Abstract Journal of Integrative Neuroscience Topical Review Models of Dendritic and Astrocy

Journal of Integrative Neuroscience
Topical Review
Models of Dendritic and Astrocytic Networks for fMRI
Roman R. Poznanski and Jorge J. Riera
CRIAMS, Claremont Graduate University, Claremont, CA 91711-3988 USA and NICHe, Tohoku University, Aobaku Sendai 980-8579 Japan Received: 19/12/2005
Abstract
In order to elucidate the relationships between hierarchical structures within the neocortical neuropil and the information carried by an ensemble of neurons encompassing a single voxel, it is essential to predict through volume conductor modeling LFPs representing average extracellular potentials, which are expressed in terms of interstitial potentials of individual cells in networks of gap-junctionally connected astrocytes and synaptically connected neurons. These relationships have been provided and can be then used to investigate how the underlying neuronal population activity can be inferred from the measurement of the BOLD signal through electrovascular coupling mechanisms across the blood-brain barrier. The importance of both synaptic and extrasynaptic transmission as the basis of electrophysiological indices triggering vascular responses between dendritic and astrocytic networks, and sequential configurations of firing patterns in composite neural networks is emphasized. The purpose of this review is to show how fMRI data may be used to draw conclusions about the information transmitted by individual neurons in populations generating the BOLD signal. Keywords: sequential configurations; astrocytes; gap-junctions; sub-assemblies; dynamic connectivity; multi-hierarchy; neocortex; functional neuroimaging; nested networks; volume conductor modeling.
1. Introduction The largest subdivision of the cerebral cortex is the neocortex, also called the isocortex. It operates by activating distinct cortical modules or columns (Mountcastle, 1957). These modules are local neural circuits of approximately 104 neurons linked together by a complex intramodular circuitry (Abeles, 1982, 1991). The modular view of neocortical organization has been explored most clearly in the posterior regions, including visual, somatosensory and auditory cortices. However, its generality to both the posterior and anterior regions is implied (Eccles, 1981; Szentagothai, 1975, 1978). That each cortical module has reciprocal connections (e.g., associative, commissural, and projection fibers) between 10 to 30 other modules (see Mountcastle, 1979) precludes an experimental evaluation of extrinsic connectivity of the neocortex with electrophysiological techniques, such as with multi-unit microelectrodes (e.g., Buzsaki, 2004). Functional neuroimaging devices, on the other hand, have a spatial resolution that is large compared with the size of neurons. Such a spatial resolution involves a high degree of interconnectedness at the systems level (Felleman and van Essen, 1991), requiring large-scale theories of extrinsic connectivity of the neocortex which have yet to be developed (see e.g., Friston et al., 1995; Stevens, 1994). An understanding of the functional organization of the neocortical module as a precisely connected but nested distributed system requires that the intrinsic organization of the neocortex be functionally sub-divided into brain regions composed of large numbers of modular elements linked together in echeloned, parallel and serial arrangements that may be compared with the idealized notion of “brain cell assemblies” first proposed by Hebb (1949). In view of the nested columnar organization of the neocortex, such neocortical assemblies (macro-scale) should be comprised of “sub-assemblies” as has been recently observed within functional columns in the rat visual cortex (Yoshimura et al., 2005). These neocortical sub-assemblies (meso-scale) can than be further sub-divided into modules or local neural circuits consisting of approximately 104 neurons. Although a large number of theoretical studies have hypothesized how cognitive events emerge through the modeling of neocortical assemblies as neuronal masses (Braitenberg, 1978; Palm, 1982; Wickelgren, 1992; Eichenbaum, 1993; Miller, 1996; Eichenbaum and Davis, 1998) little effort has been given to theoretical elucidation of the functionality of neocortical sub-assemblies in the context of an integrative theory of cognition (Poznanski, 2002b). In particular, there are no theoretical studies showing directly how one hierarchical level of neuronal organization affects another level, with each lower level manifesting an emergent behavior at the higher level.
1

Journal of Integrative Neuroscience
Abbreviations: LFPs-local field potentials; (f)MRI-(functional) magnetic resonance imaging; BOLD-blood oxygen level dependence; CBF-cerebral blood flow; PCD-primary current density; CBV-cerebral blood volume, EET- epoxyeicosatrienoic acid, PET-positron emission tomography, TCA- aerobic tricarboxylic acid, HRF-hemodynamic response function, NO- nitric oxide, BOLD-blood-oxygen level-dependent.
Brain regions within the neocortex are connected with each other at the macro-scale to form smaller sub-networks or neocortical sub-assemblies at the meso-scale, and so forth. In order to derive and implement models of neocortical sub-assemblies as nested neural networks it is necessary to replicate spatiotemporal dynamics in a continuous manner (Freeman, 2002a). This requires new methodology since the instabilities in the dynamics inherent from discretization of neurons resulting from implementation on a digital computer as used in popular software packages, such as NEURON or GENESIS precludes such an analysis. Other software packages, such as NEUROSY, exclude the biophysical properties of neurons as they consider a probabilistic representation of populations of neurons in terms of probability density clouds to predict spatiotemporal activity patterns within neocortical assemblies. Nested neural networks are attributable to Mountcastle (1978, 1979) as a fundamental design principle of the neocortex. The basic architecture of the model is illustrated in Figure l. There are two ways of conceptualizing the structure, and they are equivalent. In the bottom-up view, neurons are linked together to form functional networks, such as neocortical columns. Populations of networks join together to form larger networks, which are nested together within still larger networks, and so on. In the top-down view, the neocortex is subdivided into large overlapping networks based on anatomical and functional criteria. Smaller networks are nested within one or more of these large networks, and there is progressive nesting down to the level of individual cells. The scaling between levels of nesting is, in principle, continuous, although discrete levels do exist between neocortical columns (Sutton, 1996). The approach for evaluating nested neural networks, known as multihierarchical or multi-scale brain modeling (Chauvet, 1996; Robinson et al., 2005), provides a portal for discovering how physico-chemical mechanisms bring about neocortical integration by linking top-down and bottom-up views at the meso-scale. Multi-scale brain modeling can help to predict how the release of various neurotransmitters and the sequence and distribution of various proteins (such as ionic channels, gap-junctions, and G-proteins) affect the collective behavior of large masses of neurons. Thus, multi-scale brain modeling can provide a theoretical framework for investigating how neuronal activity in the neocortical neuropil, which is the electrophysiological trigger of vascular-metabolite demand, will be reflected in neuroimaging data-especially functional magnetic resonance imaging (fMRI) data. In order to clarify fMRI data we need macroscopic models integrated with neuronal activity associated with molecular mechanisms, and a suitable way to achieve such symbiosis is through so-called nested neural networks. The relation of neuronal activity to the fMRI signals has been previously outlined by several authors (Raichle 2001; Shulman et al., 2001; Attwell and Iadecola 2002; Iadecola 2002; Heeger and Ress, 2002; Smith et al., 2002; Hyder et al., 2002; Kim 2003; Logothetis 2003). Our review, aims to reach beyond the vasculature and metabolism associated with interpretation of fMRI data, by developing appropriate structural hierarchies for simulating fMRI signals based on microelectrode studies. Although fMRI can provide information about the functional organization of the brain by mapping brain activation associated with a variety of cognitive processes (e.g., Friston, 1997; Posner et al., 1988), it is incapable of elucidating the changes in the spatiotemporal patterns of activity across different scales. For this reason, a theoretical understanding through neuronal and brain modeling (see MacGregor, 1987) is essential to link fMRI data and neuronal/glial networks in specific regions of the brain, to reveal the cellular organization of networks involved in brain activity. The so-called “structural equation modeling” (McIntosh and Gonzalez-Lima, 1994; Taylor et al., 2000) represents one of the attempt to create such a link, which allow us to estimate from neuroimaging data the functional strengths of the connections between various brain regions during specific cognitive tasks. However, such type of models is not useful in bridging the gap between cellular organization and fMRI data because it does not consider, the physiology and anatomy of real neural and glial networks. Such an analysis is vital for interpretation of fMRI signals because imaging methods have a finite spatial and temporal resolution, and therefore the resulting signals reflect the integrated activity of hundreds of thousands of individual neurons. Characterizing neuronal activity at such a macro-scale and from a spatially and temporally averaged signal may lead to distortions if the region of the neocortex over which the average is performed contains neurons with dissimilar biophysical characteristics. However, it is possible to measure with neuroimaging methods some indexes of the neuronal activity of a piece of neocortex the size of a single voxel (see e.g., Kim et al., 2000; Esser et al, 2005), with a single voxel reflecting 30,000 or more neurons activated simultaneously (Horwitz et al., 2000). Therefore, multiple neocortical sub-assemblies are easily investigated in any resolvable region of interest (typically 1mm x 1mm x 3mm) and such meso-scale activity is more manageable to interpret in terms of micro-electrode data. In essence, the spatiotemporal dynamics of neocortical sub-assemblies, which include the modulatory envir0nment, are key in interpreting fMRI signals (over
2

Journal of Integrative Neuroscience approximately 1mm2 of neocortex), but current theoretical models fail to incorporate silent-signals not obviously visible with imaging techniques, such as field effects that can modulate the excitability of neocortical subassemblies (i.e., nonlocal interneuronal interactions at the meso-scale level of neuronal organization), and other nonclassical forms of signaling (Bullock, 1997). We discuss such silent signaling in models of neocortical subassemblies to reveal spatiotemporal patterns that may serve as a dynamic link between neuronal activity and fMRI. In essence, this will require the incorporation of a network of astrocytes, gap-junctionally and extrasynptically connected to the neural networks, yet coupled to the vascular system. What is than required is a symbiosis of micro-scale neural modeling (e.g., when simulating “sequential configurations” in neocortical subassemblies) and its interactions with astrocytic networks leading to metabolite-vascular cross-talk observed at the meso-scale as fMRI BOLD signals. This review will also draw upon dynamics in neocortical sub-assemblies, as representative of information locally carried by both nonsynaptic diffusion neurotransmission (Bach-y-Rita, 1995) and sequential configurations (MacGregor, 1991; 1993). This review presents roadmap towards modeling dendritic and astrocytic networks across different scale of structural organization-from the cellular to the systems level. It represents an attempt at explaining the functional uniqueness of brain, which is morphological rather than biochemical, and in-so doing we have avoided detailed biochemical reactions associated with metabolism because enzyme activities are not fundamentally different in brain than in other tissues and does nothing to illuminate the brain as a specialized organ. There exists recent models on brain metabolism (Aubert et al., 2001; Aubert and Costalat, 2002, 2005), and so they will not be discussed in this review. The major biochemical contribution is the synthesis of specific macromolecules (or proteins), and such morphological elements have been emphasized in the modeling of neural (and glial) networks. A shortcoming of this review is that no molecular events have been attributed in the discussion of the bloodbarrier or the biochemistry of morphogenesis. A recent contribution by Dermietzel et al. (2005) is an excellent source in this research direction. Nevertheless, we believe that by adequately covering three hierarchical levels of structural organization starting at the cellular level, we have set-up a discourse for further exploration at the molecular level in future studies.
a
c
b
Figure 1 (a) A schematic lateral view of the human brain depicting inter-connected cell assemblies traversing four brain regions
of about 1cm2 of neocortex. The linkages represent fiber pathways. (b) Within a region [e.g., the black region in (a)], there are nested sub-regions of neocortical sub-assemblies of approximately 1mm2 of neocortex. Three sub-assemblies are represented
3

Journal of Integrative Neuroscience
forming a composite neural network with connections among neurons in local neural networks that are both intra-columnar and extra-columnar. [Reproduced from modeling cortical disorders using nested networks, Sutton, J.P. In, Neural Modeling of Brain and Cognitive Disorders, 1996, J.A.Reggia, E. Ruppin, and R.S. Berndt (eds.), World Scientific Publishers, Singapore]. (c) Intra-columnar interconnections between astrocytes and neurons in the neocortex form the basis for functional modules of approximately 1mm x 1mm x 3mm of neocortex. [Reproduced from Functional Neuroscience, 2000, O.Steward, Springer, New York].
2. Inadequacies of compartment models for representing spatiotemporal dynamics One of the classical studies with which large-scale neural-modeling efforts continue to resonate is the Wilson and Cowan (1972) model. The basic concept behind the model is that neural populations are idealized as “nodes” connected to other nodes with each different node reflecting a specific grouping of either excitatory or inhibitory populations of neurons, interconnected through “synaptic weights.” The model is also known as the “firing-rate” model in which each node corresponds to a sigmoidal function representing the electrical activity (or rate of firing) of population of neurons. More recent advances consider each node to be an “integrate-and-fire” or conductance-based compartmental implementation of a single neuron rather than a whole population of neurons. Figure 2 shows a schematic illustration of a single compartment and a “point-like” neuron model connected by synaptic weights referred to as “spiking networks” in the computational neuroscience literature. Spiking networks of conductance-based or integrate-and-fire point-like neuron models to simulate large-scale neural networks that discretize space with a single compartment (i.e., node) for each neuron in the network (see e.g., Lansner and Fransen, 1995; Maass, 1997; Vollmer and Sommer, 2001; Gerstnser and Kistler, 2002; Toth and Crunelli, 2002; Alvarez and Vibert, 2002; Izhikevich et al., 2004; Sommer and Wennekers, 2005) are not nested nor distributed in-design, leaving the modeling effort non-integrative.
(a)
(b)
red= excitatory blue= inhibitory
Wilson-Cowan network comprising an excitatory spiking neuron or population of neurons (at the meso-scale) (E) and an inhibitory spiking neuron or population of neurons (at the meso-scale) (I) both coupled through synaptic weights. The parameters h, r, f, q, represent the strength of interactions between the units and P, U represent excitatory and inhibitory input respectively. [Kindly provided by Dr. Andrew Gillies, Edinburgh University]
The firing-rate model, the integrate-and-fire model, and the single compartmental model are considered to be representative of the same class of point-like neuron models. Insofar as these units have no spatial structure (i.e., they are morpholess), they are incapable of spatial differentiation, and lump all incoming synapses onto a single
Figure 2 (a) A circuit representation of the conductance-based point neuron is shown. (b) Schematic illustration of a
4

