Biotechnology core laboratories: An overview.

An assessment of the capabilities of biotechnology core facilities requires access to current data on state-of-the-art technologies, personnel, space, services, financial issues, and the demand for such facilities. Data on these topics should be useful to researchers, facility personnel, administrators, and granting agencies.To obtain such data, the Association of Biomolecular Resource Facilities (ABRF) conducted a general survey on the operation and technical capabilities of core facilities. A total of 81 ABRF core laboratories voluntarily responded to the survey. Just over 60% of the respondents were from academic institutions, with the remaining located in research institutes, industry, and one U.S. government laboratory. Fifty laboratories provided financial data, with 47 of these operating on a nonprofit basis. Four laboratories were fully self-supporting from user fees.A typical facility had three full-time staff members and occupied approximately 1100 square feet (ft(2)). The most frequently offered services were N-terminal protein sequencing, protein fragmentation, peptide synthesis and purification, amino acid analysis, DNA synthesis, and DNA sequencing. One third of the facilities provided mass analysis by matrix-assisted laser desorption and ionization (MALDI) mass spectrometry, a recently introduced service that has been offered on an average for 3 years. Another relatively new service, bioinformatics support, is offered by about one third of the responding laboratories.

Automated purification and quantification of oligonucleotides.

We have developed automated methods for the trityl-on purification and quantification of synthetic oligonucleotides. Oligonucleotide purification is by solid-phase extraction cartridges using Amberchrom CG-50 resin on an XYZ-axis robotic system. Quantification is by OD260nm using an online UV-visible spectrophotometer with sipper. The purification of 20 oligonucleotides requires 5 min of user set-up time, plus 20 min per sample of robot time. For a 15-25-mer at the 40 nmol scale of synthesis, the method gives a yield of 2.8 ODs from a load of 10.1 OD, i.e., a 28% average yield. Oligonucleotides purified by this method have proven to be successful for primers for automated DNA sequencing.

Molecular recognition of tRNA by tRNA pseudouridine 55 synthase.

Escherichia coli tRNA pseudouridine 55 synthase catalyzes pseudouridine formation at U55 in tRNA. A 17 base oligoribonucleotide analog of the T-arm was equivalent to intact native tRNA as a substrate for pseudouridine 55 synthase, viz., the features for substrate recognition by this enzyme are completely contained within the T-arm. The structures and activities of mutant tRNAs and T-arms were used to analyze substrate recognition by pseudouridine 55 synthase. The 17-mer T-arm was an excellent substrate for the synthase, while disruption of the stem structure of the 17-mer T-arm eliminated activity. Kinetic data on tRNA mutants lacking single T-stem base pairs indicated that only the 53:61 base pair, which maintains the 7 base loop size, was essential for activity. The identities of individual bases in the stem were unimportant provided base pairing was intact. A major function of the T-stem appears to be the maintainence of a stable stem-loop structure and proper presentation of the T-loop to pseudouridine 55 synthase. The 7 base T-loop could be expanded or contracted by 1 base and still retain activity, albeit with a 30-fold reduction in kcat. Kinetic analysis of T-loop mutants revealed the requirement for U54, U55, and A58, and a preference for C over U at position 56. Base substitutions at loop nonconserved position 59 or semiconserved positions 57 or 60 were well tolerated. Comparison of pseudouridine 55 synthase and tRNA (m5U54)-methyltransferase revealed that both enzymes required the stem-loop structure. However, pseudouridine 55 synthase was not stringent for a 7 base loop and recognized a consensus base sequence within the T-loop, while tRNA (m5U54)-methyltransferase recognized the secondary structure of the 7 member T-loop with only a specific requirement for U54, the T-loop substrate site. We conclude that recognition of tRNA by pseudouridine 55 synthase resides in the conformation of the T-arm plus four specific bases of the loop.

