Cyclic modes in artificial neural nets

Artificial neural nets constructed of dicrete populations of 200–1000 formal neurons have been studied through computer simulation. Among the basic assumptions of operation of these nets are the following: a) Each neuron fires at times which are integral multiples of the synaptic delay . b) It produces the appropriate PSP's after . c) All the neurons have the same refractory period and d) temporal summation occurs without decrement, for a period less than the synaptic delay. The nets were specified by a number of parameters: fraction of inhibitory neurons in the population, average number of connections to each cell, threshold for cell firing. These parameters did not determine the detailed microscopical structures of nets which was established separately on a random basis.For the range of the parameters considered in this study it was found that neural nets are capable of supporting self-maintaining activity in the form of cycling modes, characterized by a fixed period. The period of the cycles can be altered by a steady, non-cycling external input to the net. Evidence is presented that the cycling modes depend upon the statistical parameters of the net and the stimulus characteristics rather than on the detailed structure of the net. These results suggest that non-structured nerve nets may respond in specific manner to specific stimuli.

Glossary

Parameters of Neural Net Model Synaptic delay - A Total number of neurons in the netlet - h Fraction of inhibitory neurons in the netlet (in % of total number of neurons) - ⁺ Average number of axon branches emanating from anexcitatory neuron - ^– Average number of axon branches emanating from an inhibitory neuron - k ⁺ Average EPSP produced by an excitatory neuron in arbitrary units of amplitude - k ^– Average IPSP produced by an inhibitory neuron in arbitrary units of amplitude - Firing threshold of neurons in the netlet - The minimum number of ESPS's necessary to trigger a neuron in the absence of inhibitory inputs - The minimum number of ESPS's necessary to trigger a neuron in the presence of inhibitory inputs. Dynamic Parameters of the Model n An integer giving the number of elapsed synaptic delays (i.e. elapsed time) - _n The activity; i.e. the fraction of active neurons in the netlet at t=n (the actual number of active cells is given by _nA) - _n={ⁱ_n} State vector of single netlet at time n This research has been supported by NIH grants NS-8012 and NS-8498, and NSF grant GB-30498. Computation assistance was provoded by the Health Sciences Computing Facility, UCLA, sponsored by NIH Special Research Resources grant RR-3.     

Noise effects in an electronic model of a single neuron

We consider a simple electronic circuit model of a single neuron. The neuron is assumed to be driven by an external signal comprising constant (dc) and random components. In addition, the nonlinearity parameter in the circuit is assumed to fluctuate, thereby giving rise to critical behavior including the onset of hysteresis phenomena even for system parameter values that would not otherwise support such behavior. This noise-induced critical behavior is analysed, in the long time limit, through a study of the probability density function describing the neural response.     

Modelling of the active reaction of stick insects by a network of neuromimes

Depending on the activity status of the animal, the control system of the femur-tibia joint in stick insects exhibits either a resistance reflex, or the active reaction, a totally different action pattern (Bässler 1988a). Using analog electronic neuron models, several different neuronal circuits are explored that model the active reaction with all its features. The models differ in complexity, redundance and the robustness against small variations of network parameters (e.g. coupling strengths). The circuit with the highest robustness and redundancy requires interneurons with special features, such as found in real animals. When inserted into a closed loop modeling movement and sensory feedback from the periphery, this circuit produces oscillations similar to searching movements found in the real animal. In addition to intracellular recording methods, the authors propose modeling with realistic neuromimes as a complementary method in the investigation of neuronal networks which have well documented input-output relationships.     

