Testosterone modulates gene expression pathways regulating nutrient accumulation, glucose metabolism and protein turnover in mouse skeletal muscle.

Testosterone regulates energy metabolism and skeletal muscle mass in males, but the molecular mechanisms are not fully understood. This study investigated the response of skeletal muscle to castration and testosterone replacement in 8-week-old male mice. Using microarray analyses of mRNA levels in gastrocnemius muscle, 91 genes were found to be negatively regulated by testosterone and 68 genes were positively regulated. The mRNA levels of the insulin signalling suppressor molecule Grb10 and the glycogen synthesis inhibitors, protein phosphatase inhibitor-1 and phosphorylase kinase-γ, were negatively regulated by testosterone. The insulin-sensitive glucose and amino acid transporters, Glut3 and SAT2, the lipodystrophy gene, Lpin1 and protein targeting to glycogen were positively regulated. These changes would be expected to increase nutrient availability and sensing within skeletal muscle, increase metabolic rate and carbohydrate utilization and promote glycogen accumulation. The observed positive regulation of atrogin-1 (Fbxo32) by testosterone could be explained by the phosphorylation of Akt and Foxo3a, as determined by Western blotting. Testosterone prevented the castration-induced increase in interleukin-1α, the decrease in interferon-γ and the atrophy of the levator ani muscle, which were all correlated with testosterone-regulated gene expression. These findings identify specific mechanisms by which testosterone may regulate skeletal muscle glucose and protein metabolism.

Isolation of an Alu repetitive DNA binding protein and effect of CpG methylation on binding to its recognition sequence.

The structure, expression, and evolution of Alu repetitive DNA elements have been extensively studied, but the role of these sequences in the function of primate genomes has yet to be elucidated. The contribution of Alu repetitive sequences (ARS) to the structure, maintenance, or expression of the human genome is undoubtedly mediated by one or more DNA binding proteins. As part of a larger study in this laboratory to define the molecular mechanisms that result in de-repression of the glycoprotein hormone alpha-subunit (GPH alpha) gene in a variety of tumor cell types, it was found that the gene was hypermethylated in a variety of cell lines that produce alpha-subunit at high levels and significantly less methylated in cell lines where the gene is unexpressed or expressed at low levels. This is in sharp contrast to the majority of genes examined in this regard, which show an inverse correlation between methylation and expression. The analysis was extended to a group of clones isolated from a single cell line (HeLa) that were differentially methylated over the GPH alpha gene and exhibited a 400-fold range in its expression. These analyses demonstrated that methylation of a small number of CpG dinucleotides correlated with high level expression of the gene. Two of these sites are imbedded in oppositely oriented Alu repeats located in the 5'-flanking DNA and second intron. The upstream site was examined in some detail. DNase I footprint analysis demonstrated that the protein protects a region encompassing the sequence 5'-TTGAACCCGGGAG-3', and electrophoretic gel mobility shift analysis demonstrated specific binding of a protein to an oligonucleotide containing the DNase footprint sequence. Chromatography of nuclear extracts on Sephacryl S-200, heparin--agarose, and oligonucleotide--Sepharose produced an apparently homogeneous preparation of the 50-53 kDa DNA-binding protein as judged by silver staining of sodium dodecylsulfate polyacrylamide gels. The affinity-purified material was enriched 15- to 18,000-fold over crude nuclear extracts. Binding of this protein to an oligonucleotide containing the DNase-protected sequence was severely inhibited when CpG dinucleotide in the Msp I recognition site was methylated on either the sense or antisense strands. Based on its properties, this protein has been termed MeSABp50 for methylation-sensitive Alu binding protein of 50 kDa.

Sequence analysis of the human glycoprotein hormone alpha-subunit gene 5'-flanking DNA and identification of a potential regulatory element as an alu repetitive sequence.

