receptor binding and cytotoxic activity of tumour necrosis factor beta (lymphotoxin)

Single amino acid substitutions were generated in predictedhydrophilic loop regions of the human tumour necrosis factorbeta (TNF-ß) molecule, and the mutant proteins wereexpressed in Escherichia coli and purified. Mutants with singleamino acid changes at either of two distinct loop regions, atpositions aspartic acid 50 or tyrosine 108, were found to havegreatly reduced receptor binding and cytotoxic activity. Thesetwo regions in TNF-ß correspond to known loop regionswhere mutations also result in loss of biological activity ofTNF–, a related cytokine which shares the same cellularreceptors with TNF-ß. The two distinct loops at positions31-34 and 84-89 in the known three-dimensional structure ofTNF- (equivalent to positions 46–50 and 105–110respectively in TNF-ß), lie on opposite sides of theTNF- monomer. When the TNF-a monomer forms a trimer, the twoloops, each from a different subunit of the trimer, come togetherand lie in a cleft between adjacent subunits. Together, thesefindings suggest that a TNF receptor binds to a cleft betweensubunits via surface loops at amino acid residues 31–34and 84–89 in TNF–, and similarly via surface loopsincluding amino acids aspartic acid 50 and tyrosine 108 in TNF–ß.     

Mutational analysis of structure--activity relationships in human tumor necrosis factor-alpha

To determine the region of human tumor necrosis factor-alpha(TNF-), essential for cytotoxic activity against mouse L-M cells,single amino-acid-substituted TNF- mutant proteins (muteins)were produced in Escherichia coli by protein engineering techniques.An expression plasmid for TNF- was mutagenized by passage throughan E.coli mutD5 mutator strain and by oligonucleotide-directedmutagenesis. Approximately 100 single amino-acid-substitutedTNF- muteins were produced and assayed for cytotoxic activity.The cytotoxic activities of purified TNF- muteins, e.g. TNF-31T,-32Y, -82D, -85H, -115L, -141Y, -144K and -146E, were < 1%of that of parent TNF-. These results indicate that the integrityof at least four distinct regions of the TNF- molecule is requiredfor full biological activity. These regions are designated asfollows: region I, from position 30 to 32; region II, from position82 to 89; region III, from position 115 to 117; region FV, fromposition 141 to 146. In addition, TNF-141Y could not completelycompete with parent TNF- for binding to the receptor. This demonstratesthat region IV, and at least aspartk acid at position 141, mustbe involved in the TNF receptor binding site.     

Histidine-15: an important role in the cytotoxic activity of human tumor necrosis factor

The amino acids that are required for the cytotoxic activityof recombinant human tumor necrosis factor- (TNF) were investigatedby chemical modification and oligonucleotide-directed site-specificmutagenesis. TNF contains three histidine residues, locatedat positions 15, 73 and 78. The histidine-specific reagent diethylpyrocarbonate(DEP) was used to chemically modify TNF. The chemical inactivationof the in vitro cytotoxic activity of this lymphokine (usingmurine L929 target cells) was found to be time- and dose-dependent.Inactivated TNF failed to compete with fully bioactive [¹²⁵I]TNFfor human MCF-7 target cell receptors. Mutant polypeptides ofTNF were genetically engineered by oligonucoleotide-directedsite-specific mutagenesis. The cytotoxicity of a double histidinemutant, in which histidine-73 and histidine-78 were replacedwith glutamine, was not altered and was chemically inactivatedby DEP. Substituting glutamine for histidine-15 resulted in10–15% of the wild-type bioactivity. Replacing histidine-15with either asparagine, lysine or glycine resulted in a biologicallyinactive molecule. The data show that the histidine residueat position 15 is an amino acid that is required for the cytotoxicactivity of TNF.     