Journal of Integrative Neuroscience synaptic location, representing the strength of synapses from a particular neuron, referred to as “synaptic strengths”(see e.g., Senn, 2002) or “synaptic footprints’”(Rinzel et al., 1998). These older network models further assume that the post-synaptic current / conductance-change or activation can be represented directly from the pre-synaptic membrane potential via a sigmoid-type synaptic coupling (Destexhe, 1994a,b; Matsugu and Yuille, 1994) or even approximately from the pre-synaptic firing rate expressed as a sigmoidal function (Buhmann and Schulten, 1986; Gerstner et al., 1993a; Contreras-Vidal and Stelmach, 1995) or a series of delta-functions (Amit and Tsodyks, 1992; Tsodyks and Sejnowski, 1995; Hopfield and Herz, 1995; Gerstner et al., 1993b; Kalitzin et al., 1997) or a sum of double exponentials (Kudela et al., 1997; Tateno et al., 1998). Specifically, these models fail to consider the biophysical properties of synapses that are altered as a result of correlated timing of activity in preand-postsynaptic neurons, which are dependent on their spatial distribution. Additional distributed neuron models consist of a large number of compartments per neuron with some totalling thousands of compartments (e.g., Protopapas et al., 1998; Rinzel et al., 1998; DeSchutter et al., 2005; Traub et al., 2005). Such multi-compartmental models are inadequate for representing dynamic connectivity in neocortical assemblies for two major reasons. First, they are impractical for realistic simulations of large-scale neocortical activity. As expressed in the words of Traub and Jefferys (1994, p.116): “...others have modeled individual neurons with many thousands of compartments, but such models do not lend themselves to network simulations.” For instance, modeling of large-scale neural networks with 60,900 single compartment neurons interconnected with a detailed multi-compartmental model of a single neuron consisting of 4,500 compartments takes 44 hours to simulate 2 seconds real time on a workstation with 1GB of memory (Howell et al., 2000). Second, and more importantly, because all multi-compartmental models ignore or slice the continuous nature of real neurons by discretizing space a priori, they delude the dynamical state of the system (Freeman, 2000b), and therefore are incapable of capturing the essence of such infinite dynamical systems believed to play a central role in neocortical integration (Freeman, 2000b; Korn and Faure, 2003). Indeed, studies have unequivocally proven that spatially-continuous nonlinear dynamical systems (i.e., chaotic systems), when duplicated on a digital computer will fail to reproduce the dynamics of the real system (see e.g., Hayes, 2003; Young et al., 2006). Thus, finite precision of computation can significantly distort the real chaotic dynamics known to be reflective of real neocortical systems (Korn and Faure, 2003). This should be a cautionary warning to computational modelers who rely on digital computers in building large-scale neural models of neocortical columns based on compartmental models or other types of neuronal models implemented computationally a priori. A pioneering advance on such caricature models was initiated by Horwitz and colleagues, who used large-scale point-like neuron models of networks with synaptic weights to connect neural activity with fMRI (see e.g., Tagamets and Horwitz, 1997; 1998; Husain et al., 2004). Subsequently conductance-based point-like neuron models in networks became popular, employing mean local firing rates (see e.g., Deco et al., 2004). Based on similar ideas, models that are more complete have been developed in later studies (Almeida and Stetter 2002). Figure 3 illustrates the approach taken by Horwitz and colleagues with the use of large-scale modeling for interpreting human brain imaging experiments. For more details on the derivation of the connections and parameters, see Tagamets and Horwitz (1997; 1998).
Fig. 3 (a) A nested neural network model composed of connected patches of Wilson-Cowan type populations; (b) a patch of
cortex is modeled by a 9x9 group of local basic circuits, as shown in (b), and represents about 1cm2 of cortex; (c) a large-scale
5

Journal of Integrative Neuroscience
model composed of connected patches from (b) [Reproduced from The use of large-scale modeling for interpreting human brain imaging experiments, Tagamets, M.-A. and Horwitz, B., In, Exploratory Analysis and Data Modeling in Functional Neuroimaging, 2003, F.T. Sommer and A. Wichert ( eds.) MIT Press, Cambridge, MA]
3. Inadequacies of mean-field theories for representing dynamic connectivity The use of spatially coarse-grained modeling approaches in which the dynamics of local populations of neurons consist of separate populations of excitatory and inhibitory units that establish synaptic connections, and are treated as ensembles or neural masses in a mean-field formalism of nervous tissue are still used to study the dynamics of large-scale cerebral cortical networks, including piriform cortex, hippocampus and somatosensory cortex (for a review see Hasselmo and Kapur, 2001). Modeling neuronal responses from a spatially averaged signal will yield inaccuracies if the region of the neocortex over which the averaged is performed contains neurons with biophysically distinct properties. Although the macro-scale field theories of the brain are derivable from the meso-scale neural dynamics exhibited by the standard Wilson-Cowan equations (see e.g., Jirsa and Haken, 1997), it is still an open research problem how to relate such theories to real brain networks that are not classified in terms of excitatory or inhibitory synaptic connectivity patterns (Marino et al., 2005) but rather on dynamic connectivity patterns (c.f., Young et al., 1994) is still an open research problem. Neocortical sub-assemblies represent a phenomenological description of meso-scopic brain activity of thousands of neurons integrated through dynamic connectivity (see glossary for definition at different physical scales). To examine the dynamical connectivity within neocortical sub-assemblies, one turns to earlier work, on compound field potentials (i.e., integrated LFPs), which were simulated using continuum field models (Beurle, 1956; Griffith, 1963, 1965; Fisher, 1973; Wilson and Cowan, 1973). The approach entailed modeling population of neurons as lumped nodes in a one-dimensional neural sheet or layer (Wilson and Cowan, 1972; Wilson, 1999; Omurtag et al., 2000; Nykamp and Tranchina, 2000). A field potential was determined for the spread of firing activity between population of neurons as lumped models in a one-dimensional neural sheet or layer via pseudo connections and delays without taking into consideration the volume distribution of neural masses (Ventriglia, 1974; Ermentrout and Cowan, 1980; Ingber, 1982; Amari, 1983; Peretto, 1984; Mallot and Glannakopoulus, 1996; Jirsa and Haken, 1997; Tuckwell, 1998; Liley et al., 1999, 2002; Ermentrout and Kleinfeld, 2001; Renart et al., 2004). These studies of mean-field theories of spatiotemporal dynamics rely on too simplified assumptions [see Ermentrout (1998), Jirsa (2004), and Coombes (2005) for reviews]. An alternative approach to elucidating the dynamic connectivity patterns within neocortical sub-assemblies is to use principles and methods from multi-scale brain modeling (Robinson et al., 2005). Indeed, there have been some attempts at multi-scale simulation of brain networks addressing data at multiple levels (e.g., Aersten et al., 1989; Johnsen et al., 2002), the most powerful being the utilization of n-field theory to integrate contributing mechanisms across multi-scales of neural organization (Chauvet, 1993, 2002), but progress has been slow. An important aspect of neuronal geometry is its embedding in 3D space, therefore continuum models of undistributed neural aggregates of neocortical activity localized to the gray matter and oriented perpendicularly to the neocortical sheet have been developed for a volume element of neural tissue (Nunez, 1974, 1981, 1995; van Rotterdam et al., 1982; Tuckwell, 2000). The extension of the continuum models to spatially distributed neural masses began with the current source density analysis (Nicholson and Llinas, 1971; Nicholson, 1973; Nicholson and Freeman, 1975; Mitzdorf and Singer, 1977; Mitzdorf, 1985). The key assumption of this approach is that a microscopically inhomogeneous neural tissue is replaced by a macroscopically homogeneous medium with electrical properties represented as averaged (or mean-field) quantities in a volume of tissue representing the dendritic activity of a group of neurons. These simplifications result in a simple relation between the field potential and the current source density of an averaged neural mass of neurons or neuronal ensemble, but do not take into consideration the individuality of each neuron in the neuropil nor its connectivity patterns. The existence of more integrative approaches which do not assume a macroscopically homogeneous neural tissue (neurons are chosen as point source fields) are based on current flow density analysis in a volume conductor (see, e.g., Kwan and Murphy, 1974; Klee and Rall, 1977; van Rotterdam, 1980; Holsheimer et al., 1982; Feenstra et al., 1984). The consideration of trans-membrane current density of an individual neuron in a volume conductor (van Rotterdam, 1987) and its relationship with interstitial potentials requires each neuron to be represented as a “core-conductor” (i.e., a cable with extracellular sheath surrounding the core) (cf., Clark and Plonsey, 1968; Bennett et al., 1999) at the micro-scale, while at the meso-scale, as nodes, in a three-dimensional realistic, inhomogeneous volume conductor (Bennett et al., 2001). Riera et al. (2006a) proposed a canonical mass model of a cortical unit containing a layer V pyramidal call and two GABARergic interneurons. Single compartments were used to model the GABAergic interneurons. However, the layer V pyramidal cell was modeled by three different electrotonically
6

Journal of Integrative Neuroscience connected compartments. This canonical mass model allowed them not only to account for the spatial integration of basal, apical and somatic synaptic potentials but also to determine the effect these potentials produce in the neuropil. The inappropriateness of mean-field theories to capture the non-discrete, fluid-like extracellular matrix of the neuropil requires the formulation of new models of continuous neurons imbedded in the interstitial domain. The “fluidity” could be mimicked by introducing a spatial dimension, and the plasticity of synaptic connectivity along the neuronal structure could be incorporated as a template of dynamic connectivity between neurons in a micro-network at the micro-scale. However, at the meso-scale, dynamic connectivity would be determined between populations of neurons within a neocortical sub-assembly, as well as between distinct neocortical sub-assemblies; and dynamic connectivity at the macro-scale would be determined from motifs in brain networks (Sporns and Kotter, 2004). In the following sections we discuss briefly micro-scale and meso-scale modeling, while macro-scale modeling of brain networks is beyond the scope of this review and interested reader may wish to consult Sporns (2005). It is also important to note that mean-field theories for determining dynamic connectivity within neocortical subassemblies through a probabilistic representation of the 3D spatial location and distribution of synapses within the neuropil (Rao et al., 2002; Jacobs and Pittendrigh, 2002) are beyond the scope of this review. 4. Predicting spatiotemporal patterns of neural activity within the neocortical neuropil It is now well established that nonsynaptic release of neuromodulators can effect changes to neural activity (Freeman, 2005). But extrasynaptic signals between neurons due to synaptic activity in the dendritic neuropil have received little theoretical attention, because earlier models assumed the extracellular space to be isopotential. It is therefore important for spatially distributed models of single neurons to be included so that neuromodulatory mechanisms, which alter brain connectivity (Doya, 2002) through a process known as “nonsynaptic diffusion neurotransmission” (Bach-y-Rita, 1995) or “volume transmission” (Fuxe and Agnati, 1991; Zoli et al., 1998) can be conceptualized in prototype models through imbedding of the neural networks in a syncytium that interacts with the external microenvironment, namely networks of astrocytes. This can be achieved through volume conductor modeling, where both the interstitial and intracellular domains are interwoven in a unified manner, and where the signal diffuses into a local volume or neuropil (Nicholson, 1973; Nicholson and Sykova, 1998). In particular, the integration of dendritic potentials in a volume conductor of synaptic currents in local areas of the neocortex can more accurately represent the mechanisms of neural pattern formation, including how the diffusive and/or dispersive effects of dendritic trees translate into meso-scale neuronal activity. Such an approach also incorporates common synaptic plasticity effects and the nonsynaptic release of neuromodulators by nonsynaptic diffusion transmission (or volume transmission), both of which alter dynamic connectivity. This approach differs from a sequential configuration in the sense that the activity lasts for seconds, and requires both synaptic and extrasynaptic connectivity. Recent works suggest that neural activity in a “synfire chain” (i.e., a synchronized sequential configuration) is best represented by a wave (Abeles et al., 2004; Hayon et al., 2005) and/or non-wave-like propagation (Beggs and Plenz, 2004). To simulate such phenomena at the meso-scale requires new types of neural networks capable of reproducing neural pattern formation between spatially distinct sequential configurations in different neocortical sub-assemblies and the integration of dendritic potentials could facilitate this through volume transmission. Junctional synaptic matrices that could carry out interactions among sequential configurations have yet to be developed (see MacGregor, 1993), although Aiello (2002) has presented non-synaptically connected neural network that may point the way forward in this respect. Sequential configurations are coordinated firing patterns of temporal activity learned by storing associations between subsequent sequences in Hebbian synapses (MacGregor, 1991). Izhikevich (2006) recently developed spiking networks with axonal delays to investigate sequential configurations in groups of networks that he defined as “polychronization.” However, the modeling unfaithfully produces the firing patterns of neocortical neurons because it ignores the cable properties of dendrites, which in turn will determine whether a particular neuron or groups of neurons will fire spikes. Indeed, spiking networks with axonal delays superficially assume that all neurons in such networks fire spikes independently of the spatial distribution of ionic channels on dendrites, more or less in the same manner as do integrate-fire models. This is clearly not the case, as pointed out by Abeles (1994, p.46): “The classical view of the mammalian neuron has been that of an integrate and fire neuron, which resets its membrane potential after each spike, to some value well below threshold. This view is based on the pioneering studies of Sir J. Eccles on the motor neuron of the spinal cord. Such neurons show prolonged depressions in their renewal density functions, and tend to fire periodically when driven to fire at elevated rates. Most cortical neurons do not have this property!” Yet almost all earlier synfire chain models use spiking networks with either integrate-and-fire or conductance-based point neurons (see Wennekers, 2000; Levy et al., 2001; Vogels and Abbott, 2005; Sommer and Wennekers, 2005). Therefore, significant progress can be made by
7

