Ergebnis für URL: http://www.scholarpedia.org/article/Rall_model #[1]Scholarpedia (en) [2]Scholarpedia Atom feed
Rall model
From Scholarpedia
Wilfrid Rall (2009), Scholarpedia, 4(4):1369. [3]doi:10.4249/scholarpedia.1369
revision #91692 [[4]link to/cite this article]
Jump to: [5]navigation, [6]search
Post-publication activity
(BUTTON)
Curator: [7]Wilfrid Rall
Contributors:
0.40 -
[8]Frances K. Skinner
0.20 -
[9]Nick Orbeck
[10]Tobias Denninger
[11]Eugene M. Izhikevich
* [12]Dr. Wilfrid Rall, Scientist Emeritus of NIH; Present Address: RR-1, Box
689, Roseland, VA, 22967
Figure 1: Diagram of a symmetrically branched [13]dendritic tree. Mapping to an
equivalent cylinder, and to a chain of ten equal compartments, is indicated
schematically. Equivalent electric circuit, across the nerve membrane, is shown
at upper right. (Encyclopedia of [14]Neuroscience, reproduced with permissions).
The designation, Rall model, usually refers to one or more of several closely
related biophysical-mathematical models of [15]neurons that have significant
dendritic trees. The importance of the cable properties of dendrites was
emphasized in 1957, when Wilfrid Rall corrected erroneous estimates of the
membrane time constant of motoneurons in cat [16]spinal cord. During the 1960s,
cable models (a cylinder of finite length, and a ten-compartment model) were used
to predict experimental results that were successfully tested with research
collaborators at the National Institutes of Health. This research established the
significance of synaptic input to distal dendritic locations; it also explored
properties of dendritic spines, and predicted dendro-dendritic synapses in the
olfactory bulb.
Contents
* [17]1 Overview
* [18]2 d^3/2 Constraint on branch diameters
* [19]3 Equivalent cylinder model and cable equation
* [20]4 Compartmental model
* [21]5 Mitral and granule cell dendrites: olfactory bulb model
* [22]6 Dendro-dendritic synapses predicted and found
* [23]7 Solutions for input to a single branch of an idealized dendritic tree
* [24]8 Dendritic spine model and clusters of spines on branches
* [25]9 Motoneuron population models
* [26]10 Dendrites as the key to understanding soma voltage transients
* [27]11 Extracellular transients generated by a soma with passive dendrites
* [28]12 Swartz Prize, 2008
* [29]13 References
* [30]14 See also
Overview
A conceptual overview of these models can be provided by referring to Figure
[31]1. This shows an extensively branched dendritic tree. Because all of the
branches are assumed to be cylinders of uniform passive membrane, the
distribution of membrane potential along the length of every branch must obey the
cable equation (Eq.([32]1), below). Also, the branching in Figure [33]1 is
assumed to be symmetrical, and when the branch diameters satisfy a particular
(d^3/2) constraint, it was found (Rall, 1962, 1964) that dendritic locations in
this tree can be mapped to an equivalent cylinder, as indicated schematically in
Figure [34]1. For injection of current to the tree-trunk, the spread of current
and the spreading change in membrane potential from the trunk into all of the
branches, correspond to a mathematical solution of the cable equation in the
equivalent cylinder. Thus, the equivalent cylinder, of finite length, proved to
be a valuable reduced biophysical-mathematical model of the idealized dendritic
tree. With this model, it was possible to obtain analytical solutions to a number
of neurophysiologically interesting mathematical boundary value problems (Rall
1962), as illustrated and discussed below, with Eq.([35]2) and Figure [36]2.
For computational purposes, it was found advantageous to use a chain of ten equal
compartments to approximate the equivalent cylinder, as indicated in Figure
[37]1. Here, the [38]partial differential equation for the cylinder is replaced
by a system of ordinary differential equations for the chain of compartments
(Rall 1964). Compartment-1 can be viewed as the neuron soma, and compartments 2
to 10 represent increasing distance out into the dendritic tree. Here, synaptic
excitation and/or synaptic [39]inhibition can be specified to occur in particular
compartments to explore the consequences of different input locations and
different spatiotemporal patterns of synaptic input (Rall 1964).
It is noted that the simplification of collapsing distal branching was avoided in
a later mathematical treatment (Rall & Rinzel 1973; Rinzel & Rall 1974) which
provided explicit analytical solutions for input delivered to a single branch;
see later section titled: [40]7 Solutions for input to a single branch of an
idealized dendritic tree, below.
The circuit diagram, at the upper right of Figure [41]1, shows the biophysical
membrane model that was used. It consists of a membrane capacitance (per unit
area), which is in parallel with three membrane conductance (per unit area)
pathways. Subscript (r) identifies the resting membrane conductance, which is in
series with the resting membrane battery. Subscript (e, or epsilon) identifies
the variable synaptic excitatory conductance, which is in series with the
excitatory battery; this conductance is zero under resting conditions, but is
turned on by an excitatory synaptic input. Subscript (j) identifies the variable
synaptic inhibitory conductance, which is in series with the inhibitory battery;
this conductance is zero under resting conditions, but is turned on by inhibitory
synaptic input. These conductance pathways correspond to models proposed by Fatt
& Katz (1953) and Coombs, Eccles & Fatt (1955), following upon earlier insights
by Hodgkin & Katz (1949), Fatt & Katz (1951) and Hodgkin & Huxley (1952).
By specifying membrane excitability in a compartmental model of a mitral cell, it
was possible to simulate antidromic propagation of an impulse along axonal
compartments to activate the soma and dendritic compartments of this model. When
this compartmental computation was combined with a model of a mitral cell
population having spherical cortical symmetry (Rall & Shepherd 1968), it was
found possible to simulate spatiotemporal, extracellular field potentials that
had been previously recorded in the olfactory bulb of rabbit. A brief account of
how this modeling led to a prediction of dendro-dendritic synapses is provided in
two later sections titled: [42]5 Mitral and granule cell dendrites: olfactory
bulb model, and [43]6 Dendro-dendritic synapses predicted and found.
Later modeling of synaptic input to dendritic spines provided useful insights
about how such spines could be involved in neuronal plasticity, and in
integrative and logical processing in distal dendrites, as noted in a later
section titled: [44]8 Dendritic spine model and clusters of spines on branches.
d^3/2 Constraint on branch diameters
When exploring steady state solutions for a non-symmetric tree, where branch
diameters and lengths could have arbitrary values (Rall 1959), it became clear,
because the input conductance of each branch cylinder depends on the 3/2 power of
its diameter, that a special case occurs at those branch points where the two
daughter branches have diameters that satisfy the following constraint: the sum
of their 3/2 power values is equal to the 3/2 power of the parent branch
diameter. When this constraint is satisfied at all branch points of a tree, this
tree can be mapped to an equivalent cylinder (provided also that all terminal
branches end at the same electrotonic distance from the trunk). For symmetric
branching, this means that each daughter branch diameter equals 0.63 of the
parent branch diameter. (To see this intuitively, note that 0.8 is the square
root of 0.64, so the 3/2 power of 0.64 equals the cube of 0.8, namely 0.512;
similarly, the 3/2 power of 0.63 equals 0.50). A useful approximate example of
such a symmetric tree has a trunk diameter of 10 microns, with successive
daughter branch diameters of 6.3, 4.0, 2.5, 1.6, and 1.0 microns. It is
noteworthy that measurements of branch diameters in motoneurons of cat spinal
cord have found rough agreement with the d^3/2 constraint; see review article by
Rall et al., (1992) for examples. When there is consistent deviation from this
constraint, one can make use of a more general partial differential equation
(Rall 1962a) that corresponds to an equivalent taper; this has been explored in
Goldstein & Rall (1974); see also Holmes & Rall (1992).
