Kv3p4_RI2021_SC

Kv3p4_RI2021_SC#

class braincell.channel.Kv3p4_RI2021_SC(size, g_max=Quantity(4., 'mS / cm^2'), temp=Quantity(295.15, 'K'), name=None)#

Fast TEA-sensitive potassium current of the stellate cell model.

High-threshold, fast-activating and only partially inactivating potassium current, imported from the cerebellar stellate cell model of Rizza et al. (2021) [2].

Gating is an m gate of power 3 and an h gate of power 1, both written in steady-state/time-constant form and both evaluated at a junction-potential-corrected voltage \(V' = V + 11\):

\[\begin{split}\begin{aligned} m_\infty &= \frac{1}{1 + \exp(-(V' + 24) / 15.4)} \\ h_\infty &= 0.31 + \frac{0.69}{1 + \exp((V' + 5.802) / 11.2)} \end{aligned}\end{split}\]

where \(V'\) is in millivolts. Inactivation is only partial: \(h_\infty\) decays to a floor of 0.31 rather than to zero, so about 31 per cent of the conductance survives a maintained depolarisation. Both time constants are piecewise, in milliseconds:

\[\begin{split}\tau_m = 10^3 \times \begin{cases} 3 \, (3.4225 \times 10^{-5} + 0.00498 \, e^{V' / 28.29}), & V' < -35 \\ 1.2851 \times 10^{-4} + \dfrac{1}{e^{(V' + 100.7)/12.9} + e^{(V' - 56)/(-23.1)}}, & V' \ge -35 \end{cases}\end{split}\]
\[\begin{split}\tau_h = 10^3 \times \begin{cases} 0.0012 + 0.0023 \, e^{-0.141 \, V'}, & V' > 0 \\ 1.2202 \times 10^{-5} + 0.012 \, e^{-((V' + 56.3)/49.6)^2}, & V' \le 0 \end{cases}\end{split}\]

Each ladder is selected with u.math.where on the predicates V' < -35 and V' > 0, so the boundary value itself falls to the second line in both.

Parameters:

  • size (int | Sequence[int] | integer | Sequence[integer]) – Channel state shape.

  • g_max (Array | ndarray | bool | number | bool | int | float | complex | Quantity | Callable) – Maximal conductance density. Defaults to 4.0 mS/cm2, which is exactly the source mechanism’s gkbar = 0.004 mho/cm2.

  • temp (Array | ndarray | bool | number | bool | int | float | complex | Quantity) – Absolute temperature driving both gates’ q10 factor, default 22 degrees Celsius. See the temperature note below: this default is BrainCell’s own, not a value the source mechanism states.

  • name (str | None) – Optional channel name.

See also

Kv3p4_MA2020_GoC

Golgi-cell port of the same mechanism.

Kv3p4_MA2020_GrC

Granule-cell port of the same mechanism.

Kv3p4_MA2024_PC

Purkinje-cell port of the same mechanism.

Kv3p4_MA2025_BC

Basket-cell port of the same mechanism.

Notes

Ported from SC/channel/Kv3p4_RI21_SC.mod. All five cell-type ports of this mechanism carry the same equations and the same constants, and so do the five BrainCell classes: everything above is shared verbatim with Kv3p4_MA2020_GoC, Kv3p4_MA2020_GrC, Kv3p4_MA2024_PC and Kv3p4_MA2025_BC. What differs is only the deposit each was imported from, and therefore the model paper cited below.

The SC file additionally declares a g_equiv RANGE variable and assigns g_equiv = gkbar * m^3 * h in its BREAKPOINT. That is a diagnostic output for the NEURON comparison harness rather than part of the dynamics, and BrainCell does not port it.

Where the q10 factor is applied. Both gates declare q10 = 3.0 at a reference of 37 degrees Celsius, so HH.compute_derivative() scales each \((x_\infty - x)/\tau_x\) term by \(\phi = 3^{(T - 37)/10}\). The .mod file instead divides its mtau and htau by the same factor. The two forms are algebraically identical, but it means f_m_tau() and f_h_tau() return q10-free time constants rather than the mechanism’s mtau and htau.

The default temperature is BrainCell’s, not the mechanism’s. Unlike most channels in this module, this .mod file never declares a celsius value; it reads NEURON’s global instead. The temp default of 22 degrees Celsius is therefore a BrainCell choice, and because it sits 15 degrees below the gates’ reference it makes the default temperature factor \(\phi \approx 0.192\) – gating roughly 5.2 times slower than at 37 degrees Celsius. Callers reproducing a published simulation should set temp explicitly.

Provenance, and a name the sources do not support. These files are ModelDB accession 48332’s kpkj.mod renamed. The header lines “: HH TEA-sensitive Purkinje potassium current” and “: Created 8/5/02 - nwg” are the deposit’s own, “nwg” being Nathan W. Gouwens, second author of Khaliq et al. (2003) [1], and the parameters mivh = -24 mV, mik = 15.4 and hiy0 = 0.31 reproduce that paper’s Table 1 row for the current it calls K fast. The trailing : Suffix from kpkj to Kv3_4 line is a BrainCell-local addition and is not in the deposit. The model paper [2] names the deposit this parameterisation was imported from, not the origin of the equations.

That rename line is the whole basis of the Kv3p4 name, and it does not survive contact with the source. Khaliq et al. call this current K fast throughout and never name a Kv subunit; the strings “Kv3.4”, “Kv3.3” and “Kv3.1” appear nowhere in the paper, and its only mention of Kv3 at all is one Discussion sentence observing that the positive activation range is typical of the Kv3 family. The class name therefore asserts a subunit identity the cited literature does not establish. Under this project’s mismatch policy the name is documented rather than changed: read Kv3p4 as a label for the TEA-sensitive fast K current of Khaliq et al. (2003), which those authors associate with the Kv3 family, and not as a claim about Kv3.4.

Conductance default. 4.0 mS/cm2 is the deposit’s tuned value, carried across from the .mod file. It is not a value printed in either cited paper.

Import deviations: none. This mechanism carries no NMODL TABLE and was already integrated with cnexp upstream, so neither the table-removal nor the derivimplicit substitution recorded for other channels in this model applies here. Do not assume it shares the deviations of its Kv4p3 neighbour, which has three.

References

root_type#

alias of Potassium