On the Relationships between Generative Encodings,Regularity, and Learning Abilities when Evolving Plastic Artificial Neural Networks

A major goal of bio-inspired artificial intelligence is to design artificial neural networks with abilities that resemble those of animal nervous systems. It is commonly believed that two keys for evolving nature-like artificial neural networks are (1) the developmental process that links genes to nervous systems, which enables the evolution of large, regular neural networks, and (2) synaptic plasticity, which allows neural networks to change during their lifetime. So far, these two topics have been mainly studied separately. The present paper shows that they are actually deeply connected. Using a simple operant conditioning task and a classic evolutionary algorithm, we compare three ways to encode plastic neural networks: a direct encoding, a developmental encoding inspired by computational neuroscience models, and a developmental encoding inspired by morphogen gradients (similar to HyperNEAT). Our results suggest that using a developmental encoding could improve the learning abilities of evolved, plastic neural networks. Complementary experiments reveal that this result is likely the consequence of the bias of developmental encodings towards regular structures: (1) in our experimental setup, encodings that tend to produce more regular networks yield networks with better general learning abilities; (2) whatever the encoding is, networks that are the more regular are statistically those that have the best learning abilities.     

Nitric oxide synthase in the nervous system and ink gland of the cuttlefish Sepia officinalis: molecular cloning and expression

Nitric oxide (NO) signaling is involved in numerous physiological processes in mollusks, e.g., learning and memory, feeding behavior, neural development, and defence response. We report the first molecular cloning of NOS mRNA from a cephalopod, the cuttlefish Sepia officinalis (SoNOS). SoNOS was cloned using a strategy that involves hybridization of degenerate PCR primers to highly conserved NOS regions, combined with RACE procedure. Two splicing variants of SoNOS, differing by 18 nucleotides, were found in the nervous system and the ink gland of Sepia. In situ hybridization shows that SoNOS is expressed in the immature and mature cells of the ink gland and in the regions of the nervous system that are related to the ink defence system.     

Ontogeny of rhythmic motor networks in the stomatogastric nervous system

We used the lobster Homarus gammarus to study the ontogeny of neural networks involved in rhythmic behaviours. Since in the adult the neural networks belonging to the stomatogastric nervous system and controlling the rhythmic movements of the foregut are well characterised, we have studied them during ontogeny. While this foregut develops slowly throughout embryonic and larval stages, the neuronal population of these motor networks is quantitatively established since the mid-embryonic period. Moreover, in the embryo, this neural population is organised into a single functional network that displays a unique motor output. By contrast, in the adult the same neuronal elements are organised into three neural networks that express independent motor programs. Our results indicate that the multiple adult networks are partitioned progressively from a single embryonic network during development. Accepted: 23 May 1999     

Making an escape: development and function of the Drosophila giant fibre system

Flies escape danger by jumping into the air and flying away. The giant fibre system (GFS) is the neural circuit that mediates this simple behavioural response to visual stimuli. The sensory signal is received by the giant fibre and relayed to the leg and wing muscle motorneurons. Many of the neurons in the Drosophila GFS are uniquely identifiable and amenable to cell biological, electrophysiological and genetic studies. Here we review the anatomy and development of this system and highlight its utility for studying many aspects of nervous system biology ranging from neural development and synaptic plasticity to the aetiology of neural disorder.     

Characterization of a novel Sepia officinalis neuropeptide using MALDI-TOF MS and post-source decay analysis

A novel neuropeptide acting as a myosuppressor on esophagus, funnel and mantle muscular fibers has been isolated from the stellar ganglia of the mollusk cephalopod Sepia officinalis by means of HPLC analysis. Fractions were monitored using a myotropic bioassay. After three separation steps, MALDI-TOF spectrum revealed one main peak at m/z 756.6. The partial N-terminal and C-terminal digestions by exopeptidases followed by MALDI-TOF analysis allowed the determination of the nature of the two C-terminal and N-terminal amino acids. Post Source Decay fragmentation of the molecular ion accurately determined the following primary sequence: Val-Tyr-Ser-Ala-Pro-Tyr-Gly-OH. The mapping of this heptapeptide performed in ESI-MS revealed that its distribution is restricted to the stellar ganglia, the giant fibers III, and the nervous bundle containing the giant fibers II and the palleal nerve. The neuropeptide was not detected in the hemolymph suggesting a release by nerve endings next to the targets.     