Journal of Integrative Neuroscience identifying and incorporating the cellular structure of each neuron of the network in an accurate (non-discretized) manner in order to capture a true representation of the coordinated firing patterns exhibited by spatiotemporal activity in populations of neurons. The modeling of spatiotemporal patterns of neural activity through the functional integration of nested neural networks will help us to elucidate how functions are carried out by networks of interacting groups of such neocortical sub-assemblies, and how different functions correspond to different groups (Gerstein et al., 1989). The most significant manifestation of such realistic neural networks is that dynamic connectivity is malleable through both synaptic and extrasynaptic activity and through interaction with sequential configurations in neocortical subassemblies. In order to investigate dynamic connectivity, a mathematically tractable and biologically plausible model of information processing in the neuropil needs to be developed. This requires a closely packed matrix of dendrites, axons, somas and glia as a precursor condition for the occurrence of complex interactions between ion fluxes, extrasynaptic potentials and neuromodulators (see MacLennan, 1993; Chen and Nicholson, 2000). 4.1 Unifying biophysical mechanisms with cellular activity through ionic cable theory If dendritic structure is taken into account through the auspices of representing each neuron in terms of a distributed structure, instead of a point node, then axo-dendritic synaptic connections may or may not evoke an impulse (action potential) depending on the cable properties of dendrites and the sparse distribution of voltagedependent ionic channels. By specifying a unique sequential configuration for each biophysically distinct neuron in the population, ionic cable theory (Poznanski and Bell, 2000a,b) is able to incorporate such cell-level components, including the rich repertoire of voltage-dependent ionic channels present in the dendrites of neurons as depicted in Fig. 4.
Hot-spot (N)
d
Ionic channels
Iion (N*/πd)
L
V2 V1
L2
Y V3
0
X
L1
Z
L3
Figure 8.
A sc he m a tic illustra tion of a de ndritic tre e with a single bra nc h point
studded with clusters or hot-spots of ionic channels. The arrow above the hot-spot, as depicted in (a) reflects the notion of Iion representing a point source of current applied to an infinitesimal area on the cable. The symbol N denotes the number
8
Fig. 4 A schematic illustration of a dendrite of diameter d (cm) (top view) and a single branch point (bottom view)

Journal of Integrative Neuroscience
The rationale is that a continuous distribution of voltage-dependent ion-channels as expressed in cable theory is not a good approximation for dendrites with channels occurring in low densities. Furthermore, the theory gives reasonable amplifications of the synaptic potentials in neocortical neurons (Stuart and Sakmann, 1995). The application of ionic cable theory to small-scale neural networks (see Poznanski, 2001; 2002a; 2005) has not yet been extended to large-scale neural networks. A large-scale application will require the extensive use of matrix algebra and Green’s function matrices that are commonly employed in theoretical physics. This approach has three major advantages: (1) it avoids entirely the mathematical errors and uncertainties inevitable in iterative computational models which necessarily discretize time and space; (2) it provides an analytical framework for generating complete and exact solutions for network output; and (3) it allows for the explicit inclusion of cell-level biophysical properties (e.g., channel; neuronal geometries) in an analytical formalism that predicts network-level functionality. Analytical modeling, in contrast to computational modeling, gives the provision of a continuous description of the neuronal membrane as a function of position, allowing for greater feasibility in developing large-scale biophysical neural networks, without the need for ad hoc computations. The emphasis on analytical solutions rather than computational estimation is paramount because the latter methodology is unsuitable for building hierarchically plausible bridges between different levels of neural organization as discussed in the introductory chapter of the monograph by Poznanski (2001). It invariably discretizes time and/or space through iterative estimation, whole errors can propagate and bias the predications. Integration of multiple levels for understanding neural and extraneural systems is and will be based on analytical modeling. 4.2 Mathematical models of local neural networks with synaptic structure A comprehensive treatment of dendritic cables including the distribution of ionic channels in neural networks of biologically constrained sequential configuration models was undertaken in 2001 (Poznanski, 2001). A mathematical formalism of the feed-forward network of N synaptically coupled neurons (as depicted in Fig. 5a), where the presynaptic neuron whose membrane potential is Vo(x,t) provides M synaptic inputs to the postsynaptic neuron whose membrane potential is Vi(x,t), can both be represented by ionic cable equations of the form (Poznanski 2001):
of hot-spots, and N* denotes the number of ionic channels in each hot-spot per unit membrane surface of cable, represented schematically as dark spots. [Reproduced from Modeling in the Neurosciences, Second edition, Poznanski, RR, Analytical Solutions of the Frankenhaeuser-Huxley Equations Modified for Dendritic Backpropagation of a Single Sodium Spike, pp.201-226 (2005), Taylor and Francis, London].
cm Vot = (1/ri) Voxx - Vo/rm + ∑ Ioion(x,t;Vo) δ(x-xj) + Io(x,t)δ (x-xo), j=1 cm V t = (1/ri) V xx - V /rm + ∑ I iion(x,t;Vi) δ(x-xj) + Iisyn(x,t;Vi), i=1,2,...N, t > 0 j=1
i i i
Ψ
Ψ
where Iisyn is the synaptic current of the ith neuron at location x=xi, Iiion is the current density of various ionic channels imbedded in the membrane of a dendritic equivalent cable of the ith neuron at the location x = xj , I1o is the input current on the 1st neuron (only) at location x = xo, with δ(.) representing the Dirac-delta function. Poznanski (2002), as depicted in Fig. 5b, undertook a mathematical formalism of the recurrent network of two synaptically coupled neurons. The first neuron whose membrane potential is V0(x,t), provides M synaptic inputs to the post-synaptic neuron whose membrane potential is V1 (x,t). Note that Uo (x,t) reflects a quiescent neuron in the absence of synaptic activity, and V0 (x,t) reflects on the feedback from the synaptically coupled first neuron. The formulation of the system can be expressed in terms of the following equations:
CmUot = (d0/4Ri)Uoxx - Uo/Rm0 -
∑I
j =1 N
N
o
o ion(U )
δ(x-xj) + I(t) δ(x-xo) -
∑ P δ(x-x ) ,
o z z=1 N
N
t>0,
∞
9

Journal of Integrative Neuroscience
CmV1t =(d1/4Ri) V1xx - V1/Rm1 -
∑I
j =1
1
1 ion(V )
δ(x-xj) -
∑H[t-(2j-1)?t] I
j=1
1
1 syn(V )
-
∑ P δ(x-x ) ,
1 z z=1
t >0, CmVot = (d0/4Ri)Voxx - Vo/Rm0 -
∑I
j =1
N
o
o ion(V )
δ(x-xj) -
∑ H[t-2j?t] I
j=1
∞
o
0 syn(V )
-
∑ P δ(x-x ) ,
o z z=1
N
t >0,
where Ii ion (i=0,1) is the current density of various ionic channels embedded in the membrane at x = xj , Pi is the sodium-pump current density of the “i” neuron at x=xz, I in (mA/cm) is the input current density of the “0” neuron (only) at x = xo (see Fig. 5b), and ?t is the axonal delay associated with spike propagation from pre- to postsynaptic neurons in sec. Note the differences in diameter between neurons.
I
Ionic “Hot Spots”
Neuron “0”
A xon
Som a
1st 2nd
Kth
th M Synapse
I
Neuron “1”
A xon
C lusters of Ionic C hannels
1st
2nd
th K
th M Synapse
Som a
1st
2nd
Kth
th M Synapse
D endrite
Neuron “2”
A xon
A xon
Som a
1st
2nd
th K
th M Synapse
Som a
1st
2nd
Kth
th M Synapse
D endrite
A xon
S a om
Neuron “N”
A xon
Som a
(a)
(
(b)
10

Journal of Integrative Neuroscience
Fig. 5 Local neural networks: (a) schematic illustration of a feedforward excitatory neural network (homogeneous) of N synaptically coupled neurons; and (b) schematic illustration of a recurrent neural network (heterogeneous) comprising of two synaptically coupled neurons consisting of excitatory pyramidal neuron and an inhibitory interneuron. In (a) and (b), each neuron is spatially distributed with a single equivalent ionic cable representing the dendrites and coupled to a lumped isopotential soma. The ionic ‘hot-spots’ reflect voltage-dependent ion channels superimposed on a passive linear cable. [Reproduced from (a) Biophysical Neural Networks: Foundations of Integrative Neuroscience, 2001, Poznanski, RR (ed.), Mary Ann Liebert, New York and (b) the Journal of Integrative Neuroscience, vol. 1, Poznanski,RR, Dendritic Integration in a recurrent network, pp.69-99 (2002a), Imperial College Press, London].
Figure 6 illustrates the process of synaptic transfer of signals between neurons in networks that do not rely on synaptic weights but are chemically tuned. The mathematical equations for a network of N synaptically coupled dendritic neurons (as depicted in Fig. 5a) or for the network with only two neurons (N=2) (as depicted in Fig. 5b) are given below. The superscripts correspond to the ith dendritic neuron whose synaptic current is
I syn( x,t;V ) = ∑ { ∑ δ(x-xkj) gi (x,t) [Virev - Vi (x ,t)]},
i i j=1 k=1
Npre
M
i=1,2,…..N
where, Npre is the maximum number of pre-synaptic dendritic neurons that are coupled to dendritic neuron “ i ” and where gi (xkj,t) represents the synaptic conductance (nS/cm) of the ith dendritic neuron at a location governed by the kth synapse from the voltage in the pre-synaptic dendritic neuron whose membrane potential varies with space and time and is obtained at the soma of the jth dendritic neuron based on first-order kinetics (Chapeau-Blondeau and Chambert, 1995):
τg dgi / dt = - gi (xkj ,t) + β Ci (xkj , t) [gmax - gi (xkj , t) ]
where τg is the time-constant of decay of gi, β is the equilibrium constant for transmitter-receptor interaction, gmax represents the maximum conductance associated with a non-continuous release of neurotransmitter per unit membrane surface of cable (nS/cm), and Ci(xkj,t) is the concentration of neural transmitter in the synaptic cleft at the kth synapse on the ith post-synaptic dendritic neuron, released by the arrival of action potentials from the jth pre-synaptic dendritic neuron, via
τc dCi / dt = - Ci (xkj ,t) + Fi (xkj , t)
where τc is the time-constant of transmitter release in response to presynaptic depolarization, and presynaptic release of transmitter is reflected in Fi representing the mean firing-rate resulting in a saturating process of neural transmitter release at the kth synapse on the ith post-synaptic dendritic neuron, and which depends on the amount of synaptic current reaching the soma Vj (0,t) in the pre-synaptic dendritic neuron “j” governed by the Boltzman equation:
Fi (xkj ,t) = 1 / {1 + γ exp (- σ [ Vj (0 ,t - ?t) - θ ] ) }
where σ and γ are positive constants governing the net gain during synaptic transmission, and θ is a pre-synaptic threshold voltage. If the rate of neurotransmitter release is proportional to the activity of individual spikes arriving at the presynaptic terminals then to avoid the continuous representation of activity in terms of mean firing rate, discrete spiking activity that releases neurotransmitters into the synaptic cleft can be included and used to investigate the consequences of the non-uniform firing of spikes. The rate of neurotransmitter release depends upon the timing of the action potentials arriving at the presynaptic terminals, which can be represented by a series of Dirac delta functions: Nsp
11

Journal of Integrative Neuroscience
Fi (xkj ,t) =
Σ δ(t-t -?t) H [ V (0 ,t) - θ ]
K=1 k j
The concentration of the neurotransmitter in the synaptic cleft at the kth synapse on the ith post-synaptic neuron emanating from the jth pre-synaptic neuron is:
where Nsp is the total number of spikes, tk+?t is the time of arrival of the kth spike at the synapse, where tk is the time of generation of the spike at the soma or axon hillock.
Ci(t) = exp(-t/τc)∫ exp( t /τc) Fi(xkj,t)dt + k1 exp( -t /τc)
where K1 is a constant of integration determined from the initial-condition, namely C(0)=Co where Co is some predefined resting concentration. The conductance change at the kth synapse on the ith post-synaptic neuron emanating from the jth pre-synaptic neuron is:
gi(t) = (gmaxβ / τg ) exp ( - {
∫
(1+β) / τg Ci(t) dt } )
∫
Ci(t) exp( ∫ (1+β) / τg Ci(t) dt ) dt + k2 exp( - {
∫ (1+β) / τg Ci(t) dt } )
where K2 is a constant of integration determined from the initial-condition, namely g(0)=0.
V pre
F
PRESYNAPTIC SIGNAL
PRESYNAPTIC TRANSMITTER
C g
Isyn
POSTSYNAPTIC ACTIVATION RELEASED AND BOUND TRANSMITTER
Fig. 6 A schematic illustration depicts the synaptic release of neurotransmitter. The presynaptic release of neurotransmitter (F) is represented either as: (i) a mean firing-rate in terms of a sigmoidal dependence on the pre-synaptic voltage (Vpre) resulting in a continuous saturating process of neurotransmitter release; or (ii) in terms of discrete spiking activity where the amount of neurotransmitter release into the synaptic cleft is driven by an individual spike activity. The concentration (C) of released neurotransmitter causes a change in the conductance of the post-synaptic neuron (g) resembling an exponential function as a result of the bound neurotransmitter on the post-synaptic membrane. The conductance change in turn invokes a synaptic current ( Isyn) in the post-synaptic neuron. [Reproduced from Carpenter, G.A. and Grossberg, S. (1991) Distributed hypothesis testing, attention shifts, and neurotransmitter dynamics during self-organization of brain recognition codes. In, Nonlinear Dynamics and Neuronal Networks. (H.G. Schuster, ed.) VCH: Weinheim.]
4.3 Biophysically plausible associative learning rules for the storage and recall of activity patterns
12