Experimental analysis of a neural system: Two modeling approaches

The retinal neural system in the catfish which transforms light intensity temporal variations into the horizontal cell potential is experimentally analyzed and modeled by two distinct methods. The first method involves testing the system with gaussian white-noisemodulated light intensity and the subsequent derivation of a mathematical model in terms of a Wiener functional series. The second method involves testing of the system by step and sinewave stimuli and the postulation of a set of nonlinear differential equations which are designed to fit these stimulus-response data. In this latter approach, the differential equations describe the usually assumed dynamic behavior of the component subsystems, such as photoreceptor and horizontal cell membranes in terms of properties of membrance resistance and capacitance. The system behavior is found to exhibit certain small signal nonlinearities such as dynamic asymmetry in the response as well as certain large signal nonlinearities. The two modeling approaches and the resulting models are compared and it is found that the functional model derived from the white-noise experiment, while it does not attempt to describe the underlying system structure as the differential equation does, produced, in general, more satisfactory results as far as the input-output behavior of the system is concerned. It is suggested that combination of the two approaches could be very fruitful in modeling a particular system.     

The occurrence and identification of intracellular polyglucose storage granules inMethylococcus NCIB 11083 grown in chemostat culture on methane

The accumulation of intracellular storage granules (0.03–0.5 m) byMethylococcus NCIB 11083 when grown under conditions of ammonia limitation with methane as the sole source of carbon and energy was inversely proportional to the dilution rate. The isolated material was composed entirely of glucose residues and the infra-red spectrum exhibited characteristic absorption bands at 925 cm^-1, 845 cm^-1 and 745±4 cm^-1, indicating the presence of (14) glycosidic linkages. The polymer dissolved in hot water to give an opalescent solution that formed a violet iodine complex with an absorption maximum at 550 nm, identical to that observed with reference amylopectin. The percentage of the polysaccharide released as maltose by the action of - and -amylases was 55–64% and 80–90% respectively, values very similar to those obtained by the action of these enzymes on reference amylopectin and glycogen. Methylation analysis indicated that the average interior and exterior chain lengths of the polymer were 2.7 and 10.0 glucose units respectively and confirmed that theMethylococcus polyglucose is a branched polymer composed of units joined by 14 and 16 linkages. The number average molecular weight of the polymer is 2–4.5×10⁵. The stored polymer was metabolised by the organism and its metabolism resulted in the synthesis of protein.Abbreviations GLC gas liquid chromatography - MS mass spectroscopy - PAAN peracetylated aldononitriles     

On the character of and distance between states in a binary switching net

Behavioral examination of binary switching net models has typically concerned itself with an examination of their cyclical character. This article considers two less frequently discussed behavioral variables-the density of 1's in net states and the (Hamming) distance between net states. These variables are studied under fully random nets and under nets controlled at levels of internal homogeneity, forcibility or threshold. A collection of theoretical and simulated results is presented.     

Color categorization and color constancy in a neural network model of V4

We develop a neural network model that instantiates color constancy and color categorization in a single unified framework. Previous models achieve similar effects but ignore important biological constraints. Color constancy in this model is achieved by a new application of the double opponent cells found in the blobs of the visual cortex. Color categorization emerges naturally, as a consequence of processing chromatic stimuli as vectors in a four-dimensional color space. A computer simulation of this model is subjected to the classic psychophysical tests that first uncovered these phenomena, and its response matches psychophysical results very closely.     

Abbau der 4-Chlorbenzoesäure durch eineArthrobacter-Species

Zusammenfassung EineArthrobacter-Species, die 4-Chlorobenzoesäure als einzige Kohlenstoffquelle verwerten kann, gibt beim Wachstum auf dieser Verbindung 4-Hydroxybenzoesäure und Protocatechusäure ins Medium ab. Der weiter Abbau des aromatischen Ringes erfolgt durch meta-Spaltung. Beim Wachstum derArthrobacter-Species auf Benzoesäure trit im Medium cis,cis-Muconsäure auf. In diesem Fallewird also der ortho-Weg eingeschlagen. Die Enzyme für beide Abbauwege sind induzierbar.

---

Degradation of 4-Chlorobenzoic acid by anArthrobacter species

AnArthrobacter sp. growing on 4-Chlorobenzoic acid as its sole source of carbon excretes 4-hydroxybenzoic acid and protocatechuic acid into the culture medium. Protocatechuic acid is further attacked by meta-cleavage. During growth of theArthrobacter sp. on benzoic acid cis-cis muconic acid can be isolated from the medium, suggesting the involvement of the ortho-cleavage pathway. The enzymes both for the meta- and the ortho-cleavage pathway are inducible.