The nucleotide sequence of the human glycoprotein hormone alpha-subunit (GPHalpha) gene 5'-flanking DNA was determined from -1637 to +49 relative to the cap site (+1). Comparison of the upstream sequence of the human gene with those of rhesus and mouse demonstrates regions with variable identity. When the 1.7 kb fragment was used to drive the expression of chloramphenicol acetyltransferase (CAT) in transiently transfected HeLa cells, it was found that CAT activity was elevated about 3-fold when the fragment was truncated from -1637 to -846, suggesting the presence of a negative regulatory element in the distal 5'-flanking DNA. This overlaps an Alu repetitive sequence (ARS) located between nucleotides -1330 and -1007. Gel mobility shift and DNase protection analyses identified a protein binding site centered around -1100 in the ARS second monomer. The GPHalpha upstream ARS was cloned in both orientations in positions upstream and downstream from the bacterial CAT gene under control of the herpes simplex virus thymidine kinase (tk) promoter. DNA-mediated transient transfection of these plasmids revealed a marked inhibition (79-82%) of CAT production by the ARS when it was cloned upstream from the tk promoter and in the same orientation as that found in the GPHalpha 5'-flanking DNA. Smaller decreases (29-57%) were produced by the ARS cloned upstream from the tk promoter in the reverse orientation. In marked contrast, the Alu repetitive element had little or no effect when cloned in either orientation downstream from the tk-CAT gene. Introduction of a second ARS downstream from the CAT reporter gene in vectors already containing an ARS upstream from the tk promoter significantly reduced the strong negative effect elicited by the upstream repetitive element. When compared to the Blur 8 Alu element, the GPHalpha upstream ARS differs markedly with respect to its effect on tk-CAT expression in transient assays and as a substrate for DNA binding proteins present in HeLa nuclear extracts. Together, the transient expression results demonstrate that ARS elements can influence expression of nearby class II promoters. The extent of this effect depends on element position and orientation, cell type, the particular ARS (e.g., GPHalpha or Blur 8), and whether copies were present both upstream and downstream from the transcription unit.

MspI restriction fragment length polymorphism at the glycoprotein hormone alpha-subunit locus. Association of certain genotypes with neoplasia.

A restriction fragment length polymorphism (RFLP) in the human glycoprotein hormone common alpha-subunit gene has been identified and partially characterized in normal lymphocytes and placentae, established tumor cell lines, and tumor biopsy samples. High molecular weight DNA was digested with the restriction endonuclease MspI, separated by electrophoresis in agarose gels, transferred to nylon membranes by the method of Southern, and hybridized to 32P-labeled human chorionic gonadotropin alpha-subunit cDNA. After autoradiography, bands were detected at 5.3, 3.3, 2.1, 1.6, 0.8 and 0.6 kbp. Presence of the 5.3, 3.3 and 0.6 kbp bands was invariant and uninformative. Patterns missing the 0.8 kbp band and both the 2.1 and 1.6 kbp bands are consistent with separate alleles that occur in placental and lymphocyte DNA with frequencies of 0.44 (15/34) and 0.06 (2/34), respectively. Presence of all three bands (2.1, 1.6 and 0.8 kbp) is indicative of heterozygosity, occurring at a frequency of 0.50 (17/34). Additional restriction patterns, not yet observed in DNA isolated from term placentae or circulating lymphocytes, were detected in DNA obtained from tumor cell lines and fresh tumor tissues at frequencies of 0.79 (15/19) and 0.59 (10/17), respectively. Thus, particular alpha-subunit genotypes are disproportionately represented in tumor-derived DNA, occurring at frequencies 10- to 13-times higher than would be predicted from their occurrence in normal tissue. Paired normal and tumor tissues from the same individual exhibited identical hybridization patterns, suggesting that this RFLP may be representative of a predisposition toward a variety of neoplasias rather than indicative of a change in DNA structure at or near this locus as a result of tumor development.

Cloning, expression and characterization of a cDNA encoding a lipase from Rhizopus delemar.

A lambda gt11 cDNA library was constructed in Escherichia coli using poly(A)-selected mRNA from the fungus, Rhizopus (Rp.) delemar. Lipase-producing members of the library were identified by means of a phenotypic score wherein the release of fatty acids by lipase causes a characteristic color change in the growth medium. One such isolate contained a 1287-bp insert (LIP cDNA) which hybridizes to 1.25- to 1.35-kb mRNA species from Rp. delemar. The lipase produced in E. coli containing the LIP cDNA exhibits the same substrate selectivity as the authentic fungal enzyme, hydrolyzing ester bonds at the stereospecific numbering (sn) sn-1 and sn-3, but not the sn-2, positions of triglycerides. The complete nucleotide sequence of the LIP cDNA was determined. By reference to the N-terminal sequence of authentic Rp. delemar lipase, the lipase-encoding region was identified within this fragment. The LIP cDNA encodes a putative preprolipase consisting of a 26-amino-acid(aa) signal sequence, a 97-aa propeptide, and a 269-aa mature enzyme. The predicted mature lipase has the same molecular weight and aa composition as that of Rp. delemar, is highly homologous to that produced by the fungus Rhizomucor miehei, and contains the consensus pentapeptide (Gly-Xaa-Ser-Yaa-Gly) which is conserved among lipolytic enzymes. It is concluded that the LIP cDNA is an essentially full-length analogue of the lipase-encoding gene of Rp. delemar. The lipase encoded by the LIP cDNA occupies a cytoplasmic location when synthesized in E. coli. Unprocessed forms of the lipase accumulate in E. coli.