Journal of Integrative Neuroscience Synaptic plasticity is believed to contribute to learning by way of both static synaptic strength (i.e., synaptic weights) and dynamic synaptic strength or Hebb-type synaptic plasticity (i.e., synapses change as a result of activity through NMDA receptors). Learning deals with the biophysical properties of synapses that are altered as a result of correlated timing of activity in pre- and postsynaptic neurons. This interaction of current and previous response is known as “associative learning” that can emerge in nested neural networks if one incorporates a variety of biological factors at the cellular level, including: (1) the relationship between patterns of output activity (backpropagating spikes) and dendritic activity (NMDA receptors, Ca+2 influx, Na+- Ca+2 exchange, persistent Na+, and A-type K+ channels); (2) cellular-level calcium influx through NMDA receptor channels involving diffusion modeling of the synaptic cleft (Bernard et al., 1994); and (3) the effects of ambient glutamate and the time course of glutamate release on NMDA conductance (Holmes and Levy, 1990). If expressed analytically, these mechanisms can be incorporated as a way to approximate the synaptic physiology over any number of dendritic/somatic loci of any number of individual neurons. The synaptic dynamics would provide a credible indication for any emergent “learning rule” associated with sequential configuration models. A simple first attempt at modeling associative memory in terms of Hopfield networks given by Lansner and Fransen (1995). Fransen and Lansner (1998) refined their original model, replacing the single-unit model of a cortical column by a circuit of Hodgkin-Huxley type models. However, in order to model sequential configurations possessing the capability for spike-timing dependent plasticity (Markram et al. 1997; Song et al., 2000), a new approach is necessary that is based on integrative methods, incorporating both the spatial structure of single neurons essential for Hebbian principles of learning in the brain (Tsai et al., 1994), and a learning rule that is independent of synaptic weights. The Hebbian rule depends on the timing of pre- and postsynaptic spikes, which depends on the spatial distance a spike needs to backpropagate to the synaptic junction; unfortunately, synaptic weights ignore such spatial parameters.
13

Journal of Integrative Neuroscience
Fig. 7 The effect of boosting on the shape of the dendritic excitatory postsynaptic spike-like potentials (EPSLPs) mediated by persistent sodium hot-spots distributed (a) close to the soma, and (b) uniformly distributed along the stimulated neuron as a result of synaptic feedback from the non-stimulated neuron "0" with inputs distributed peripherally along the dendritic cable of the stimulated neuron (M = 24). The curves are for dendritic spikes at: xp = 0.9 L (solid line) and without hot-spots (dashed line), and without backpropagating electrotonic potentials (dotted line) for a suprathreshold (current) input at Xo = 0 (soma). The key parameters were Ra = 200 ?cm, Rmo = 40,000 ?cm2, Rm1 = 50,000 ?cm2, Lo = 800 μm, L1 = 1000 μm, do = 2.5 μm, d1 = 3.7 μm, λ0 = 0.1118 cm, λ1 = 0.152 cm, g0NaP = 0.229 μS/cm, g1NaP = 0.155 μS/cm, β = 0.3 mA, and α = 0.25/msec. The values of β and α were selected arbitrarily to yield a response at the soma of approximately 9.0 m V (in the absence of sodium persistent hot-spots). [Reproduced from the Journal of Integrative Neuroscience, vol.1, Poznanski, R.R., Dendritic Integration in a Recurrent Network, pp.69-99 (2002a), Imperial College Press, London].
As shown by Poznanski (2002a), the spatial distribution of ionic channels can produce a broadening in the time course of the backpropagating spike (such broadening is of course absent in models without spatial distribution of ionic channels)(cf. Fig. 7a and 7b), which would allow a larger time window for the pre-synaptic signal associate itself with the postsynaptic signal. Such broadening is not accounted for in models without spatial structure, despite the fact that it is an essential mechanism of spike-timing plasticity and learning (Bi and Poo, 1998). That work shows analytically, using a two-neuron recurrent excitatory network model (see Figure 5b), that “weakly” excitable dendrites enable the dendritic spike to propagate passively along its arbor without too much decrement in its amplitude if the spatial distribution between clusters (or hot-spots) of voltage-dependent ionic channels is small, approximating an almost uniform density distribution of voltage-gated ion channels. Poznanski also demonstrated that feedback in a system of two neurons could maintain peak amplitude in the post-synaptic neuron (see Fig. 7). A preliminary study has commenced with the development of a minimal model for backpropagation of action potentials (BAPs) in a dendritic cable with Na+ channels (see Poznanski, 2004) has made it feasible for the first time to analytically solve the Frankenhaeuser-Huxley equations in reproducing complete sets of BAPs. In the future, this work will be further developed by including synaptic input as a conductance change expressed in terms of an alpha function and by explicitly considering voltage-gated K+ currents (e.g., IK, IA, Ih ); including their proper densities, for relevance in spike trains and repolarization during BAP activation in cortical and hippocampal pyramidal neurons (Migliore et al., 1999; Bekkers, 2000). Further work is needed, using Green’s function matrices as a notation, to extend the modeling based on ionic cable theory to local neuronal networks. This would entail modeling a large-scale network of neurons with synaptic structure. 4.4. Volume conductor modeling of ionic diffusion in the neuropil: a phenomological approach With respect to integrating the functional relationships between single biophysically realistic neurons and their collective population properties, spatially coarse-grained modeling approaches are intrinsically unable to provide credible non-arbitrary simulations and spatially fine-grained modeling approaches (governing information processing at the individual neuron level) cannot adequately explain brain activity in terms of a “spatial mean field” as measured with imaging modalities. Therefore, dynamic connectivity (resulting from recurrent interconnections within a neural network and/or through volume transmission where the signal diffuses into a local volume or neuropil) should be conceptualized in prototype models by embedding the neural and glial networks in a volume conductor that interacts with the external micro-environment. One way is to combine the interstitial potentials surrounding individual cells in the neocortical sub-assemblies as a spatial average of the responses through LFPs. Volume conductor modeling of LFPs at the meso-scale can be interpreted in terms of cellular activity (e.g., extrasynaptic signals) at the single neuron or molecular level by a Nernst-Planck analysis for determining the concentration of each ion in the neuropil. Such multi-scale brain modeling captures the essence of intra-cortical pathways exhibited throughout the neocortex. The change in ionic concentrations leads to changes in Nernst potentials, resting membrane potentials, and synaptic transmission. As a first approximation, the effects of such ionic composition changes can be neglected, but a full understanding of ionic diffusion in the extracellular space requires that the dynamics described by time-dependent Nernst-Planck equations and those described by ionic cable theory be included in a single unified theory. The contribution from glial syncytium (Bennett et al., 2005; Spray et al., 2006) is also required for a complete model of extrasynaptically connected nested neural networks. The utilization of electrodiffusion modeling in a volume conductor is crucial to the development of realistic large-scale diffusive neural network models and their interaction within the neuropil. For illustrative purposes, we shall propose a phenomological approach similar to Hodgkin-Huxley’s heuristic
14

Journal of Integrative Neuroscience approach is adopted where membrane offers a certain resistance to the current flow for each ion type, without showing the details of electrical and diffusive forces through the time-dependent Nernst-Planck equation. We propose a diffusive network of coupled neurons, each of which is represented by a model that includes spatially distributed units (or dendritic cables) with ionic channels and extrasynaptic receptors. These neurons are synaptically connected in groups forming neocortical sub-assemblies. The incorporation of ionic cable theory with electrodiffusion theory is a worthwhile and daunting challenge (see Tuckwell, 2000). We shall consider neocortical sub-assemblies consisting of non-recurrent (feed-forward) topologies consisting of excitatory neurons represented by spatially distributed units (or dendritic cables with ionic channels), interregionally connected both synaptically and extrasynaptically in populations or groups to form neocortical sub-assemblies, as depicted in Fig. 8. A novel approach to build models of dendritic and astrocytic networks, which in turn would allow for the individual characterization of neurons and astrocytes in the neuropil requires a simplification in the geometry, which can be made by assuming that the interstitial and intracellular domains are linked everywhere by a membrane such that the outflow of current from one domain is equal to the inflow of current to the other domain as stipulated by Kirchoff’s current law and the continuity of current with the convention that the positive current direction is out through the membrane, can be described by:
15

Journal of Integrative Neuroscience
Fig. 8: A schematic illustration of a neocortical sub-assembly (sub-network or composite neural network) as a nested neural
network consisting of a system of interacting neurons represented by a populations or groups of local neocortical neurons in which “sequential configurations” and “spatiotemporal patterns” in functional columns are manifested. Groups α and β are both interconnected synaptically as shown by dashed lines, and extrasynaptically through diffusive volume signaling as indicated by the arrows. Each different color denotes different neuromodulators. Two extrasynaptically coupled neurons consisting of excitatory pyramidal neurons (blue color) and astrocytes (red color) form a local neural network (i.e., a local heterogeneous population or group of neurons). Each neuron is spatially distributed, with a single dendritic cable representing the dendrites, and coupled to a lumped isopotential soma. The ionic hot-spots reflect voltage-dependent ion channels superimposed on a passive cable. The neurons are connected via synaptic axo-dendritic connections and extrasynaptically between groups in the neocortical sub-assembly. -σ ∫Se?Φe ds =
∑
M
∫Mj IMj ds +
∑
N
∫Nk INk ds
16

Journal of Integrative Neuroscience
j=1 k=1
where Se is the surface of the extracellular medium on the face of the volume conductor, M is the number of dendritic cables in the neuropil, N is the number of astrocytic cables in the neuropil, σ is the conductivity tensor of extracellular medium (S/cm), Φe is the local field potential (LFP) or an “average” extracellular potential (assumed that all variety of cations and anions are grouped together) in mV, Mj is the membrane surface of the jth dendritic cable, Nk is the membrane surface of the kth astrocytic cable, S is the surface enclosing a volume conductor, ds is the element of the surface identified below the integral sign (cm2), and IMj is the transmembrane current density of the jth dendritic cable (A/cm2) given by
RMj IMj = (1/Rmj)(φj - φe) + Cm ?[φj - φe]/?t +
∑I
r=1
PMj
j ion
+
∑I
p=1
j
syn
where φj is the intracellular potential of the jth dendritic cable (mV), φe is the interstitial potential encompassing the jth dendritic cable (mV), RMj is the total number of ionic channels of the jth dendritic cable, PMj is the total number of synaptic connections of the jth dendritic cable, Rmj is the membrane surface resistivity of the jth dendritic cable (? cm2), Cm is the membrane capacitance (F/cm2), Ijion is the ionic current flowing into the j th dendritic cable (A/cm2), and Ijsyn is the synaptic current flowing into the jth dendritic cable (A/cm2). The expressions for Ijion and for Ijsyn are complicated nonlinear functions of intracellular and interstitial potentials that reflect the neuronal responses of individual cells in a large-scale biophysical neural network. For the network of astrocytes the transmembrane current of the kth astrocytic cable (A/cm2) is given by
INk = (1/Rmk)(φk - φe) + Cm ?[φk - φe]/?t + ∑ Ikion +
r=1
RNk
∑I
p=1
PNk
k
gap
where φk is the intracellular potential of the kth astrocytic cable (mV), φe is the interstitial potential encompassing the jth astrocytic cable (mV), RNk is the total number of ionic channels of the kth astrocyte, PNk is the total number of electrical connections through gap-junctions of the kth astrocyte , Rmk is the membrane surface resistivity of the kth astrocytic cable (which in mammals is much smaller that that of the Rmj) (? cm2), Ikion is the ionic current flowing into the kth astrocytic cable (A/cm2), and Ikgap is the gap-junctional current flowing into the kth astrocytic cable (A/cm2). The expressions for Ikion and for Ikgap are complicated nonlinear functions of intracellular and interstitial potentials that reflect the responses of individual astrocytes in a large-scale astrocytic network. In order to investigate the dynamic connectivity of such a system it is now feasible to use Green’s function matrices for a large-scale simulation of approximately 105 neurons arranged as a hypercolumn (i.e., meso-scale). For instance, the scalar analysis (Poznanski, 2001; 2005) can be extended to matrix analysis by determining Green’s function matrices for a system of neuronal cables with ionic channels by taking Laplace transforms and obtaining a linear system of coupled ordinary differential equations. The integrated membrane potentials arising from spiking activity in the dendrites of n simulated neocortical neurons are taken as the sum of depolarizing potentials of excitatory neurons by utilizing a theorem from Tuckwell (1988) on Green’s function matrices in order to explicitly evaluate eAt where A is an n x n diagonal matrix containing the potentials of n neurons in a population. In particular, the use of matrix theory will be used to simplify the notation of large-scale neural networks in the same manner as tensor notation is used in relativity theory. Furthermore, through the application of n-field theory (Chauvet, 1993; 2002) it is possible to capture the dynamical connectivity within neocortical subassemblies. Analytical solutions of such infinite-dimensional nonlinear dynamical systems will require extensive use of Green’s functions together with matrix theory and asymptotic analysis. The work in this direction is currently in progress. 5. Estimating spatiotemporal patterns from functional neuroimaging The main signal in MRI comes from the H2 atoms in water, which are present in all brain tissues. These atoms are highly magnetizable after exposing them to a powerful, constant and uniform magnetic field and their atomic
17