Equivalent cylinder model and cable equation
The first model was a cylinder of finite length, consisting of uniform passive
nerve membrane (Rall, 1962a). In this reduced model, a dendritic tree was
collapsed into a single cylinder; the neuron soma lies at one end, \(X=0\ ,\)
while the dendritic terminals lie at the other end, \(X=L\ .\) Because passive
membrane is assumed, the spatiotemporal distribution of [45]membrane potential
along the cylinder must obey a [46]partial differential equation, known as the
cable equation (a derivation is provided in Rall 1977, pages 64-67); this can be
expressed \[\tag{1} \frac{\partial V}{\partial T} = - V + \frac{\partial^2
V}{\partial X^2} \]
where \(V = V_m - E_r\) represents the departure of the membrane potential,
\(V_m\ ,\) from its resting value, \(E_r\) (and \(V_m\) is intracellular voltage
\(V_i\) minus extracellular voltage \(V_e\)). Also, \[X=x/\lambda\ ,\] with
\(\lambda = \sqrt{r_m/r_i} = \sqrt{(R_m/R_i)(d/4)}\) \[T=t/\tau_m\ ,\] with
\(\tau_m = r_mc_m = R_mC_m\ .\) Here, \(d\) is the cylinder diameter, while
\(r_i\) is the intracellular (core) resistance per unit length of the cylinder
and \(c_m\) and \(r_m^{-1}\) are the membrane capacity and membrane conductance,
respectively, per unit length of the membrane cylinder; \(R_m\) and \(C_m\) apply
to unit membrane area, while\(R_i\) is the volume resistivity of the
intracellular medium.
Figure 2: Theoretical voltage transients computed for analytical solutions based
on Eq.([47]2), for a tree (or equivalent cylinder) with L=1. Both sets of curves
show the time-course of membrane depolarization at three locations X=0, X=0.5, &
X=1. The upper three curves (A) show the computed response to a step increase in
synaptic excitatory conductance, (from zero to twice the resting conductance),
applied to only the distal half of the cylinder. The inset (B) shows the computed
result when the same conductance step is turned off at T=0.2; the upper sharper
curve is at X=1, while the lower rounded curve is at X=0 (Rall 1962). Note that
for this square synaptic conductance transient, symmetry implies the sharper
shape would represent a theoretical EPSP (at the soma) for synaptic input to the
proximal half of the tree or cylinder, while the delayed and rounded shape
represents a theoretical EPSP (at the soma) for synaptic input to the distal half
of the tree. (Reproduced with permissions from Rall (1962)).
For "sealed ends", meaning no current flows out of either end, the mathematical
boundary conditions are \(\partial V/\partial X = 0\ ,\) at both \(X = 0\) and
\(X = L\ .\) Using the classical method known as [48]separation of variables, a
general solution of this [49]boundary value problem can be expressed \[\tag{2}
V(X,T) = \sum_{n=0}^\infty B_n \cos(\alpha_n X) e^{-(1+\alpha_n^2)T} \]
where \(n\) is any positive integer, or zero, and \[\tag{3} \alpha_n = n\pi/L \]
The values of \(\alpha_n^2\) are known as the [50]eigenvalues of the [51]boundary
value problem. The \(B_n\) are constants to be determined from the initial
condition.
A related boundary value problem distinguishes between two regions of the
cylinder, corresponding to proximal and distal regions of the dendritic tree,
where one region has a uniform [52]synaptic excitatory input turned on for a
short time. Once this synaptic input was turned off, the whole cylinder obeys
Eq.([53]2). Computations using these analytical solutions demonstrated how the
transient shape of a synaptic potential (EPSP at the soma) depends upon distal
versus proximal locations of synaptic input: for distal input, the EPSP rises
more slowly to a later and broader peak (of lesser amplitude), compared to an
earlier, sharper peak for proximal dendritic input locations (Rall, 1962a), as
illustrated in Figure [54]2.
It is noteworthy that when the expression ([55]3) is introduced into the
theoretical expression for the exponent in Eq.([56]2), we obtain an expression
that implies a set of time constants, \(\tau_n\) (for the passive cylinder with
sealed ends), as follows \[\tag{4} \tau_0/\tau_n = 1 + (n\pi/L)^2 \]
Note that \(\tau_0=\tau_m\) governs the decay of a uniformly distributed membrane
potential, while the \(\tau_n\ ,\) for \(n>0\ ,\) have been called equalizing
time constants, because they govern the more rapid decay of non-uniformly
distributed components of the membrane potential. Because Eq.([57]4) can be
rearranged to provide the expression \[\tag{5} L =
\frac{n\pi}{\sqrt{\tau_0/\tau_n-1}} \]
one can see that the value of \(L\) can be calculated from the ratio of two time
constants. This has provided a very useful method for estimating the electrotonic
length, \(L\ ,\) from analysis of experimental transients, as discussed and
illustrated in Rall (1969).
Compartmental model
The next model was a chain of ten equal compartments (Rall, 1964), see Figure
[58]1. This can be regarded as an approximation to the cylinder of finite length.
Here compartment-1 represents the neuron soma, while compartments-2 through 10
represent increasing distance out into the dendritic tree. The continuous
variation of \(V\) with \(X\) in the first model is replaced by stepwise
differences in \(V\) between adjacent, connected regions of lumped membrane. This
model offers valuable computational advantages. In 1962, Rall made use of a
computer program, SAAM, that had been developed at NIH by Mones Berman for
studies in metabolic kinetics, to solve the equations. Today computer programs
designed for neuroscience, such as [59]NEURON, developed by Hines, and
[60]GENESIS, developed by Wilson and Bower are used for this purpose.
It became possible to compute EPSP shapes for many different dendritic locations
of synaptic input (Rall, 1964), as illustrated in Figure [61]3, for square
(on-off) synaptic conductances.
Figure 3: EPSP shapes computed with the ten-compartment model. Here each
compartment corresponded to 0.2 units of electrotonic length; with the soma as
compartment-1, this implies approximately L=1.8 for the dendritic tree. Each of
these EPSP shapes (at the soma), resulted from a square pulse of excitatory
conductance (equal to the resting conductance, with a duration of 0.25T)
delivered only to two adjacent compartments, as indicated in the figure
(reproduced with permissions, Rall, 1964).
