Probing the altered specificity and catalytic properties of mutant subtilisin chemically modified at position S156C and S166C in the S1 pocket.

A series of chemically modified mutants (CMMs) of subtilisin B. lentus (SBL) were generated employing the combination of site-directed mutagenesis and chemical modification. This strategy entails the mutation of a selected active site residue to cysteine and its subsequent modification with a methanethiosulfonate reagent CH3SO2S-R, where R may be infinitely variable. The present study was undertaken to evaluate the changes in specificity and pH-activity profiles that could be induced by modification of S156C and S166C in the S1 pocket of SBL with a representative range of side chain modifications, namely R=-CH3, -CH2C6H5, -CH2CH2NH3+ and CH2CH2SO3 . The side chain of S156C is surface exposed and well solvated while that of S166C points into the pocket. Kinetic evaluation of the CMMs with suc-AAPF-pNA as substrate showed that the kcat/K(M)s changed very little for the S156C CMMs, but varied by up to 11-fold for the S166C CMMs. pH-Activity profiles were also determined, and showed that a negatively or positively charged side chain modification increased or decreased respectively, the pKa of the catalytic triad histidine for both modification sites but with more dramatic changes for the interior pointing S166C than for the solvent exposed S156C site. As an additional probe of altered specificity, inhibition of the CMMs by a representative series of 5 boronic acid transition state analogue inhibitors was determined. The K(I)s observed ranged from a 3.5-fold improvement over the WT value, to a 12-fold decrease in binding. Overall, greater variability in all the parameters measured, activity, pKa, and boronic acid binding resulted from modification at the inward pointing 166 site than at the solvent-exposed 156 site.     

Engineering substrate preference in subtilisin: structural and kinetic analysis of a specificity mutant

Bacillus subtilisin has been a popular model protein for engineering altered substrate specificity. Although some studies have succeeded in increasing the specificity of subtilisin, they also demonstrate that high specificity is difficult to achieve solely by engineering selective substrate binding. In this paper, we analyze the structure and transient state kinetic behavior of Sbt160, a subtilisin engineered to strongly prefer substrates with phenylalanine or tyrosine at the P4 position. As in previous studies, we measure improvements in substrate affinity and overall specificity. Structural analysis of an inactive version of Sbt160 in complex with its cognate substrate reveals improved interactions at the S4 subsite with a P4 tyrosine. Comparison of transient state kinetic behavior against an optimal sequence (DFKAM) and a similar, but suboptimal, sequence (DVRAF) reveals the kinetic and thermodynamic basis for increased specificity, as well as the limitations of this approach. While highly selective substrate binding is achieved in Sbt160, several factors cause sequence specificity to fall short of that observed with natural processing subtilisins. First, for substrate sequences which are nearly optimal, the acylation reaction becomes faster than substrate dissociation. As a result, the level of discrimination among these substrates diminishes due to the coupling between substrate binding and the first chemical step (acylation). Second, although Sbt160 has 24-fold higher substrate affinity for the optimal substrate DFKAM than for DVRAF, the increased substrate binding energy is not translated into improved transition state stabilization of the acylation reaction. Finally, as interactions at subsites become stronger, the rate-determining step in peptide hydrolysis changes from acylation to product release. Thus, the release of the product becomes sluggish and leads to a low k(cat) for the reaction. This also leads to strong product inhibition of substrate turnover as the reaction progresses. The structural and kinetic analysis reveals that differences in the binding modes at subsites for substrates, transition states, and products are subtle and difficult to manipulate via straightforward protein engineering. These findings suggest several new strategies for engineering highly sequence selective enzymes.     

Probing the S1 specificity pocket of the aminopeptidases that generate antigenic peptides

ERAP1 (endoplasmic reticulum aminopeptidase 1), ERAP2 and IRAP (insulin-regulated aminopeptidase) are three homologous enzymes that play critical roles in the generation of antigenic peptides. These aminopeptidases excise amino acids from N-terminally extended precursors of antigenic peptides in order to generate the correct length epitopes for binding on to MHC class I molecules. The specificity of these peptidases can affect antigenic peptide selection, but has not yet been investigated in detail. In the present study we utilized a collection of 82 fluorigenic substrates to define a detailed selectivity profile for each of the three enzymes and to probe structural and functional features of the S1 (primary specificity) pocket. Molecular modelling of the three S1 pockets reveals substrate-enzyme interactions that are critical determinants for specificity. The substrate selectivity profiles suggest that IRAP largely combines the S1 specificity of ERAP1 and ERAP2, consistent with its proposed biological function. IRAP, however, does not achieve this dual specificity by simply combining structural features of ERAP1 and ERAP2, but rather by an unique amino acid change at position 541. The results of the present study provide insights on antigenic peptide selection and may prove valuable in designing selective inhibitors or activity markers for this class of enzymes.     