Effects of ketoacidosis on rat apolipoprotein A1 gene expression: a link with acidosis but not with ketones.

To determine if ketoacidosis contributes to reduced apolipoprotein A1 (apoA1) expression in insulin-deficient diabetic rats, we examined the regulation of apoA1 gene expression in response to changes in ambient pH or ketone body concentrations. Hepatic apoAI mRNA levels were reduced 42% in diabetic rats relative to nondiabetic controls (means+/-s.d.; 321.8+/-43.7 vs 438.7+/-58.8 arbitrary units; P<0.03). Neither endogenous apoA1 mRNA nor transcriptional activity of the rat apoA1 gene promoter (from -474 to -7) were altered by sodium butyrate or isobutyramide (0.3 mM to 10 mM) in Hep G2 or Caco-2 cells. Rat hepatic and intestinal apoA1 mRNA levels, and plasma apoA1 concentration, were not altered 24 h after isobutyramide administration (500 mg/kg by gavage). When the effect of altering ambient pH within a wide range commonly encountered in vivo was studied, acidosis (pH 6.7), relative to alkalosis (pH 7.9), decreased apoAI mRNA levels relative to glyceraldehyde-3-phosphate dehydrogenase mRNA by 47% in Hep G2 cells (P<0.025) and by 24% in Caco-2 cells (P<0.017). Acidosis did not alter cytomegalo virus (CMV)-beta-galactosidase activity, or the activity of the simian virus (SV40) early-region promoter, in either cell line transfected with the respective constructs. The lowering of ambient pH was associated with a graded reduction in apoAI promoter activity. At pH 6.7, apoAI promoter activity was reduced by 75% compared with promoter activity at pH 7.9. These observations indicate that acidosis, but not ketosis, contributes to the reduction in apoA1 expression during diabetic ketoacidosis by down-regulating apoAI promoter activity.

Regulation of apoA1 gene expression with acidosis: requirement for a transcriptional repressor.

Serum apolipoprotein A(1) (apoA(1)) concentration is inversely correlated with the risk of premature atherosclerosis. Serum apoA(1) concentrations are regulated, in part, at the transcriptional level. ApoA(1) mRNA is synthesized primarily in the liver and small intestine, under the direction of a number of signaling molecules and tissue-specific regulatory elements. Previously, we demonstrated that extracellular acidosis suppresses apoA(1) mRNA levels at the level of transcription. Here we demonstrate that intracellular acidosis, in the absence of extracellular pH changes, represses apoA(1) promoter activity. Repression occurs through a pH responsive element (pH-RE) located within the apoA(1) gene promoter. Acidosis increases the specific DNA binding activity of a putative repressor protein within the immediate 5'-flanking region of the apoA(1) gene. The cis-element that binds the putative repressor protein contains a negative thyroid hormone response element (nTRE) located 3' and adjacent to the apoA(1) TATA box. Mutation of the nTRE/pH-RE abrogates protein binding and alters the activity of reporter genes controlled by this element. Repression by acidosis did not require de novo mRNA and protein synthesis. Inhibition of tyrosine kinase activity and diacylglycerol-stimulated protein kinase C (PKC) signaling pathways with tyrophostin A47 and phorbol myristate acetate, respectively, did not affect the repression of apoA(1) promoter activity with acidosis. These results suggest that transcriptional repression of the apoA(1) gene by alterations in ambient pH is associated with enhanced DNA binding activity of a repressor protein, through a mechanism which appears to be independent of de novo mRNA and protein synthesis, tyrosine kinase activity, or PKC activation.

Humoral response and memory formation in carp after injection of Aeromonas hydrophila bacterin.