Figure 4: Two EPSP shape-index loci, at left, based on theoretical EPSP shapes
like those at right. Here, the synaptic conductance was not square, but a brief
transient function, F(T) = (T/ Tp) exp (1 - T/ Tp) where Tp equals the time of
the peak value. For a peak time, Tp = 0.04 of the membrane time constant, this
transient is shown at right as the dotted curve in mid-diagram. Lower right shows
three computed EPSPs, each for an input to a single compartment, indicated by the
number in the triangle; note that the peak amplitudes were normalized to aid
shape comparison. Two shape index values: "time to peak" and "half width" were
measured and then plotted in a two-dimensional plot (at left); each point in this
plot represents a different EPSP shape. The dotted line is a shape-index locus,
for single compartment inputs, using the same input transient. The solid line
represents a contrasting shape index locus obtained when the synaptic input was
delivered uniformly over the entire neuron surface; only the time course of the
input transient was changed. (Reproduced with permissions from Rall et al. (1967)
and Rall (1977)).
Later such computed EPSP shapes were refined by using a transient time course to
govern the synaptic conductance, in order to enhance the comparison with
experimental EPSP shapes (Rall, 1967, Rall et al, 1967). For each EPSP shape, two
shape index values, time to peak and half-width, could be used to construct
two-dimensional shape index plots, leading to theoretical shape index loci (for
different input locations and for different synaptic input transients); see
Figure [62]4.
Figure 5: Computed contrast between opposite spatiotemporal sequences of
excitatory synaptic input. Input sequence, ABCD, (at upper left) produced the
earliest transient time-course, while input sequence, DCBA, (at upper right)
produced the delayed transient with the higher peak. The component inputs are the
same as in Figure [63]3; the successive time intervals were one quarter of the
membrane time constant, as indicated in the figure. The dotted curve shows the
control transient, where spatiotemporal pattern is eliminated by setting the
synaptic conductance amplitude 1/4 as large, over all four locations, for the
full four time periods. (Reproduced with permission, Rall (1964)).
These theoretical loci provided explicit theoretical predictions that could be
compared with experimental EPSP shapes. A successful NIH research collaboration
with Bob Burke, Tom Smith, Phil Nelson & K. Frank, (Rall et al 1967) persuaded
most neurophysiologists about the significant contribution of dendritic synapses
in [64]motoneurons of cat [65]spinal cord (after years of denial by Eccles). The
agreement of experimental EPSP shapes with theoretical predictions was further
demonstrated by Julian Jack and colleagues at Oxford (Jack et al, 1971), by
Mendell and Henneman (1971) at Harvard, and by Iansek and Redman (1973) in
Australia. Later a remarkable experiment by Redman and Walmsley (1983), in
Australia, succeeded in combining electrophysiology and histology (in the
identical neuron) to show agreement between two actual known synaptic input
locations (for two individual afferent fibers) and the theoretical input
locations implied by the two different recorded EPSP shapes.
This compartmental model was also used to demonstrate the significantly different
results obtained when comparing/contrasting spatiotemporal sequences of synaptic
input (Rall, 1964); see Figure [66]5. Such differences could have relevance for
pattern discrimination and for movement detection.
Different computations explored nonlinear interactions between synaptic
excitation and synaptic inhibition (Rall, 1962, 1964). Synaptic inhibition was
shown to be less effective when it was located distal to the synaptic excitation;
it was almost equally effective when placed in the same compartment as the
synaptic excitation or at the soma compartment.
With regard to more general [67]compartmental models, it must be emphasized that
the ten compartment model, used above, is a very simple, passive special case.
Compartmental models can be designed to go beyond the constraint of passive
membrane properties and unbranched cylinders. Different non-linear and
[68]voltage dependent membrane properties can be specified for any compartment
(e.g. the model in the next section below); compartments can be of different
size, and also, dendritic branching can be represented explicitly (Rall, 1964).
Mitral and granule cell dendrites: olfactory bulb model
A more complicated model, combining compartmental and cortical modeling, was used
by Rall and Shepherd (1968) to simulate the spatiotemporal [69]extracellular
(field) potentials that had been recorded in experiments with the [70]olfactory
bulb of rabbit by (Phillips, Powell & Shepherd, 1963). This involved
[71]synchronous antidromic activation of a very large population of [72]mitral
cells, arranged in an almost spherically symmetrical cortex, with their primary
dendrites aligned radially outward. To model mitral cell activation, a
compartmental model was used; this had three small axonal compartments, and a
larger soma compartment, all of which had specified active membrane properties.
(Note that these active properties depended on a new mathematical model that was
designed to generate conductance transients similar to those of the
Hodgkin-Huxley model; this was computationally economical; also the values of the
H-H parameters were not known for neurons of rabbit). These active properties
provided propagation of an antidromic action potential from the axonal
compartments to the soma. The dendritic compartments had passive membrane in some
computations, and active membrane in other computations. Because of the large
population of simultaneously activated mitral cells in the olfactory bulb, this
population was idealized as a spherically symmetrical cortical layer in which the
extracellular current generated by the mitral cells flows radially, producing
spherical equipotential contours whose transient values could be computed.
However, because the symmetry of the actual olfactory bulb is not perfect, and
because it is "punctured" by the lateral olfactory tract, it was necessary to
consider a small secondary extracellular current whose extra-bulbar path includes
the reference electrode. Thus the computational model included a "potential
divider" correction, which resulted in agreement with the experimental data; see
Rall and Shepherd (1968) for the details.
There is also a very large population of [73]granule cells whose dendrites
intermingle with dendrites of the mitral cells in the external plexiform layer
(EPL) of the olfactory bulb; the granule cells also extend dendrites into a
deeper layer. A granule cell was also represented by a chain of compartments;
these compartments had passive membrane and they were assumed to extend from the
EPL, deep into the granular layer (GRL) of the olfactory bulb. Computations with
this granule cell model also required a "potential divider" correction, to
simulate the effect of a distant reference electrode.
Dendro-dendritic synapses predicted and found
The successful simulation of the experimental field potentials implied that the
granule cell population must receive synaptic excitatory input within the
external plexiform layer (EPL) of the bulb, at the very time that the mitral
secondary dendrites are depolarized in the EPL. This suggested that the
depolarized mitral cell dendrites must deliver synaptic excitation to the granule
cell dendrites in the EPL; this would imply dendro-dendritic excitatory synaptic
contacts, which were not yet known to exist. Also, because the mitral cells were
subsequently inhibited, this suggested that synaptic inhibition must be delivered
by the granule cell dendrites to the mitral cell dendrites in the EPL; this
implies dendro-dendritic inhibitory synaptic contacts, which were not yet known
to exist. Only seven months later (March 1965), both kinds of
[74]dendro-dendritic synapses were found by Tom Reese and Milton Brightman in
their independent electron-microscopic research, at NIH. Because the theoretical
model and the experimental results agreed so well, a joint paper was prepared and
submitted to Science in 1965; their referee found it not to be of general
interest. The paper was later accepted for publication in Experimental Neurology
(Rall et al, 1966). Dendro-dendritic synapses were controversial for a while, but
they have since been found in other regions of the CNS. It should be noted that
these synapses provide a new pathway for lateral inhibition; also such synapses
provide for graded interactions that do not make use of an all-or none impulse.