Probing the specificity of the S1 binding site of subtilisin Carlsberg with boronic acids

The binding properties and limitations of the key S1 site of subtilisin Carlsberg have been probed with boronic acid inhibitors bearing structurally varied substituents ranging from small alkyl to large aromatic groups. The data permit structural features favoring, and disfavoring, good S1 binding to be clarified. In addition, applications of electrostatic energy calculations have identified a hitherto unsuspected region of positive potential in the fundamentally hydrophobic S1 pocket, whose interactions with electronegative substituents of inhibitors can make significant binding contributions.     

Trifunctional agents as a design strategy for tailoring ligand properties: irreversible inhibitors of A1 adenosine receptors

The 1,3-phenylene diisothiocyanate conjugate of XAC (8-[4-[[[[(2-aminoethyl)amino]carbonyl]methyl]- oxy]phenyl]-1,3-dipropylxanthine, a potent A1 selective adenosine antagonist) has been characterized as an irreversible inhibitor of A1 adenosine receptors. To further extend this work, a series of analogues were prepared containing a third substituent in the phenyl isothiocyanate ring, incorporated to modify the physiochemical or spectroscopic properties of the conjugate. Symmetrical trifunctional cross-linking reagents bearing two isothiocyanate groups were prepared as general intermediates for cross-linking functionalized congeners and receptors. Xanthine isothiocyanate derivatives containing hydrophilic, fluorescent, or reactive substituents, linked via an amide, thiourea, or methylene group in the 5-position, were synthesized and found to be irreversible inhibitors of A1 adenosine receptors. The effects of the 5-substituent on water solubility and on the A1/A2 selectivity ratio derived from binding assays in rat brain membranes were examined. Inhibition of binding of [3H]-N6-(2-phenylisopropyl)-adenosine and [3H] CGS21680 (2-[2-[4-carboxyethyl)phenyl]ethyl]amino] adenosine-5'-N-ethylcarboxamide) at central A1 and A2 adenosine receptors, respectively, was measured. A conjugate of XAC and 1,3,5-triisothiocyanatobenzene was 894-fold selective for A1 receptors. Reporter groups, such as fluorescent dyes and a spin-label, were included as chain substituents in the irreversible binding analogues, which were designed for spectroscopic assays, histochemical characterization, and biochemical characterization of the receptor protein.     

Non-peptidic phenyl-based thrombin inhibitors: exploring structural requirements of the S1 specificity pocket with amidines

We expand the structural requirements and structure-activity relationship of a novel class of non-peptidic aryl-based thrombin inhibitors through exploration of the S1 specificity pocket of thrombin using flexible and constrained amidines. The most active compound of this class is 11 with Ki = 69 nM, which is ca. 15-fold less potent than constrained guanidine 5.     

A strategy for in vivo screening of subtilisin E reaction specificity in E. coli periplasm

We developed a protocol for efficient expression of the functional serine protease, subtilisin E, in Escherichia coli periplasm that permits direct in vivo measurement of the enzyme's catalytic activity. Activity assays and SDS-PAGE/Western blot analysis showed that the levels of expressed subtilisin varied and were correlated with both the culture conditions and the induction procedures. The highest level of subtilisin expression was achieved at 0.10-0.15% (w/v) of arabinose as inducer and a temperature of 20-22 degrees C, and was ca. eightfold higher as compared to the expression level at 30 degrees C. Cultivation of bacterial cells to a steady state of balanced growth before induction was required for uniform subtilisin expression in cell cultures growing in wells of microtiter plates. Amidase and esterase cell-based kinetic assays on microtiter plates were developed based on the direct measurement of subtilisin activity in vivo. Intact E. coli cells displaying wild-type, dimethylformamide-resistant, and temperature-resistant subtilisins were assayed on N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide and N-acetyl-Phe-p-nitrophenyl ester for their amidase and esterase activity, respectively. Additionally, the periplasmic fractions were isolated from the three E. coli strains expressing the respective subtilisins and tested for amidase activity. The amidase activity of the three subtilisins was ca. 15-fold higher than the esterolytic activity when measured in both the intact cells and in the periplasmic fractions. The strategy combining periplasmic expression of subtilisins with two cell-based kinetic assays permits rapid screening of subtilisin mutant libraries for desired activities.     