The humoral immune response of carp (Cyprinus carpio) upon a bacterial fish pathogen Aeromonas hydrophila was studied in relation to memory formation. After a single intramuscular (i.m.) injection of formalin killed A. hydrophila cells (F-Ah), maximum serum Ab (Ab) titers were observed at day 20. Distinct titers were still seen at day 360 in the groups injected with a medium or high antigen (Ag) dose (10(7) respectively 10(9) F-Ah). The effect of a second immunization with a high Ag dose was studied in fish primed 1, 3, 8 or 12 months earlier with 10(5), 10(7) or 10(9) F-Ah. The height of the secondary Ab response was positively correlated with the height of the priming Ag dose. Challenge with a low Ag dose (10(6) F-Ah) gave the best results with 10(7) F-Ah primed animals. The highest secondary responses were obtained with combinations of corresponding priming and challenge dosages. It is concluded that fish are able to form immunological memory to this bacterial Ag. However, optimal memory levels are reached after a relative long period (3-8 months).

Induction of the apolipoprotein AI gene by fasting: a relationship with ketosis but not with ketone bodies.

Apolipoprotein AI (apoAI) expression is inversely related to the incidence of atherosclerosis. ApoAI expression is also influenced by the nutritional state and diabetes. We used both cell culture and animal models to examine the effect of fasting and ketoacidosis on apoAI gene expression. Two days of food deprivation in rats increased hepatic and intestinal apoAI mRNA by 2.6- and 2.3-fold, respectively (P < .05). The absolute concentration of plasma apoAI did not change. However, the plasma apoAI concentration relative to the plasma concentration of serum proteins was increased 23% (P < .05). In fasting rats, there was a significant positive correlation between the serum beta-hydroxybutyrate concentration and hepatic or intestinal apoAI mRNA level. Despite this correlation, changes in apoAI mRNA are probably not mediated by ketone bodies, since neither hepatic nor intestinal apoAI mRNA levels were altered in rats maintained on a ketogenic diet for 10 days or treated with isobutyramide, an orally active ketone analog. In addition, the activity of the rat apoAI promoter was not altered in Hep G2 cells treated with isobutyramide or fatty acids or exposed to hypoglycemic conditions, while dexamethasone increased promoter activity 1.9-fold (P < .05). These data indicate that metabolic changes other than ketone bodies, such as an increase in plasma glucocorticoids, may account for starvation-induced expression of apoAI.

In vitro comparison of cefoperazone/sulbactam with selected antimicrobials against 300 bacteroides isolates. Inhibitory activity and time-kill kinetic studies.

The susceptibilities of 258 Bacteroides fragilis group isolates and 42 other Bacteroides species isolates against cefoperazone, cefoperazone/sulbactam (2:1 ratio) and selected other antimicrobials were determined by broth microdilution method. All isolates were susceptible to cefoperazone/sulbactam, ampicillin/sulbactam, ticarcillin/clavulanate, metronidazole, and imipenem. Other antibiotics showed variables levels of resistance (5-30%). Killing curves with the cef/sulb against selected B. fragilis group isolates were performed and showed excellent bactericidal activity at two to four times the minimum inhibitory concentration (MIC) after 12 hr incubation, even against the isolates with high cefoperazone MICs (greater than or equal to 64 micrograms ml). There was no regrowth at 24 hr. Cefoperazone/sulbactam is a compound with excellent inhibitory and bactericidal activity against B. fragilis group isolates.

Multiple polypeptide hormone expression in pancreatic islet cell carcinomas derived from phosphoenolpyruvatecarboxykinase-SV40 T antigen transgenic rats.

Transgenic rats carrying a PEPCK-SV40 large T-antigen (TAg) transgene rapidly develop numerous pancreatic islet cell neoplasms, the cells of which express TAg. Although many of the larger neoplasms contain relatively undifferentiated cells, many tumors contain areas of well-differentiated cells with abundant endoplasmic reticulum (ER) and secretory granules for endocrine hormones like those observed in normal pancreatic islets. In the well-differentiated lesions, glucagon-producing alpha-cells, insulin-producing beta-cells, and somatostatin-producing delta-cells are readily identifiable morphologically under the electron microscope. Beta-cells were observed in all normal and hyperplastic islets, and nests of these cells were scattered throughout the larger neoplasms. These nests varied from small clusters of epithelium-like cells that stain intensely for insulin, to sheets of small, basophilic cells that stain more diffusely for the hormone. Alpha-cells were also present in all of the normal and hyperplastic islets, but in larger hyperplastic islets, the peripheral localization was absent. Larger neoplasms contained many nests of glucagon-expressing cells, as well as scattered glucagon-producing single cells. Delta-cells were rarely observed in the hyperplastic islets and in the neoplasms. Blood-glucose levels were unaltered in the transgenic animals relative to their nontransgenic litter mates. Thus although these islet cell neoplasms express several polypeptide hormones, there is no obvious clinical effect of such expression in vivo.