Coupled Genomic Evolutionary Histories as Signatures of Organismal Innovations in Cephalopods

How genomic innovation translates into organismal organization remains largely unanswered. Possessing the largest invertebrate nervous system, in conjunction with many species‐specific organs, coleoid cephalopods (octopuses, squids, cuttlefishes) provide exciting model systems to investigate how organismal novelties evolve. However, dissecting these processes requires novel approaches that enable deeper interrogation of genome evolution. Here, the existence of specific sets of genomic co‐evolutionary signatures between expanded gene families, genome reorganization, and novel genes is posited. It is reasoned that their co‐evolution has contributed to the complex organization of cephalopod nervous systems and the emergence of ecologically unique organs. In the course of reviewing this field, how the first cephalopod genomic studies have begun to shed light on the molecular underpinnings of morphological novelty is illustrated and their impact on directing future research is described. It is argued that the application and evolutionary profiling of evolutionary signatures from these studies will help identify and dissect the organismal principles of cephalopod innovations. By providing specific examples, the implications of this approach both within and beyond cephalopod biology are discussed.     

An approach to the study of neural activity during behaviour in insects

A method is described for easily recording neural activity from the central nervous system of a suspended cockroach while the animal can still walk, groom, and perform other acts. Recordings revealed that some giant fibre units are active during spontaneous walking. The preparation should be especially useful in further studies of the neural basis of behaviour.     

Rhythmic behaviour and pattern-generating circuits in the locust: Key concepts and recent updates

There is growing recognition that rhythmic activity patterns are widespread in our brain and play an important role in all aspects of the functioning of our nervous system, from sensory integration to central processing and motor control. The study of the unique properties that enable central circuits to generate their rhythmic output in the absence of any patterned, sensory or descending, inputs, has been very rewarding in the relatively simple invertebrate preparations. The locust, specifically, is a remarkable example of an organism in which central pattern generator (CPG) networks have been suggested and studied in practically all aspects of their behaviour. Here we present an updated overview of the various rhythmic behaviours in the locust and aspects of their neural control. We focus on the fundamental concepts of multifunctional neuronal circuits, neural centre interactions and neuromodulation of CPG networks. We are certain that the very broad and solid knowledge base of locust rhythmic behaviour and pattern-generating circuits will continue to expand and further contribute to our understanding of the principles behind the functioning of the nervous system and, indeed, the brain.     

A model biological neural network: the cephalopod vestibular system

Artificial neural networks (ANNs) have become increasingly sophisticated and are widely used for the extraction of patterns or meaning from complicated or imprecise datasets. At the same time, our knowledge of the biological systems that inspired these ANNs has also progressed and a range of model systems are emerging where there is detailed information not only on the architecture and components of the system but also on their ontogeny, plasticity and the adaptive characteristics of their interconnections. We describe here a biological neural network contained in the cephalopod statocysts; the statocysts are analogous to the vertebrae vestibular system and provide the animal with sensory information on its orientation and movements in space. The statocyst network comprises only a small number of cells, made up of just three classes of neurons but, in combination with the large efferent innervation from the brain, forms an 'active' sense organs that uses feedback and feed-forward mechanisms to alter and dynamically modulate the activity within cells and how the various components are interconnected. The neurons are fully accessible to physiological investigation and the system provides an excellent model for describing the mechanisms underlying the operation of a sophisticated neural network.     

SITE OF BIOSYNTHESIS OF BRAIN-SPECIFIC PROTEINS IN THE GIANT FIBRE SYSTEM OF THE SQUID

The synthesis of brain-specific proteins has been examined in perikaryal and axonal regions of the giant fibre system of the squid. After in vitro incubation of stellate ganglia, stellate nerves and isolated giant axons with radioactive amino acids, the labelled soluble proteins have been extracted from the giant fibre lobe, the axoplasm and the axonal sheath of the giant axon and have been separated by gel electrophoresis on a continuous system. In addition, they have been challenged with antisera prepared against the cephalopod brain-specific proteins L1 and L2 and the resulting precipitate has been resolved by sodium dodecyl sulphate-gel electrophoresis. Synthesis of these two proteins appears to be restricted to the giant fibre lobe, while an additional discrete protein band (L5) also becomes clearly labelled in the isolated giant axon. Radioactive components migrating in the region of the L1 and L2 proteins are synthesized in the isolated giant axon. They can be distinguished from tbese proteins on the basis of electrophoretic and immunochemical criteria.     