Custom chemical microarray production and affinity fingerprinting for the S1 pocket of factor VIIa

The goal of this study was to explore the applicability of surface plasmon resonance (SPR)-based fragment screening to identify compounds that bind to factor VIIa (FVIIa). Based on pharmacophore models virtual screening approaches, we selected fragments anticipated to have a reasonable chance of binding to the S1-binding pocket of FVIIa and immobilized these compounds on microarrays. In affinity fingerprinting experiments, a number of compounds were identified to be specifically interacting with FVIIa and shown to fall into four structural classes. The results demonstrate that the chemical microarray technology platform using SPR detection generates unique chemobiological information that is useful for de novo discovery and lead development and allows the detection of weak interactions with ligands of low molecular weight.     

The interaction of a tyrosyl residue and carboxyl groups in the specific interaction between Streptomyces subtilisin inhibitor and subtilisin BPN'. A chemical modification study.

An ultraviolet absorption difference spectrum that is typical of a change in ionization state (pKa 9.7 leads to greater than 11.5) of a tyrosyl residue has been observed on the binding between Streptomyces subtilisin inhibitor (SSI) and subtilisin BPN' [EC 3.4.21.14] at alkaline pH, ionic strength 0.1 M, at 25 degrees C (Inouye, K., Tonomura, B., and Hiromi, K., submitted). When the complex of SSI and subtilisin BPN' is formed at an ionic strength of 0.6 M and pH 9.70, the characteristic features of the protonation of a tyrosyl residue in the difference spectrum are diminished. These results suggest that the pKa-shift of a tyrosyl residue observed at alkaline pH and lower ionic strength results from an electrostatic interaction. Nitration of tyrosyl residues of SSI and of subtilisin BPN' was performed with tetranitromethane (TNM). By measurements of the difference spectra observed on the binding of the tyrosyl-residue-nitrated SSI and the native subtilisin BPN', and on the binding of the native SSI and the tyrosyl-residue-nitrated subtilisin BPN' and alkaline pH, the tyrosyl residue in question was shown to be one out of the five tyrosyl residues of pKa 9.7 of the enzyme. This tyrosyl residue was probably either Tyr 217 or Tyr 104 on the basis of the reactivities of tyrosyl residues of the enzyme with TNM and their locations on the enzyme molecule. Carboxyl groups of SSI were modified by covalently binding glycine methyl ester with the aid of water-soluble carbodiimide, in order to neutralize the negative charges on SSI. In the difference spectrum which was observed on the binding of subtilisin BPN' and the 5.3-carboxyl-group-modified SSI at alkaline pH, the characteristic features of the protonation of a tyrosyl residue were essentially lost, and the difference spectrum is rather similar to that observed on the binding of the native SSI and the enzyme at neutral pH. This phenomenon indicates that the pKa of a tyrosyl residue of the enzyme is shifted upwards by interaction with carboxyl group(s) of SSI on the formation of the enzyme-inhibitor complex.     

Amidino-substituted aromatic heterocycles as probes of the specificity pocket of trypsin-like proteases.

The effect of pargyline on the uptake of acetaldehyde (in the presence of pyrazole) by isolated rat liver cells was studied after incubating the liver cells for 0, 10, 30, 45, and 60 min with 0.40, 1.30, and 2.6 mm pargyline. Without any incubation period, pargyline had no effect on acetaldehyde uptake. With increasing time of incubation, there was a progressive increase in the extent of inhibition of acetaldehyde uptake by pargyline. This suggests the possibility that pargyline is metabolized to the effective inhibitor or the incubation period allows pargyline to reach its site(s) of action. Pargyline was also a more effective inhibitor of the uptake of lower concentrations of acetaldehyde, e.g., 0.167 mm, than of higher concentrations (1.0 mm) of acetaldehyde, especially after short incubation periods or when pyrazole was omitted from the reaction medium. After a 20- to 30-min incubation period, pargyline inhibited the control rate of ethanol oxidation by the liver cells, as well as the accelerated rate of ethanol oxidation found in the presence of pyruvate or an uncoupling agent. Pargyline had no effect on hepatic oxygen consumption. During ethanol oxidation, a time-dependent release of acetaldehyde into the medium was observed. Pyruvate, by increasing the rate of ethanol oxidation, increased the output of acetaldehyde five- to tenfold. Pargyline increased the output of acetaldehyde two- to threefold, despite decreasing the rate of ethanol metabolism by the liver cells. These data indicate that pargyline inhibits the low K_m aldehyde dehydrogenase in intact rat liver cells and that this enzyme plays the major role in oxidizing the acetaldehyde which arises during the metabolism of ethanol. Although most of the acetaldehyde generated during the oxidation of ethanol is removed by the liver cells in an effective manner, changes in the activity of aldehyde dehydrogenase or the rate of acetaldehyde generation significantly alter the hepatic output of acetaldehyde.     