Recognition of the T-arm of tRNA by tRNA (m5U54)-methyltransferase is not sequence specific.

tRNA (m5U54)-methyltransferase (RUMT) catalyzes the methylation of U54 of tRNAs. In contrast to enzymes which recognize a particular tRNA, RUMT recognizes features common to all tRNAs. We have shown that these features reside in the T-arm of tRNA and constructed a minimal consensus sequence for RUMT recognition and catalysis (Gu et al., 1991b). Here, we have mutated each conserved T-loop residue and conserved T-stem base pair to bases or base pairs which are not observed in Escherichia coli tRNA. The substrate specificity of RUMT for 30 in vitro synthesized T-arm mutants of tRNAPhe and 37 mutants of the 17-mer analog of the T-arm derived from tRNA1Val was investigated. A 2-5 base pair stem was essential for recognition of the T-arm by RUMT, but the base composition of the stem was unimportant. The 7-base size of the T-loop maintained by the stem was essential for RUMT recognition. For tRNA, most base substitutions in the 7-base loop did not eliminate RUMT activity, except for any mutation of the methyl acceptor U54 and the C56G mutation. The effect of base and base pair mutations on Kcat or the rate of methylation by RUMT was more striking than the effect on the Kd for binding to RUMT. In comparison with mutations in the T-loop of intact tRNA, base mutation in the T-loop of the 17-mer T-arm had a more deleterious effect on binding and methylation. Surprisingly, recognition of tRNA by RUMT appears to reside in the three-dimensional structure of the seven-member T-loop rather than in its primary structure.

Interaction of tRNA (uracil-5-)-methyltransferase with NO2Ura-tRNA.

tRNA in which uracil is completely replaced by 5-nitro-uracil was prepared by substituting 5-nitro-UTP for UTP in an in vitro transcription reaction. The rationale was that the 5-nitro substituent activates the 6-carbon of the Ura heterocycle towards nucleophiles, and hence could provide mechanism-based inhibitors of enzymes which utilize this feature in their catalytic mechanism. When assayed shortly after mixing, the tRNA analog, NO2Ura-tRNA, is a potent competitive inhibitor of tRNA-Ura methyl transferase (RUMT). Upon incubation, the analog causes a time-dependent inactivation of RUMT which could be reversed by dilution into a large excess of tRNA substrate. Covalent RUMT-NO2Ura-tRNA complexes could be isolated on nitrocellulose filters or by SDS-PAGE. The interaction of RUMT and NO2Ura-tRNA was deduced to involve formation of a reversible complex, followed by formation of a reversible covalent complex in which Cys 324 of RUMT is linked to the 6-position of NO2Ura 54 in NO2Ura-tRNA.

Delta 6-desaturase: improved methodology and analysis of the kinetics in a multi-enzyme system.

A new method of assay for the delta 6-desaturation of linoleic acid was developed. This method, which uses HPLC for separation of the fatty acid substrate and product, exhibited a lower coefficient of variation (0.3%) than the reported TLC method (3.5%), and avoided the step of methylation of the saponified fatty acid substrate and product. Using this new method of assay, the kinetics of the delta 6-desaturase in a multi-enzyme system were analysed. A number of factors that could have striking effects on desaturase kinetics were investigated, including the effect of (i) endogenous microsomal linoleic acid on total substrate concentration, and (ii) the pre-reaction catalysed by acyl-CoA synthetase and competing reactions catalysed by lysophospholipid acyltransferase and acyl-CoA hydrolase. Endogenous free linoleate in the hepatic microsomes was found to be 2.9 +/- 1.0 microM (0.5 mg microsomal protein/ml), which was comparable to added substrate concentrations (1.8 to 7.9 microM). The kinetics of the delta 6-desaturase were dissected from the kinetics of the above mentioned pre-reaction and competing reactions through a combination of experimental approaches and computer modeling. From computer modeling, a Km and Vmax of 1.5 microM and 0.63 nmol/min were calculated for the delta 6-desaturase, compared to Km and Vmax of 10.7 microM and 0.08 nmol/min calculated directly from data uncorrected for endogenous substrate. It was concluded that lysophospholipid acyltransferase, acyl-CoA synthetase and endogenous linoleic acid significantly affect the kinetic measurements of hepatic microsomal delta 6-desaturase. These results have implications for kinetic analyses of all desaturates in microsomal systems.

Kinetic mechanism of human erythrocyte acidic isoenzyme rho.