Dem Andenken an Professor Bernhauer gewidmet     

Asymmetric diffusion into the postsynaptic neuron: An extremely efficient mechanism for removing excess GABA from synaptic clefts on the Deiters' neurone plasma membrane

Microdissected Deiters' neuron plasma membranes have been used for studying the passage of GABA through the membrane both in the inward and outward direction. Working with 0.2 mM GABA in the compartment simulating the outside of the neurone and with 2.0 mM GABA in the one simulating the inside we found a net transport of GABA towards the inside. This mechanism does not require a Na⁺ ion gradient across the membrane. The nature of the transport process involved was studied by determining the rate of [³H]-GABA inward passage as a function of GABA concentration (1 nM–800 M) on the outward side of the membrane. The results have shown that until 50 M a diffusion process (v=D₁×C, where D₁=3.1×10^–11 1/m²×sec) is the sole mechanism involved. Above 50 M a second diffusion process is activated v=D₂×(C–50×10^–6), where D₂=2.8×10^–11 1/m²×sec. Taking in account both inward and outward directed diffusion, one can calculate 16 M as the equilibrium concentration of GABA on the outward side of the membrane. From a kinetic point of view, these diffusion processes are able to reduce GABA concentration in a synaptic cleft from 3 mM to 20 M within 3 sec. These diffusion systems are discussed as extremely efficient in removing the excess of released GABA in the synaptic cleft.     

Organization of the nervous system of physonectid siphonophores

Summary An antiserum to the sequence Arg-Phe-amide (RFamide) was used to stain the nervous systems of various physonectid siphonophores. In the stem of Nanomia bijuga, this antiserum stained an ectodermal nerve net, which was interrupted, at regular intervals, by transverse collars of neurons. Injection of Lucifer yellow into the giant axon of the stem showed that this axon was dye-coupled to an ectodermal nerve net that resembled the RFamide-positive network. Ectodermal nets of neurons were also found in the pneumatophore, gastrozooids, tentacles and tentilla. At the junctions of the pneumatophore, the gastrozooids, the dactylozooids and the gonozooids with the stem, and at the junctions of tentacles and tentilla, collars or rings of neurons occurred. The stem was connected to the phyllozooids and nectophores by muscular lamellae, which were bordered by chains of neurons. At the margin of the nectophores, an immunoreactive nerve ring was found. Connected to this ring and located in theseitliche Zapfen (sidely-located patche), were two agglomerations of nerve cells. On the upper side of the bell margin, positioned at 90° relative to the seitliche Zapfen, a delta-shaped neuronal structure was found. This structure was connected to the nerve ring and was associated with a muscle, which ran a short distance along the exumbrellar surface.The nervous systems of Agalma elegans, Forskalia edwardsi, Forskalia leuckarti and Halistemma rubrum resembled that of Nanomia bijuga in all major respects.     

Effects of ambient and local cutaneous temperatures on hand blood flow

The hand blood flow ( ) was investigated in response to a wide range of general and local cutaneous thermal stimuli (0–36°C and 4–42°C respectively), the local stimulus consisting of a thermostatically controlled water bath for the right hand (T_w), and the general stimulus, the ambient room temperature (T_a). was measured at the right wrist by strain gauge plethysmography; it was seen to respond more significantly to variations in T_w than to those in T_a at cold to comfortable ambient temperatures (T_a<22°C). A paradoxical vasodilatation was observed at T_w=4°C (Lewis' hunting phenomenon). The graphs of versus T at average to high local cutaneous temperatures (T_w > 33°C) are remarkably similar, except for an upward shift at successively higher values of T_w. The slope (or vasomotor reactivity) is interpreted as being controlled by variations in T_a. The curves exhibited maximum values at T_a = 31°C. Their subsequent decrease could represent a thermoregulatory adaptation to environment-organism heat transfer, the relative vasoconstriction tending to reduce the transfer. Although the qualitative response was the same for both sexes, the absolute value of was generally greater in male than in female subjects.     