Epitope analysis of human insulin and intact proinsulin

Residues belonging to epitopes on human insulin that were recognizedby a panel of three monoclonal antibodies were located usingmutated insulins and insulins from a number of different animalspecies. Epitopes on human proinsulin recognized by two monoclonalantibodies were also identified using partially processed proinsulinspecies. Epitopes were located on the C-A and B-C junctionsof proinsulin and on the N-termini of the A-and B-chains andthe central region of the B-chain of human insulin. Antibodiesthat bound proinsulin were found to induce conformational changesin the prohormone. The presence of a well-defined interactionbetween the C-peptide portion and the N-terminus of the A-chainof the insulin moiety of intact proinsulin has also been demonstrated.The relevance of these studies to the development of two-siteassays for the measurement of partially processed proinsulinspecies in human sera is also discussed     

The functional importance of Leu15 of human epidermal growth factor in receptor binding and activation

The biological importance of Leu15 of epidermal growth factor(EGF) is suggested by its conservation through evolution, itscritical location in the domain–domain interface of EGFand its close proximity to Arg41, a residue that is crucialfor receptor binding and activation. Mutagenesis of Leu15 ofhuman EGF (hEGF) was employed to examine the role of this residuein the ligand-receptor interaction. The relative receptor affinitiesof the hEGF variants, as determined by radioreceptor competitionassays, varied depending on the amino acid substitution. TheL15F, L15W and L15V hEGF analogues had receptor affinities 45,26 and 18% respectively of wild type hEGF. The L15A and L15Ranalogues displayed receptor affinities of only 2.4 and 1.6%relative to wild type hEGF. No binding of the L15E analoguewas detected. The relative agonist activities, as measured byreceptor tyrosine kinase stimulation assays, generally followeda similar trend. The L15F, L15W and L15V analogues stimulatedthe receptor kinase to a level (V_max) similar to that for wildtype hEGF. A striking difference was observed between the L15Aand L15R variants; although having similar binding affinities,the L15A mutant activated the receptor to only {small tilde}5%of the wild type V_max in contrast to 53% for the L15R mutant¹H-NMR analysis of the L15R and L15A mutants showed only minorstructural alterations that were not sufficient to account forthe dramatic losses in binding and agonist activities. The resultsindicate that both the size and hydrophobicity of the -branchedaliphatic side chain of Leu15 of hEGF are important in the formationof a catalytically active ligand–receptor complex.     

pH-sensitive interactions between IgG and a mutated IgG-binding protein based upon two B domains of Protein A from Staphylococcus aureus

A fusion protein, consisting of the N-terminal 81 amino acidsfrom an inactive bovine DNase I (Q38,E39–E38,Q39) andtwo sequential synthetic IgG-binding domains based upon domainB of Protein A from Staphylococcus aureus has been shown tobind to porcine IgG with a similar affinity and pH profile toProtein A. The same residue in each B domain (Tyr111 and Tyr169)has been mutated by cassette mutagenesis to Ser, Glu, His, Lysor Arg and the effect of the mutation on binding interactionswith porcine IgG investigated. The evidence presented suggeststhat the interactions at the B domain are highly sensitive tothe presence of a charged residue.     

Structural study of mutants of Escherichia coli ribonuclease HI with enhanced thermostability

Systematic replacement of the amino acid residues in Escherichiacoli ribonuclease HI with those in the thermophilic counterparthas revealed that two mutations, His62–Pro (H62P) andLys95Gly (K95G), increased the thermostability of the protein.These single-site mutant proteins, together with the mutantproteins His62Ala (H62A), Lys95Asn (K95N) and Lys95Ala (K95A),were crystallized and their structures were determined at 1.8Å resolution. The crystal structures of these mutant proteinsreveal that only the local structure around each mutation siteis essential for the increase in thermostability. For each mutantprotein, the stabilization mechanism is considered to be asfollows: (i) H62P is stabilized because of a decrease in theentropy of the unfolded state, without a change in the nativebackbone structure; (ii) K95G is stabilized since the straincaused by the left-handed backbone structure in the typical3:5 type loop is eliminated; and (iii) K95N is slightly stabilizedby a hydrogen bond formed between the side-chain N-atom of themutated aspargine residue and the main-chain carbonyl oxygenwithin the same residue.     