Solutions for input to a single branch of an idealized dendritic tree
An enriched model was used to explore the consequences of input to a single
branch of an extensively branched dendritic tree. By using superposition methods,
the analytical solutions of the cylinder model (see earlier section titled: [75]3
Equivalent cylinder model and cable equation above) were combined to provide
solutions for the case of idealized branching; the branching was assumed
symmetric, with branch diameters that satisfy the d^3/2 constraint. This neuron
model allowed for several equal dendritic trees, \(N\) in number; each tree was
equivalent to a cylinder of electrotonic length, \(L\ ,\) and had \(M\) orders of
symmetric branching. For input to a single branch of one of these trees, steady
state solutions (Rall & Rinzel, 1973), and transient solutions (Rinzel & Rall,
1974), were provided, illustrated and discussed. Insights were provided regarding
input resistance and impedance values at different input locations in the tree,
as well as voltage attenuation in different branches of the tree.
Dendritic spine model and clusters of spines on branches
One reason to explore solutions for input to a single branch was an interest in
[76]dendritic spines, together with the question: when a spine head receives a
synaptic input, how much of the potential generated in the spine head membrane is
delivered to the tree-branch where the usually narrow (high resistance) spine
stem is attached? It was found important to consider the ratio of spine-stem
resistance to the input resistance of the tree-branch (Rall & Rinzel, 1971). For
a ratio of 0.01 or less, steady input to the spine head is delivered to the
branch without significant attenuation. For a ratio of 100 or more, the voltage
delivered to the branch is negligible. For an intermediate range, from 0.1 to 10,
steady voltage attenuation ranges approximately between 10% and 90%. This was
identified as an "operating range" for possible [77]synaptic plasticity and
learning.
Related models were used to explore the effects of excitable spines (Miller et
al. 1985), and of clusters of spines (both passive and excitable) on distal
branches of dendritic trees (Rall and Segev, 1987); see also (Segev and Rall,
1988) and (Shepherd et al, 1985). Examples were provided for various kinds of
synaptic interactions, supporting a concept of possible logical processing in the
distal dendrites of a neuron.
Motoneuron population models
An entirely different mathematical model was used by Rall in his 1953 Ph.D.
thesis. This involved a comparison of theoretical predictions with the
input-output relation for the mono-synaptic reflex in a motoneuron population in
cat spinal cord. This input-output relation had been demonstrated and discussed
by Lloyd (1943, 1945), for the spinal segmental level.
Two probabilistic models (differing in the definition of the threshold) were
developed and tested. The simplest assumed that a motoneuron discharges when the
number of simultaneously activated synapses on this motoneuron reaches or exceeds
a specified threshold number. The other model assumed that a smaller number of
synapses could succeed, provided that they were concentrated in a subregion
(zone) of the motoneuron's receptive surface. It was found that the experimental
data could be fitted by both models. An important test was to fit a set of four
input-output curves obtained, from a single preparation, at four different levels
of reflex excitability; only one parameter of the theoretical model (threshold
value) needed to be adjusted to fit the four curves.
More detail can be found in the original publications (Rall, 1955a, 1955b); also,
a useful summary appeared later in a book chapter (Rall, 1990), and as Appendix
A.2 in another book (Segev, Rinzel and Shepherd, 1995), where Appendix A.1, by
Julian Jack, is also relevant.
In collaboration with Cuy Hunt at the Rockefeller Institute, it was found
possible to fit data on the "firing index" distribution in a motoneuron
population. It sufficed to assume a normal probability distribution of firing
thresholds (Rall and Hunt, 1956).
Note that dendrites had not been distinguished from the motoneuron soma in these
early theoretical models; modeling of dendritic cable properties began in 1957;
see next section.
Dendrites as the key to understanding soma voltage transients
In a note to Science (Rall, 1957), it was pointed out that the rapid voltage
transient recorded in response to an applied current step from a single cat
spinal motoneuron, by means of the recently introduced glass micro-electrode, was
being misinterpreted, because the cable properties of the dendrites had been
neglected. By assuming this transient to be a single exponential, Eccles and
others implicitly assumed that they were recording from a "soma without
dendrites". If , on the other hand, one assumes the dendrites to be dominant, one
expects a significantly different transient function, that is much closer to
those known to cable theory for non-myelinated axons. The actual problem is a
soma with significant dendrites.
Rall prepared a detailed analysis of this intermediate problem, and submitted it
to the Journal of General Physiology, in 1958. A negative referee persuaded the
editors to reject this MS. The fact that Eccles was this referee was obvious from
the many marginal notes found on the returned MS. Fortunately, K. Frank and W.
Windle, editors of a new journal, Experimental Neurology, encouraged Rall to
expand this MS into two papers: one paper (Rall, 1959) solved the steady state
problem for general dendritic branching, and included estimation of membrane
resistivity from experimental data; the other paper (Rall, 1960) solved the
transient problem, using Laplace transforms, and included estimation of the
membrane time constant from experimental data.
It was necessary to work with poorly matched data: electrophysiological data from
several sets of impaled motoneurons together with anatomical measurements from
different sets of histological motoneurons, in order to estimate values for the
underlying model parameters. Assuming uniform passive membrane over dendrites and
soma, and assuming that the larger input resistance values correspond to the
smaller motoneuron dimensions, it was possible to estimate a range of values from
1000 to 8000 \(\Omega\)-cm^2, with a mean of around 5000, for the membrane
resistivity; this was significantly greater than values, of 400 to 600, estimated
by Eccles and his collaborators for their "standard motoneuron". Also, their
"standard" model provided a value of 2.3 for the ratio of combined dendritic
input conductance to the somatic input conductance, whereas Rall estimated an
extreme range from 10 to 47, with a mid-range from 21 to 35, for this important
ratio (Rall, 1959).
It is now clear that the "standard motoneuron" significantly under-estimated the
importance of the dendrites. Note that estimating a membrane resistivity that was
ten times too small, Eccles obtained values for the dendritic length constant
that were 3.2 times too small, leading to exaggerated values for L; for example,
a dendritic tree with an actual L=2, would be misunderstood to have 6.4 as its
value of L. Such errors contributed to assertions by Eccles that distal dendritic
synapses could be dismissed as virtually ineffective. In Rall's modeling, the
electrotonic length of the dendritic tree has been in the range from L=1 to L=2,
and distal dendritic synapses have been shown to be effective (Rall et al.,
1967). This range of values for cat spinal motoneurons has been confirmed by
several studies that compared anatomically based estimates with independent
estimates based on Eq.([78]5) above. Note also, that later considerations of
synaptic amplification by excitable dendritic spines, and by distal clusters of
such spines (section above titled: [79]8 Dendritic spine model and clusters of
spines on branches) add to the efficacy of synaptic input to distal dendritic
branches.