Polysaccharide and protein secretion by grass microhairs

Summary Secretory activities of bicellular microhairs from grasses belonging to the subfamilies Chloridoideae, Arundinoideae, Panicoideae, and Bambusoideae, and including the chloridoid, panicoid and Enneapogon microhair morphological types, have been investigated. Light microscopic histochemistry indicated that all microhairs studied secrete polysaccharide and protein (or glycoprotein), including those which also secrete salt. Localization of polysaccharide at ultrastructural level using periodic acid-thiocarbohydrazidesilver proteinate staining revealed that in panicoid type microhairs dictyosomes are involved in polysaccharide secretion, whereas in the chloridoid and Enneapogon types partitioning membranes seem to be involved instead.Abbreviations Ag silver precipitates representing localization of polysaccharide - BC basal cell - C cuticle - CC cap cell - CH cuticular chamber - CN system of membrane bound channels and vesicles - CP chloroplast - CW cell wall - D dictyosomes - M mitochondria - N nucleus - PTM partitioning membranes - RER rough endoplasmic reticulum - S secretory material - St starch grain - US unstained dictyosome cisternae - V vesicle     

Rotational diffusion tensor of nucleic acids from 13C NMR relaxation

Rotational diffusion properties have been derived for the DNA dodecamer d(CGCGAATTCGCG)₂ from ¹³C R₁ and R₁ measurements on the C₁, C₃, and C₄ carbons in samples uniformly enriched in ¹³C. The narrow range of C-H bond vector orientations relative to the DNA axis make the analysis particularly sensitive to small structural deviations. As a result, the R₁/R₁ ratios are found to fit poorly to the crystal structures of this dodecamer, but well to a recent solution NMR structure, determined in liquid crystalline media, even though globally the structures are quite similar. A fit of the R₁/R₁ ratios to the solution structure is optimal for an axially symmetric rotational diffusion model, with a diffusion anisotropy, D_||/D, of 2.1±0.4, and an overall rotational correlation time, (2D_||+4D)^–1, of 3.35 ns at 35 °C in D₂O, in excellent agreement with values obtained from hydrodynamic modeling.     

On a response characteristic of a mathematical neuron model

A mathematical neuron model in the form of a nonlinear difference equation is proposed and its response characteristic is investigated.If a sequence of pulses with a fixed frequency is applied to the neuron model as an input, and the amplitude of the input pulses is progressively decreased, the firing frequency of the neuron model, regarded as the output, also decreases. The relationship between them is quite complicated, but a mathematical investigation reveals that it takes the form of an extended Cantor's function. This result explains the unusual and unsuspected phenomenon which was found by L. D. Harmon in experimental studies with his transistor neuron models.Besides this, as an analogue of our mathematical neuron model, a very simple circuit composed of a delay line and a negative resistance element is presented and discussed.     

Mathematical properties of responses of a neuron model

Recently, Nagumo and Sato proposed a mathematical neuron model in the form of a nonlinear difference equation and investigated its response characteristic. The result showed that the input-output relationship of the neuron model is quite complicated and takes the form of an extended Cantor's function. It also explained the unusual and unsuspected phenomenon found by Harmon in experimental studies with his transistor neuron model. — In this paper, a fraction representation of a sequence of pulses is proposed. A mathematical treatment of the same neuron model based on the representation gives the same result as in the previous paper. Moreover, many mathematical properties, including the one where the ratio of the number of 1's contained in a cycle of a sequence to the length of the cycle gives any rational number between 0 and 1, were obtained by investigating sequences generated by the model.     