Biphenyl/diphenyl ether renin inhibitors: filling the S1 pocket of renin via the S3 pocket

Structure-based design led to the discovery of a novel class of renin inhibitors in which an unprecedented phenyl ring filling the S1 site is attached to the phenyl ring filling the S3 pocket. Optimization for several parameters including potency in the presence of human plasma, selectivity against CYP3A4 inhibition and improved rat oral bioavailability led to the identification of 8d which demonstrated antihypertensive efficacy in a transgenic rat model of human hypertension.     

Conformational interactions between alpha and beta subunits in the F1 ATPase of Escherichia coli as shown by chemical modification of uncA401 and uncD412 mutant enzymes

In contrast to wild-type F1 adenosine triphosphatase, the beta subunits of soluble ATPase from Escherichia coli mutant strains AN120 (uncA401) and AN939 (uncD412) were not labeled by the fluorescent thiol-specific reagents 5-iodoacetamidofluorescein, 2-(4'-iodoacetamidoanilino)naphthalene-6-sulfonic acid or 4-[N-(iodoacetoxy)ethyl-N-methyl]amino-7-nitrobenzo-2-oxa-1,3-diazole. The mutation in the alpha subunit (uncA401) of F1 ATPase thus influences the accessibility of the single cysteinyl residue in the beta subunit. Following reaction of ATPase with 4-chloro-7-nitrobenzo-2-oxa-1,3-diazole or N,N'-dicyclohexylcarbodiimide, the alpha and beta subunits of the uncA401, but not of the uncD412 mutant F1 ATPase were intensely labeled by a fluorescent thiol reagent. The mutation in the beta subunit (uncD412) thus influences the accessibility of the cysteinyl residues in the alpha subunit. In other work [Stan-Lotter, H. and Bragg, P.D. (1986) Arch. Biochem. Biophys. 248] we have shown that 4-chloro-7-nitrobenzo-2-oxa-1,3-diazole and 2-(4'-iodoacetamidoanilino)naphthalene-6-sulfonic acid react with a different beta subunit from that labeled by N,N'-dicyclohexylcarbodiimide. This asymmetry with respect to modification by 4-chloro-7-nitrobenzo-2-oxa-1,3-diazole and N,N'-dicyclohexylcarbodiimide was seen in both mutant enzymes. In addition, the modification of one beta subunit of the uncA401 F1 ATPase induced the previously unreactive sulfhydryl group of another beta subunit to react with 2-(4'-iodoacetamidoanilino-naphthalene-6-sulfonic acid. These results provide evidence for at least three types of conformational interactions of the major subunits of F1 ATPase: from alpha to beta, from beta to alpha, and from beta to beta. As in wild-type ATPase, labeling of membrane-bound unc mutant ATPase by a fluorescent thiol reagent modified the alpha subunits. This suggests that a conformational change of yet a different type occurs when the enzyme binds to the membrane.     

On the possibilities and limitations of rational protein design to expand the specificity of restriction enzymes: a case study employing EcoRV as the target

