Machine learning approaches for phenotype-genotype mapping: predicting heterozygous mutations in the CYP21B gene from steroid profiles

OBJECTIVE: Non-linear relations between multiple biochemical parameters are the basis for the diagnosis of many diseases. Traditional linear analytical methods are not reliable predictors. Novel nonlinear techniques are increasingly used to improve the diagnostic accuracy of automated data interpretation. This has been exemplified in particular for the classification and diagnostic prediction of cancers based on expression profiling data. Our objective was to predict the genotype from complex biochemical data by comparing the performance of experienced clinicians to traditional linear analysis, and to novel non-linear analytical methods. DESIGN AND METHODS: As a model, we used a well-defined set of interconnected data consisting of unstimulated serum levels of steroid intermediates assessed in 54 subjects heterozygous for a mutation of the 21-hydroxylase gene (CYP21B) and in 43 healthy controls. RESULTS: The genetic alteration was predicted from the pattern of steroid levels with an accuracy of 39% by clinicians and of 64% by linear analysis. In contrast, non-linear analysis, such as self-organizing artificial neural networks, support vector machines, and nearest neighbour classifiers, allowed for higher accuracy up to 83%. CONCLUSIONS: The successful application of these non-linear adaptive methods to capture specific biochemical problems may have generalized implications for biochemical testing in many areas. Nonlinear analytical techniques such as neural networks, support vector machines, and nearest neighbour classifiers may serve as an important adjunct to the decision process of a human investigator not 'trained' in a specific complex clinical or laboratory setting and may aid them to classify the problem more directly.     

The nine residue plasminogen-binding motif of the pneumococcal enolase is the major cofactor of plasmin-mediated degradation of extracellular matrix, dissolution of fibrin and transmigration

The glycolytic enzyme alpha-enolase represents one of the nonclassical cell surface plasminogen-binding proteins of Streptococcus pneumoniae. In this study we investigated the impact of an internal plasminogen-binding motif of enolase on degradation of extracellular matrix and pneumococcal transmigration. In the presence of host-derived plasminogen activators (PA) tissue-type PA or urokinase PA and plasminogen S. pneumoniae expressing wild-type enolase efficiently degraded Matrigel or extracellular matrix (ECM). In contrast, amino acid substitutions in the nine residue plasminogen-binding motif of enolase significantly reduced degradation of ECM or Matrigel by mutated pneumococci. Similarly, recombinant wild-type enolase but not a mutated enolase derivative that lacks plasminogen-binding activity efficiently degraded ECM and Matrigel, respectively. In particular, bacterial cell enolase-bound plasmin potentiated dissolution of fibrin or laminin and transmigration of pneumococci through a fibrin matrix. In conclusion, these results provide evidence that the enolase is the major plasminogen-binding protein of pneumococci and that the nine residue plasminogen-binding motif of enolase is the key cofactor for plasmin-mediated pneumococcal degradation and transmigration through host ECM.     

Staphylococcus aureus interactions with the endothelium: the role of bacterial "secretable expanded repertoire adhesive molecules" (SERAM) in disturbing host defense systems

The intravascular manifestation of Staphylococcus aureus infection is often associated with a severe, and sometimes catastrophic disease. Many host factors contribute to endothelial tropism of S.aureus including subendothelial matrix proteins, endothelial cell receptors, and platelets that are engaged together with S. aureus cell wall adhesins such as the fibronectin binding proteins. Recently, the role of secreted staphylococcal factors that were initially identified by virtue of their binding function with host proteins and ligands, has been reappraised in this regard. Among these, bacterial proteins without significant homology among each other, coagulase (Coa), the extracellular fibrinogen binding protein (Efb), the extracellular matrix binding protein (Emp), or the extracellular adhesive protein (Eap), are the most prominent ones to be associated with endovascular disease. Newly discovered interactions with host components may account for profound effects on immunmodulation and wound healing which are summarized in this short review and which ascribe an important role of these molecules in acute and chronic endo- and extravascular staphylococcal disease. Further research in the complex functional role of these "secretable expanded repertoire adhesive molecules" (SERAM) may not only help to increase our understanding in the pathogenesis of S. aureus infection but can specify novel targets for preventive or therapeutic strategies.     