Biotinylated probes in colony hybridization.

Colony hybridization is a procedure that allows the detection of cells containing nucleic acid sequences of interest (1). In this method, microbial colonies grown on, or transferred to, a supporting membrane are lysed and their nucleic adds denatured to single strands and fixed in place on the membrane. The membrane is then exposed to a similarly denatured "probe" sequence, which is identical or homologous to all or part of the target sequence, under conditions favoring reannealling. Probe sequences hybridize to complementary sequences on the membrane. Positive hybridization events are then detected by determining the presence and location of probe sequences on the membrane.

Biotinylated probes in colony hybridization.

Colony hybridization is a procedure that allows the detection of cells containing nucleic acid sequences of interest (1). In this method, microbial colonies grown on, or transferred to, a supporting membrane are lysed and their nucleic adds denatured to single strands and fixed in place on the membrane. The membrane is then exposed to a similarly denatured "probe" sequence, which is identical or homologous to all or part of the target sequence, under conditions favoring reannealling. Probe sequences hybridize to complementary sequences on the membrane. Positive hybridization events are then detected by determining the presence and location of probe sequences on the membrane.

Troglitazone stimulates acetyl-CoA carboxylase activity through a post-translational mechanism.

Troglitazone, a thiazolidinedione, is known to act as an insulin sensitizer. The various effects of the drug include stimulation of glucose utilization and inhibition of gluconeogenesis and fatty acid oxidation. We studied the effect of troglitazone treatment on rat liver acetyl-CoA carboxylase (ACC), the key enzyme that catalyzes the formation of malonyl-CoA, the rate-limiting step in the synthesis of long chain fatty acids. Treatment of rats with troglitazone for 18 days resulted in more than 200% increase in the activity of hepatic acetyl-CoA carboxylase (1.01+/-0.14 and 2.33+/-0.28 mU/mg supernatant protein for control and troglitazone-treated rats, respectively) (p<0.001). The expression of acetyl-CoA carboxylase mRNA, as studied by RNAse protection assay, was not significantly different between the two groups of animals. The ACC from control and troglitazone-treated groups was purified by avidin-affinity chromatography. The purified enzyme migrated as a major protein band (Mr 262,000) on SDS-polyacrylamide gels. Troglitazone treatment was associated with increased citrate sensitivity of ACC. The specific activity of the purified preparation in troglitazone-treated rats was increased by 67% (2.5 vs. 1.5 U/mg). Quantitation of alkali-labile phosphate content of the purified preparation revealed 5.66+/-0.17 and 6.29+/-0.13 mol Pi/mol subunit of 262 Kda for control and troglitazone-treated rats, respectively (P<0.01). The subtle increase in phosphate content does not explain the observed activation of the enzyme. It is possible that additional mechanisms such as troglitazone related rearrangement of the occupancy of select phosphate binding sites or altered binding of the biotin cofactor may also contribute to the observed activation of ACC.

Monosaccharide-enriched diets cause hyperleptinemia without hypophagia.

To determine the effect of monosaccharide-enriched diets on plasma leptin and food consumption, body weight, food intake, and serum glucose, insulin, and leptin concentrations were measured in rats maintained on a 10-d course of 60% glucose or 60% fructose diet. The serum leptin concentration in rats fed a high-glucose diet (7.60 +/- 0.6 ng/mL) or a high-fructose diet (5.12 +/- 0.8 ng/mL) was significantly increased compared with that in control rats (2.45 +/- 0.10 ng/mL; P < 0.001). To ascertain that the observed effect was related to hyperinsulinemia, a group of rats were infused with exogenous insulin or rendered insulin resistent with a high-fat diet. When hyperinsulinemia was induced with exogenous infusion, the serum leptin was increased (5.56 +/- 0.23 ng/mL; P < 0.001). High-fat feeding was associated with increased serum leptin (12.1 +/- 1.4 ng/mL) and insulin levels. The increased serum leptin concentration was not associated with decreased food intake. We conclude that monosaccharide-enriched diets, probably through hyperinsulinemia or relative or absolute insulin resistance, cause hyperleptinemia, which does not appear to change feeding behavior.