The restriction endonuclease EcoRV has been characterized in structural and functional terms in great detail. Based on this detailed information we employed a structure-guided approach to engineer variants of EcoRV that should be able to discriminate between differently flanked EcoRV recognition sites. In crystal structures of EcoRV complexed with d(CGGGATATCCC)(2) and d(AAAGATATCTT)(2), Lys104 and Ala181 closely approach the two base pairs flanking the GATATC recognition site and thus were proposed to be a reasonable starting point for the rational extension of site specificity in EcoRV [Horton,N.C. and Perona,J.J. (1998) J. Biol. Chem., 273, 21721-21729]. To test this proposal, several single (K104R, A181E, A181K) and double mutants of EcoRV (K104R/A181E, K104R/A181K) were generated. A detailed characterization of all variants examined shows that only the substitution of Ala181 by Glu leads to a considerably altered selectivity with both oligodeoxynucleotide and macromolecular DNA substrates, but not the predicted one, as these variants prefer cleavage of a TA flanked site over all other sites, under all conditions tested. The substitution of Lys104 by Arg, in contrast, which appeared to be very promising on the basis of the crystallographic analysis, does not lead to variants which differ very much from the EcoRV wild-type enzyme with respect to the flanking sequence preferences. The K104R/A181E and K104R/A181K double mutants show nearly the same preferences as the A181E and A181K single mutants. We conclude that even for the very well characterized restriction enzyme EcoRV, properties that determine specificity and selectivity are difficult to model on the basis of the available structural information.     

Glucosamine-derived phospholipids in Escherichia coli. Structure and chemical modification of a triacyl glucosamine 1-phosphate found in a phosphatidylglycerol-deficient mutant

Certain Escherichia coli mutants defective in phosphatidylglycerol biosynthesis accumulate novel glucosamine-derived phospholipids. We previously demonstrated that the simplest of these substance (lipid X) is a diacylglucosamine 1-phosphate bearing beta-hydroxymyristoyl groups at positions 2 and 3 (Takayama, K., Qureshi, N., Mascagni, P., Nashed, M. A., Anderson, L., and Raetz, C. R. H. (1983) J. Biol. Chem. 258, 7379-7385). We now report the structural characterization of a triacylglucosamine 1-phosphate (designated lipid Y) that is also found in these mutants. Hydrolyzates of Y contain 2 mol of beta-hydroxymyristate and 1 mol of palmitate/mol of glucosamine. In the lipid, one of the beta-hydroxymyristates is amide-linked at position 2, while the two other fatty acyl groups are ester-linked. Fast atom bombardment mass spectrometry is used to confirm that Y is a monosaccharide derivative and that the molecular weight of Y as the free acid (C50H96NO13P) is 950.29. Analysis of Y by proton NMR spectroscopy at 200 MHz reveals that the anomeric configuration is alpha. Further, one of the esterified fatty acid residues is attached to the 3 OH of the sugar, while the second is linked to an OH moiety of a hydroxymyristate. The 4 and 6 OH groups of the sugar are unsubstituted, as in E. coli lipid X. To establish the precise location of each esterified fatty acyl residue, we subjected Y to a very mild alkaline hydrolysis in the presence of triethylamine. This resulted in the selective removal of a single hydroxymyristoyl group. The triethylamine-treated derivative (lipid Y) has a molecular weight of 723. NMR spectroscopy of Y shows that the 3 OH of the sugar is no longer substituted, while the beta OH of the remaining amide-linked hydroxymyristate is still esterified with palmitate. On the basis of these findings, we propose that lipid Y has the same fundamental structure as lipid X, except for the additional presence of a palmitoyl moiety on the N-linked hydroxymyristate. Presumably, lipid Y is synthesized from X by a selective acylation reaction.     

On the substrate specificity of cathepsins B1 and B2 including a new fluorogenic substrate for cathepsin B1.

Cathepsin B1 from bovine spleen exhibited its greatest rates of hydrolysis on peptide β-naphthylamide (βNA) derivatives containing paired basic residues, i.e., Cbz-Arg-Arg-βNA, t-Boc-Lys-Lys-βNA, and t-Boc-Lys-Arg-βNA. Internal peptide bonds were not attacked. At its pH 6.5 optimum, cathepsin B1 hydrolyzed Cbz-Arg-Arg-βNA (K_m 0.18 mM) 64 times faster than Bz-DL-Arg-βNA (K_m 3.3 mM or 1.6 mM for the L isomer) and was therefore chosen to replace the latter as a more soluble and sensitive substrate for the assay of cathepsin B1. Although cathepsin B2 had no action on the β-naphthylamide substrates, it did manifest carboxypeptidase activity by attacking COOH-terminal residues exposed by the action of cathepsin B1. At its pH 5.0 optimum, cathepsin B2 behaved as a SH-dependent, non-specific carboxypeptidase by releasing COOH-terminal amino acids from a variety of Cbz-Gly-X substrates and polypeptides such as glucagon, Val-Leu-Ser-Glu-Gly, and penta-lysine.     

