Upstream subsite interactions for oligonucleotide binding with ribonuclease T1.

Ribonuclease T1 (RNase T1) specifically cleaves RNA at guanylyl residues. Previous studies revealed the presence of an enzyme-subsite interaction for adenosine residues of ApGpC and ApGpU substrates (Osterman and Walz (1979) Biochemistry 10, 1984-1988). The binding of ApG and 2'-deoxyadenylyl-(3',5')-guanosine (dApG) with RNase T1 was studied in the pH range 5-9 using ultraviolet difference spectroscopy. The association constants for these dinucleoside monophosphates showed the same pH dependence both of which differed at low pH values with that for the methyl phosphoester of 5'-GMP (MepG). This difference suggested that binding of the adenosine group is strongly dependent on the deprotonation of an enzyme/ligand group with a pKa value of < or = 4.8. delta G zero for ApG binding minus that for MepG at pH > 6 yielded a delta delta G of -1.17 +/- 0.10 kcal/mol which is a measure of the contribution of the adenosine moiety to binding. ApG bound more tightly than dApG with a mean delta delta G value of -0.73 +/- 0.10 kcal/mol which demonstrated the involvement of the adenosine 2'-OH group in binding. These and other comparisons indicated that delta delta G for maximal binding the adenine base per se was -0.44 kcal/mol. delta delta G for binding pdApdG minus that for dApdG (-0.94 kcal/mol) suggested an enzyme subsite for the phosphomonoester group of former ligand.

Relaxation kinetics of ribonuclease T1 binding with guanosine and 3'-GMP.

Temperature-jump relaxation kinetic studies were undertaken at 25 degrees C with ribonuclease T1 (RNase T1) alone and in the presence of guanosine (Guo) and 3'-guanylic acid (3'-GMP). No relaxations were observed in the absence of ligands and only one process was observed in their presence which reflected a simple on-off reaction in both cases. Apparent association rate constants, k(on), and dissociation rate constants, k(off), were evaluated at several pH values and their ratios, k(on)/k(off), were contrasted with independently determined values of the equilibrium association constant, Ka(eq). The value of k(on)/k(off) for Guo was significantly greater than Ka(eq), whereas Ka(eq) was significantly greater than k(on)/k(off) for 3'-GMP. The simplest interpretation of the result for Guo is that free RNase T1 undergoes a relatively slow undetected isomerization and Guo can bind only with one isomer. 3'-GMP can be considered to bind with the same preference, but in this case the initial enzyme complex undergoes a relatively slow undetected isomerization. These results are consistent with a recent NMR study which suggested that RNase T1 binding with Guo and 3'-GMP are coupled to slow exchange processes in a ligand dependent manner (Shimada, I. and Inagaki, F. (1990) Biochemistry 29, 757-764). It is tentatively concluded that binding of Guo and 3'-GMP at the active site of RNase T1 is limited to a sub-population of conformers involving the base-recognition site and that the phosphomonoester group of the nucleotide can engage in additional conformationally linked interactions at the adjacent catalytic site.

Adult-male-specific S-warfarin (11S-OH) and progesterone (20 beta-OH) keto-reductases in rat hepatic microsomes are not identical.

Adult-male-specific reductase activities in rat hepatic microsomes use NADPH to reduce S-warfarin and progesterone to their 11S-OH and 20 beta-OH products, respectively (Apanovitch et al. (1992) Biochem. Biophys. Res. Commun. 184, 338-346). When microsomes were treated with increasing concentrations of detergent, S-warfarin (11S-OH) reductase (SW(11S)R) activity was subject to monophasic activation by Triton X-100, monophasic inhibition by sodium cholate, and, activation followed by inhibition with either CHAPS or dodecyl-beta-D-maltoside. A non-dialyzable, heat-sensitive factor in rat and rabbit sera activates microsomal SW(11S)R activity six- to eight-fold. Similar detergent inhibitions but no detergent or serum activations were observed for progesterone (20 beta-OH) reductase (P(20 beta)R) activity. A significant amount of SW(11S)R activity was lost during purification regardless of whether the detergent used for solubilization was activating or inhibiting. Octyl-Sepharose, hydroxyapatite, DEAE-cellulose and carboxymethyl matrices were used to partially purify SW(11S)R. P(20 beta)R activity co-purified with SW(11S)R and the most purified fraction contained two major and several minor polypeptides. Partially purified SW(11S)R is activated by detergents, serum, and salt. These and previous results indicate that SW(11S)R and P(20 beta)R are not identical even though they are both adult male-specific, integral membrane proteins apparently having their active sites exposed on the cytoplasmic surface of the endoplasmic reticulum.

