Simulation studies of data classification by artificial neural networks: Potential applications in medical imaging and decision making

Artificial neural networks are being investigated in the field of medical imaging as a means to facilitate pattern recognition and patient classification. In the work reported here, the effects of internal structure and the nature of input data on the performance of neural networks were investigated systematically using computer-simulated data. Network performance was evaluated quantitatively by means of receiver operating characteristic analysis and compared with the performance of an ideal statistical decision maker. We found that the relatively simple neural networks investigated in this study can perform at the level of an ideal decision maker. These simple networks were also found to learn accurately even when the training data are extremely unbalanced with respect to the prevalence of actually positive cases and to differentiate input data patterns by recognizing their unique characteristics.     

Relationship between production of epidermal growth factor receptors, gene amplification, and chromosome 7 translocation in variant A431 cells

Synthesis of the epidermal growth factor (EGF) receptor has been analyzed in a series of variant A431 human epidermoid carcinoma cell clones reported to contain different amounts of EGF binding sites. The amount of EGF receptor protein, quantitated by immunoaffinity chromatography, and EGF receptor mRNA, quantitated by cDNA hybridization, were closely correlated to the extent of EGF receptor gene amplification. This correlation existed in variants selected for reduced EGF receptors and in revertants from those variants with increased EGF receptors. There was also a correlation between the frequency of translocation of chromosome 7, containing the EGF receptor gene, and EGF receptor protein. These results support gene amplification as the mechanism enhancing A431 cell EGF receptor protein and determining growth responses.     

T-cell acute lymphoblastic leukemia with add(1)(p36) and del(12)(p11) following acute myelocytic leukemia with partial deletion of 9p

We describe the case of a 40-year-old man whose disease was initially diagnosed as acute myelocytic leukemia. The patient achieved remission with chemotherapy, but relapsed shortly afterwards with an acute T-cell lymphoblastic leukemia. He died of intracranial bleeding. Karyotyping analysis showed a del(9p?) as a common abnormality in the leukemic cells at onset and relapse. Fluorescence in situ hybridization analysis demonstrated allelic loss of the CDKN2A gene in cells from both stages of the disease. At relapse the leukemia cells had additional abnormalities such as add(1)(p36) and del(12)(p11). We postulate that the loss of CDKN2A is involved in leukemogenesis but does not determine the lineage of the leukemic cells. Instead, abnormalities of genes at 1p36, 12p11, or both may be involved in driving a lymphoid phenotype.     

Statistical and integrative approach for constructing biological network maps

A goal of systems biology is to build a concrete biochemical network map, which provides an important instruction to trace the pathways of interest or to understand the mechanism of a biological system. In the postgenomic era, not only the concrete biochemical maps, but also postgenomic maps (mRNA coexpression and protein-protein interaction networks) have been extensively produced. In the biochemical map, the individual reactions are reliable, but the number of the reactions is limited, because molecular biology requires extensive experiments to verify them. By contrast, postgenomic data provide much information regarding interactions, but are coarse-grained. To expand the biochemical network, an intuitional approach, which superposes postgenomic data on the map one by one, has been carried out, but it is not effective when a large amount of the coarse-grained data is handled. In order to effectively integrate such postgenomic interactions into a biochemical map, a statistical approach would be suitable rather than intuition. In this article, we proposed a novel statistical approach that integrates postgenomic interaction networks into the biochemical network, predicting novel pathways. A statistical correlation for such different types of networks identifies functional modules; subsequently the superposition of the different networks on the functional modules predicts inter-modular relations, which are the key pathways to construct a large-scale biochemical network.     

Kanamycin ototoxicity in glutamate transporter knockout mice

Glutamate-aspartate transporter (GLAST), a powerful glutamate uptake system, removes released glutamate from the synaptic cleft and facilitates the re-use of glutamate as a neurotransmitter recycling system. Aminoglycoside-induced hearing loss is mediated via a glutamate excitotoxic process. We investigated the effect of aminoglycoside ototoxicity in GLAST knockout mice using the recorded auditory brainstem response (ABR) and number of hair cells in the cochlea. Kanamycin (100 mg/mL) was injected directly into the posterior semicircular canal of mice. Before the kanamycin treatment, there was no difference in the ABR threshold average between the wild-type and knockout mice. Kanamycin injection aggravated the ABR threshold in the GLAST knockout mice compared with the wild-type mice, and the IHC degeneration was more severe in the GLAST knockout mice. These findings suggest that GLAST plays an important role in preventing the degeneration of inner hair cells in aminoglycoside ototoxicity.     

A novel STK11 germline mutation in two siblings with Peutz-Jeghers syndrome complicated by primary gastric cancer

Patients with Peutz-Jeghers syndrome (PJS) are known to be at risk of gastric cancer (GC), and the STK11 gene is a susceptibility gene for PJS. However, as no cases of PJS with GC in which a STK11 germline mutation has been identified have ever been reported and other susceptibility genes have also been suggested to be involved in PJS, the relation between STK11 germline mutations and GC in PJS is still unknown. In this study, we used sequencing analysis to investigate the STK11, CDH1, and TP53 loci for a germline mutation in two siblings with PJS with primary GC. A novel type of the STK11 germline mutation, c.890delG, encoding a truncated protein (p.Arg297fsX38) was identified, but no germline mutations of the CDH1 and TP53 genes were detected. No inactivation of the wild-type allele by somatic mutation or chromosomal deletion or hypermethylation at the 5'-CpG site of STK11 was detected in the GC. This is the first report of a STK11 germline mutation in a PJS patient with GC and should contribute to establishing correlations between the STK11 germline mutations and GC in PJS patients.     