3D modeling, ligand binding and activation studies of the cloned mouse {delta}, {micro} and {kappa} opioid receptors

Refined 3D models of the transmembrane domains of the cloned, µ and opioid receptors belonging to the superfamilyof G-protein coupled receptors (GPCRs) were constructed froma multiple sequence alignment using the alpha carbon templateof rhodopsin recently reported. Other key steps in the procedurewere relaxation of the 3D helix bundle by unconstrained energyoptimization and assessment of the stability of the structureby performing unconstrained molecular dynamics simulations ofthe energy optimized structure. The results were stable ligand-freemodels of the TM domains of the three opioid receptors. Theligand-free receptor was then used to develop a systematicand reliable procedure to identify and assess putative bindingsites that would be suitable for similar investigation of theother two receptors and GPCRs in general. To this end, a non-selective,`universal' antagonist, naltrexone, and agonist, etorphine,were used as probes. These ligands were first docked in allsites of the model opioid receptor which were sterically accessibleand to which the protonated amine of the ligands could be anchoredto a complementary proton-accepting residue. Using these criteria,nine ligand–receptor complexes with different bindingpockets were identified and refined by energy minimization.The properties of all these possible ligand–substratecomplexes were then examined for consistency with known experimentalresults of mutations in both opioid and other GPCRs. Using thisprocedure, the lowest energy agonist–receptor and antagonist–receptorcomplexes consistent with these experimental results were identified.These complexes were then used to probe the mechanism of receptoractivation by identifying differences in receptor conformationbetween the agonist and the antagonist complex during unconstraineddynamics simulation. The results lent support to a possibleactivation mechanism of the mouse opioid receptor similar tothat recently proposed for several other GPCRs. They also allowedthe selection of candidate sites for future mutagenesis experiments.     

High-level bacterial expression, purification and characterization of human calreticulin

To investigate its cellular function and role in autoimmunedisease pathogenesis, we have bacterlally expressed human calreticulin,a major calcium-binding protein in the endoplasmic reticulumand a human autoantigen. This is the First report describingthe heterologous expression of calreticulin from any source.The recombinant calreticulin constituted {small tilde}32% ofthe soluble Escherichia coli proteins, and was purified to apparenthomogeneity by ion exchange and hydrophobic liquid chromatography.As does the bona fide protein, the recombinant calreticulinbinds calcium and undergoes changes in its conformation uponZn²⁺ binding. We take this as a strong indication that the foldingof the E.coli-expressed calreticulin is very similar, if notidentical, to that of the authentic protein. Moreover, the bacteriallyexpressed calreticulin readily reacted with anti-human and anti-rabbitantibodies, and the anti-recombinant calreticulin antibodiesimmunoreacted with HeLa calreticulin. The availability of thisexpression system will allow us to carry out site-specific anddeletion mutagenesis analysis in structure-function studiesof calreticulin.     

The S variant of human {alpha}1-antitrypsin, structure and implications for function and metabolism

The S variant of the human ₁-antitrysin with E-264 – V,is responsible for a mild ₁-antitrypsin deficiency quite commonin the European population. S protein specifically cleaved atthe susceptible peptide bond was crystallized and its crystalstructure determined and refined to 3.1 Å resolution.The S variant crystallizes isomorphous to the normal M variant.The difference Fourier electron density map shows the E –V change as outstanding residual density. In addition, smallstructural changes of the main polypeptide chain radiate fromthe site of mutation and affect parts far removed from it. Bythe mutation, internal hydrogen bonds and salt linkages of E-264to Y-38 and K-487, respectively, are lost. They cause the far-reachingslight distortions and are probobly related to the reduced thermalstability of the S mutant. They may also be responsible forslower folding of the polypeptide chain and the clinical symptomsof ₁-antitrypsin deficiency. In a theoretical study by moleculardynamics methods simulations of the M and S proteins were madeand the results analysed with respect to structrual and dynamicproperties and compared with the experimental results. Thereis a significant correlation between experimental and theoreticalresults in some respects.     

Structural domains of P450-containing monooxygenase systems

All known P450-containing monooxygenase systems share commonstructural and functional domain architecture. Apart from P450itself, these systems can comprise several fundamentally differentprotein components or domains, all of which are shared by othermulticomponent/multidomain enzyme systems with various functions:FAD flavoprotein or domain, FMN domain, Fe₂S₂ ferredoxin, Fe₃S₄ferredoxin, and cytochrome b₅. Either FMN domain, ferredoxinsor cytochrome bs serve as the electron transport intermediatebetween the FAD domain and P450. The molecular evolution ofboth P450-containing systems and of each particular componentdoes not follow phylogeny in general. Gene fusion and horizontalgene transfer events can lead to the appearance of novel redoxchains in the same manner that artificial chimeric proteinscan be constructed by humans. Recent studies using genetic andprotein engineering techniques to investigate the separate domainsand their interaction are described.     