Cytochrome P-450 polypeptides in pulmonary microsomes from rats.

Pulmonary microsomal polypeptides from different strains of rats were resolved using two-dimensional electrophoresis and were further characterized by in situ peptide mapping. Triton X-114 detergent separation was used to enrich cytochromes P-450 (P-450) and other integral membrane proteins from pulmonary microsomes, and these were directly compared with corresponding polypeptides from hepatic microsomes. The results demonstrated that P-450b and epoxide hydrolase were present in the lungs of male and female rats and that their expression in this tissue was independent of phenobarbital treatment. P-450e, which is co-induced with P-450b in the liver, was not detected in pulmonary microsomes under any condition. Four other pulmonary microsomal polypeptides were characterized and preliminary evidence suggested that they represent unique isozymic forms of P-450 with three of them being related to P-450b.

Comparisons of highly purified hepatic microsomal cytochromes P-450 from Holtzman and Long-Evans rats.

The present study describes the purification and characterization of strain variant forms of a major phenobarbital-inducible microsomal hemoprotein, cytochrome P-450b, from Holtzman and Long-Evans rats. The strain variant hemoproteins cannot be resolved by sodium dodecyl sulfate gel electrophoresis, but can be partially separated in two-dimensional isoelectric focusing SDS gels. If, however, sodium tetradecyl sulfate is incorporated into the one-dimensional gel system, separation of the cytochromes P-450b is achieved. Minor structural differences are detected in the peptides of the cytochromes P-450b following limited proteolysis by Staphylococcus aureus V8 protease, cleavage by cyanogen bromide, or reverse-phase high-pressure liquid chromatography of tryptic peptides. The strain variant cytochromes P-450b are immunochemically and spectrally indistinguishable. The optical spectra of the ferric and ferrous hemoproteins are identical, as are the CO- and ethylisocyanide-reduced difference spectra. Ferrous cytochromes P-450b from both rat strains effectively bind metyrapone with equivalent affinities. In addition, the cytochromes P-450b do not differ in their catalytic activities toward benzphetamine, hexobarbital, benzo [a]pyrene, zoxazolamine, 7-ethoxycoumarin, estradiol-17 beta and testosterone. Cytochrome P-450c, the predominant isozyme inducible in rat liver by 3-methylcholanthrene, was purified from Holtzman and Long-Evans rats. Cytochromes P-450c from both rat strains are indistinguishable based on electrophoretic, immunological, spectral and catalytic properties. Minor structural differences in the cytochromes P-450c were revealed in the reverse-phase high-pressure liquid chromatographic profiles of the tryptic peptides of these hemoproteins, but not in the peptides generated by limited proteolysis or cleavage with cyanogen bromide.

Spectrophotometric titration of a single carboxyl group at the active site of ribonuclease T1.

Low-pH-induced difference spectra for ribonuclease T1, which were determined using a reference solution at pH 6, consisted of a shorter wavelength component from 270 to 285 nm that relfected an ionization having a pKa of 3.54 and a longer wavelength component above 285 nm that reflected an ionization having a pKa of 4.29. The temperature dependence of the pKa value for data at 300 nm is consistent with its representing the dissociation of a carboxyl group. In addition, the pKa determined at this wavelength significantly decreased at lower ionic strength. Similar experiments which were conducted using catalytically inactive gamma-carboxymethyl-Glu-58-ribonuclease T1 gave difference spectra having only the shorter wavelength component and were characterized by a single pKa of 3.53. It is concluded that the longer wavelength component of the difference spectra is due to the ionization of Glu-58. The pKa determined for this residue in the present study agrees with one found previously from kinetic studies which supports a role for Glu-58 in catalysis. Furthermore, the results suggest a model for the interaction of Glu-58 with histidine and tryptophan residues at the active site.

Ribose recognition by ribonuclease T1: difference spectral binding studies with guanosine and deoxyguanosine.