Studies of the M15 β-Galactosidase Complementation Process

M15 -Galactosidase was activated by heat-denatured wild-type -galactosidase, urea, and heat-denatured wild-type -galactosidase, a peptide made up of residues 6–44 of -galactosidase and CB2, the peptide that is normally used for complementation (residues 3–92 of -galactosidase). In each case roughly equal activation levels were attained. Heat-denatured wild-type -galactosidase was present as a finely divided visible white precipitate both before and after complementation. The heat-denatured protein by itself did not migrate on native PAGE and both the protein and the activity that occurred as a result of the complementation also remained at the point of application. The N-terminal ends of the heat-denatured wild-type -galactosidase must have been available for complementation and must have been mobile enough to allow tetramer to form despite being aggregated. -Galactosidase denatured by both urea and heat resulted in a streak of interacting protein on the native PAGE. Upon activation, a streak (indicating that interaction was still occurring) was still present, but it moves more slowly. Complementation using a peptide called XP (made up of residues 6–44 plus an additional nine C-terminal amino acids) resulted in three discrete forms of active enzyme at ratios of peptide to M15 -galactosidase monomer of less than 1:1. The fastest migrating of the three bands predominated at ratios near 1:1. A single active tetrameric form of M15 -galactosidase was formed with CB2. In both of these last two cases an active slow-moving diffuse band also formed (possibly a dimer of the tetramer). A quantitation of the amount of peptide bound to M15 -galactosidase by titration with XP and with CB2 and by using gel filtration after an excess of fluorescent-labeled XP was added showed that peptide bound in a 1:1 ratio (peptide/monomer) when full activity was achieved. These fluorescent studies also showed that peptide initially bound to dimer and that the tetramer was then formed.     

Initial Stages of Trisporic Acid Synthesis in Blakeslea trispora

Biotransformation of -carotene with enzyme preparations isolated from the mycelium of Blakeslea trispora resulted in the formation of its hydroxylated metabolite and apocarotenals, products of oxidative degradation of this compound. Based on its spectral, chromatographic, and chemical properties, the -carotene derivative was identified as 4-hydroxy--carotene (isocryptoxanthine). One of the products of oxidative degradation of -carotene, -apo-13-carotenone, was modified in the presence of enzyme preparations from Blakeslea trispora to form trisporic acid precursors. -Apo-13-carotenone transformation proceeded more rapidly than -carotene oxidation at the carbon atom at position 4. The data suggest that, under oxidative stress, oxidative degradation of -carotene into -apo-13-carotenone leads to the formation of considerable amounts of trisporic acids.     

Die Brutzeiten einiger Gänsearten und ihrer Bastarde in identischen Bedingungen

Zusammenfassung An Hand von 229 Brutbeginn-Daten von freilebenden Gänsen, die während der Jahre 1956–1966 in Seewiesen (Obb.) (48°N, 11°11E) brüteten, wurden die mittleren Brutbeginn-Daten von 5 Gänsearten und von Artbastarden bestimmt. Es zeigte sich, daß die untersuchten Arten unter diesen Bedingungen in derselben Reihenfolge brüteten, wie ihre Artgenossen in freier Wildbahn. Die mittleren Brutbeginn-Termine wurden allerdings um so mehr vorverlegt, je später die Art normalerweise brütet (Abb. 1). , die mit artfremden verpaart waren, brüteten zur selben Zeit wie ihre Artgenossen, die mit artgleichen verpaart waren (Abb. 1). GraugansxSchneegans-Bastard-, die mit Schneegantern verpaart waren, begannen meist nach den Graugänsen, aber stets vor den Schneegänsen zu brüten (Abb. 1, 2). Das intermediäre Brüten dieser wird als starkes Argument für die Richtigkeit der Hypothese gewertet, nach welcher die artspezifisch verschiedenen Brutzeiten wenigstens zum Teil genetisch bedingt sind. In der Diskussion wird die Frage kritisch erörtert, wie weit schon allein die Tatsache, daß die verschiedenen Arten über Generationen hinweg in derselben Reihenfolge wie ihre wildlebenden Artgenossen zu brüten beginnen, als Beweis für derartige genetische Unterschiede angesehen werden kann.