Repertoires of aggregation-resistant human antibody domains

We recently described a method for the generation of a large human domain antibody repertoire involving combinatorial assembly of CDR building blocks from a smaller repertoire comprising a high frequency of aggregation-resistant antibody domains. Here we show that the frequency of aggregation-resistant domains in the combinatorial repertoire remained high. Furthermore, one of the antigen-binding domains selected from the combinatorial repertoire retained its binding properties through 25 cycles of thermal denaturation, suggesting that antibody domains can be created that rival the heat-resistance of thermophilic proteins such as Taq polymerase.     

Context dependence of phenotype prediction and diversity in combinatorial mutagenesis

Two different combinatorial mutagenesis experiments on the light-harvestingII (LH2) protein of Rhodobacter capsulatus indicate that heuristicrules relating sequence directly to phenotype are dependenton which sets or groups of residues are mutated simultaneously.Previously reported combinatorial mutagenesis of this chromogenicprotein (based on both phylogenetic and structural models) showedthat substituting amino acids with large molar volumes at Gly^ß31caused the mutated protein to have a spectrum characteristicof light-harvesting I (LH1). The six residues that underwentcombinatorial mutagenesis were modeled to lie on one side ofa transmembrane -helix that binds bacteriochlorophyll. In asecond experiment described here, we have not used structuralmodels or phylogeny in choosing mutagenesis sites. Instead,a set of six contiguous residues was selected for combinatorialmutagenesis. In this latter experiment, the residue substitutedat Gly^ß31 was not a determining factor in whetherLH2 or LH1 spectra were obtained; therefore, we conclude thatthe heuristic rules for phenotype prediction are context dependent.While phenotype prediction is context dependent, the abilityto identify elements of primary structure causing phenotypediversity appears not to be. This strengthens the argument forperforming combinatorial mutagenesis with an arbitrary groupingof residues if structural models are unavailable.     

Structure-function analysis of human IL-1{alpha}: identification of residues required for binding to the human type I IL-1 receptor

Using oligonucleotide-directed mutagenesis, the binding siteon human interleukin-1 (IL-1) for the human type I IL-1 receptor(IL-1R) has been analyzed. Substitution of seven amino acids(Arg12, Ile14, Asp60, Asp61, Ile64, Lys96 and Trp109) resultedin a significant loss of binding to the receptor. Based on crystallographicinformation, the side chains of these residues are clusteredin one region of IL-1 and exposed on the surface of the protein.Five of the residues in the IL-1 binding site align with thebinding residues previously determined in human IL-1ß,demonstrating that the type I IL-1R recognizes homologous regionsin both ligands. Unexpectedly, only three of the aligned residuesare identical between IL-1 and IL-1ß. These observationssuggest that the composition of contact residues in the bindingsite is unique for each ligand–receptor complex in theIL-1 system.     

Homology modeling study of the human interleukin-7 receptor complex

Following a recent model of human interleukin-7 (IL-7), we presenthere a modeling study of the extracellular part of the humanIL-7 receptor complex, including the IL-7 specific (IL-7R) andthe common gamma (_c) chains. The investigation is based on structuralhomology to the complex of human growth hormone (hGH) boundto its receptor (hGHR). For domain 1 of IL-7R two differentmodels are presented which differ in the alignment to hGHR inthree regions. However, these differences affect binding toIL-7 in only one region, at the interface between loop EF ofdomain 1 of IL-7R and helix C of IL-7. The disulfide patternin domain 1 of IL-7R is predicted to deviate from that observedin hGHR in that the C'–E disulfide (hGHR) is replacedby a C-C' cross-link. The prediction for the _c chain is comparedwith two previous studies. The models of the complex provideinsight into the binding of IL-7 to its receptor and have implicationsfor the suggestion of mutagenesis experiments and the designof (ant)agonists.     

The BC loop in poliovirus coat protein VP1: an ideal acceptor site for major insertions

In poliovirus the BC loop of the VP1 coat protein is hypervariableand can accommodate numerous foreign sequences introduced bygenetic engineering. This paper examines the characteristicsof the VP1 BC loop of the picornaviruses to see why this loopis so variable and particularly favourable for the insertionof foreign sequences leading to the formation of chimeric particles.The characteristics which make this loop distinctive can beused to find equally permissive loops in other proteins.     