The binding of ribonuclease T1 with guanosine (Guo) and deoxyguanosine (dGuo) was studied in experiments employing ultraviolet difference spectroscopy in the pH range 3-9 at 0.2 M ionic strength and 25 degrees C. Similar experiments were also conducted with psi-carboxymethyl-glutamate-58 ribonuclease T1 at pH 5.0. At most pH values the characteristic difference spectrum and association constant were obtained. The binding constant for dGuo was approximately 550 M-1 and did not significantly vary in the pH range 3.5-9.0. The binding constant for Guo increased from pH 3.5 to 5.0, was constant between pH 5.0 and 7.0 (approximately 3200 M-1), and decreased at higher pH values. The binding of Guo and dGuo with ribonuclease T1 could also be distinguished in terms of the wavelength for maximal difference absorbance, lambdamax, between pH 5.0 and 7.0. At higher and lower pH values, lambdamax for Guo approached that found fr dGuo. On the other hand, the value of the binding constant (approximately6500 M-1) and the nature of the difference spectra for Guo and dGuo binding with lambdamax-carboxymethyl-glutamate-58-ribonuclease T1 at pH 5.0 were identical. These results suggest that the discrete interaction of the Guo 2'-hydroxyl group with ribonuclease T1 involves the lambda-carboxylate of glutamate-58 and an imidazolium group at the active site.

Subsite interactions of ribonuclease T1: binding studies of dimeric substrate analogues.

Ultraviolet difference spectral binding studies of ribonuclease T1 with pGp, ApG, CpG, UpG, DGpdA, dGpdC, dGpdG, dGpdT, dTpdG, pdApdG, pdTpdG, pdGpdA, pdGpdG, pdGpdT, c(pdGpdA), and c(pdGpdG) were conducted at pH 5.0, 0.2 M ionic strength and 25 degrees C. Under these conditions, the characteristic difference spectrum and association constant for (1:1) ribonuclease T1 binding were determined for each ligand. The binding of guanosine and deoxyguanosine containing ligands could be distinguished by the shapes of their difference spectra. The results indicated that the guanine moiety of each ligand was bound at the enzyme's primary recognition site. Evidence of a specific enzyme subsite for binding the adenine moiety of ApG and pdApdG is presented. The proposal of a specific enzyme subsite for binding the 5'-phosphate group of a complexed guanosine moiety (Sawada, F., Samejima, T., and Saneyoshi, M. (1973), Biochim. Biophys. Acta 299, 596) is not supported in the present work. Preliminary evidence for the existence of two additional enzyme subsites and the effect of oligomer conformation on enzyme binding are also discussed.

Polymorphisms of four hepatic cytochromes P-450 in twenty-eight inbred strains of rat.

Two-dimensional gel electrophoresis of hepatic microsomes from phenobarbital-treated animals was used to analyze electrophoretic/regulatory polymorphisms for cytochromes P-450b, P-450e, P-450g, and P-450h in 28 inbred strains of rat. Previous studies with outbred rats revealed the existence of four electrophoretic variants for P-450b, two for P-450e, and three for P-450h as well as two regulatory alleles for P-450g. With the exception of one allozymic form of P-450h, all of these alleles as well as a novel (null) allele for P-450e were found to be homozygous in at least two of the inbred strains tested. Eight phenotypes for combinations of these four cytochromes P-450 were observed. Inbred strains were identified that can be used in studies on the structure/function of unique cytochrome P-450-allozymes and in genetic crosses to map the four distinct cytochrome P-450 genes.

At least six forms of extremely homologous cytochromes P-450 in rat liver are encoded at two closely linked genetic loci.

A subpopulation of phenobarbital-induced cytochrome P-450 in rat liver has been shown to consist of four closely related forms of the enzyme that appeared to be strain-related. In the present study, polypeptides composing this family were analyzed by two-dimensional electrophoresis of hepatic microsomes from 64 individual phenobarbital-treated rats. The animals surveyed included both sexes from four inbred and five outbred strains/colonies and F1 progenies from 10 crosses. Two new members of this polypeptide family were identified on the basis of their unique electrophoretic behavior and peptide maps. Eight phenotypes were observed that consisted of two to four member polypeptides. The six closely related cytochromes P-450 were found to be encoded at two genetic loci with at least four alleles at the P-450b locus and at least two alleles at the P-450e locus. Most colonies of outbred strains were characterized by polymorphism at one or both of these loci, and in no case did they contain unique alleles. Analyses of parents and their F1 progenies indicated that the P-450b and P-450e loci are closely linked on the same autosome and are expressed codominantly. Furthermore, the products of these loci appear to be coordinately regulated. The extreme homology between P-450b and P-450e genes, their high degree of polymorphism, and their close linkage suggest that they are subject to the same genetic mechanisms that maintain these features in other multigene families.

Genetic and developmental diversity of hepatic cytochromes P450. Warfarin and progesterone metabolism by hepatic microsomes from four inbred strains of rat.

