The Chinese bamboo leaf odorous frog (Rana (Odorrana) versabilis) and North American Rana frogs share the same families of skin antimicrobial peptides.

The Chinese bamboo leaf odorous frog (Rana (Odorrana) versabilis) and the North American pickerel frog (Rana palustris) occupy different ecological niches on two different continents with no overlap in geographical distribution. R. palustris skin secretions contain a formidable array of antimicrobial peptides including homologs of brevinin-1, esculentin-1, esculentin-2, ranatuerin-2, a temporin and a family of peptides considered of unique structural attributes when isolated, palustrins 1–3. Here we describe the structures of mature peptides and precursors of eight putative antimicrobial peptides from the skin secretion of the Chinese bamboo leaf odorous frog (Rana (Odorrana) versabilis). Each peptide represents a structural homolog of respective peptide families isolated from R. palustris, including two peptides identical in primary structure to palustrin 1c and palustrin 3b. Additionally, two peptides were found to be structural homologs of ranatuerin 2B and ranatuerin 2P from the closely-related North American species, Rana berlandieri (the Rio Grande leopard frog) and Rana pipiens (the Northern leopard frog), respectively. Both palustrins and ranatuerins have hitherto been considered unique to North American ranid frogs. The use of primary structures of amphibian skin antimicrobial peptides is thus questionable as a taxonomic device or alternatively, the micro-evolution and/or ancestry of ranid frogs is more highly complex than previously thought.

Cloning of maximakinin precursor cDNAs from Chinese toad, Bombina maxima, venom.

Using a novel technique that we have developed for cloning of amphibian skin secretion peptide cDNAs from lyophilized samples, we report here that maximakinin (DLPKINRKGP-bradykinin) is encoded by two different cDNAs, named BMK-1 and BMK-2, containing either four tandem repeat sequences or a single copy. The open reading frames of both precursor cDNAs were found to be 152 and 116 amino acid residues, respectively. These data provide evidence that the structural diversity of peptides in amphibian skin secretions arising from molecular evolutionary events, can be mediated by parallel diversity in encoding mRNAs that in itself may reflect serial gene duplications.

Identification and molecular cloning of novel trypsin inhibitors from the dermal venom of the Oriental fire-bellied toad (Bombina orientalis) and the European yellow-bellied toad (Bombina variegata).

The structural diversity of polypeptides in amphibian skin secretion probably reflects different roles in dermal regulation or in defense against predators. Here we report the structures of two novel trypsin inhibitor analogs, BOTI and BVTI, from the dermal venom of the toads, Bombina orientalis and Bombina variegata. Cloning of their respective precursors was achieved from lyophilized venom cDNA libraries for the first time. Amino acid alignment revealed that both deduced peptides, consisting of 60 amino acid residues, including 10 cysteines and the reactive center motif, -CDKKC-, can be affirmed as structural homologs of the trypsin inhibitor from Bombina bombina skin.

Cloning of the (Thr6)-phyllokinin precursor from Phyllomedusa sauvagei skin confirms a non-consensus tyrosine 0-sulfation motif.

Nine bradykinin-related peptides were identified in Phyllomedusa sauvagei skin secretion using QTOF MS/MS fragmentation sequencing. The major peptides were (Thr6)-bradykinin, (Hyp3, Thr6)-bradykinin, (Thr6)-phyllokinin and (Hyp3, Thr6)-phyllokinin. The phyllokinins occurred in both sulfated and non-sulfated forms. All (Thr6)-substituted bradykinins/phyllokinins could be generated from a common precursor by differential post-translational processing and modification. The open-reading frame of the cloned precursor cDNA consisted of 62 amino acid residues with a single bradykinin/phyllokinin coding sequence located at the C-terminus. Structural features included a Glu-Arg processing site at the N-terminus of the bradykinin/phyllokinin domain and the absence of an acidic amino acid residue adjacent to the C-terminal Tyr residue in the phyllokinins. However, the neutral amino acid residue (Ile) at position -1 and the basic amino acid residue (Arg) at position -2 from the Tyr residue, constitute a sulfation motif previously identified only in a protochordean.

Bradykinin-related peptides, including a novel structural variant, (Val1)-bradykinin, from the skin secretion of Guenther's frog, Hylarana guentheri and their molecular precursors.