The kinetic mechanism has been determined for human glutathione S-transferase rho (rho), an isoenzyme related to the human pi (pi) isoenzyme. The kinetic mechanism was investigated by both non-linear regression studies and the analysis of primary and secondary plots, utilizing initial rate and product inhibition data. It was concluded that human isoenzyme rho obeys a random sequential Bi-Bi rapid equilibrium mechanism with the formation of an enzyme-substrate-product (enzyme-CDNB-conjugate) dead-end complex. The values of KCDNB, KGSH and Kconjugate were 0.70 +/- 0.11, 0.12 +/- 0.02 and 0.016 +/- 0.004 mM, respectively. Comparison of the kinetic mechanism and kinetic parameters obtained for glutathione S-transferase isoenzyme rho with other class pi isoenzymes showed similarities at the primary kinetic level.

Biotechnology core facilities: trends and update.

A survey of 128 biotechnology core facilities has provided data on the finances, services, space requirements, and personnel. An average facility had four full-time personnel and 7.5 major instrument systems, and occupied 969 sq. ft. Average total income was $244,000/year, but annual user fee income was only $125,000. Typically, facilities required substantial institutional support or grants. Cost recovery (user fee income divided by total income) averaged 49%. During the last 5 years user fee income, total income, and cost recovery have increased. In-house charges for protein sequencing and peptide synthesis increased approximately 30%, while oligonucleotide synthesis charges decreased by 74%. The costs (charges corrected for subsidy from non-user fee income) for most services did not significantly change, except that oligonucleotide synthesis costs decreased by 25% in 1992. DNA synthesis had the highest throughout per month (116 samples), followed by amino acid analysis (86 samples) and DNA sequencing (67 samples). Other services averaged from 5 to 60 samples. DNA synthesis and purification were the services used by the greatest number of principal investigators. A number of services including DNA sequencing, mass spectrometry, capillary electrophoresis, RNA synthesis, electroblotting, and carbohydrate analysis have been introduced in the last 3 years. Although these services are characterized by high levels of methods development and non-user runs, they are offered by twice the percentage of facilities as in 1989, and are increasingly contributing to facility income.

Automated purification of synthetic oligonucleotides.

We have automated the trityl-on purification of oligonucleotides by use of an XYZ axis robotic solid-phase extraction system. This greatly decreased the preparation time required for oligonucleotide purification. After about 15 min for set up of the samples and instrument, the oligonucleotides are automatically purified with a 15-min run time per sample. Thus, for example, the purification of 15 oligonucleotides requires only about 15 min of preparation time and 4 h of machine time. Yields and purity are equivalent to manual methods.

Reversible inhibition of rat hepatic glutathione S-transferase 1-2 by bilirubin.

The inhibition of rat hepatic glutathione (GSH) S-transferase 1-2 by bilirubin exhibited pseudo first-order kinetics with k(obs) values of 0.0214 +/- 0.0005 and 0.040 +/- 0.008 sec-1 at 4 and 8 microM bilirubin, when followed to 72 and 84% completion respectively. These correspond to calculated second-order rate constants of 5.3 +/- 0.1 x 10(3) and 5.0 +/- 1.0 x 10(3)/M.sec. The extent of inhibition of the transferase increased with bilirubin concentration, with half-maximal inhibition at 4 microM bilirubin. Inhibition was reversed by 10-fold dilution of bilirubin or by increasing the pH from 6.0 to 7.4. Premixing 0.2 to 0.5 microM albumin, hemoglobin or aldolase with bilirubin prevented inhibition of GSH S-transferase 1-2. Protection by these proteins occurred at a selected high concentration (0.2 to 0.4 microM) at which they reduced free bilirubin to concentrations (less than 0.5 microM) that did not inhibit isoenzyme 1-2 significantly. No protection was afforded by a selected low protein concentration (0.001 to 0.01 microM) which did not strikingly reduce bilirubin levels in solution. We conclude that bilirubin inhibition of GSH S-transferase 1-2 appears to be a second-order process; the reaction is clearly first-order with respect to GSH S-transferase and appears also to be first-order with respect to bilirubin. It is proposed that (a) inhibition of GSH S-transferase 1-2 results from slow, reversible bilirubin binding, and (b) added proteins appear to prevent GSH S-transferase inhibition by binding high molar ratios of bilirubin.