The transient analysis (Rall, 1960) provided the basis for a new semi-log plot to
estimate the membrane time constant. Instead of plotting the log of the slope,
dV/dt. against t (as would be correct for a single exponential transient), one
must plot the log of the product: (dV/dt times the square root of t), against t.
Examples were provided and discussed. Also, because these analytical solutions
were obtained for the case of very long dendrites, it should be noted that the
early part of this transient, at the soma, is essentially the same for a cylinder
of length, L. This can be understood by considering a long cylinder with equal
inputs at X=0 and X=2L. Then symmetry causes the slope, dV/dX, to be zero at X=L,
and the early solution at X=0 is only slightly effected by the contribution
(spread back) from the distant input at X=2L.
This method of semi-log plotting was applied to some of the transients published
by Eccles and his collaborators; larger values for the membrane time constant
were obtained, and their implications were discussed (Rall, 1960). Suffice it to
say that the special hypotheses, promoted for several years by Eccles, about two
phases of synaptic current, were eventually abandoned. A useful review of these
issues is provided by Redman and Jack, on pages 27-33, in a book edited by Segev,
Rinzel and Shepherd (1995), which also contains reprints of (Rall, 1957, 1959 and
1960).
Extracellular transients generated by a soma with passive dendrites
At the first International Biophysics Congress, held in Stockholm in 1961, Rall
(1962b) presented the results of early computations with an IBM 650 mainframe
computer. Assuming an action potential in the soma membrane, and passive
electrotonic spread into passive (dendrite) cylinders, the resulting field of
extracellular potential was computed, for the case of one cylinder, and then for
the case of seven cylinders, for the moment of peak action potential. It was
found that the equipotential contours near the soma are almost spherical. Then,
with the simplifying assumption of complete radial symmetry, the extracellular
transient was computed (at several radial locations), during the timecourse of
the action potential at the soma. The resulting extracellular voltage transient
was found to be diphasic (-,+), in agreement with the experimental observations
of Nelson and Frank (1964) near motoneurons of cat spinal cord.
It was significant that this Rall model generated the kind of (-,+) diphasic
transient that others had claimed was evidence for the propagation of action
potentials in dendrites. But in this computation, the dendrites were explicitly
passive. This meant that such a diphasic record could not be claimed to provide
evidence for impulse propagation in dendrites. A discussion of this issue can be
found in Nelson and Frank (1964).
The physical intuitive explanation of this computed result is the following: the
negative extracellular peak is generated by extracellular current flowing
radially toward the soma (during sodium ion current inward across the soma
membrane), while the positive extracellular peak is generated by extracellular
current flowing radially away from the soma (during potassium ion current outward
across the soma membrane, which very rapidly repolarizes the soma, relative to
passive continued depolarization of dendritic membrane); see also pages 577& 578
in (Rall, 2006).
Swartz Prize, 2008
At the Annual Meeting of the Society for Neuroscience, held in November of 2008,
the inaugural Swartz Prize for Theoretical and [80]Computational Neuroscience was
awarded to Wilfrid Rall (now age 86).
References
Some of the referenced publications can be found reprinted in a book edited by
Segev, Rinzel, and Shepherd (1995). An autobiographical chapter (Rall, 2006) has
been published recently by the Society for Neuroscience.
* Coombs J.S., Eccles J.C., and P. Fatt. (1955) The electrical properties of
the motoneuron membrane. J.Physiol. 130: 291-325.
* Fatt P., and B. Katz. (1951) An analysis of the end-plate potential recorded
with an intracellular electrode. J. Physiol. 115: 320-370.
* Fatt P., and B. Katz. (1953) The effect of inhibitory nerve impulses on a
crustacean muscle fibre. J. Physiol. 121: 374-389.
* Goldstein S.S., and W. Rall. (1974) Changes of action potential shape and
velocity for changing core conductor geometry. Biophys. J. 14: 731-757.
* Hodgkin A.L., and B. Katz. (1949) The effect of sodium ions on the electrical
activity of a crustacean nerve fibre. J. Physiol. 108: 37-77.
* Hodgkin A.L., and A.F. Huxley. (1952) A quantitative description of membrane
current and its application to conduction and excitation in nerve. J.
Physiol. 117: 500-544.
* Holmes W.R., and W. Rall. (1992) Electrotonic length estimates in neurons
with dendritic tapering or somatic shunt. J. Neurophysiol. 68: 1421-1437.
* Iansek R., and S.J. Redman. (1973) An analysis of the cable properties of
spinal motoneurones using a brief intracellular current pulse. J. Physiol.
234: 613-636.
* Jack J.J.B, Miller S., Porter R., and S.J. Redman. (1971) The time course of
minimal excitatory post-synaptic potentials evoked in spinal motoneurones by
group Ia afferent fibres. J. Physiol. 215: 353-380.
* Lloyd D.P.C. (1943) Reflex action in relation to pattern and peripheral
source of afferent stimulation. J. Neurophysiol. 6: 111-120.
* Lloyd D.P.C. (1945) On the relation between discharge zone and subliminal
fringe in a motoneuron pool supplied by a homogeneous presynaptic pathway.
Yale J. Biol. Med. 18: 117-121.
* Mendell L.M., and E. Henneman. (1971) Terminals of single 1a fibers: Location
density and distribution within a pool of 300 homonymous motoneurons. J.
Neurophysiol. 34: 171-187.
* Miller J.P., Rall W., and J. Rinzel. (1985) Synaptic amplification by active
membrane in dendritic spines. [81]Brain Res. 325: 325-330.
* Nelson P.G., and K. Frank. (1964) Extracellular potential fields of single
spinal motoneurons. J. Neurophysiol. 27: 913-927.
* Phillips C.G., Powell T.P.S., and G.M. Shepherd. (1963) Responses of mitral
cells to stimulation of lateral olfactory tract in the rabbit. J. Physiol.
168: 65-88.
* Rall W. (1955a) A statistical theory of monosynaptic input-output relations.
J. Cell. Comp. Physiol. 46: 373-411.
* Rall W. (1955b) Experimental monosynaptic input-output relations in the
mammalian spinal cord. J. Cell. Comp. Physiol. 46: 413-437.
* Rall W., and C.C. Hunt (1956) Analysis of reflex variability in terms of
partially correlated excitability [82]fluctuation in a population of
motoneurons. J. Gen. Physiol. 39: 397-422.
* Rall W. (1957) Membrane time constant of motoneurons. Science 126: 454.
* Rall W. (1959) Branching dendritic trees and motoneuron membrane resistivity.
Exp. Neurol. 1: 491-527.