Journal of Integrative Neuroscience spins are able to precess around particular directions determined by the local characteristic of the magnetic field. The aligned hydrogen nuclei assume a temporary non-aligned high-energy state when is perturbed transversally with a low radio-frequency magnetic stimulation. The MRI signal is determined by the relaxation properties of these nucleus after the stimulation ceases. The signal produced by the H2 atoms will be affected by the relative concentration of oxy (diamagnetic) and deoxy (paramagnetic) hemoglobin. An increase of the deoxy-hemoglobin will produce a distortion (attenuation) in the H2-MRI signal, which is used as a source of contrast in fMRI studies. Hence, the BOLD signal is directly related to the level of oxygenated hemoglobin in the blood (i.e. The BOLD effect). 5.1 The Balloon model The main BOLD effect was though to reflect a locus of increased energy utilization, but nowadays we know that it is mostly related to a sizeable departure of the CBF from the resting states locked to the stimulus (see Attwell and Iadecola, 2002 for a critical discussion). The amount of oxygen carried by hemoglobin in hundred of postcapillary venous compartments within an elementary volume of about mm3 (voxel) determines the degree of distortion of the intrinsic H2-MRI signal. Hence, the BOLD effect can be modeled by describing the mechanical expansion of these compartments due to a speedy increase of the CBF. The Windkessel theory (Mandeville et al., 1999) establishes the interrelationship between outflow and volume dynamics during post-capillary venous inflation, where the evolution of the mean transit time τ 0 of the venous compartments determines the underlying dynamics. The mean transit time is defined as the ratio of the volume of the pots-capillary venous compartment and the blood flow getting into it during baseline condition. This parameter produces alterations in the time scale of the BOLD signal, slowing down its dynamic with respect to the CBF when it is increased. The stiffness exponent or Grubb’s parameter α is closely related to the flow-volume relationship; hence, it determines the degree of non-linearity of the BOLD response. The inverse of parameter α relates linearly to the laminar flow
γ = 2 and the diminished volume reserve at high pressure β > 1 (i.e. 1 α = γ + β ). As the CBF f ( t ) and CBV ν ( t ) will change dynamically, this parameter will also vary. In a steady-state condition the CBF
and CBV both appeared to reach a plateau. The value reported in the literature from animal studies α = 0.38 ± 0.10 (Mandeville et al., 1999) seems to be very stable during steady-state stimulation condition. This result is identical to that obtained previously by Grubb et al. (1974) using PET measurements during hypercapnia. Buxton et al. (1998) proposed a differential equation system to describe changes in the CBV and the relative concentration of deoxi-hemoglobin q ( t ) due to the expansion of large amount of post-capillary venous, which was named the balloon model.
The magnitude E ( t ) represents the oxygen extraction fraction at the capillary bed and E0 is its baseline value. Friston et al. (2000) extended the balloon model with two additional differential equations that relate the CBF to a blood flow-inducing signal s ( t ) . These equations may represent an implicit low-pass filter associated with vasodilations and vasoconstrictions at the level of pre-capillary arterioles. They introduced a signal decay term and a feedback auto-regulatory mechanism with respective time constants τ s and τ f in accordance with the frequency of vasomotor signals the neuronal activity. The symbol external force could be delayed
dν ( t ) 1α = f ( t ) ?ν ( t ) dt dq ( t ) f ( t ) E ( t ) (1?α ) α τ0 = ? q ( t )ν ( t ) dt E0
τ0
? = τ s?2 + 4τ ?1 4π ≈ 0.104 Hz . f
ε
The blood flow-inducing signal in the
extended balloon approach is perturbed linearly by an external force u ( t ) , which is supposed to be correlated with stands for the synaptic efficacy. Riera et al. (2006a) pointed out that such
τh
with respect to the neuronal activity and they also introduced a baseline value
u0 to consider external random fluctuations associated with resting states.
18

Journal of Integrative Neuroscience
ds ( t ) s ( t ) ( f ( t ) ? 1) = ε ( u ( t ? τ h ) ? u0 ) ? ? τs τf dt df ( t ) = s (t ) dt
Now we know that spontaneous rhythmic vasomotions depend critically on two main factors (Haddock and Hill 2005): a) intracellular Ca2+ dynamics within the smooth muscle cells and b) smooth muscle and endothelial cell coupling via electrical gap-junction that synchronize Ca2+ oscillations. It is good to say at this moment, that NO levels in the endothelium might play an important role in the regulation of this vasomotion through cyclic GMP (Ignarro et al., 1999). In the original balloon model, it is assumed that the oxygen extraction fraction by the capillary bed satisfies
? , which is only valid for steady-state conditions and a zero concentration of oxygen in ? the tissue. In this particular case, it is assumed that E ( t ) does not depend dynamically on the oxidative
metabolism, which must increase considerably during transient stimulation. Even for this simple case, it is well known that the waveform of the BOLD signal varies with E0 . For instance, the brief initial dip, rarely observed, is attributed to a very light increase in this parameter. During stimulation condition in humans, the values reported in the literature for this parameter are in the range of 0.20 ≤ E0 ≤ 0.55 (Friston et al., 2000; Riera et al., 2004a). Using a deoxi-hemoglobin dilution model, Hoge et al. (1999) showed that BOLD signal is highly sensitive to shifts in CMRO2 at a given level of CBF in steady-state conditions. Recently, Zheng et al. (2002) introduced an active oxygen extraction fraction at the capillary bed depending explicitly of the oxidative metabolism.
1 f (t ) ? ? ? E0 ? ? = ? E ( t ) + (1 ? g ) ?1 ? ? 1 ? ? f ( t ) dt ? ? 1 ? g0 ? ? ? ? The magnitude g is introduced to characterize the ratio of tissue and plasma oxygen concentration at the arterial
E ( t ) = ?1 ? (1 ? E0 ) ?
1 f (t )
? dE ( t )
end, which dynamically determines the mean total blood oxygen concentration in the capillary bed. The constant ? stands for the capillary transit time. Several detailed models have been proposed to account for the oxygen transport phenomena at the capillary bed (Hudetz, 1999; Fantini, 2002) as well as its consumption by oxidative metabolism (Binzoni et al., 1999), but still this is a subject focus of attention of theoretical neurobiologists. Figure 9 summarizes the main physiological processes implicated in the balloon model. The BOLD effect can be mimicked from the vascular states of the balloon model by the equation (Buxton et al., 1998):
? ? ? q (t ) ? BOLD ( t ) = V0 ? k1 (1 ? q ( t ) ) + k2 ?1 ? ? + k3 (1 ? ν ( t ) ) ? ? ν (t ) ? ? ? ? ? ? ? The factors k1 , k 2 and k3 are dimensionless and correspond to how the (extra/intra) vascular systems and the
changing balance effect contribute to the BOLD effect, respectively. Their values depend on the characteristic of the fMRI recording system. In the particular case of a 1.5 T scanner with an echo-time of around 40 ms, these factors can be evaluated empirically by the expressions: k1 ? 7 E0 , k2 ? 2 and k3 ? 2 E0 ? 0.2 (Ogawa et al., 1993; Boxerman et al., 1995). The resting blood volume fraction V0 is usually set up around 0.02. Even with all these progresses, several coupled physiological processes that help the brain to maintain an appropriate energy flow and consumption during neuronal activity must be considered to properly characterize the complex neurovascular and neurometabolic coupling mechasnims, and also the new partner, the emerging neurobarrier coupling mechanism for an adaptative glucose transport based on energetic demands (see for a review Leybaert, 2005). Henceforth, we will discuss new findings and insights of how these coupling mechanisms are triggered during neuronal activity.
19

Journal of Integrative Neuroscience
Fig. 9: A schematic illustration of the balloon model underlying the BOLD effect. The micro-vascular building block includes three parts: a) the arteriole, responsible in controlling of CBF via vascular and/or perivascular structures, b) the capillary bed, where the extraction of oxygen takes place and c) the post-capillary venous, the main contributor to the BOLD signal.
5.2 The neurovascular coupling It is known that several neurovascular mechanisms co-exist to accurately control blood flow in the vascular networks. These mechanisms could be local or global in nature. Unfortunately, how these mechanisms are triggered during neuronal activity remains a mystery. The existence of a parallel electrical network for global vascular control has been conjectured since the early discovery of perivascular structures. Examples are the smooth muscle bands at the arterioles and plastic strips at the pre-capillary branching points, which could be both stimulated from basal forebrain (Ach), thalamus (Glu), locus coeruleus (NA) and Raphe nucleous (5-HT); (Attwell and Iadecola, 2002; Hamel 2006). In the case of the local neurovascular coupling two parallel systems could play an unquestionably role: the neuron-astrocytes pathway (Koehler et al., 2006; Metea and Newman, 2006) and the micro-networks of vasomotor GABAergic interneurons (Cauli et al., 2004). However, we must not discard the possibility that this coupling to be mediated directly by volume transmission involving either diffusible by-product of neuronal spiking activity or metabolites in a plexus of sythase-expressing neurons (e.g. NO, Philippides et al., 2005). The local neuronal regulation of vascular sphincters is a recent result; hence, this hypothesis needs to be corroborated in the near future to be conclusive. Interesting points in this type of coupling are (Cauli et al. 2004): a) vasoactive intestinal peptide and NO synthase were expressed in GABAergic interneurons inducing dilations and b) somatostatin was expressed in GABAergic interneurons inducing constrictions. Although the mechanisms in the neuron-astrocyte pathway are not completely understood, experimental evidence suggests a remarkable role being played by some vasoactive substances (e.g. cyclooxygenase-2 (COX-2) metabolite (prostanoid), EETs, adenosine, or cell derived NO, K+, H+, CO2), which could be synthesized and released by different cells during neuronal activation. It has been shown that Ca2+ signaling is crucial for this coupling in several segments in the neuron-astrosyte pathway (Filosa et al., 2004), e.g. astrocytes, smooth muscle and epithelial cells. For example, an increase of astrocyte intracellular Ca2+ is followed by a decrease of Ca2+ oscillations in the smooth muscle cells and arteriole dilation (Koehler et al., 2006). Ca2+ signals in the astrocytes could be influenced through some neurotransmitter receptors (e.g., GABAb, mGLuR, AMPA, NMDA, A2B) or by an intrinsic mechanism via the (EETs). Prostaglandins, EETs, arachidonic acid and K+ (Paulson and Newman, 1987) are the best candidate mediators between astrocyte end-feet (see Fig. 10-a) and smooth muscle cells. An astrocyte communicates with other astrocytes through intracellular Ca2+ waves and intracellular diffusion of chemical messengers. Therefore, astrocytes could also modulate neuronal synaptic activity and coordinate activity across networks of neurons by releasing neurotransmitters (Ca2+-dependent release of glutamate) or other extracellular signal molecules (volume transmission, see e.g., Arague et al., 2001; Fields and Stevens-Graham, 2002). A functional role of the neuronastrocytes gap-junction could be the transfer of small molecules or metabolites from a particular astrocyte to other
20

基因治疗的概念

基因治疗的概念 基于修饰活细胞遗传物质而进行的医学干预。细胞可以体外修饰，随后再注入患者体内；或将基因治疗产品直接注入患者体内，使细胞内发生遗传学改变。这种遗传操纵的目的可能会预防、治疗、治愈、诊断或缓解人类疾病。 基因治疗的类型 根据目的 基因治疗：治疗或预防疾病 基因增强：增强人的性状和能力 根据干预对象 体细胞：基因转入体细胞内 生殖细胞：基因转入生殖细胞或早期胚胎。 体细胞基因治疗、生殖细胞基因治疗、 体细胞基因增强、生殖细胞基因增强 治疗简要过程 1. 体外法(ex vivo)：患者细胞经体外遗传处理后，再移入患者体内进行治疗。 2. 体内法(in vivo)：把基因通过载体（如病毒）直接转入患者病患部位，进行治疗。 3. 基因治疗使用的载体 （1）病毒 1.1 反转录病毒：RNA病毒，其基因可整入受体基因组 1.2 腺病毒：双链DNA病毒，形成核外DNA，不遗传，需多次施用 1.3 腺相关病毒：单链DNA病毒，可近100%插入19号染色体的特殊位点。感染人和灵长类，不致病，宿主免疫原性小。携带DNA小，难以制造 1.4 病毒壳蛋白形成的假病毒：“魔法子弹” （2）非病毒载体：易于制造、低免疫，但感染和表达少 2.1 裸DNA ：表达效率低；电转移和基因枪转移 2.2 多核苷酸：合成的小分子核酸，使体内基因的表达失活 2.3 脂质体和多聚体：用脂质体和多聚体分子包裹DNA，使DNA被细胞吞入，提高转移效率 （3）复合方法：如脂质体加病毒 （4 ）多晶体：含有核酸（DNA、RNA）的多晶体可以增强基因的转移效率、毒性小 用腺病毒进行基因治疗 Fragile-X 的基因治疗 癌症的基因治疗 基因在癌细胞中表达，引起癌细胞凋亡，从而达到治杀死癌细胞，治愈癌症的目的。 今又生Ad-p53腺病毒注射液 小分子核酸的抑制基因表达和治疗的机制 从DNA到蛋白质，小分子核酸进入细胞后剪切mRNA使基因不能翻译

研究基因功能的4大绝招

研究基因功能的“四大绝招”（初步总结版） 生命科学的研究有很大一部分集中于研究基因及其产物的功能。到底有哪些方法可以用来研究基因功能呢？本文初步总结为“四大绝招”。 第一招：患得患失 得，指的是基因的过表达（Overexpression）；失，指的是基因敲除（Knockout）或者低表达（Under-expression）。例如研究的假说是基因A与记忆力呈正相关，那么可以这样设计实验：先过表达基因A，预期结果是记忆力增强，再降低基因A的表达水平，或者完全敲除（如果不是lethal的话），预期结果是记忆力减弱。一个高质量的实验设计，一般应该“患得患失”，两方面的实验都要做。 我们不仅要“患得患失”，还要“斤斤计较”。因为过表达或低表达的水平不同，表型改变可能也不同，甚至看不到表型改变。例如，用RNAi Knockdown一个基因的表达水平的70%也许看不到任何变化，但是Knockdown 90%就能观察到表型改变了。所以，当过表达或者低表达研究基因却没有得到预期结果的时候，就需要考虑基因表达水平的变化是否不足。 有时即使完全敲除一个基因也看不到任何表型的改变，此时也不能下“研究基因与研究表型无关”的结论。这就好比一个桥有十个桥墩，如果只去除掉桥墩4，在非过负荷的情况下，桥可能不会倒塌，可以正常通车。可是，如果先去除掉桥墩5，再去除桥墩4，桥就会倒塌了。我们能够认为桥墩4是无用的吗？当然不能。桥墩在这里好比处于同一个通路具有相似功能的基因，桥是否可以通车好比基因的表型是否正常。所以，在敲除一个基因A看不到表型改变的情况下，可以在基因B（与A的功能具有相似性，或者是上下游的基因）敲除的动物模型上敲除基因A，观察敲除后是否有变化。 有时候，全身性的过表达、低表达或者基因敲除会出现我们不想要的结果。例如，全身性的基因敲除会致命。为了解决这个问题，现已开发出许多种组织特异性和时间特异性的过表达、低表达和基因敲除技术，使得基因调控更加准确。 第二招：上下求索 基因需要经过转录为RNA、翻译为蛋白质至少两步才能发挥功能。所以，研究一个基因的功能，就可以在DNA、RNA和蛋白质的水平分别进行研究。DNA 的水平相同，不代表RNA的水平相同；同样，RNA的水平相同，不代表在蛋白质的水平相同。哪怕就是在RNA水平，还有不同的剪切的可能。 一个基因翻译成蛋白质之后，常常需要同其它的基因及其产物相互作用才能发