Alteration of chain length selectivity of a Rhizopus delemar lipase through site-directed mutagenesis.

The coding sequences of the Rhizopus delemar lipase and prolipase were altered by oligonucleotide-directed mutagenesis to introduce amino acid substitutions. The resulting mutant enzymes, synthesized by the bacterial host Escherichia coli BL21 (DE3), were tested for their ability to hydrolyze the triglycerides triolein (TO), tricaprylin (TC) and tributyrin (TB). Mutagenesis and lipase gene expression were carried out using plasmid vectors derived from previously described recombinant plasmids [Joerger and Haas (1993) Lipids 28, 81-88] by introduction of the origin of replication of bacteriophage f1. Substitution of threonine 83 (thr83), a residue thought to be involved in oxyanion binding, by alanine essentially eliminated lipolytic activity toward all substrates examined (TB, TO and TC). Replacement of thr83 with serine caused from two- to sevenfold reductions in the activity toward these substrates. Introduction of tryptophan (trp) at position 89, where such a residue is found in closely related fungal lipases, reduced the specific activity toward the three triglyceride substrates. For the mutagenesis of residues in the predicted acyl chain binding groove, mutagenic primers were designed to cause the replacement of a specific codon within the prolipase gene with codons for all other amino acids. Phenylalanine 95 (phe95), phe112, valine 206 (val206) and val209, were targeted. A phenotypic screen was successfully employed to identify cells producing prolipase with altered preference for olive oil, TC or TB. In assays involving equimolar mixtures of the three triglycerides, a prolipase with a phe95-->aspartate mutation showed an almost twofold increase in the relative activity toward TC. Substitution of trp for phe112 caused an almost threefold decrease in the relative preference for TC, but elevated relative TB hydrolysis. Replacement of val209 with trp resulted in an enzyme with a two- and fourfold enhanced preference for TC and TB, respectively.

Overexpression of a Rhizopus delemar lipase gene in Escherichia coli.

A cloned complementary deoxyribonucleic acid encoding the precursor polypeptide of an extracellular lipase from the fungus Rhizopus delemar was altered by site-directed mutagenesis to generate deoxyribonucleic acid fragments that specifically code for the polypeptides of the proenzyme and the mature form of the lipase. Attempts to produce these polypeptides in enzymatically active form in Escherichia coli revealed toxic effects toward the host. Therefore the polypeptides were expressed as inactive and insoluble forms in the cytoplasm of E. coli BL21 (DE3) cells using plasmid vector pET11-d. With this tightly regulated high-level expression system, lipase and prolipase polypeptides were produced to estimated levels of up to 21% and 15%, respectively, of total cellular protein. The insoluble polypeptides were solubilized in 8 M urea. Refolding into active forms was achieved by treatment with the redox system cystine/cysteine and dilution. Refolded mature lipase was purified to homogeneity by affinity and ion exchange chromatography. The enzyme had a specific activity comparable to that of lipase from the fungal culture. The quantities of pure enzyme obtained from a 1-L culture of E. coli exceeded those obtained from the fungal culture by a factor of at least 100. Refolded recombinant prolipase was purified essentially to homogeneity and had a specific activity similar to that of the mature enzyme. Its pH optimum was 7.5, rather than the pH 8 determined for recombinant mature lipase and for the enzyme purified from the fungal culture. Recombinant prolipase retained activity after 15 min incubation at 65 degrees C, while mature lipase retained activity only up to 45 degrees C.

Altered acyl chain length specificity of Rhizopus delemar lipase through mutagenesis and molecular modeling.