* Rall W. (1960) Membrane potential transients and membrane time constant of
motoneurons. Exp. Neurol. 2: 503-532.
* Rall W. (1962a) Theory of physiological properties of dendrites. Ann. N.Y.
Acad. Sci. 96: 1071-1092.
* Rall W. (1962b) Electrophysiology of a dendritic neuron model. Biophys. J.
2:145-167.
* Rall W. (1964) Theoretical significance of dendritic trees for neuronal
input-output relations. In Neural Theory and Modeling, ed. R.F. Reiss.
Stanford Univ. Press.
* Rall W., Shepherd G.M., Reese T.S., and M.W. Brightman. (1966)
Dendro-dendritic synaptic pathway for inhibition in the olfactory bulb.
Exptl. Neurol. 14:44-56.
* Rall W. (1967) Distinguishing theoretical synaptic potentials computed for
different soma-dendritic distributions of synaptic input. J. Neurophysiol.
30:1138-1168.
* Rall W., Burke R.E., Smith T.R., Nelson P.G., and K. Frank. (1967) Dendritic
location of synapses and possible mechanisms for the monosynaptic EPSP in
motoneurons. J. Neurophysiol. 30: 1169-1193.
* Rall W., and G.M. Shepherd. (1968) Theoretical reconstruction of field
potentials and dendro-dendritic synaptic interactions in olfactory bulb. J.
Neurophysiol. 31: 884-915.
* Rall W. (1969) Time constants and electrotonic length of membrane cylinders
and neurons. Biophys. J. 9: 1483-1508.
* Rall W., and J. Rinzel. (1971) Dendritic spine function and synaptic
attenuation calculations. Program and Abstracts Soc . Neurosci. First Annual
Mtg. p.64.
* Rall W., and J. Rinzel. (1973) Branch input resistance and steady attenuation
for input to one branch of a dendritic neuron model. Biophys. J. 13: 648-688.
* Rall W. (1977) Core conductor theory and cable properties of neurons. In
Kandel, E.R., J.M. Brookhardt, and V.M. Mountcastle eds. Handbook of
physiology, cellular biology of neurons. Bethesda, MD: American Physiological
Society; 39-97.
* Rall W., and I. Segev. (1987) Functional possibilities for synapses on
dendrites and dendritic spines. In Synaptic Function, ed. G.M. Edleman, W.E.
Gall, and W.M. Cowan. New York: Wiley; 605-636.
* Rall W. (1990) Perspectives on neuron modeling. In Binder M.D. and L.M.
Mendell, eds. The segmental motor system. New York: Oxford University Press.
* Rall W., Burke R.E., Holmes W.R., Jack J.J.B., Redman S.J., and I. Segev.
(1992) Matching dendritic neuron models to experimental data. Physiol. Rev.
72: S159-S186.
* Rall W. (2006) Chapter in "The History of Neuroscience in Autobiography Vol.
5". ed. L.R. Squire, Society for Neuroscience; Elsevier Academic Press.
* Redman S.J., and B. Walmsley. (1983) The time course of synaptic potentials
evoked in cat spinal motoneurones at identified group Ia synapses. J.
Physiol. 343: 117-133.
* Rinzel J., and W. Rall. (1974) Transient response in a dendritic neuron model
for current injected at one branch. Biophys. J. 14: 759-790.
* Segev I., and W. Rall. (1988) Computational study of an excitable dendritic
spine. J. Neurophysiol. 60: 499-523.
* Segev I., Rinzel J., and G.M. Shepherd. (1995) The Theoretical Foundation of
Dendritic Function. Selected papers of Wilfrid Rall, with commentaries.
M.I.T. Press
* Shepherd G.M., Brayton R.K., Miller J.P., Segev I., Rinzel J., and W. Rall.
(1985) Signal enhancement in distal cortical dendrites by means of
interaction between active dendritic spines. Proc. Nat. Acad. Sci. 82;
2192-2195.
Internal references
* Ian Gladwell (2008) [83]Boundary value problem. [84]Scholarpedia, 3(1):2853.
* Frances K. Skinner (2006) [85]Conductance-based models. Scholarpedia,
1(11):1408.
* Peter Jonas and Gyorgy Buzsaki (2007) [86]Neural inhibition. Scholarpedia,
2(9):3286.
* Rodolfo Llinas (2008) [87]Neuron. Scholarpedia, 3(8):1490.
* Ernst Niebur (2008) [88]Neuronal cable theory. Scholarpedia, 3(5):2674.
* Ted Carnevale (2007) [89]Neuron simulation environment. Scholarpedia,
2(6):1378.
* Andrei D. Polyanin, William E. Schiesser, Alexei I. Zhurov (2008) [90]Partial
differential equation. Scholarpedia, 3(10):4605.
* Robert E. Burke (2008) [91]Spinal cord. Scholarpedia, 3(4):1925.
* Arkady Pikovsky and Michael Rosenblum (2007) [92]Synchronization.
Scholarpedia, 2(12):1459.
See also
[93]Conductance-based models, [94]Dendritic processing , [95]Electrophysiology,
[96]Hodgkin-Huxley model, [97]Neuron, [98]Spines, [99]Synaptic transmision,
[100]Synapse
Sponsored by: [101]Dr. Wilfrid Rall, Scientist Emeritus of NIH; Present Address:
RR-1, Box 689, Roseland, VA, 22967
Sponsored by: [102]Eugene M. Izhikevich, Editor-in-Chief of Scholarpedia, the
peer-reviewed open-access encyclopedia
Sponsored by: [103]Frances K. Skinner, Toronto Western Research Institute,
University Health Network, and University of Toronto
[104]Reviewed by: [105]Anonymous
Accepted on: [106]2009-04-28 16:06:42 GMT
Retrieved from
"[107]http://www.scholarpedia.org/w/index.php?title=Rall_model&oldid=91692"
[108]Categories:
* [109]Computational neuroscience
* [110]Models of neurons
* [111]Eponymous
Personal tools
* [112]Log in
Namespaces
* [113]Page
* [114]Discussion
Variants
Views
* [115]Read
* [116]View source
* [117]View history
Actions
Search
____________________ (BUTTON) Search
Navigation
* [118]Main page
* [119]About
* [120]Propose a new article
* [121]Instructions for Authors
* [122]Random article
* [123]FAQs
* [124]Help
Focal areas
* [125]Astrophysics
* [126]Celestial mechanics
* [127]Computational neuroscience
* [128]Computational intelligence
* [129]Dynamical systems
* [130]Physics
* [131]Touch
* [132]More topics
Activity
* [133]Recently published articles
* [134]Recently sponsored articles
* [135]Recent changes
* [136]All articles
* [137]List all Curators
* [138]List all users
* [139]Scholarpedia Journal
Tools
* [140]What links here
* [141]Related changes
* [142]Special pages
* [143]Printable version
* [144]Permanent link
* [145][twitter.png?303]
* [146][gplus-16.png]
* [147][facebook.png?303]
* [148][linkedin.png?303]
* [149]Powered by MediaWiki [150]Powered by MathJax [151]Creative Commons
License
* This page was last modified on 21 October 2011, at 04:15.