基因生物免疫调理疗法指南

＂基因生物免疫调理疗法＂治愈结肠炎 导语：众所周知结肠炎是一种常见的、多发的肠道疾病，但是由于传统的治疗方法起到的效果微乎其微使结肠炎患者们苦恼不以。21世纪以来随着生物医学的发展，西安京华胃肠医院专家家们发现了一种结肠炎最新疗法——＂基因生物免疫疗法＂ 什么是＂基因生物免疫调理疗法＂ “基因生物免疫调理疗法”是应用基因工程原理，以便利快捷的治疗手段激活患者自身原始细胞，而原始的细胞，具有自我更新、多向分化和高度增殖的能力，在特定的条件下，可分化成为一种后多种体细胞。而溃疡性结肠炎的治疗根本在于修复损伤的结肠粘膜组织，使其恢复正常的生理功能。 采用基因生物免疫调理疗法治疗结肠炎，第一步要取患者自身感染的细胞，运用基因、免疫工程进行人工干预的方式让其产生抗体，把这种具有杀灭病毒能力的免疫血清蛋白抗体再提取出来通过灭菌后注射患者体内，这种细胞可以不断的分泌很多免疫调节因子，调节人体自身免疫系统的平衡，减轻患者炎症反应；第二步细胞植入体内可以不断的分泌很多修复因子，修复组织器官损伤细胞，同时激活患者体内原有的细胞在病灶部位分化出新组织细胞，代替坏死组织细胞，从而使肠粘膜组织恢复正常达到治疗的目的。在结肠炎，溃疡性结肠炎等病毒性疾病的治疗方面有着不可替代的功效和必需性作用。基因生物免疫调理疗法针对性强，不会对肠、内脏造成伤害，无毒副作用，绿色高效。 激发自身免疫，＂基因生物免疫调理疗法＂突破结肠炎疗瓶颈 针对近年来结肠炎发病率上升，国内治疗结肠炎疗效不佳，治愈率偏低，西安京华胃肠中医院胃肠诊疗中心与英国国家医疗总局、美国联邦医院结肠炎防治中心、德国医学院结肠炎治疗中心合作，联合众多国内外顶级胃肠病专家共同科技攻关，通过不断探索、研究，经过十六年的临床实验以及不懈努力，采用最先进的技术理念，集祖国医理论和最先进的纳米技术、基因调控技术，成功开创治疗结肠炎最新疗法＂基因生物免疫调理疗法＂。随后，2011年中国海复旦大学研究的治疗性疫苗也宣告三期临床完成。两大技术各依据不同机理，创造性的分别从不同角度攻克自身免疫治疗的关键技术难题。其中“基因生物免疫调理疗法”由于可以有效激发患者自身免疫，以其治疗便利，费用相对平价，康复最为快速，成为目前为世界多数医疗机构所推广采用。

TCR基因重组免疫激活疗法

TCR基因重组免疫激活疗法 生物基因工程技术引领医学界新变革 性病是危害人类最严重、发病最广泛的一种传染病，2000年全球统计数据：患有性病人数高达4亿多人口，并因其可控性低和传染性强导致性病的扩散速度急剧上升。直到2007年国际医疗组织将基因工程引领应用到性病治疗上，得到了异乎寻常的效果，其控制性疾病的发病和传播有效率达到了100%。 随后经过中国性病学家3年的医学科研，“TCR基因重组免疫激活疗法”于2007年成功问世，截止目前，治愈湿疣疱疹患者高达17万余例且无一例复发，并且首位接受治疗的患者，无需任何药物控制已经见证了6年未复发的奇迹！【TCR基因重组免疫激活疗法】能在3-4天内彻底清除体内HPV/HSV病毒，从而达到治愈的目的，具有快速治愈，无复发，标本兼治等优点。这一科研成果引起了国内外医学界的轰动，令全球医学专家们欣喜若狂。 该技术综合了基因工程学、细胞学、病理学、药理学等多门学科，突破以往治疗的弊端，能彻底根除尖锐湿疣、生殖器疱疹的复发难题，打破了传统疗法占据主导地位的格局，迅速收获了“性病诊疗金标准”的美誉。 TCR基因重组免疫激活疗法四个步骤 第一步：提取自体病毒，进行生物解析 临床提取患者自体1-2粒病毒疣(疱)体，在实验室进行生物解析，确定个体病毒亚型，并提纯个体抗原，为个体特异性治疗提供科学依据。 第二步：生物减毒、灭活技术，进行特异性抗体培养 经实验室进行生物减毒、灭活、提纯、活性封存等一系列手段，针对患者个体病毒型，培养出专门适用于患者个体的特异性抗体制剂T细胞。 第三步：T细胞回输，歼灭体内病毒 根据患者体内检测出的病毒数量，注入相应量的抗体制剂T细胞，对潜伏在体内血液、皮肤基底层、生发层细胞和神经节内的病毒或病菌进行杀灭，经临床观察，1-3天即可快速见效! 第四步激活免疫，杜绝复发。 刺激机体产生保护性免疫来防止病毒再感染，杜绝湿疣、疱疹病毒复发。 -------------TCR基因重组免疫激活疗法—清除HPV、HSV六大优势：------ 1. 基因培植，免疫排毒：完全有自体基因提取、培植，从而达到更有针对性的进行免疫激发，促进机体自身修复，加之TCR基因重组免疫激活疗法潜移默化的对机体的影响下，全面做到无副作用自体免疫排毒! 2. 自体免疫，深层拔毒：利用生物基因工程自身基因培养免疫细胞，通过对自身免疫基因的培养与激活，达到深层拔毒，彻底治愈的目的 3. 抗体繁殖，彻底清毒：利用TCR基因重组免疫激活疗法对患者体内病毒的控制和抗体细胞植入的迅速繁殖，更有效的杀灭病毒并不断吞噬体内残余病毒，控制病毒永不复发。

研究新基因的功能

基因功能的研究思路主要包括: 1.基因的亚细胞定位和时空(发育期或梯度药物处理浓度, 不同组织/器官)表达谱; 2.基因在转录水平的调控(可以通过genome walking PCR或通过已有的资源库寻找该基因的启动子等转录调控区域, 通过单杂交或ChIP等技术, 寻找该基因的转录调控蛋白) 3.细胞生化水平的功能研究(也就是蛋白蛋白作用复合体的寻找验证,具体方法有酵母双杂交, GST pulldown, co-IP, BRET, FRET, BiFc等等,对该基因的表达产物做一个细胞信号转导通路的定位) 4.gain-of-function & loss-of-function: 也就是分别在细胞和个体水平,做该基因的超表达和knockdown(或knockout), 从表型分析该基因的功能. 功能研究应从完整的分子-细胞-个体三个层次研究, 综合分析. 关于基因的表达和定位，可以这样去做： 1. mRNA水平检测基因表达：选择表达目的基因的组织/细胞（发育不同时期、机体不同部位、加处理因素...），提取RNA，反转录，做RT-PCR或real time RT-PCR，检测基因的表达情况/变化。 （或者以northern blot、Rnase protection assay方法，检测基因的mRNA表达情况/变化。）2. 蛋白质水平检测基因表达：选择相应的组织/细胞，以Western blot、免疫组化（OR免疫荧光)检测目的蛋白的表达。 3. 检测目的蛋白的细胞定位：将目的基因克隆至带荧光标签（如GFP）的表达载体，在适合的模式细胞中表达，在活细胞中观察蛋白的细胞定位。 1 首先应当表达该基因，原核基因最好原核表达，真核基因最好真核表达。 2 蛋白质的功能首先观察这种蛋白是膜蛋白还是分泌型蛋白，通过软件分析都可以预测。 3 功能分析可以通过基因树，探究此基因与其他基因的同源性，然后用表达的蛋白进行分析。 4 再有考虑到蛋白相互作用往往介导了蛋白的功能，可以应用酵母双杂交技术和噬菌体展示技术筛选能与此基因表达的蛋白相互作用的蛋白。 5 发现与其相互作用的蛋白后可以通过与其相互作用蛋白的功能推测。 6 可以通过体外过表达此基因观察各种信号通路的改变从而推测其功能

自体DNA免疫自愈疗法”核心原理

自体DNA免疫自愈疗法”核心原理 “自体DNA免疫自愈疗法”核心原理：人体具有天生的自愈功能，免疫力就是证明。“自体DNA免疫自愈疗法”产生的自体DNA免疫抗体能利用和激发人体固有的自愈体系，来加强受损细胞基因的自我修复能力，使紊乱的内分泌系统、神经系统、免疫系统的功能恢复平衡，从而从根本上治愈疾病。采用免疫自愈疗法，能使疣体自然脱落，长效杀菌，主动杀毒，彻底清毒，全面排毒，系统修复、主动免疫、愈后免疫。“自体DNA免疫自愈疗法”核心在于激发自愈力 自体DNA免疫自愈疗法相关介绍 自体DNA免疫自愈疗法：近年来我国性传播疾病发病率居高不下，尤其是生殖器疱疹和尖锐湿疣的高感染、高复发已成为公共卫生领域的难题。最新医学研究表明：生殖器疱疹、尖锐湿疣疾病久治不愈根本原因是“由于皮肤深层皮肤黏膜层中有潜伏和游离的致病病毒没有被彻底清除”。因此想要治愈就要彻底清除病毒。而目前国际上只有“自体DNA免疫自愈疗法”这一根治法才能做到这一点。 “自体DNA免疫自愈技术是国际高端生物基因技术，具有安全、快捷、短程高效的特点。该疗法缔造“全球治疗湿疣疱疹突破400万例不复发”之神奇，是治疗湿疣疱疹的不二选择。 “自体DNA免疫自愈疗法”科研背景： 随着医学的发展，人们越来越多地依赖于药物“代替”身体器官的抗病能力，人体自身的自愈力也受到了削弱，逐渐丧失了本应属于自己的健康，尤其是各种病毒感染。预防医学界的专家们认为，现代医学理念的“疾病治疗”主要是依靠各类药物的作用，而各类药物在发挥作用的同时，其副作用又是以损坏患者部分机体功能并加速其衰老为代价，来寻求患者病灶部位暂时的平衡。即使非常先进的现代医学，也并不能从真正意义上治好疾病，其结果往往是药物的副作用加速了生命体细胞组织的老化。世界卫生组织(WHO)呼吁，要摆脱“对药物的依赖”，拥有真正的健康就应从增强人体自身自愈力着手。修缮人体各器官功能，帮助机体维持并恢复自主健康的能力。这是人类命运的呼唤，也将成为未来医学发展的趋势。 “自体DNA免疫自愈疗法”简要介绍 “自体DNA免疫自愈疗法”是我院专家团独家引进的国际最新科研成果，本疗法以“自体免疫激活作为核心治疗”实现快速杀毒排毒，并利用“病毒免疫技术”使病患机体具有“自体免疫基因抗体”，产生特有的抗毒免疫力，从快速唤醒和激活被病毒抑制的“人体自愈能力”，使病毒基因彻底逆转，更能达到从深层次修复自身免疫细胞活性、彻底消除隐匿潜伏尖锐湿疣病毒和游离病毒，达到快速自愈、根治尖锐湿疣不复发的目的 “自体DNA免疫自愈疗法”治疗过程 自体DNA免疫自愈疗法：近年来我国性传播疾病发病率居高不下，尤其是生殖器疱疹和尖锐湿疣的高感染、高复发已成为公共卫生领域的难题。最新医学研究表明：生殖器疱疹、尖锐湿疣疾病久治不愈根本原因是“由于皮肤深层皮肤黏膜层中有潜伏和游离的致病病

基因治疗

基因治疗 【摘要】研究发现，以基因为基础，从疾病和健康的角度考虑，人类疾病大多直接或间接地与基因相关，故有“基因病”概念产生。根据这一概念，人类疾病大致可分为三类：单基因病、多基因病和获得性基因病。随着现代生物科学的发展，基因工程已在多个领域得到广泛应用。基因治疗是利用基因工程技术向有功能缺陷的人体细胞补充相应功能基因，以纠正或补偿其疾病缺陷，从而达到治疗疾病的目的。基因治疗作为治疗疾病的一种新手段，已经在肿瘤、感染性疾病、心血管疾病和艾滋病等疾病的治疗方面取得进展。它在一定程度上改变了人类疾病治疗的历史进程，被称为人类医疗史上的第四次革命。本文就基因治疗的载体以及基因治疗在肿瘤、艾滋病治疗方面取得的成就作出介绍，并就基因治疗的现状和问题对基因治疗的未来作出展望。 【关键词】基因治疗、载体、肿瘤、p53、IAP、艾滋病、CCR5 【正文】 一、基因治疗背景及概念 1990年9月，美国政府批准实施世界上第一例基因治疗临床方案，对一名患有重度联合免疫缺陷症（SCID）的女童进行基因治疗并获得成功，从而开创了医学的新纪元。自此以来，基因治疗已从单基因疾病扩大到多基因疾病，从遗传性疾病扩大到获得性疾病，给人类的医疗事业带来革命性变革。 基因治疗（gene therapy）是指通过一定的方式，将正常的功能基因或有治疗作用的DNA 序列导入人体靶细胞去纠正基因突变或表达失误产生的基因功能缺陷，从而达到治疗或缓和人类遗传性疾病的目的，它是治疗分子疾病最有效的手段之一。 基因治疗包括体细胞基因治疗和生殖细胞基因治疗。但由于用生殖细胞进行治疗会产生伦理道德问题，因此通常采用体细胞作为靶细胞。其基本内容包括基因诊断、基因分离、载体构建和基因转移四项。根据功能及作用方式，用于基因治疗的基因可分为三大类：（1）正常基因：可通过同源重组方式置换病变基因或依靠其表达产物弥补病变基因的功能，常用于矫正各种基因缺陷型的遗传病；（2）反义基因：通过其与病毒激活因子编码基因互补，或与肿瘤mRNA互补，从而阻断其表达，常用于治疗病毒感染或肿瘤疾病；（3）自杀基因：能将无毒的细胞代谢产物转变为有毒的化合物，用于治疗癌症。 二、基因治疗载体