The brainstem reticular formation is a small-world, not scale-free, network

Recently, it has been demonstrated that several complex systems may have simple graph-theoretic characterizations as so-called 'small-world' and 'scale-free' networks. These networks have also been applied to the gross neural connectivity between primate cortical areas and the nervous system of Caenorhabditis elegans. Here, we extend this work to a specific neural circuit of the vertebrate brain--the medial reticular formation (RF) of the brainstem--and, in doing so, we have made three key contributions. First, this work constitutes the first model (and quantitative review) of this important brain structure for over three decades. Second, we have developed the first graph-theoretic analysis of vertebrate brain connectivity at the neural network level. Third, we propose simple metrics to quantitatively assess the extent to which the networks studied are small-world or scale-free. We conclude that the medial RF is configured to create small-world (implying coherent rapid-processing capabilities), but not scale-free, type networks under assumptions which are amenable to quantitative measurement.     

Increased axonal excitability during exposure to hyperbaric oxygen

The effect of high oxygen pressure on neural function was studied in the isolated nervous system of the cockroach. Intracellular and extracellular action potentials were recorded from single giant axons during exposure to 7 ATA (atmosphere absolute) (1 ATA = 0.1 MPa) of oxygen. Axonal excitability was measured as changes in stimulus strength-duration relationship. Initially, a transient increase in the rheobase current was observed followed by a significant decline to 75% of air control values. This decrease was accompanied by a parallel increase in the membrane time constant. The results demonstrate that hyperbaric oxygen increases axonal excitability. Such changes are consistent with the epileptogenic properties of high oxygen pressure.     

集成的专家系统和神经网络应用于大熊猫生境评价

充分了解大熊猫生境的时空格局及其变化,对有效保护大熊猫非常重要.绘制生境图既是野生动物生境评价和监测的一个有效方式,也是一个必要的步骤.新发展起来的人工智能方法(包括专家系统和神经网络方法),在模拟复杂系统过程中能够同时综合定性和定量信息,并可集成于GIS中,有助于大熊猫复杂生境的制图及评价.为了对大熊猫生境进行评价,本文建立了一个较全面的综合制图方法,将专家系统、神经网络和多类型数据全部集成在GIS环境下.结果表明,采用专家系统和神经网络集成方法绘制的大熊猫生境图的精度达到80％以上,高于单一的专家系统方法、神经网络方法和传统的最大似然法制图的精度.Z统计方法也证实了新建立的专家系统和神经网络集成方法要显著好于3种单一方法.     

Rapid neural circuit switching mediated by synaptic plasticity during neural morphallactic regeneration

The aquatic oligochaete, Lumbriculus variegatus (Lumbriculidae), undergoes a rapid regenerative transformation of its neural circuits following body fragmentation. This type of nervous system plasticity, called neural morphallaxis, involves the remodeling of the giant fiber pathways that mediate rapid head and tail withdrawal behaviors. Extra- and intracellular electrophysiological recordings demonstrated that changes in cellular properties and synaptic connections underlie neurobehavioral plasticity during morphallaxis. Sensory-to-giant interneuron connections, undetectable prior to body injury, emerged within hours of segment amputation. The appearance of functional synaptic transmission was followed by interneuron activation, coupling of giant fiber spiking to motor outputs and overt segmental shortening. The onset of morphallactic plasticity varied along the body axis and emerged more rapidly in segments closer to regions of sensory field overlap between the two giant fiber pathways. The medial and lateral giant fibers were simultaneously activated during a transient phase of network remodeling. Thus, synaptic plasticity at sensory-to-giant interneuron connections mediates escape circuit morphallaxis in this regenerating annelid worm.     

