Structural and functional analysis of a bacterial cellulase by proteolysis

CenA is an endo-beta 1,4-glucanase from the cellulolytic bacterium Cellulomonas fimi. It is a bifunctional enzyme comprising an amino-terminal cellulose-binding domain and a carboxyl-terminal catalytic domain joined by a short sequence of prolyl and threonyl residues (the Pro-Thr box). Additional structural and functional information was revealed by a detailed analysis of the products generated by proteolytic cleavage of a nonglycosylated form of CenA. An extracellular C. fimi protease attacked nonglycosylated CenA at the junctions between the Pro-Thr box and the two functional domains. A stable "core" peptide (p30), corresponding to the catalytic domain, remained after extensive proteolysis. p30 was resistant to further attack even in the presence of 2-mercaptoethanol plus urea or dithiothreitol, but treatment in the presence of sodium dodecyl sulfate allowed complete fragmentation to small peptides. Stable peptides, identical, or closely related to p30, were generated by alpha-chymotrypsin or papain. These results indicated that the catalytic domain adopts a tightly folded conformation affording protection from proteolytic attack. In contrast, the cellulose-binding domain showed a relatively loose conformation. Progressive proteolytic truncation from the amino terminus was apparent during incubation with alpha-chymotrypsin or papain, or with C. fimi protease under reducing conditions. Affinity for cellulose was retained by products missing up to 64 amino-terminal amino acids. The remaining carboxyl-proximal region of the cellulose-binding domain with affinity (47 amino acids) contained sequences highly conserved in analogous domains from other bacterial endo-beta 1,4-glucanases. By analogy with other systems, the properties of the Pro-Thr box are consistent with an elongated conformation. The results of this investigation suggest that CenA has a tertiary structure which resembles that of certain fungal cellulases.     

A covalently constrained congener of the Saccharomyces cerevisiae tridecapeptide mating pheromone is an agonist

An analog of alpha-factor, the Saccharomyces cerevisiae tridecapeptide mating pheromone (Trp-His-Trp-Leu-Gln-Leu-Lys-Pro-Gly-Gln-Pro-Met-Tyr), in which the side chains of Lys7 and Gln10 were covalently linked, was synthesized using solid phase methodologies. The yield of the purified cyclic analog cyclo7,10[Nle12]alpha-factor was 30%, and its structure was verified by amino acid analysis, peptide sequencing, fast atom bombardment-mass spectrometry, and proton nuclear magnetic resonance spectroscopy. Cyclo7,10[Nle12]alpha-factor caused growth arrest and morphological alterations in S. cerevisiae MATa cells qualitatively identical to those induced by linear pheromone and was one-fourth to one-twentieth as active as the linear alpha-factor depending upon the S. cerevisiae strain tested. Consistent with the relative activities of the linear and cyclic peptides, binding competition studies indicated that cyclo7,10[Nle12]alpha-factor had approximately 20-40-fold less affinity for the alpha-factor receptor. Hydrolysis of the cyclic peptide by the target cells did not lead to opening of the ring and was less rapid than that of linear alpha-factor. The alpha-factor antagonist des-Trp1-[Ala3,Nle12]alpha-factor reversed the activity of the cyclic analog, and cyclo7,10[Nle12]alpha-factor was not active at the restrictive temperature in a temperature-sensitive receptor mutant. These results support the conclusion that the cyclic alpha-factor occupies the same binding site within the receptor as is occupied by the natural pheromone. The cyclic alpha-factor represents a rare example of an agonist among covalently constrained congeners of small linear peptide messengers.     

The fragmentation of 670A MeV neon-20 as a function of depth in water. II. One-generation transport theory

The results of an experiment to study the interaction of a beam of 670A MeV neon ions incident on a water column set to different thicknesses were compared with a "first principles" transport calculation in the straight-ahead approximation. This calculation assumes that the nuclear interactions of the incident particles lead to a secondary particle with the velocity of the incident projectile at the interaction point moving in the direction of the incident projectile. Subsequent nuclear interactions of the fragments were taken into account partially, by calculating the nuclear attenuation of the fragments in the residual material, but were not taken into account as a source of further nuclear interaction products. Fluence spectra were calculated per unit incident neon fluence for 14 absorber thicknesses. The acceptance for each fragment was calculated based on a knowledge of the material in the beam and of the beam extraction energy. The theoretical spectra were multiplied by the calculated acceptance and convoluted with the LET resolution associated with the experiment. The stopping power used in the transport calculation was found to predict a range approximately 1.6% shorter than that given by experiment; this small difference resulted in significant discrepancies between theory and experiment in the stopping region. For particles not stopping in the absorber, the transport calculation was accurate to within 30% for depths less than approximately 15 cm; the effects of tertiary particles become significant at greater depth.     