Receptors and transporter for serotonin in Merkel cell-nerve endings in the rat sinus hair follicle. An immunohistochemical study

Serotonin (5-HT) has been a candidate for neurotransmitters in cutaneous type I mechanoreceptors (i.e., Merkel cell-nerve endings). Although recent electrophysiological studies have suggested the presence of the 5-HT2 and 3 receptors in the Merkel cell-nerve endings, the histological localization of these receptors are obscure. We thus immunohistochemically examined the presence of 5-HT1, 2, 3 receptors in Merkel cell-nerve endings in sinus hair follicles of the rat whisker pad. We also studied the immunohistochemical localization of the 5-HT transporter to confirm the site of 5-HT secretion. For this purpose, we used antibodies for the 5-HT1A, 5-HT1B, 5-HT2A, 5-HT2C and 5-HT3 receptors, and for the 5-HT transporter, as well as antibodies for cytokeratin 20 (as a marker of Merkel cells) and neurofilament H (a marker of type I sensory nerve terminals). The immuno-stained sections were analyzed under a laser-scanning microscope. It was found that the sensory nerve terminals in the Merkel cell-nerve endings showed strong positive immunoreactions of 5-HT1A and 1B receptors but not 5-HT2A, 2C, and 3 receptors. Furthermore, both the Merkel cells and related axon terminals showed strong immunoreactions of the 5-HT transporter. These findings support the idea that 5-HT molecules are released from the Merkel cells during mechanical reception and indirectly regulate neural actions of sensory neurons via 5-HT1 receptors. The localization of the 5-HT transporter found in this study also suggests a possibility that axon terminals in the Merkel cell-nerve endings also release 5-HT.     

Bcl11b is required for differentiation and survival of alphabeta T lymphocytes

The gene Bcl11b, which encodes zinc finger proteins, and its paralog, Bcl11a, are associated with immune-system malignancies. We have generated Bcl11b-deficient mice that show a block at the CD4-CD8- double-negative stage of thymocyte development without any impairment in cells of B- or gammadelta T cell lineages. The Bcl11b-/- thymocytes showed unsuccessful recombination of V(beta) to D(beta) and lacked the pre-T cell receptor (TCR) complex on the cell surface, owing to the absence of Tcrb mRNA expression. In addition, we saw profound apoptosis in the thymus of neonatal Bcl11b-/- mice. These results suggest that Bcl11b is a key regulator of both differentiation and survival during thymocyte development.     

Active matrix metalloproteinases in the tear fluid of individuals with vernal keratoconjunctivitis

Corneal epithelial lesions distinguish vernal keratoconjunctivitis (VKC) from other ocular allergic diseases. Such lesions result from degradation of the corneal epithelial basement membrane, which comprises mostly type IV collagen and laminin. Matrix metalloproteinase 2 (MMP-2) and MMP-9 catalyze the degradation of these 2 extracellular matrix proteins. The possible role of MMP-2 and MMP-9 in the pathogenesis of corneal lesions associated with VKC was investigated by assaying tear fluid for the presence of these enzymes. Tear fluid was collected from 6 eyes of 6 patients with active VKC, 14 eyes of 14 patients with active allergic conjunctivitis, and 6 eyes of 6 nonallergic healthy volunteers. Gelatin zymography revealed that the tear fluid of healthy volunteers contained inactive proforms of both MMP-2 and MMP-9 but not the active forms of these enzymes. Active forms of MMP-2 or MMP-9 were detected in a minority of patients with allergic conjunctivitis. However, with the exception of one individual for whom active MMP-9 was not detected, tear fluid from all patients with VKC contained both proforms and active forms of MMP-2 and MMP-9. These results implicate MMP-2 and MMP-9 in the pathogenesis of corneal epithelial disorders associated with VKC.     

Possible contribution of follicular interleukin-1beta to nitric oxide generation in human pre-ovulatory follicles

The aim of this study was to investigate the relationships between follicular nitric oxide (NO) metabolite concentrations and several related variables, with special reference to follicular interleukin- 1beta (IL-1beta). The follicular fluid from the leading and secondary follicles was collected individually from 20 women undergoing in-vitro fertilization (IVF) treatment, and the concentrations of nitrite (NO2-) and nitrate (NO3-) were determined fluorometrically using 2,3- diaminonaphthalene. Both follicular nitrite (r = 0.42, P < 0.01) and nitrate (r = 0.49, P < 0.001) were found to be significantly correlated with follicular IL-1beta concentrations. There were also significant positive correlations between follicular nitrate and the number of oocytes retrieved (P < 0.01) and serum oestradiol concentration on the day of human chorionic gonadotrophin (HCG) administration (P < 0.05). When follicular cells were incubated in vitro with 10 ng/ml of IL-1beta for 24 h, nitrate generation was significantly (P < 0.01) elevated compared with the control. In conclusion, our study demonstrates that follicular IL-1beta and the number of developing follicles are significant variables that affect follicular NO concentrations, and points to the possible contribution of IL-1beta to NO generation in human preovulatory follicles.