Subsite specificity of trypanosomal cathepsin L-like cysteine proteases. Probing the S2 pocket with phenylalanine-derived amino acids.

The S2 subsite of mammalian cysteine proteinases of the papain family is essential for specificity. Among natural amino acids, all these enzymes prefer bulky hydrophobic residues such as phenylalanine at P2. This holds true for their trypanosomal counterparts: cruzain from Trypanosoma cruzi and congopain from T. congolense. A detailed analysis of the S2 specificity of parasitic proteases was performed to gain information that might be of interest for the design of more selective pseudopeptidyl inhibitors. Nonproteogenic phenylalanyl analogs (Xaa) have been introduced into position P2 of fluorogenic substrates dansyl-Xaa-Arg-Ala-Pro-Trp, and their kinetic constants (Km, kcat/Km) have been determined with congopain and cruzain, and related host cathepsins B and L. Trypanosomal cysteine proteases are poorly stereoselective towards D/L-Phe, the inversion of chirality modifying the efficiency of the reaction but not the Km. Congopain binds cyclohexylalanine better than aromatic Phe derivatives. Another characteristic feature of congopain compared to cruzain and cathepsins B and L was that it could accomodate a phenylglycyl residue (kcat/Km = 1300 mM-1.s-1), while lengthening of the side chain by a methylene group only slightly impaired the specificity constant towards trypanosomal cysteine proteases. Mono- and di-halogenation or nitration of Phe did not affect Km for cathepsin L-like enzymes, but the presence of constrained Phe derivatives prevented a correct fitting into the S2 subsite. A model of congopain has been built to study the fit of Phe analogs within the S2 pocket. Phe analogs adopted a positioning within the S2 pocket similar to that of the Tyr of the cruzain/Z-Tyr-Ala-fluoromethylketone complex. However, cyclohexylalanine has an energetically favorable chair-like conformation and can penetrate deeper into the subsite. Fitting of modeled Phe analogs were in good agreement with kinetic parameters. Furthermore, a linear relationship could be established with logP, supporting the suggestion that fitting into the S2 pocket of trypanosomal cysteine proteases depends on the hydrophobicity of Phe analogs.     

Mutants of immunotoxin anti-Tac(dsFv)-PE38 with variable number of lysine residues as candidates for site-specific chemical modification. 1. Properties of mutant molecules

Chemical modification of proteins with substances such as poly(ethylene glycol) can add useful properties to proteins. Currently PEGylation is done in a random manner utilizing amino residues dispersed throughout a protein. For proteins such as immunotoxins, which have several different functional domains, random modification leads to inactivation. To determine if we could produce an immunotoxin with a diminished number of lysine residues so that chemical modification could be restricted to certain regions of the protein, we chose the recombinant immunotoxin anti-Tac(dsFv)-PE38 that has 13 lysine residues in the Fv portion and 3 in the toxin. We prepared a series of mutants with 0-12 lysines in the Fv and 0 or 3 in the toxin. Almost all of these molecules retain full biological activity. Our data indicate that replacement of lysine residues can be achieve without loss of biological potency. These molecules are a useful starting point to carry out site-specific PEGylation experiments.     

Expression of the mouse dihydrofolate reductase cDNA in B. subtilis: a system to select mutant cDNAs coding for methotrexate resistant enzymes.

With the aim to obtain a cDNA coding for a mammalian methotrexate resistant dihydrofolate reductase (Dhfr) a plasmid ( pQS1 ) harboring the mouse wild type Dhfr cDNA was constructed and used to transform a methotrexate sensitive bacteria: B. subtilis. A plasmid, pQS4 , expressing large amount of Dhfr in both E. coli and B. subtilis was isolated through a two steps selection with two substrate analogues, trimethoprim followed by methotrexate. This new plasmid has a 54 bp duplication including the beta-lactamase promoter and a deletion of 564 bp removing the 5' end of the beta-lactamase coding region. These changes create a new -35 region TTGAAA and a potentially stronger binding site for both E. coli and B. subtilis 16S ribosomal RNA. pQS4 transformed B. subtilis were then grown in the presence of high level of methotrexate and resistant mutants isolated. One of them, pQS6 , which codes for an enzyme about 50 times more resistant to methotrexate than the wild type Dhfr was sequenced. It shows that a point mutation replaces the glutamine residue at position 35 by a proline.