Multiple bradykinin-related peptides including a novel bradykinin structural variant, (Val1)-bradykinin, have been identified from the defensive skin secretion of Guenther's frog, Hylarana guentheri by a tandem mass spectrometry method. Subsequently, four different preprobradykinin cDNAs, which encoded multiple bradykinin copies and its structural variants, were consistently cloned from a skin derived cDNA library. These preprobradykinin cDNAs showed little structural similarity with mammalian kininogens and the kininogens from the skin of toads, but have regions that are highly conserved in the kininogens from another ranid frog, Odorrana schmackeri. Alignment of these preprobradykinins revealed that preprobradykinin 1, 2 and 3 may derive from a single gene by alternative exon splicing.

Helokinestatin: a new bradykinin B2 receptor antagonist decapeptide from lizard venom.

Synthetic bradykinin antagonist peptides/peptoids have been powerful tools for delineating the roles of kinins in both normal physiology and in pathological states. Here, we report the identification of a novel, naturally occurring bradykinin B2 receptor antagonist peptide, helokinestatin, isolated and structurally characterized from the venoms of helodermatid lizards—the Gila monster (Heloderma suspectum) and the Mexican beaded lizard (Heloderma horridum). The primary structure of the peptide was established by a combination of microsequencing and mass spectroscopy as Gly-Pro-Pro-Tyr-Gln-Pro-Leu-Val-Pro-Arg (Mr 1122.62). A synthetic replicate of helokinestatin was found to inhibit bradykinin-induced vasorelaxation of phenylephrine pre-constricted rat tail artery smooth muscle, mediated by the B2 receptor sub-type, in a dose-dependent manner. Natural selection, that generates functional optimization of predatory reptile venom peptides, can potentially provide new insights for drug lead design or for normal physiological or pathophysiological processes.

Novel brevinins from Chinese piebald odorous frog (Huia schmackeri) skin deduced from cloned biosynthetic precursors

Antimicrobial peptides represent the most characterized and diverse class of peptides within the defensive skin secretions of anuran amphibians. With an ever expanding database of primary structures, the current accepted rules for nomenclature have become increasingly difficult to apply to peptides whose primary structural attributes are either unique or that fall between those that define existing groups. An additional factor that adds to the confusion is the regular re-classification or revision of existing taxa. In the present study, we have identified five new antimicrobial peptide homologs in the defensive skin secretion of the Chinese piebald odorous frog, Huia schmackeri (formerly Rana (Odorrana) schmackeri), by cloning of their respective biosynthetic precursors. As these peptides are obvious homologs of the brevinin-1 and brevinin-2 families we have named these in accordance: (1) brevinin-1HS1, (2) brevinin-2HS1, (3) brevinin-2HS2, (4) brevinin-2HS3 and (5) brevinin-1HS2. The reasons for adopting these names are discussed. It is clear that with an ever-increasing number of amphibian skin antimicrobial peptides appearing in the literature that a consistent nomenclature scheme needs to be established.

Advanced glycation of fibronectin impairs vascular repair by endothelial progenitor cells: Implications for vasodegeneration in diabetic retinopathy

PURPOSE. Vascular repair by marrow-derived endothelial progenitor cells (EPCs) is impaired during diabetes, although the precise mechanism of this dysfunction remains unknown. The hypothesis for the study was that progressive basement membrane (BM) modification by advanced glycation end products (AGEs) contributes to impairment of EPC reparative function after diabetes-related endothelial injury.

METHODS. EPCs isolated from peripheral blood were characterized by immunocytochemistry and flow cytometry. EPC interactions on native or AGE-modified fibronectin (AGE-FN) were studied for attachment and spreading, whereas chemotaxis to SDF-1 was assessed with the Dunn chamber assay. In addition, photoreactive agent-treated monolayers of retinal microvascular endothelial cells (RMECs) produced circumscribed areas of apoptosis and the ability of EPCs to “endothelialize” these wounds was evaluated.

RESULTS. EPC attachment and spreading on AGE-FN was reduced compared with control cells (P < 0.05–0.01) but was significantly restored by pretreatment with Arg-Gly-Asp (RGD). Chemotaxis of EPCs was abolished on AGE-FN but was reversed by treatment with exogenous RGD. On wounded RMEC monolayers, EPCs showed clustering at the wound site, compared with untreated regions (P < 0.001); AGE-FN significantly reduced this targeting response (P < 0.05). RGD supplementation enhanced EPC incorporation in the monolayer, as determined by EPC participation in tight junction formation and restoration of transendothelial electric resistance (TEER).