---

Summary In 229 cases onset of breeding was recorded from free-living geese of 5 species and of some hybrids of these species, kept in Seewiesen/Obb. (48° N, 11° 11E) from 1956 to 1966. It was found that the species under these conditions bred in the same seasonal sequence as did wild birds. The mean breeding times, however, were found to be advanced in relation to the onset of breeding in the wild (Fig. 1). This was especially evident in the case of late-breeding species. paired with of another species came into breeding condition at the same time as paired with of the same species (Fig. 1). GraylegxSnowgoose hybrid paired with Snowgoose in most cases started to breed later than Greyleg geese but always earlier than the mean breeding time for Snowgeese (Fig. 1, 2). This intermediate breeding time is taken as a strong argument for the hypothesis that the species specific differences in breeding times are, at least in part, genetic in origin. The question as to the extent to which the differences in breeding times alone, persisting for generations in the same sequence as those of wild birds, can be attributed to genetic differences between the species, is critically discussed.

     

The NGF complex from the African ratMastomys natalensis

The 7S NGF complex from the male mouse submaxillary gland consists of the , and subunits in the ratio ₂₂. The (NGF) subunit contains all the known biolocial activity of 7S NGF. The and subunits are both members of glandular kallikrein gene family, yet only subunit has protease activity. The subunit plays a role in the processing of preproNGF to its mature form, while the role of the subunit is not yet understood. Despite the fact that 7S NGF has been extensively characterized, no other NGF complex has been characterized, nor have the or subunits been observed in tissues which express NGF. We have therefore purified and characterized the NGF complex from the submaxillary glands of the ratMastomys natalensis in order to more fully understand the roles of the and subunits. The NGF complex from M. natalensis contains subunits similar to those found in mouse 7S NGF. Although similar, there are significant differences between mouse and M. natalensis NGF complexes, especially in the degree of post-translational modification of the and NGF subunits, the expression of esterase activity and the ease with which the complexes dissociate. Evidence is presented that suggests that the NGF complex from M. natalensis may consist of subunits in the ratio ₂. The amino acid sequence of the M. natalensis NGF suggests some, but not all, ways in which these differences arise.Special issue dedicated to Dr. Lawrence Austin     

Protein kinases expressed by interstitial cells of Cajal

Interstitial cells of Cajal (ICC) are involved in the generation of electrical rhythmicity of intestinal muscle and in the transduction of neural inputs in the gut. Although the expression of receptors for neurotransmitters and hormones and some second messengers have been investigated in ICC, the protein kinases present in these cells have not been well documented. This study has demonstrated the immunohistochemical localisation of PKA, PKC and PKC in ICC that were identified by the known ICC marker, c-Kit, in the guinea-pig gut. Other PKCs, PKC , , , , , and , and Ca²⁺-calmodulin-dependent protein kinase II were not localised in ICC. Double labelling studies were conducted on longitudinal muscle–myenteric plexus and external muscle–myenteric plexus preparations of the oesophagus, stomach (fundus, corpus and antrum), duodenum, distal ileum, caecum, proximal and distal colon, and rectum. The three protein kinases were detected in c-Kit-immunoreactive ICC at the level of the myenteric plexus (IC-MY), in the muscle (IC-IM) and at the level of the deep muscular plexus (IC-DMP) in the small intestine. PKA was found in over 90% of IC-IM in all regions examined, and in over 90% of IC-MY in the gastric body and antrum and throughout the small and large intestines. PKC was in the majority of ICC in the gastric body and antrum and in the small intestine, but was largely absent from ICC in the oesophagus, proximal stomach and large intestine. PKC occurred in the majority of ICC in all regions except the rectum. The intensity of staining was greatest for PKA, with PKC giving comparatively weak labelling of ICC. PKA was also detected in myenteric neurons, smooth muscle, macrophages and fibroblast-like cells. PKC labelling occurred in large, multipolar neurons throughout the small and large intestine, as well as in lymph vessels and in capillaries. It is concluded that PKA, PKC and PKC are all present in ICC, with the differences in their localisations suggesting specific roles for each in ICC function.