Explanation of the non-hyperbolic kinetics of the glutathione S-transferases by the simplest steady-state random sequential Bi Bi mechanism.

We have demonstrated that the simplest steady-state random sequential Bi Bi mechanism is sufficient to explain the previously reported non-hyperbolic kinetics of glutathione S-transferase 3-3 [Pabst MJ et al., J Biol Chem 249: 7140-7150, 1974; Jakobson I et al., Biochem J 177: 861-868, 1979]. The metabolism of 1-chloro-2,4-dinitrobenzene by rat liver glutathione S-transferase isoenzymes 2-2 and 3-3 and of 1,2-dichloro-4-nitrobenzene by isoenzyme 3-4 was shown to exhibit non-hyperbolic kinetics, which are best fit by the simplest steady-state random sequential Bi Bi mechanism. Neither more complex steady-state mechanisms nor the superimposition of product inhibition or enzyme memory on the simplest steady-state mechanism was necessary to generate non-hyperbolic kinetics for the glutathione S-transferases.

Bifunctional thymidylate synthase-dihydrofolate reductase in protozoa.

Protozoa contain thymidylate synthase (TS) and dihydrofolate reductase (DHFR) on the same polypeptide. In the bifunctional protein, the DHFR domain is on the amino terminus, TS is on the carboxyl terminus, and the two domains are separated by a junction peptide of varying size depending on the source. The native protein is composed of a dimer of two such subunits and is 110-140 kDa. Most studies of the bifunctional TS-DHFR have been performed with the protein from anti-folate resistant strains of Leishmania major, which show amplification of the TS-DHFR gene and overproduction of the bifunctional protein. The Leishmania TS-DHFR has also been highly expressed in heterologous systems. There appears to be extensive communication among domains and channeling of the H2folate product of TS to DHFR. Anti-folates commonly used to treat microbial infections are poor inhibitors of L. major DHFR. However, selective inhibition of L. major vs. human DHFR does not appear difficult to achieve, and selective inhibitors are known. The TS-DHFR from Plasmodium falciparum has also been cloned and has recently been expressed in Escherichia coli, albeit in small amounts. Interestingly, pyrimethamine-resistant strains of P. falciparum all have a common point mutation in the DHFR coding sequence (Thr/Ser 108 to Asn), which causes decreased binding of the folate analog. It is suggested that if an appropriate inhibitor of the pyrimethamine-resistant P. falciparum DHFRs can be found, it may serve in combination with pyrimethamine as an antimalarial regimen with low propensity for the development of resistance. In the future, we project that we will have a detailed knowledge of the structure and function of TS-DHFRs, and have the essential tools necessary for a molecular-based approach to drug design.

Thymidylate synthase-dihydrofolate reductase in protozoa.

In protozoa, thymidylate synthase (TS) and dihydrofolate reductase (DHFR) exist on the same polypeptide. The DHFR domain is on the amino terminus, TS is on the carboxy terminus, and the domains are separated by a junction peptide of varying size depending on the source. The native protein is a dimer of two such subunits and is 110-140 kDa. Most studies of bifunctional TS-DHFR have been performed with the protein from anti-folate resistant strains of Leishmania major, which show amplification of the TS-DHFR gene and overproduction of the bifunctional protein. The Leishmania TS-DHFR has also been highly expressed in heterologous systems. There is extensive communication between domains, and channeling of the H2folate product of TS to DHFR. Anti-folates commonly used to treat microbial infections are poor inhibitors of L. major DHFR. However, selective inhibitors of L. major vs human DHFR have been found. The TS-DHFR from Plasmodium falciparum has also been cloned and sequenced. Interestingly, pyrimethamine-resistant strains of P. falciparum have a common point mutation in the DHFR coding sequence which causes decreased binding of the folate analog. A detailed knowledge of the structure and function of protozoan TS-DHFRs will soon be available.

A rapid equilibrium random sequential bi-bi mechanism for human placental glutathione S-transferase.