功能基因研究

吉基 吉凯基因 2014 让我们用心，换取您的放心！

常见问题 ◆创新性不够。 ◆立题依据不充分。 ◆实验设计不合理。 ◆如何做好预实验。 ◆如何获得一个好基因。 让我们用心，换取您的放心！

典型基金精要 A基因通过调控B信号通路影响C肿瘤的D功能

A基因通过调控B信号通路影响C肿瘤的D功能 1. 相关性研究: A--C 组织水平：肿瘤组织样本基因的表达情况 临床水平：基因表达水平与各种临床特点（恶性程度，转移与否，耐药性，生存率等）的相关性 2.功能研究: A--D 细胞水平：生长，凋亡，转移，侵润，血管新生，耐药 细胞水平：生长凋亡转移侵润血管新生耐药 动物水平：成瘤，转移，药物敏感 3. 机制研究: A--B 3机制研究:A B 分子水平：相互结合，表达调控，翻译后修饰前，降解调控，剪切调控，胞内定位，激酶信号传导等 3

肿瘤基因功能研究流程推进 第一步：由特定肿瘤找出相关基因 肿瘤 表达检测 基因常规方法：表达芯片，等 目的：通过筛选，找出在肿瘤组织中有表达，和肿瘤的 目的通过筛选找出在肿瘤组织中有表达和肿瘤的 临床特征有相关性的基因 功能意义：研究的应用性以及临床相关性 意义研究的应用性以及临床相关性 机制 让我们用心，换取您的放心！

肿瘤基因功能研究流程推进 第二步确定候选基因的生物学功能 肿瘤第二步：确定候选基因的生物学功能基本逻辑：改变基因状态后检测细胞模型、动物模型 的表型变化 基因基因操作基因操作常规方法：过表达，RNAi 功能检测功能检测方向：增殖凋亡，转移，血管新生等 功能目标确定候选肿瘤相关基因的生物学功能机制 目标：确定候选肿瘤相关基因的生物学功能 意义：研究的重要性让我们用心，换取您的放心！

自体基因免疫抗体疗法

自体基因免疫抗体技术（也称自体基因免疫自愈技术或自体基因免疫自愈疗法）是由北京京仁医院专家团从美国引进的国际高端生物基因技术，即根据致病病毒的特点，将患者自体病毒抗体种植病患体内，快速提高自愈力杀灭病毒，同时，以生物制剂、特效药物直接作用于病灶处，在局部造成高浓度药离子，快速杀灭病毒;破坏病毒的基因生物链，防止病毒的“半死不活”和“死灰复燃”，是尖锐湿疣、生殖器疱疹、淋病、梅毒等的克星。本疗法的最大特点是提高抗毒免疫力，实现疾病快速“自愈”，大大降低了长期和大量药物的副作用，是一种绿色疗法。 一、核心原理：、淋病、梅毒、疱疹等的克星。 人体具有天生的自愈功能，免疫力就是证明。自体基因免疫抗体技术利用和激发人体固有的自愈体系，来加强受损细胞基因的自我修复能力，使紊乱的内分泌系统、神经系统、免疫系统的功能恢复平衡，从而从根本上治愈疾病。采用自体基因免疫抗体技术，能使疣体自然脱落，长效杀菌，主动杀毒，彻底清毒，全面排毒，系统修复、主动免疫、愈后免疫。 二、治疗过程： (利用患者的自体抗原提升自愈力) 1)将患者所感染的人类的乳头瘤病毒进行生物学分型检测，明确找出病毒DNA类别、型号、数量、感染部位、程度; 2)将患者所感染的病毒提取出来，并根据分子生物学原理，经过较复杂的培养、提纯、灭活、减毒等特殊工序，使之成为人体能够使用的病毒抗原免疫制剂; 3)采取高度保守的病毒核壳蛋白的编码基因获得对不同亚型病毒的交叉保护治疗，从而避开了病毒抗原的病变，提高抗体产生牵引的抗体能力; 4)该疗法能够诱导机体保护性能免疫，发病时能主动免疫，防止病毒再次侵入机体; 5)选择特定型别的抗原肽刺激抗原提呈细胞，从而制作成特定型别的病毒抗原免疫制剂;再进一步使用传统常规辅助疗法，激活免疫。 6)抗原免疫制剂注射到患者体内后，使特定型别的病毒抗原人为的传导给免疫活性细胞，在抗原刺激下，免疫活性细胞被致敏，利用“超导无痛靶向性排毒法”利用其靶向游走特性及免疫性，深入病灶局部，彻底杀灭细胞内病毒和组织深层的游离病毒，从而达到彻底治愈的目的。 三、三点优势： 优势一：激活自体免疫系统产生耐菌抗体其治病机理为自身主动免疫，提取病毒，经培养、灭活、减毒、提纯、合成等特殊的处理，产生免疫性很强的特异性抗体，使人体产生主动免疫的能力，从而有效的避免了梅毒再感染的机会。以最简便的方法了解病毒，培养人力免疫能力，防其病毒再次感染。 优势二：破坏病毒的基因生物链PSD自体激活免疫疗法能有效的干扰病毒的基因表达过程，破坏病毒的生物链，导致病毒无法复制，杜绝了病毒的再生，从根本上保证了通过治愈后极低的复发率。从病灶里直接干扰病毒生物链。并结合仪器与药物结合治疗，效果极佳。 优势三：结合特效中西药物彻底驱除病毒同时吸取祖国中医学之精华，结合临床经反复实践，运用特效中西药物，生物制剂、特效药物直接作用于病灶处，在局部造成高浓度药离子，能够快速杀灭病毒，清除病毒产生的毒物在最短的时间内将患者体内病毒排出，同时使机体产生免疫抗体，彻底根治梅毒杜绝复发。以中西医结合，制定高强度药剂对病毒根源覆

基因治疗

基因治疗 名词定义 中文名称：基因治疗 英文名称：gene therapy 定义1：将正常基因导入造血干细胞或其他组织细胞，以纠正其特定的遗传性缺陷，从而达到治疗目的的方法。 应用学科：免疫学（一级学科）；免疫病理、临床免疫（二级学科）；肿瘤免疫（三级学科） 定义2：在基因水平上治疗疾病的方法。包括基因置换、基因修正、基因修饰、基因失活、引入新基因等。 应用学科：生物化学与分子生物学（一级学科）；方法与技术（二级学科） 定义3：在基因水平上治疗疾病的方法。其手段包括基因置换、基因修正、基因修饰、基因失活、引入新基因等。 应用学科：细胞生物学（一级学科）；细胞生物学技术（二级学科） 定义4：将缺陷基因的野生型拷贝引入患者细胞内以治疗疾病的方法。 应用学科：遗传学（一级学科）；经典遗传学（二级学科） 基因治疗（gene therapy）是指将外源正常基因导入靶细胞，以纠正或补偿因基因缺陷和异常引起的疾病，以达到治疗目的。也就是将外源基因通过基因转移技术将其插入病人的适当的受体细胞中，使外源基因制造的产物能治疗某种疾病。从广义说，基因治疗还可包括从DNA水平采取的治疗某些疾病的措施和新技术。 基本信息 gene therapy 遗传病的基因治疗(gene therapy)是指应用基因工程技术将正常基因引入患者细胞内，以纠正致病基因的缺陷而根治遗传病。纠正的途径既可以是原位修复有缺陷的基因，也可以是用有功能的正常基因转入细胞基因组的某一部位，以替代缺陷基因来发挥作用。基因是携带生物遗传信息的基本功能单位，是位于染色体上的一段特定序列。将外源的基因导入生物细胞内必须借助一定的技术方法或载体，目前基因转移的方法分为生物学方法、物理方法和化学方法。腺病毒载体是目前基因治疗最为常用的病毒载体之一。基因治疗目前主要是治疗那些对人类健康威胁严重的疾病，包括：遗传病（如血友病、囊性纤维病、家庭性高胆固醇血症等）、恶性肿瘤、心血管疾病、感染性疾病（如艾滋病、类风湿等）。 基因治疗是将人的正常基因或有治疗作用的基因通过一定方式导入人体靶细胞以纠正基因 的缺陷或者发挥治疗作用，从而达到治疗疾病目的的生物医学高技术。基因治疗与常规治疗方法不同：一般意义上疾病的治疗针对的是因基因异常而导致的各种症状，而基因治疗针对的是疾病的根源--异常的基因本身。基因治疗有二种形式：一是体细胞基因治疗，正在广泛使用；二是生殖细胞基因治疗，因能引起遗传改变而受到限制。 基因治疗的靶细胞主要分为两大类：体细胞和生殖细胞，目前开展的基因治疗只限于体细胞。生殖细胞的基因治疗是将正常基因直接引入生殖细胞，以纠正缺陷基因。这样，不仅可使遗传疾

光动基因二联免疫诱导疗法介绍

光动基因二联免疫诱导疗法介绍 光动基因二联免疫诱导疗法是当前临床治疗性病最理想的国际诊疗技术，是国内少数几家医疗机构能够操作的治疗技术，是美国FDA认可的权威性病治疗方法。基因诱导自体疫苗埋植疗法是法国医学专家在《2007国际性病诊疗高峰论坛》治疗要求的框架下研究的一种治疗新型技术由“ALA光动力疗法和基因诱导自体免疫法”、以及多套尖端诊疗检测系统组成。光动基因二联免疫诱导疗法最适合当前病毒变异特性的一种新疗法，是目前临床治疗性病的最优化方案。 传统疗法存在的不足 1.不规范治疗导致尖锐湿疣治疗难度增加。 临床治疗不规范治疗和对抗生素的不合理使用，加剧了病毒的抗药性。药物治疗的作用减弱，给病毒复制、病情发展提供了机会。同时由于多种药物的协同作用，不合理搭配不但可降低药物的治疗效用，而且还可能起到负面效果。 2.传统诊疗技术难以确定HPV亚型分类。 传统的诊疗技术很难准确诊断出尖锐湿疣的亚型病毒，不能在治疗方法上准确指导治疗实施，在药效发挥中不能精确诱导对病毒的杀灭和清除。给病毒的复制和存活提供了条件，使得治疗效果更不理想。 3.病毒出现新亚型，抗病毒无特效药。 在临床中抗病毒类药物种类有限，而且没有很好特效的药。在近年来，随着尖锐湿疣等生殖感染的发病率上升，治疗方法和措施的不规范合理，致使病毒出现很多新的变异亚型，对于传统有限的抗病毒药物产生明显的耐药性，治疗药效减弱，甚至失效。传统抗病毒药物已不能有效杀灭病毒。 4.病毒感染抑制和影响免疫系统。 HPV的发病本身和免疫抑制或免疫缺陷有关，但在发病后HPV病毒本身对身体的免疫系统有进一步的抑制和影响，致使正常的免疫功能受到限制，对疾病的防御和病损范围的控制作用减弱，病毒的复制加速，病情逐渐加重，直接影响疾病的治疗效果和康复过程。目前临床中常用的治疗方法很难进行双向治疗作用，只能对常规病毒进行有限的杀灭，左右着整个治疗进度和效果。 5.多合并其它致病微生物潜伏感染。 HPV发病往往是和其它致病菌潜匿隐藏发病，也可见于炎症等疾病之后。这些潜在的与HPV没有直接关系的疾病或致病菌，成为HPV在体内蔓延复制和复发的帮凶。在临床由于诊疗技术和治疗方法的缺陷很容易被忽视，是传统治疗方法不能解决的困难之一。 6.传统治疗方法副作用大，威胁健康。

研究基因功能的“七大绝招”与“三板斧”

研究基因功能的“七大绝招”与“三板斧” 2011.12.6-12.8 （说明：本文是对大约一年前的博文《研究基因功能的“四大绝招”（初步总结版）》的修改和补充，仍存在继续修改的可能。本文可能适合的读者是生物医学的科研新手和非生物医学领域的人。） 生命科学的研究有很大一部分集中于研究基因及其产物的功能。到底有哪些方法可以用来研究基因功能呢？本文初步总结为“七大绝招”和“三板斧”。掌握了这“七大绝招”和“三板斧”，设计实验更容易，看文献听学术报告也更轻松。 第一招：天地人合 无论是学习还是研究，必须遵循的一个原则是“从生活中来，到生活中去”。学习的时候，如果与日常生活中熟悉的、简单的事情结合起来理解，就可以化繁为简，化难为易。学习的目的是为了应用，学到的东西，必须应用到日常生活中去。 怎样研究基因的功能？要回答这个问题，我们先看怎样研究人的功能。假如你是男生，喜欢上了一个“女生”，可是这个“女生”长得扑朔迷离，帅气中带着妩媚，羞涩中透出豪爽。所以，你面临的第一个问题就是：TA到底是男还是女？第二个问题是：TA是不是学生？如果是的话，是本校的吗？你不认识TA的任何朋友，所以也没法打听。为了回答这几个问题，你决定翘课跟踪TA。你发现，白天的大多数时间，TA去了本校教学楼的教室。你守在教室的洗手间旁边，观察TA下课的时候上洗手间，去的是男洗手间还是女洗手间。高兴的是，你发现 TA去了女洗手间（终于松了一口气）。不过你比较小心，为防万一，你又跟 踪她，看她晚上回宿舍去的是男生宿舍还是女生宿舍。不出所料，她去了本校的女生宿舍。这下你终于放心了：她是本校的一个女生。