Antiallodynic effects of NMDA glycine(B) antagonists in neuropathic pain: possible peripheral mechanisms

NMDA receptors are implicated in central sensitisation underlying chronic pain, and NMDA antagonists have a potential for the treatment of neuropathic pain. Functional NMDA receptors are also present on primary afferents, where they may play a role in pro-nociceptive plasticity. The importance of this mechanism in neuropathic pain remains unclear. In the present work, we have compared in models of chronic pain the effects of NMDA antagonists at the glycine(B) site with different central access. L-701,324 (the centrally active antagonist) and 5,7-dichlorokynurenic acid (5,7-DCK, known to have limited central access) were tested after systemic administration in rats in the formalin test and in two models of neuropathic pain. The ability of these compounds to exert central actions (sedation, ataxia) was tested in the open field locomotion test; central NMDA antagonism in vivo was tested in anaesthetised rats on responses of spinal cord neurones to iontophoretic NMDA. Both L-701,324 (2.15-21.5 mg/kg i.p.) and 5,7-DCK (10-46.4 mg/kg i.v.) dose-dependently inhibited Phase II of formalin-evoked behaviour. Likewise, both compounds reversed cold allodynia in the chronic constriction injury model and tactile allodynia in animals with spinal nerve ligation. However, only L-701,324 was able to inhibit neuronal responses to NMDA in the antihyperalgesic dose range; 5,7-DCK was inactive on NMDA responses up to 46.4 mg/kg i.v. or 68.1 mg/kg i.p. Consistent with the lack of inhibition of central NMDA-evoked activity, 5,7-DCK did not alter spontaneous behaviour in the open field test, whereas it was significantly inhibited by L-701,324. Thus, peripheral NMDA receptors may substantially contribute to the efficacy of NMDA antagonists in neuropathic pain.     

Inhibitors of cation-chloride-cotransporters affect hypoxic/hypoglycemic injury in hippocampal slices

Electroneutral cation-chloride cotransporters are abundantly expressed in the brain and are involved in the regulation of the intracellular Cl(-) concentration and thus gamma-aminobutyric acid-dependent inhibition of neuronal excitability. As yet there is little evidence whether or not Na(+)-K(+)-2Cl(-) or K(+)-Cl(-) cotransporters are involved in neuronal hyperexcitability and death in cerebral ischemia. In this study, by measuring propidium iodide staining in organotypic hippocampal slice cultures from young rats and population spike recovery in acutely isolated hippocampal slices from adult rats after a hypoxic/hypoglycemic insult, we were able to assess if cation-chloride cotransport inhibitors reduce neuronal injury. The Na(+)-K(+)-2Cl(-) cotransport inhibitor bumetanide in the range of 1-10 microM reduced neuronal damage in the slice cultures by 25%, but did not affect population spike recovery in acutely isolated slices. In contrast the K(+)-Cl(-) cotransport inhibitor [(dihydroindenyl)oxy] alkanoic acid (DIOA, 100 microM) significantly diminished the restitution of the population spikes from 33% before to 8% after hypoxia/hypoglycemia and increased the damage in the slice cultures by 60%. Consequently, our data suggest that the Na(+)-K(+)-2Cl(-) cotransporter may contribute to neuronal injury and that the activity of the K(+)-Cl(-) cotransporters is an intrinsic protective mechanism of neurons against ischemic damage.     

Estrogen as an adjuvant therapy to antipsychotics does not prevent relapse in women suffering from schizophrenia: results of a placebo-controlled double-blind study

The expected therapeutic effect of estrogen as an adjunct treatment to antipsychotics in women suffering from schizophrenia for relapse prevention was to be tested under real-life conditions. A multicenter, randomized, placebo-controlled, double-blind, cross-over study based on an A-B-A-B (and/or B-A-B-A) design was applied. Forty-six hypoestrogenic women with schizophrenia hospitalized for the first time or repeatedly were included in the study. Their average age was 37.9 and they had been suffering from schizophrenia for 8.4 years. During the drug treatment phases, they received a three-phase estrogen-gestagen combination drug (17beta-estradiol+norethisterone acetate) in addition to an antipsychotic drug. Significant effects of the adjuvant hormone replacement therapy on the estradiol levels could be observed, and high and low levels of estradiol prevailed in the active drug and placebo phases, respectively. We did not find any difference either in defined relapse events or in the psychopathology between estradiol replacement and placebo phases. Neither did the required antipsychotic doses or the tolerance data differ between the two phases. Thus, the results of our study do not confirm the hypothesis that a combined estradiol/antipsychotic therapy is superior to an antipsychotic monotherapy for relapse prevention.