Site-directed mutagenesis study on the roles of evolutionally conserved aspartic acid residues in human glutathione S-transferase P1-1

The evolutionally conserved aspartyl residues (Asp57, Asp98and Asp152) in human glutathione S-transferase P1-1 were replacedwith alanine by site-directed mutagenesis to obtain the mutants(D57A, D98A and D152A). The replacement of Asp98 with alanineresulted in a decrease of the affinity for S-hexyl-GSH-agarose,a 5.5-fold increase of the K_m^GHS and a 2.9-fold increase ofthe I₅₀ of S-hexyl-GSH for GSH–CDNB conjugation. Asp98seems to participate in the binding of GSH through hydrogenbonding with the -carboxylate of the -glutamyl residue of GSH.The k_cat of D98A was 2.6-fold smaller than that of the wild-type,and the pK_a of the thiol group of GSH bound in D98A was {smalltilde}0.8 pK units higher than those in the wild-type. Asp98also seems to contribute to the activation of GSH to some extent.On the other hand, most of the kinetic parameters of D57A andD152A were similar to those of the wild-type. However, the thermostabilitiesof D57A and D152A were significantly lower than that of thewild-type. Asp57 and Asp152 seem to be important for maintainingthe proper conformation of the enzyme.     

Combinatorial exploration of the catalytic site of a drug-resistant dihydrofolate reductase: creating alternative functional configurations

We have applied a global approach to enzyme active site exploration, where multiple mutations were introduced combinatorially at the active site of Type II R67 dihydrofolate reductase (R67 DHFR), creating numerous new active site environments within a constant framework. By this approach, we combinatorially modified all 16 principal amino acids that constitute the active site of this enzyme. This approach is fundamentally different from active site point mutation in that the native active site context is no longer accounted for. Among the 1536 combinatorially mutated active site variants of R67 DHFR we created, we selected and kinetically characterized three variants with highly altered active site compositions. We determined that they are of high fitness, as defined by a complex function consisting jointly of catalytic activity and resistance to trimethoprim. The k(cat) and K(M) values were similar to those for the native enzyme. The favourable Delta(DeltaG) values obtained (ranging from -0.72 to -1.08 kcal/mol) suggest that, despite their complex mutational pattern, no fundamental change in the catalytic mechanism has occurred. We illustrate that combinatorial active site mutagenesis can allow for the creation of compensatory mutations that could not be predicted and thus provides a route for more extensive exploration of functional sequence space than is allowed by point mutation.     

Structural interpretation of site-directed mutagenesis and specificity of the catalytic subunit of protein kinase CK2 using comparative modelling

The catalytic subunit of protein kinase casein kinase 2 (CK2),which has specificity for both ATP and GTP, shows significantamino acid sequence similarity to the cyclin-dependent kinase2 (CDK2). We constructed site-directed mutants of CK2 and useda three-dimensional model to investigate the basis for the dualspecificity. Introduction of Phe and Gly at positions 50 and51, in order to restore the pattern of the glycine-rich motif,did not seriously affect the specificity for ATP or GTP. Weshow that the dual specificity probably originates from theloop situated around the position His115 to Asp120 (HVNNTD).The insertion of a residue in this loop in CK2 subunits, comparedwith CDK2 and other kinases, might orient the backbone to interactwith the base A and G; this insertion is conserved in all knownCK2. The mutant N118, the design of which was based on the modelling,showed reduced affinity for GTP as predicted from the model.Other mutants were intended to probe the integrity of the catalyticloop, alter the polarity of a buried residue and explore theimportance of the carboxy terminus. Introduction of Arg to replaceAsn189, which is mapped on the activation loop, results in amutant with decreased k_cat, possibly as a result of disruptionof the interaction between this residue and basic residues inthe vicinity. Truncation at position 331 eliminates the last60 residues of the subunit and this mutant has a reduced catalyticefficiency compared with the wild-type. Catalytic efficiencyis restored in the truncation mutant by the replacement of apotentially buried Glu at position 252 by Lys, probably owingto a higher stability resulting from the formation of a saltbridge between Lys252 and Asp208.