Hepatic microsomes from immature and sexually mature male and female ACI/SEGHsd, F344/NHsd, SHR/NHsd, and WKY/NCrl inbred rats were used to study cytochrome P450 (P450)-catalyzed oxidations of progesterone and both enantiomers of warfarin. Strains were selected on the basis of their different, homozygous allelic compositions for CYP2C11 and CYP2C13 (Rampersaud and Walz, 1992), but no correlations with the microsomal activities were observed. However, correlations were made regarding catalytic activities and the developmental control of CYP2A1 and CYP2C11 levels in microsomes. Other correlations were found for reactions of both warfarin enantiomers at the same atom or for a given enantiomer at different positions, and these appear to involve several P450 isozymes. Strain-dependent activity differences mainly involved the SHR/NHsd and WKY/NCrl strains. WKY/NCrl rats were the most unique strain, having low levels of CYP2C11 in adult males compared with the other inbreds but relatively high S-warfarin 4'- and 6-hydroxylase activities in immature animals of both sexes and adult females. These results suggest that the regulation and/or allozymic composition of hepatic P450s are different for WKY/NCrl rats, which makes them a poor choice as "normotensive controls" in comparison with hypertensive SHR rats.

Spermine stabilization of folded ribonuclease T1.

The interaction of ribonuclease T1 with tetraprotonated spermine (SPM4+), Mg2+, phosphate and other ionic ligands at pH 6.0 was investigated in binding experiments at 25 degrees C and/or by their effects on the midpoint temperature for thermal unfolding of the enzyme. SPM4+ binding with the native protein at 25 degrees C was characterized by an association constant of approximately 2 x 10(4) M-1. This ligand also binds to the unfolded protein but with a approximately 35-fold lower affinity. Phosphate binds at the active site whereas Mg2+ and SPM4+ cations compete for binding at a polyanionic locus that probably involves residues Glu-28, Asp-29, and Glu-31 at the C-terminal end of the alpha-helix. Steady-state kinetic studies using minimal RNA substrates demonstrated that SPM4+ binding with the enzyme does not affect its catalytic activity. SPM4+ also preferentially binds with the folded form of the disulfide-reduced enzyme which has the same or slightly enhanced catalytic properties compared with native ribonuclease T1. The unfolding rate for the native protein in 8 M urea was approximately 8-fold lower in the presence of 0.05 M SPM4+. SPM4+ appears to increase the amplitude of an unobserved fast phase(s) for refolding of the native enzyme. A single kinetic phase characterized refolding of the reduced enzyme which was slightly faster than the slowest refolding phase for the native form.

Mapping of genes for cytochromes P-450b, P-450e, P-450g, and P-450h in the rat. Demonstration of two gene clusters with P450-b,e localized on chromosome 1.

Inbred ACI, WF, and RCS rats having characteristic markers for albino (c), hemoglobin beta-chain (Hbb), and pink-eyed dilution (p) on chromosome 1 and expressing variants for hepatic cytochromes P-450b, P-450e, P-450g, and P-450h were used in genetic mapping studies for these hemoproteins. The results of WF X (ACI X WF)F1 and RCS X (WF X RCS)F1 backcrosses revealed the existence of two gene clusters designated the P450-b,e and P450-g,h loci. The linkage map P450-b,e-p-c-Hbb on rat chromosome 1 was demonstrated and found to be congruent with Coh(P450-b,e)-p-c-Hbb on mouse chromosome 7. P450-g,h is not linked with P450-b,e and the other markers tested on rat chromosome 1. It appears that close genetic linkage, rather than common functional/regulatory properties, typify members of cytochrome P-450 families/subfamilies.

Probing functional perfection in substructures of ribonuclease T1: double combinatorial random mutagenesis involving Asn43, Asn44, and Glu46 in the guanine binding loop.

