No abstract

Serratia marcescens, a gram-negative enteric bacterium, is capable of secreting a number of proteins extraceliularly. The types of activity found in the growth media include proteases, chitinases, a nuclease, and a lipase. Genetic studies have been undertaken to investigate the mechanisms used for the extracellular secretion of these exoproteins by S. marcescens. Many independent mutations affecting the extracellular enzymes were isolated after chemical and transposon mutagenesis. Using indicator media, we have identified loci involved in the production or excretion of extraceliular protease, nuclease, or chitinase by S. marcescens. None of the mutations represented general extracellular-excretion mutants; in no case was the production or excretion of multiple exoproteins affected. A variety of loci were identified, including regulatory mutations affecting nuclease and chitinase expression. A number of phenotypically different protease mutants arose. Some of them may represent different gene products required for the production and excretion of the major metalioprotease, a process more complex than that for the other S. marcescens exoproteins characterized to date.The movement of molecules from the site of synthesis to a new location is a fundamental property of biological systems. Protein export to the cell envelope has been the object of intensive study in Escherichia coli (4), and mutations affecting this process have been identified (18,30). Many of these mutations are conditional, demonstrating that protein export to the cell envelope is essential to the cell. The extracellular secretion (or excretion) of some proteins into the growth medium can also be achieved by some bacteria (29, 32), although enteric bacteria as a group are not renowned for their ability to excrete proteins. In fact, E. coli only excretes proteins when it carries extrachromosomal elements specifying exoproteins. For some proteins, such as a-hemolysin, bacteriocins, and toxins (14,19,22,26,32), the release of the exoprotein usually requires the presence of other gene products; this implies that E. coli is not normally endowed with a general extracellular secretory system. Since E. coli is such a limited system for studying extracellular proteins, a range of other organisms has been investigated (29,32). Usually the exoproteins these organisms excrete are either toxins or degradatory proteins like proteases and nucleases.Some steps, such as signal sequence (31) recognition and processing, may be common to both envelope and extracellular protein translocation. Genes required for the export of both membrane-bound and extracellular proteins are essential to the cell, but genes required only for excretion appear not to be essential. By isolating mutants that are defective only in the excretion of extracellular proteins, it may be possible to separate this mechanism from envelope secretion.Mutations in Pseudomonas aeruginosa strains defective in the excretion of certain extracellular proteins have been isolated (41) formation of many but not ...

We present an approach to predicting protein structural class that uses amino acid composition and hydrophobic pattern frequency information as input to two types of neural networks: (1) a three-layer back-propagation network and ( 2 ) a learning vector quantization network. The results of these methods are compared to those obtained from a modified Euclidean statistical clustering algorithm. The protein sequence data used to drive these algorithms consist of the normalized frequency of up to 20 amino acid types and six hydrophobic amino acid patterns. From these frequency values the structural class predictions for each protein (all-alpha, all-beta, or alpha-beta classes) are derived. Examples consisting of 64 previously classified proteins were randomly divided into multiple training (56 proteins) and test (8 proteins) sets. The best performing algorithm on the test sets was the learning vector quantization network using 17 inputs, obtaining a prediction accuracy of 80.2%. The Matthews correlation coefficients are statistically significant for all algorithms and all structural classes. The differences between algorithms are in general not statistically significant. These results show that information exists in protein primary sequences that is easily obtainable and useful for the prediction of protein structural class by neural networks as well as by standard statistical clustering algorithms.

No abstract

We have developed a computerized search pattern for recognition of the three-dimensional redox site of thioredoxins based on primary and predicted secondary structure. This pattern, developed in the ARIADNE protein expert system, is used to search for thioredoxin-like tertiary structural motif among proteins for which the only structural information is the primary sequence. The pattern was trained on 102 protein sequences (25 functionals and 77 controls); it matches all 25 members of the functional set under cutoff conditions that include only 2 members of the control set, for a sensitivity of 1.0 and a specificity of 0.97. The pattern matches only one of the two thioredoxin-like domains in protein disulfide isomerases (PDIs) and their analogues, suggesting that the C-terminal domain is more structurally similar to thioredoxin than the N-terminal domain. The Escherichia coli DsbA protein, a possible PDI analogue, appears to be more structurally similar to the N-terminal thioredoxin-like domain of PDIs. Thioredoxin-like redox functionality has been proposed for lutropin and follitropin, in part on the basis of their having -Cys-X-Pro-Cys- sequences. None match our pattern; all lack a predicted alpha-helix pattern element immediately after the active site. Hypothetical proteins in the National Biomedical Research Foundation Protein Identification Resource database were searched for matches to the pattern. The most interesting match was a hypothetical protein (161 residues) from the third open reading frame in the Staphylococcus aureus mer operon, which is involved in mercury detoxification. The match to our pattern and the hydrophobicity distribution in aligned elements of secondary structure not in our pattern strongly suggest that it has thioredoxin-like structure.