Double-reciprocal plots of initial-rate data for the conjugation of 1-chloro-2,4-dinitrobenzene (CDNB) and GSH by human placental GSH S-transferase pi were linear for both substrates. Computer modelling of the initial-rate data using nonlinear least-squares regression analysis favoured a rapid equilibrium random sequential bi-bi mechanism, over a steady-state random sequential mechanism or a steady-state or rapid equilibrium ordered mechanism. KGSH was calculated as 0.125 +/- 0.006 mM, KCDNB was 0.87 +/- 0.07 mM and alpha was 2.1 +/- 0.3 for the rapid equilibrium random model. The product, S-(2,4-dinitrophenyl)glutathione, was a competitive inhibitor with respect to GSH, and a mixed-type inhibitor toward CDNB (KP = 18 +/- 3 microM). The observed pattern of inhibition is consistent with a rapid equilibrium random mechanism, with a dead-end enzyme.CDNB.product complex, but inconsistent with the inhibition patterns of other bireactant mechanisms. Since rat liver GSH S-transferase 3-3 acts via a steady-state random sequential mechanism [1], while human placental GSH S-transferase and perhaps also rat liver GSH S-transferase 1-1 [2] exhibit rapid equilibrium random mechanisms, we conclude that the kinetic mechanism of the GSH S-transferases is isoenzyme-dependent.

Site-directed inactivation of human lung acidic glutathione S-transferase by 1-chloro-2,4-dinitrobenzene in the absence of glutathione.

Human lung acidic glutathione S-transferase is irreversibly inhibited by 1-chloro-2,4-dinitrobenzene (CDNB) in the absence of the co-substrate glutathione (GSH). The time-dependent inactivation is pseudo-first-order and demonstrates saturation kinetics, suggesting that inactivation occurs from an EI complex. The Ki was 0.14 mM; and kobs was 0.32 min-1 at 0.6 mM CDNB. The enzyme was protected against CDNB inactivation by GSH. The other two classes of glutathione S-transferase, the basic and near-neutral, are not significantly inactivated by CDNB. Incubation with [14C]CDNB indicated covalent binding to all three classes of transferase. One peptide fraction was found to be radiolabelled in both the basic and acidic transferases when these were incubated with [14C]CDNB and GSH, cleaved with cyanogen bromide, and chromatographed by HPLC. Incubation in the absence of GSH yielded one and two additional labelled peptide fractions for the basic and acidic transferases, respectively. Our results suggest that while CDNB arylates all three classes of human transferases, only the acidic transferase possesses a specific GSH-sensitive CDNB binding site, binding to which leads to time-dependent inactivation.

Halothane: inhibition and activation of rat hepatic glutathione S-transferases.

Multiple halothane anesthesias (1.25 MAC for 1 hr on 3 alternate days) of male Long-Evans rats initially decreased by up to 30% and subsequently increased to up to 185% liver cytosolic glutathione S-transferase activity toward 1-chloro-2,4-dinitrobenzene, 3,4-dichloro-1-nitrobenzene and trans-4-phenyl-3-buten-2-one and glutathione peroxidase activity. Halothane rapidly and reversibly activated hepatic cytosolic glutathione S-transferases and purified isoenzyme 1-2 but not isoenzymes 1-1 and 3-3. At high concentrations of halothane (ca. 22 mM), maximal activation was ca. 25%. Halothane, enflurane, isoflurane and methoxyflurane, but not the halothane metabolite 1-chloro-2,2-difluoroethylene, inhibited a mixture of liver cytosolic glutathione S-transferases with time (ca. 30% inhibition/15 min). The inhibition exhibited pseudo-first order kinetics (kobs = 0.13 min-1) and an I50 for halothane of greater than or equal to 15 mM. Halothane inhibited glutathione S-transferases 3-3, 3-4, and 4-4 by 50-60%, but did not affect isoenzymes 1-1 and 1-2. The ability of halothane to diminish hepatic glutathione S-transferase activity in vivo may in part reflect the time-dependent inhibition of glutathione S-transferase isoenzymes containing the 3- and 4-subunits.