Combinatorial random mutageneses involving either Asn43 with Asn44 (set 1) or Glu46 with an adjacent insertion (set 2) were undertaken to explore the functional perfection of the guanine recognition loop of ribonuclease T(1) (RNase T(1)). Four hundred unique recombinants were screened in each set for their ability to enhance enzyme catalysis of RNA cleavage. After a thorough selection procedure, only six variants were found that were either as active or more active than wild type which included substitutions of Asn43 by Gly, His, Leu, or Thr, an unplanned Tyr45Ser substitution and Glu46Pro with an adjacent Glu47 insertion. Asn43His-RNase T(1) has the same loop sequence as that for RNases Pb(1) and Fl(2). None of the most active mutants were single substitutions at Asn44 or double substitutions at Asn43 and Asn44. A total of 13 variants were purified, and these were subjected to kinetic analysis using RNA, GpC, and ApC as substrates. Modestly enhanced activities with GpC and RNA involved both k(cat) and K(M) effects. Mutants having low activity with GpC had proportionately even lower relative activity with RNA. Asn43Gly-RNase T(1) and all five of the purified mutants in set 2 exhibited similar values of k(cat)/K(M) for ApC which were the highest observed and about 10-fold that for wild type. The specificity ratio [(k(cat)/K(M))(GpC)/(k(cat)/K(M))(ApC)] varied over 30 000-fold including a 10-fold increase [Asn43His variant; mainly due to a low (k(cat)/K(M))(ApC)] and a 3000-fold decrease (Glu46Ser/(insert)Gly47 variant; mainly due to a low (k(cat)/K(M))(GpC)) as compared with wild type. It is interesting that k(cat) (GpC) for the Tyr45Ser variant was almost 4-fold greater than for wild type and that Pro46/(insert)Glu47 RNase T(1) is 70-fold more active than the permuted variant (insert)Pro47-RNase T(1) which has a conserved Glu46. In any event, the observation that only 6 out of 800 variants surveyed had wild-type activity supports the view that functional perfection of the guanine recognition loop of RNase T(1) has been achieved.

The interaction of ribonuclease T 1 with DNA.

The interaction of RNase T1 with calf thymus DNA was studied using uv difference spectroscopy and the effect of the enzyme on DNA melting. There was no indication of RNase T1 binding with native DNA. A prominent difference spectrum for RNase T1 binding with denatured DNA (d-DNA) was observed at pH 5, 25 degrees and low ionic strength (mu = .01 M) which was depressed at higher ionic strength and pH. The normalized difference spectrum at mu = .01 M, pH 5 and 25 degrees can be interpreted as indicating an interaction of an exposed guanine residue directly with the enzyme and a coupling of this process with the "melting" of short folded segments of d-DNA. The apparent association constant calculated per M guanine residues was 2.4 X 10-4 M-1 under these conditions. The results are discussed in reference to comparable studies on the interaction of RNase T1 with RNA and small guanine ligands.

Exocyclic-keto reductase activities for progesterone and S-warfarin in hepatic microsomes from adult male rats.

Hepatic microsomes from adult male rats representing six inbred strains catalyzed quantitatively significant, NADPH dependent reductions of progesterone to the 20 beta (20R) alcohol and S-warfarin to its 11S-OH product. Microsomes from mature females and immature rats of both sexes were essentially devoid of these activities. Two strains of rat evidenced about 21% of these activities compared with the other strains and both activities were 25-81% repressed by treatment of rats with phenobarbital (PB). An excellent linear correlation was demonstrated for the two activities considering sex, age, NADPH much greater than NADH preference, PB-repression and strain differences. However, detergent latency (71%) and resistance to trypsinolysis were only observed for the keto-reductase activity with S-warfarin. Microsomes also catalyzed the reduction of progesterone to its 20 alpha-OH derivative but this activity preferred NADH greater than NADPH, was induced 2.7-fold by PB and was essentially independent of age, sex and animal strain. Furthermore, unlike the 20 beta-OH activity, this reduction was resistant to proteolytic inactivation.

Phenobarbital induction of rat liver cytochromes P-450b and P-450e. Quantitation of specific RNAs by hybridization to synthetic oligodeoxyribonucleotide probes.

Cytochromes P-450b and P-450e are extremely homologous and immunochemically indistinguishable proteins that are coordinately induced by phenobarbital in rat liver. To assess the effect of phenobarbital on mRNA levels for each of these hemoproteins we performed solution hybridization and Northern blot experiments with synthetic oligodeoxynucleotide probes of defined sequence. Our data demonstrate that phenobarbital administration to rats resulted in marked increases in levels of hepatic mRNA for both cytochrome P-450b and cytochrome P-450e, with a 4- to 5-fold greater accumulation of P-450b mRNA vis à vis P-450e mRNA. The level of hepatic mRNA increased from less than 3 molecules/cell of each mRNA in untreated rats, to 630 and 130 molecules/cell for P-450b and P-450e, respectively, in phenobarbital-treated rats. Data obtained in Northern blot hybridization experiments demonstrated that the size of the mRNAs for each protein were identical, being approximately 1800 bases in length.