* This page has been accessed 94,648 times.
* "Rall model" by [152]Wilfrid Rall is licensed under a [153]Creative Commons
Attribution-NonCommercial-ShareAlike 3.0 Unported License. Permissions beyond
the scope of this license are described in the [154]Terms of Use
* [155]Privacy policy
* [156]About Scholarpedia
* [157]Disclaimers
References
Visible links:
1. http://www.scholarpedia.org/w/opensearch_desc.php
2. http://www.scholarpedia.org/w/index.php?title=Special:RecentChanges&feed=atom
3. http://dx.doi.org/10.4249/scholarpedia.1369
4. http://www.scholarpedia.org/w/index.php?title=Rall_model&action=cite&rev=91692
5. http://www.scholarpedia.org/article/Rall_model#mw-head
6. http://www.scholarpedia.org/article/Rall_model#p-search
7. http://www.scholarpedia.org/article/User:Wilfrid_Rall
8. http://www.scholarpedia.org/article/User:Frances_K._Skinner
9. http://www.scholarpedia.org/article/User:Nick_Orbeck
10. http://www.scholarpedia.org/article/User:Tobias_Denninger
11. http://www.scholarpedia.org/article/User:Eugene_M._Izhikevich
12. http://www.scholarpedia.org/article/User:Wilfrid_Rall
13. http://www.scholarpedia.org/article/Neuron
14. http://www.scholarpedia.org/article/Neuroscience
15. http://www.scholarpedia.org/article/Neuron
16. http://www.scholarpedia.org/article/Spinal_cord
17. http://www.scholarpedia.org/article/Rall_model#Overview
18. http://www.scholarpedia.org/article/Rall_model#d3.2F2_Constraint_on_branch_diameters
19. http://www.scholarpedia.org/article/Rall_model#Equivalent_cylinder_model_and_cable_equation
20. http://www.scholarpedia.org/article/Rall_model#Compartmental_model
21. http://www.scholarpedia.org/article/Rall_model#Mitral_and_granule_cell_dendrites:_olfactory_bulb_model
22. http://www.scholarpedia.org/article/Rall_model#Dendro-dendritic_synapses_predicted_and_found
23. http://www.scholarpedia.org/article/Rall_model#Solutions_for_input_to_a_single_branch_of_an_idealized_dendritic_tree
24. http://www.scholarpedia.org/article/Rall_model#Dendritic_spine_model_and_clusters_of_spines_on_branches
25. http://www.scholarpedia.org/article/Rall_model#Motoneuron_population_models
26. http://www.scholarpedia.org/article/Rall_model#Dendrites_as_the_key_to_understanding_soma_voltage_transients
27. http://www.scholarpedia.org/article/Rall_model#Extracellular_transients_generated_by_a_soma_with_passive_dendrites
28. http://www.scholarpedia.org/article/Rall_model#Swartz_Prize.2C_2008
29. http://www.scholarpedia.org/article/Rall_model#References
30. http://www.scholarpedia.org/article/Rall_model#See_also
31. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig1.gif
32. http://www.scholarpedia.org/article/Rall_model#Eq-1
33. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig1.gif
34. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig1.gif
35. http://www.scholarpedia.org/article/Rall_model#Eq-2
36. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig2.gif
37. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig1.gif
38. http://www.scholarpedia.org/article/Partial_differential_equation
39. http://www.scholarpedia.org/article/Neural_Inhibition
40. http://www.scholarpedia.org/article/Rall_model#Solutions_for_input_to_a_single_branch_of_an_idealized_dendritic_tree
41. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig1.gif
42. http://www.scholarpedia.org/article/Rall_model#Mitral_and_granule_cell_dendrites:_olfactory_bulb_model
43. http://www.scholarpedia.org/article/Rall_model#Dendro-dendritic_synapses_predicted_and_found
44. http://www.scholarpedia.org/article/Rall_model#Dendritic_spine_model_and_clusters_of_spines_on_branches
45. http://www.scholarpedia.org/w/index.php?title=Membrane_potential&action=edit&redlink=1
46. http://www.scholarpedia.org/article/Partial_differential_equation
47. http://www.scholarpedia.org/article/Rall_model#Eq-2
48. http://www.scholarpedia.org/w/index.php?title=Separation_of_variables&action=edit&redlink=1
49. http://www.scholarpedia.org/article/Boundary_value_problem
50. http://www.scholarpedia.org/w/index.php?title=Eigenvalues&action=edit&redlink=1
51. http://www.scholarpedia.org/article/Boundary_value_problem
52. http://www.scholarpedia.org/w/index.php?title=Synaptic_Transmission&action=edit&redlink=1
53. http://www.scholarpedia.org/article/Rall_model#Eq-2
54. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig2.gif
55. http://www.scholarpedia.org/article/Rall_model#Eq-3
56. http://www.scholarpedia.org/article/Rall_model#Eq-2
57. http://www.scholarpedia.org/article/Rall_model#Eq-4
58. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig1.gif
59. http://www.scholarpedia.org/article/NEURON_Simulation_Environment
60. http://www.scholarpedia.org/article/GENESIS
61. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig3.gif
62. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig4.gif
63. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig3.gif
64. http://www.scholarpedia.org/w/index.php?title=Motoneurons&action=edit&redlink=1
65. http://www.scholarpedia.org/article/Spinal_cord
66. http://www.scholarpedia.org/article/Rall_model#fig:Rall_Fig5.gif
67. http://www.scholarpedia.org/article/Neuronal_Cable_Theory
68. http://www.scholarpedia.org/w/index.php?title=Electrophysiology&action=edit&redlink=1
69. http://www.scholarpedia.org/article/Local_Field_Potential
70. http://www.scholarpedia.org/w/index.php?title=Olfactory_bulb&action=edit&redlink=1
71. http://www.scholarpedia.org/article/Synchronization
72. http://www.scholarpedia.org/w/index.php?title=Mitral_cell&action=edit&redlink=1
73. http://www.scholarpedia.org/w/index.php?title=Granule_cells&action=edit&redlink=1
74. http://www.scholarpedia.org/w/index.php?title=Dendro-dendritic_synapses&action=edit&redlink=1
75. http://www.scholarpedia.org/article/Rall_model#Equivalent_cylinder_model_and_cable_equation
76. http://www.scholarpedia.org/w/index.php?title=Dendritic_spines&action=edit&redlink=1
77. http://www.scholarpedia.org/w/index.php?title=Synaptic_plasticity&action=edit&redlink=1
78. http://www.scholarpedia.org/article/Rall_model#Eq-5