Tissue distribution of cocaine in the pregnant rat

Cocaine hydrochloride was administered by single intraperitoneal (IP) doses to pregnant rats at day 18 or 19 of gestation. Plasma and tissue cocaine and norcocaine concentrations were measured by high-pressure liquid chromatography. Pharmacokinetic analysis of concentration versus time data showed rapid distribution of cocaine and its metabolite to maternal and fetal tissues. The area under the cocaine concentration versus time curve (AUC) in fetus compared to maternal plasma was 3.33. The half-life of cocaine in the maternal plasma and fetus was 46 and 55 minutes, respectively, similar to values reported for cocaine elimination half-life in human plasma. The order of cocaine concentrations was placenta greater than fetal liver greater than maternal heart greater than whole fetus greater than fetal brain greater than maternal brain = maternal plasma. Norcocaine concentrations were usually less than 20% of cocaine concentrations in plasma and tissues. These results support extensive fetal exposure to cocaine following administration to pregnant rodents. Pharmacodynamic studies of cocaine in pregnancy should consider the effects of the drug on the developing fetus.     

Recombinant Treponema pallidum rare outer membrane protein 1 (Tromp1) expressed in Escherichia coli has porin activity and surface antigenic exposure.

We recently reported the cloning and sequencing of the gene encoding a 31-kDa Treponema pallidum subsp. pallidum rare outer membrane porin protein, designated Tromp1 (D. R. Blanco, C. I. Champion, M. M. Exner, H. Erdjument-Bromage, R. E. W. Hancock, P. Tempst, J. N. Miller, and M. A. Lovett, J. Bacteriol. 177:3556-3562, 1995). Here, we report the stable expression of recombinant Tromp1 (rTromp1) in Escherichia coli. rTromp1 expressed without its signal peptide and containing a 22-residue N-terminal fusion resulted in high-level accumulation of a nonexported soluble protein that was purified to homogeneity by fast protein liquid chromatography (FPLC). Specific antiserum generated to the FPLC-purified rTromp1 fusion identified on immunoblots of T. pallidum the native 31-kDa Tromp1 protein and two higher-molecular-mass oligomeric forms of Tromp1 at 55 and 80 kDa. rTromp1 was also expressed with its native signal peptide by using an inducible T7 promoter. Under these conditions, rTromp1 fractionated predominantly with the E. coli soluble and outer membrane fractions, but not with the inner membrane fraction. rTromp1 isolated from the E. coli outer membrane and reconstituted into planar lipid bilayers showed porin activity based on average single-channel conductances of 0.4 and 0.8 nS in 1 M KCl. Whole-mount immunoelectron microscopy using infection-derived immune serum against T. pallidum indicated that rTromp1 was surface exposed when expressed in E. coli. These findings demonstrate that rTromp1 can be targeted to the E. coli outer membrane, where it has both porin activity and surface antigenic exposure.     

Formation of extrachromosomal circular DNA in HeLa cells by nonhomologous recombination.

Extrachromosomal circular DNA (eccDNA) generated from chromosomal DNA is found in all mammalian cells and increases with cell stress or aging. Studies of eccDNA structure and mode of formation provide insight into mechanisms of instability of the mammalian genome. Previous studies have suggested that eccDNA is generated through a process involving recombination between repetitive sequences. However, we observed that approximately one half of the small eccDNA fragments cloned from HeLa S3 cells were composed entirely of nonrepetitive or low-copy DNA sequences. We analyzed four of these fragments by polymerase chain reaction and nucleotide sequencing and found that they were complete eccDNAs. We then screened a human genomic library with the eccDNAs to isolate the complementary chromosomal sequences. Comparing the recombination junctions within the eccDNAs with the chromosomal sequences from which they were derived revealed that nonhomologous recombination was involved in their formation. One of the eccDNAs was composed of two separate sequences from different parts of the genome. These results suggest that rejoining of ends of fragmented DNA is responsible for the generation of a substantial portion of the eccDNAs found in HeLa S3 cells.