The acyl binding site of Rhizopus delemar prolipase and mature lipase was altered through site-directed mutagenesis to improve lipase specificity for short- or medium-chain length fatty acids. Computer-generated structural models of R. delemar lipase were used in mutant protein design and in the interpretation of the catalytic properties of the resulting recombinant enzymes. Molecular dynamics simulations of the double mutant, val209trp + phe112trp, predicted that the introduction of trp112 and trp209 in the acyl binding groove would sterically hinder the docking of fatty acids longer than butyric acid. Assayed against a mixture of triacylglycerol substrates, the val209trp + phe112trp mature lipase mutant showed an 80-fold increase in the hydrolysis of tributyrin relative to the hydrolysis of tricaprylin while no triolein hydrolysis was detected. By comparison, the val94Trp mutant, predicted to pose steric or geometric constraints for docking fatty acids longer than caprylic acid in the acyl binding groove, resulted in a modest 1.4-fold increase in tricaprylin hydrolysis relative to the hydrolysis of tributyrin. Molecular models of the double mutant phe95asp + phe214arg indicated the creation of a salt bridge between asp95 and arg214 across the distal end of the acyl binding groove. When challenged with a mixture of triacylglycerols, the phe95asp + phe214arg substitutions resulted in an enzyme with 3-fold enhanced relative activity for tricaprylin compared to triolein, suggesting that structural determinants for medium-chain length specificity may reside in the distal end of the acyl binding groove. Attempts to introduce a salt bridge within 8 A of the active site by the double mutation leu146lys + ser115asp destroyed catalytic activity entirely. Similarly, the substitution of polar Gln at the rim of the acyl binding groove for phe112 largely eliminated catalytic activity of the lipase.

Lipase-catalyzed fractionation of conjugated linoleic acid isomers.

The abilities of lipases produced by the fungus Geotrichum candidum to selectively fractionate mixtures of conjugated linoleic acid (CLA) isomers during esterification of mixed CLA free fatty acids and during hydrolysis of mixed CLA methyl esters were examined. The enzymes were highly selective for cis-9,trans-11-18:2. A commercial CLA methyl ester preparation, containing at least 12 species representing four positional CLA isomers, was incubated in aqueous solution with either a commercial G. candidum lipase preparation (Amano GC-4) or lipase produced from a cloned high-selectivity G. candidum lipase B gene. In both instances selective hydrolysis of the cis-9,trans-11-18:2 methyl ester occurred, with negligible hydrolysis of other CLA isomers. The content of cis-9, trans-11-18:2 in the resulting free fatty acid fraction was between 94 (lipase B reaction) and 77% (GC-4 reaction). The commercial CLA mixture contained only trace amounts of trans-9,cis-11-18:2, and there was no evidence that this isomer was hydrolyzed by the enzyme. Analogous results were obtained with these enzymes in the esterification in organic solvent of a commercial preparation of CLA free fatty acids containing at least 12 CLA isomers. In this case, G. candidum lipase B generated a methyl ester fraction that contained >98% cis-9,trans-11-18:2. Geotrichum candidum lipases B and GC-4 also demonstrated high selectivity in the esterification of CLA with ethanol, generating ethyl ester fractions containing 96 and 80%, respectively, of the cis-9,trans-11 isomer. In a second set of experiments, CLA synthesized from pure linoleic acid, composed essentially of two isomers, cis-9,trans-11 and trans-10,cis-12, was utilized. This was subjected to esterification with octanol in an aqueous reaction system using Amano GC-4 lipase as catalyst. The resulting ester fraction contained up to 97% of the cis-9,trans-11 isomer. After adjustment of the reaction conditions, a concentration of 85% trans-10,cis-12-18:2 could be obtained in the unreacted free fatty acid fraction. These lipase-catalyzed reactions provide a means for the preparative-scale production of high-purity cis-9,trans-11-18:2, and a corresponding CLA fraction depleted of this isomer.

Isolation and identification of a degradation product in a capsule formulation containing the elastase inhibitor, DMP 777.

An unexpected degradation product, greater than 0.10%, was observed in a DMP 777 capsule formulation stored at 40 degrees C/75% r.h. for 3 months and 25 degrees C/60% r.h. for 2 years. The degradant of interest was prepared in quantity by refluxing the drug substance in dilute acid. A preparative HPLC method was developed to separate the various degradants and to collect each as a separate fraction. Each fraction was analyzed by the analytical HPLC gradient test method to assure positive identification of each peak and to correlate each peak to the original capsule sample. Key isolated degradation products were used for structure elucidation with mass spectrometry and NMR. The major degradant of interest in the capsule formulation was found to be a carboxylic acid resulting from the acid hydrolysis of an amide bond.