79. http://www.scholarpedia.org/article/Rall_model#Dendritic_spine_model_and_clusters_of_spines_on_branches
80. http://www.scholarpedia.org/article/Encyclopedia_of_Computational_Neuroscience
81. http://www.scholarpedia.org/article/Brain
82. http://www.scholarpedia.org/article/Fluctuations
83. http://www.scholarpedia.org/article/Boundary_value_problem
84. http://www.scholarpedia.org/article/Scholarpedia
85. http://www.scholarpedia.org/article/Conductance-based_models
86. http://www.scholarpedia.org/article/Neural_inhibition
87. http://www.scholarpedia.org/article/Neuron
88. http://www.scholarpedia.org/article/Neuronal_cable_theory
89. http://www.scholarpedia.org/article/Neuron_simulation_environment
90. http://www.scholarpedia.org/article/Partial_differential_equation
91. http://www.scholarpedia.org/article/Spinal_cord
92. http://www.scholarpedia.org/article/Synchronization
93. http://www.scholarpedia.org/article/Conductance-based_models
94. http://www.scholarpedia.org/w/index.php?title=Dendritic_processing&action=edit&redlink=1
95. http://www.scholarpedia.org/w/index.php?title=Electrophysiology&action=edit&redlink=1
96. http://www.scholarpedia.org/w/index.php?title=Hodgkin-Huxley_model&action=edit&redlink=1
97. http://www.scholarpedia.org/article/Neuron
98. http://www.scholarpedia.org/w/index.php?title=Spines&action=edit&redlink=1
99. http://www.scholarpedia.org/w/index.php?title=Synaptic_transmision&action=edit&redlink=1
100. http://www.scholarpedia.org/w/index.php?title=Synapse&action=edit&redlink=1
101. http://www.scholarpedia.org/article/User:Wilfrid_Rall
102. http://www.scholarpedia.org/article/User:Eugene_M._Izhikevich
103. http://www.scholarpedia.org/article/User:Frances_K._Skinner
104. http://www.scholarpedia.org/w/index.php?title=Rall_model&oldid=61810
105. http://www.scholarpedia.org/article/User:Anonymous
106. http://www.scholarpedia.org/w/index.php?title=Rall_model&oldid=61818
107. http://www.scholarpedia.org/w/index.php?title=Rall_model&oldid=91692
108. http://www.scholarpedia.org/article/Special:Categories
109. http://www.scholarpedia.org/article/Category:Computational_neuroscience
110. http://www.scholarpedia.org/article/Category:Models_of_neurons
111. http://www.scholarpedia.org/article/Category:Eponymous
112. http://www.scholarpedia.org/w/index.php?title=Special:UserLogin&returnto=Rall+model
113. http://www.scholarpedia.org/article/Rall_model
114. http://www.scholarpedia.org/w/index.php?title=Talk:Rall_model&action=edit&redlink=1
115. http://www.scholarpedia.org/article/Rall_model
116. http://www.scholarpedia.org/w/index.php?title=Rall_model&action=edit
117. http://www.scholarpedia.org/w/index.php?title=Rall_model&action=history
118. http://www.scholarpedia.org/article/Main_Page
119. http://www.scholarpedia.org/article/Scholarpedia:About
120. http://www.scholarpedia.org/article/Special:ProposeArticle
121. http://www.scholarpedia.org/article/Scholarpedia:Instructions_for_Authors
122. http://www.scholarpedia.org/article/Special:Random
123. http://www.scholarpedia.org/article/Help:Frequently_Asked_Questions
124. http://www.scholarpedia.org/article/Scholarpedia:Help
125. http://www.scholarpedia.org/article/Encyclopedia:Astrophysics
126. http://www.scholarpedia.org/article/Encyclopedia:Celestial_Mechanics
127. http://www.scholarpedia.org/article/Encyclopedia:Computational_neuroscience
128. http://www.scholarpedia.org/article/Encyclopedia:Computational_intelligence
129. http://www.scholarpedia.org/article/Encyclopedia:Dynamical_systems
130. http://www.scholarpedia.org/article/Encyclopedia:Physics
131. http://www.scholarpedia.org/article/Encyclopedia:Touch
132. http://www.scholarpedia.org/article/Scholarpedia:Topics
133. http://www.scholarpedia.org/article/Special:RecentlyPublished
134. http://www.scholarpedia.org/article/Special:RecentlySponsored
135. http://www.scholarpedia.org/article/Special:RecentChanges
136. http://www.scholarpedia.org/article/Special:AllPages
137. http://www.scholarpedia.org/article/Special:ListCurators
138. http://www.scholarpedia.org/article/Special:ListUsers
139. http://www.scholarpedia.org/article/Special:Journal
140. http://www.scholarpedia.org/article/Special:WhatLinksHere/Rall_model
141. http://www.scholarpedia.org/article/Special:RecentChangesLinked/Rall_model
142. http://www.scholarpedia.org/article/Special:SpecialPages
143. http://www.scholarpedia.org/w/index.php?title=Rall_model&printable=yes
144. http://www.scholarpedia.org/w/index.php?title=Rall_model&oldid=91692
145. https://twitter.com/scholarpedia
146. https://plus.google.com/112873162496270574424
147. http://www.facebook.com/Scholarpedia
148. http://www.linkedin.com/groups/Scholarpedia-4647975/about
149. http://www.mediawiki.org/
150. http://www.mathjax.org/
151. http://creativecommons.org/licenses/by-nc-sa/3.0/deed.en_US
152. http://www.scholarpedia.org/article/Rall_model
153. http://creativecommons.org/licenses/by-nc-sa/3.0/deed.en_US
154. http://www.scholarpedia.org/article/Scholarpedia:Terms_of_use
155. http://www.scholarpedia.org/article/Scholarpedia:Privacy_policy
156. http://www.scholarpedia.org/article/Scholarpedia:About
157. http://www.scholarpedia.org/article/Scholarpedia:General_disclaimer
Hidden links:
159. http://www.scholarpedia.org/article/File:Rall_Fig1.gif
160. http://www.scholarpedia.org/article/File:Rall_Fig1.gif
161. http://www.scholarpedia.org/article/File:Rall_Fig2.gif
162. http://www.scholarpedia.org/article/File:Rall_Fig2.gif
163. http://www.scholarpedia.org/article/File:Rall_Fig3.gif
164. http://www.scholarpedia.org/article/File:Rall_Fig3.gif
165. http://www.scholarpedia.org/article/File:Rall_Fig4.gif
166. http://www.scholarpedia.org/article/File:Rall_Fig4.gif
167. http://www.scholarpedia.org/article/File:Rall_Fig5.gif
168. http://www.scholarpedia.org/article/File:Rall_Fig5.gif
169. http://www.scholarpedia.org/article/Rall_model
170. http://www.scholarpedia.org/article/Rall_model
171. http://www.scholarpedia.org/article/Main_Page
Usage: http://www.kk-software.de/kklynxview/get/URL
e.g. http://www.kk-software.de/kklynxview/get/http://www.kk-software.de
Errormessages are in German, sorry ;-)