CONCLUSIONS. AGE-modification of vascular substrates impairs EPC adhesion, spreading, and migration; and alteration of the RGD integrin recognition motif plays a key role in these responses. The presence of AGE adducts on BM compromises repair by EPC with implications for vasodegeneration during diabetic microvasculopathy.

Neuroprotective actions of a histidine analogue in models of ischemic stroke.

Histidine is a naturally occurring amino acid with antioxidant properties, which is present in low amounts in tissues throughout the body. We recently synthesized and characterized histidine analogues related to the natural dipeptide carnosine, which selectively scavenge the toxic lipid peroxidation product 4-hydroxynonenal (HNE). We now report that the histidine analogue histidyl hydrazide is effective in reducing brain damage and improving functional outcome in a mouse model of focal ischemic stroke when administered intravenously at a dose of 20 mg/kg, either 30 min before or 60 min and 3 h after the onset of middle cerebral artery occlusion. The histidine analogue also protected cultured rat primary neurons against death induced by HNE, chemical hypoxia, glucose deprivation, and combined oxygen and glucose deprivation. The histidine analogue prevented neuronal apoptosis as indicated by decreased production of cleaved caspase-3 protein. These findings suggest a therapeutic potential for HNE-scavenging histidine analogues in the treatment of stroke and related neurodegenerative conditions.

Prediction- and simulation-error based perceptron training: Solution space analysis and a novel combined training scheme

Previous papers have noted the difficulty in obtaining neural models which are stable under simulation when trained using prediction-error-based methods. Here the differences between series-parallel and parallel identification structures for training neural models are investigated. The effect of the error surface shape on training convergence and simulation performance is analysed using a standard algorithm operating in both training modes. A combined series-parallel/parallel training scheme is proposed, aiming to provide a more effective means of obtaining accurate neural simulation models. Simulation examples show the combined scheme is advantageous in circumstances where the solution space is known or suspected to be complex. (c) 2006 Elsevier B.V. All rights reserved.

Computationally efficient sequential learning algorithms for direct link resource-allocating networks

Computionally efficient sequential learning algorithms are developed for direct-link resource-allocating networks (DRANs). These are achieved by decomposing existing recursive training algorithms on a layer by layer and neuron by neuron basis. This allows network weights to be updated in an efficient parallel manner and facilitates the implementation of minimal update extensions that yield a significant reduction in computation load per iteration compared to existing sequential learning methods employed in resource-allocation network (RAN) and minimal RAN (MRAN) approaches. The new algorithms, which also incorporate a pruning strategy to control network growth, are evaluated on three different system identification benchmark problems and shown to outperform existing methods both in terms of training error convergence and computational efficiency. (c) 2005 Elsevier B.V. All rights reserved.

Improving neural network training solutions using regularisation

This paper describes the application of regularisation to the training of feedforward neural networks, as a means of improving the quality of solutions obtained. The basic principles of regularisation theory are outlined for both linear and nonlinear training and then extended to cover a new hybrid training algorithm for feedforward neural networks recently proposed by the authors. The concept of functional regularisation is also introduced and discussed in relation to MLP and RBF networks. The tendency for the hybrid training algorithm and many linear optimisation strategies to generate large magnitude weight solutions when applied to ill-conditioned neural paradigms is illustrated graphically and reasoned analytically. While such weight solutions do not generally result in poor fits, it is argued that they could be subject to numerical instability and are therefore undesirable. Using an illustrative example it is shown that, as well as being beneficial from a generalisation perspective, regularisation also provides a means for controlling the magnitude of solutions. (C) 2001 Elsevier Science B.V. All rights reserved.

Support vector machine classification for large data sets via minimum enclosing ball clustering

Support vector machine (SVM) is a powerful technique for data classification. Despite of its good theoretic foundations and high classification accuracy, normal SVM is not suitable for classification of large data sets, because the training complexity of SVM is highly dependent on the size of data set. This paper presents a novel SVM classification approach for large data sets by using minimum enclosing ball clustering. After the training data are partitioned by the proposed clustering method, the centers of the clusters are used for the first time SVM classification. Then we use the clusters whose centers are support vectors or those clusters which have different classes to perform the second time SVM classification. In this stage most data are removed. Several experimental results show that the approach proposed in this paper has good classification accuracy compared with classic SVM while the training is significantly faster than several other SVM classifiers.