因为她去女洗手间和女生宿舍，提示她是女的。因为她白天去本校教室，晚上去本校女生宿舍，提示她是本校女生。上述事例告诉我们：一个人什么时候，在哪里活动可以提示其身份。同样，基因表达的时间和部位，常常可以提示其功能。例如，如果基因A在胚胎发育过程中表达，成年后不表达，则提示该基因与发育有关。基因B在在大脑的海马中表达，而海马与记忆有关，那么这个基因可能与记忆有关。 接下来的问题就是：她是哪个系的？她有什么兴趣爱好？你发现，她有两个个形影不离的好朋友。你恰好有同学认识她的这两个朋友。同学告诉你：她两个好朋友都是中文系的，都喜欢打羽毛球。这时，你基本上就可以认为这个女生是中文系的，爱好之一是羽毛球了。 以上所述可以归纳为三点：天时，地利，人和。“天时”指基因及其产物什么时候表达，“地利”指的是基因及其产物表达于哪个部位。“人和”可以进一步引申为两点：1. 近朱者赤，近墨者黑。就是说一个人会与他经常接近的人相似。基因也是如此，相互作用的一些基因常常具有类似的功能，它们为了完成同一个功能而通力合作。例如，如果实验发现蛋白C与蛋白D相互结合，而基因D是某一信号通路的受体，则基因C可能也是该信号通路的成员。2. 近猪者吃，近墨者喝。就是说一个人不仅会经常接触与自己类似的人，还经常接触自己的工作对象。例如，一个经常与学生接触的人，既可能是学生（近朱者赤），也可能是老师（近猪者吃）。同理，与一个基因接触的其它基因或者物质，也可能是其作用的对象。例如，如果蛋白质F与DNA结合在一起，则提示其对DNA发挥作用，可能参与DNA复制、转录等。 “合”指的是合理。亿万年的进化使不合理的基因基本上都被淘汰了，所以存在的基因其功能必定是合理的，至少具有合理性的一面。合理的表现就是能够促进个体的生存和繁衍。合理遵循两个原则：一，经济原则。生物不会浪费物质和能量在一个无用或者冗余的基因上。凡是一个基因可以实现的功能，没有必要用两个基因。二，有效原则。基因的功能应该促进而不是损害生物的生存和繁衍。根据合理性原则，无需任何实验证据，我们不难想到非编码DNA序列是有用的，不是无用的垃圾。 第二招：患得患失

基因诱导抗病毒免疫综合疗法

尖锐湿疣疯狂袭击全军最新基因免疫抗病毒科研成果 让病毒停止不前！ 据医学报道：尖锐湿疣是目前我国发病率最高的性传播疾病之一，每年有数十万人被感染，已经是继梅毒、淋病之后，又一高发性的性传染疾病，正在侵蚀人们的幸福，值得引起广大人民关注。近年，由于患病人数比例不断增加，产生了一系列问题，家庭破裂、性生活不和谐，离婚率普遍增加。专家强烈呼吁让患者不在备受煎熬，及时送患者健康幸福已成头等大事，对此北京解放军93462医院的专家邀请业内人士齐聚一堂，共同商讨研发新技术，希望又一次复燃，备受患者关注。 基因诱导抗病毒免疫综合疗法横空出世 目前，尖锐湿疣的治疗主要采用传统治疗手段，药物治疗只能缓解病情发展，还不能彻底从根本解决问题。近几年也出现了一些新技术，如冷冻治疗、激光治疗、微波治疗、电烧治疗、手术治疗等，虽然可以有效的去掉皮肤表面的疣体，但并不能准确到达疣体的根部，对无症状感染区域的病毒也无法准确的识别清除。导致病毒继续潜伏在体内，过段时间随着病毒的繁殖增生，就会重新出现症状。 自从我国科技迅速发展，医疗技术取得巨大突破。北京解放军93462医院泌外中心作为全军临床新技术应用定点单位，在尖锐湿疣的临床、科研、防治上处于绝对领先地位。通过引进全军最新基因免疫抗病毒科研成果，率先开展了针对尖锐湿疣的基因免疫抗病毒技术临床研究应用，并先后获得了中国医学科学院、中国军事医学科学院、中国医药生物技术协会、第二军医大学等多家医学机构的支持，形成了当前我国最为先进的治疗：基因诱导特异性抗病毒免疫综合疗法，以规范的治疗过程、显著的疗效、温馨至诚的服务，解除了众多尖锐湿疣患者的痛苦和烦恼，赢得了众多患者和家人的好评。 据了解，基因诱导特异性抗病毒免疫综合疗法原理就是：首先采用患者自身疣体，将病原疣体及其代谢物质，经过人工减毒、灭活后，利用基因工程等方法制成特异性疫苗。这种疫苗可以有效地激活自身的免疫系统，产生一定抗病毒引自，具有主动清除病毒能力。然后，采用患者自身疣体，将病原疣体及其代谢物质，经过人工减毒、灭活后，利用基因工程等方法制成特异性疫苗。这种疫苗可以有效地激活自身的免疫系统，产生一定抗病毒引自，具有主动清除病毒能力。最后关键一步是：重建机体免疫系统，增强机体抗病修复能力，真正做到综合治疗。杜绝复发。 北京解放军93462医院专家表示：为了能够更好地为尖锐湿疣患者服务，我们有一支强大的专家团队，我们在科研路上全力以赴，用最好的医术为患者带来健康，最终为的就是能够振兴我国的性病防治技术，在性病防治领域取得更辉煌的成就，让人们的性福生活不再受疾病困扰。 京城专线已开通快速咨询乐无忧 北京解放军93462医院属于高科技现代化大型公立医院，是国家重点性病治疗中心、全军湿疣疱疹治疗中心、全军新技术应用基地，凭借58年的医学实力和全军深厚的背景资源，

基因免疫渗透新疗法解析

基因免疫渗透新疗法解析 “基因免疫渗透新疗法”采用“纳米微粒技术”提取的生物浓缩因子，从外阴白斑发病根源着手，靶向作用于病灶组织，使生物浓缩因子彻底渗透病灶组织并被高效吸收，阻断外阴白斑浸淫途径，同时疏肝理气、滋阴降燥、消炎化瘀、解邪祛痛止痒，治疗效果更稳定，中药分子促进网状内皮系统的吞噬作用，提高细胞免疫和体液免疫功能，促进蛋白质合成和能量代谢，双向调节机体免疫功能提高表皮细胞的应激能力，促进免疫细胞因子的产生,增强NK细胞活性，达到标本兼治之功效。 技术原理 “基因免疫渗透新疗法”是以纳米微粒技术提取的药物浓缩生物因子，直接靶向注射于外阴病损区，通过反复熏洗按摩使生物因子彻底渗透到病损组织，阻断皮损浸润途径，以激活细胞自我修复功能，加快细胞核的分裂与再生，促使局部受损细胞恢复原有活力。通过使用特种光照射病损部，促进病损处对药物的吸收，加快蛋白质的合成和细胞的新陈代谢，促使病损细胞正常化，达到止痒、祛除白斑、恢复外阴皮肤红润与光泽、富有弹性、外观饱满的效果，且不易复发。 “基因免疫渗透新疗法”是在总结古代经方的基础上结合现代医学科技研制成针对外阴白斑顽疾的一套完善的治疗体系。该体系是由三大特色检查、四大核心技术组成，通过精准检查、仪器疏导、口服中成药、外用药熏洗巩固四个步骤，根据每位患者不同的体质、病情、病因，采取最佳诊疗组合，成功解决外阴白斑久治不愈、易复发的难题。 治疗过程 1、精确检查 利用CHIP1000组织切片分析仪、TBV220免疫系统检测仪、DS-3C微量元素分析仪等现代化仪器，对患者的病损组织以及免疫机能进行检测，精确查找病因、定位发病源头；

2、仪器疏导 运用国际尖端仪器对外阴白斑患者病损局部进行治疗，在仪器作用下，能更加有效的促进靶向到病损处的生物浓缩因子快速渗透到病灶组织内部，吞噬消融病变细胞，并有效组织细胞间的营养供给，使病变细胞失去活性，让外阴病损部位恢复正常。 3、口服中成药 口服中成药能通过血液循环，使药物达到机体各个部位，对机体具有疏肝理气、滋阴降火、软坚散结、消炎化瘀、解邪祛痛止痒，使治疗效果更加稳定，中药分子促进网状内皮系统的吞噬作用，提高细胞免疫和体液免疫功能，促进蛋白质合成和能量代谢，使病损部位营养供给恢复正常。 4、熏洗巩固 外用药熏洗，是在高温高压环境下利用纳米微粒技术提取的生物浓缩因子，外敷于患者外阴病损部位，进行反复熏洗按摩。使生物浓缩因子能迅速渗透到病灶组织，对外阴白斑治疗起到巩固作用，铲除病根，有效杜绝复发。 疗法优势 1、因病施治，专人专方 根据每个患者的病因、病情以及身体状况等因素，为患者量身定制一套最佳诊疗方案，实现辩证施治，专人专方。 2、作用迅速，疗效显著 治疗当日见效，7天左右痛痒可止、皲裂愈合、粗糙、角化、溃疡、水肿的外阴皮肤变得光滑，21天左右因疾病产生的萎缩、增生、硬化等现象恢复正常；30天左右所有症状消失达到临床康复目的。 3、绿色安全，无副作用

肌肉疾病基因功能研究介绍

肌肉疾病基因功能研究介绍 我们常遇到的肌肉疾病主要包括杜氏肌营养不良症（Duchenne Muscular Dystrophy, DMD），Becker型肌营养不良(Becker Muscular Dystrophy, BMD) , 脊髓性肌萎缩(spinalmuscularatrophy, SMA)和肌萎缩性侧索硬化症（Amyotrophic lateral sclerosis, ALS）等。 DMD和BMD都属于X染色体隐性遗传，Duchenne型发病率最高，病情最严重，是遗传性肌萎缩症中最有代表性的疾病。DMD也是人体最大的基因，由DMD基因功能丧失引起，可以通过基因治疗技术手段治疗。SMA是常染色体隐性遗传，由SMN基因缺陷、运动神经元丧失引起，可以通过基因治疗的方式来治愈，多发生在婴幼儿。ALS也称为运动神经元病（MND），它是上运动神经元和下运动神经元损伤之后，导致包括球部（即延髓支配的这部分肌肉）、四肢、躯干、胸部腹部的肌肉逐渐无力和萎缩，病因至今不明。 神经肌肉萎缩患者大部分是缓慢起病，神经肌肉萎缩患者往往出现四肢肌肉无力、萎缩，也可有麻木、易疲劳、肌肉疼痛等症状;有的患者还有眼肌、咽喉肌的无力，出现视物重影、眼球活动障碍、发音和吞咽困难等症状，严重者无法吞咽，甚至可能因呼吸肌无力而死亡。 1、这些疾病中主要集中在DMD和SMA疾病，模型分别为： DMD: DMD小鼠模型mdx小鼠，原理为：编码抗肌萎缩蛋白基因的第3185位基因自发突变，由C-T，致使原本编码谷氨酰胺的密码子变成了终止密码子，基因表达提前终止，导致DMD功能缺陷。 SMA: SMA模型鼠，通常是SMN1和SMN2双基因缺陷小鼠。 ALS：SOD1突变相关的ALS小鼠模型（比如突变SOD1 转基因小鼠等）

首个抗癌“基因疗法”以天价上市,改造患者自身免疫细胞攻克癌症

首个抗癌“基因疗法”以天价上市，改造患者自身免疫细胞攻 克癌症 今日，全球首个针对癌症的基因治疗药物Kymriah 获 批上市。Kymriah 是制药巨头诺华（Novartis）在全球获批的首个CAR-T 疗法，目前主要用于儿童和青年的白血病和骨癌的治疗——通过利用患者自身的免疫T 细胞，将其改 造后输回患者体内，可有效杀伤癌细胞。在临床试验中， 63 名接受Kymriah 治疗的急性淋巴性白血病儿童及青年 患者，83 ％在治疗后三个月内病情得到显著缓解，6 个月、12 个月的患者存活率分别为89 ％和79 ％。Kymriah 单次治疗定价为47.5 万美元，约合313 万人民币，价格十分高昂，但诺华承诺，如果患者在接受治疗后的一个月之内没有见到疗效，将全额退款。CAR-T 疗法的具体原理是什么，为什么它能够在对晚期白血病及淋巴瘤患者的临床试验中展现出惊人的疗效？另一方面，CAR-T 疗法 可能引发严重的副作用，甚至可能导致患者死亡——在通往大范围应用的路上，这种疗法还有哪些路要走？ 在《环球科学》2017 年4 月刊的《CAR-T 疗法终结癌症？》一文中，3 位宾夕法尼亚大学的医学专家为我们带来精彩讲解：

撰文小艾弗里· 波西（Avery Posey, Jr. 宾夕法尼亚大学佩雷尔曼医学院讲师）卡尔· 朱恩（Carl June 宾夕法尼亚大学佩雷尔曼医学院教授）布鲁斯· 莱文（Bruce Levine 宾夕法尼亚大学佩雷尔曼医学院教授）翻译杨小鲁宾夕法尼亚大学佩雷尔曼医学院教授几十年来，肿瘤免疫学家一直试图利用人体自身的免疫系统来治疗癌症。然而，在早期研究中，大多数调动人体内部的潜能来抗击癌症的尝试，结果都令人失望。失败的原因是研究人员没有大幅提升免疫系统中的主力军——一种名为T 细胞的关键免疫细胞的作战能力。而只有提高T 细胞识别和攻击肿瘤细胞的能力，免疫系统才有可能在抗癌战役中取胜，否则，无益于要求免疫细胞乘着纸飞机，拿着气枪去战斗。 科学家第一次发现大幅提升T 细胞大军战斗力有助抗击癌症，是20世纪80年代。当时，研究人员想从患者身上提取T 细胞，在实验室培养扩增，然后再输回病人体内，增强患者免疫系统的抗癌能力。这种方法虽然对一些病人有效，但效果持续时间往往不长：注入病人体内的T 细胞会越来越少，不久就消失殆尽。 为了解决这个问题，多个研究团队开始从不同的角度进行攻关。其中，我们和同事提出的一种解决方案，现在已经在临床试验中被广泛应用。20世纪90年代中期，我们中的两个