Cephalopods in neuroscience: regulations,research and the 3Rs

Cephalopods have been utilised in neuroscience research for more than 100 years particularly because of their phenotypic plasticity, complex and centralised nervous system, tractability for studies of learning and cellular mechanisms of memory (e.g. long-term potentiation) and anatomical features facilitating physiological studies (e.g. squid giant axon and synapse). On 1 January 2013, research using any of the about 700 extant species of “live cephalopods” became regulated within the European Union by Directive 2010/63/EU on the “Protection of Animals used for Scientific Purposes”, giving cephalopods the same EU legal protection as previously afforded only to vertebrates. The Directive has a number of implications, particularly for neuroscience research. These include: (1) projects will need justification, authorisation from local competent authorities, and be subject to review including a harm-benefit assessment and adherence to the 3Rs principles (Replacement, Refinement and Reduction). (2) To support project evaluation and compliance with the new EU law, guidelines specific to cephalopods will need to be developed, covering capture, transport, handling, housing, care, maintenance, health monitoring, humane anaesthesia, analgesia and euthanasia. (3) Objective criteria need to be developed to identify signs of pain, suffering, distress and lasting harm particularly in the context of their induction by an experimental procedure. Despite diversity of views existing on some of these topics, this paper reviews the above topics and describes the approaches being taken by the cephalopod research community (represented by the authorship) to produce “guidelines” and the potential contribution of neuroscience research to cephalopod welfare.     

Coarse-grained reverse engineering of genetic regulatory networks

We have modeled genetic regulatory networks in the framework of continuous-time recurrent neural networks. A method for determining the parameters of such networks, given expression level time series data, is introduced and evaluated using artificial data. The method is also applied to a set of actual expression data from the development of rat central nervous system.     

Filming the glial dreams: real-time imaging of cannabinoid receptor trafficking in astrocytes

How does the brain process incoming information and produce thoughts? These questions represent, to all likelihood, the most challenging matters ever faced by natural sciences, matters which may never be fully comprehended. The evolution of the nervous system that, in about billion of years, brought into existence the human brain progressed through an ever-increasing complexity of neural networks. This evolution began from the diffuse nervous system, in which primordial neurons were able to sense the environmental inputs and convey them to effector organs and to the neighbouring neurons. At the next evolutionary stage the conglomerates of neuronal cell bodies, the ganglia, appeared, thus forming the primitive centralized nervous system. The developments which ensued went through a continuous increase in complexity of neuronal conglomerates, which eventually formed the central nervous system, which attained maximal perfection in mammals. In this issue of ASN NEURO, Osborne et al. have described details of real-time imaging of cannabinoid receptor trafficking in astrocytes, a technique that will help to elucidate the role of these receptors in the ever-increasing complex neural networks.     

The structure and function of tenascins in the nervous system.

The tenascins are a family of large extracellular matrix glycoproteins that comprise five known members. Three of these, tenascin-C (TN-C) tenascin-R (TN-R) and tenascin-Y (TN-Y) are expressed in specific patterns during nervous system development and are down-regulated after maturation. The expression of TN-C, the best studied member of the family, persists in restricted areas of the nervous system that exhibit neuronal plasticity and is reexpressed after lesion. Numerous studies in vitro suggest specific roles for tenascins in the nervous system involving precursor cell migration, axon growth and guidance. TN-C has been shown to occur in a large number of isoform variants generated by combinatorial variation of alternatively spliced fibronectin type III (FNIII) repeats. This finding indicates that TN-C might specify neural microenvironments, a hypothesis supported by recent analysis of TN-C knockout animals, which has begun to reveal subtle nervous system dysfunctions.     

One messenger shared by two systems: How cytokines directly modulate neurons

Cytokines are small, secreted proteins that are known for their roles in the immune system. An accumulating body of evidence indicates that cytokines also work as neuromodulators in the central nervous system (CNS). Cytokines can access the CNS through multiple routes to directly impact neurons. The neuromodulatory effects of cytokines maintain the overall homeostasis of neural networks. In addition, cytokines regulate a diverse repertoire of behaviors both at a steady state and in inflammatory conditions by acting on discrete brain regions and neural networks. In this review, we discuss recent findings that provide insight into how combinatorial codes of cytokines might mediate neuro-immune communications to orchestrate functional responses of the brain to changes in immunological milieus.