graphs in extracelllular stimulation

[shwetakgp]

can anybody suggest me where can i find the morphology any primate retinal ganglion cell? like diameter of soma, number of dendrites etc..
nd i also wanted to know that when we create a soma or axon in cell through cell builder is that mylinated?

[shwetakgp]

dear ted,
i sent an email about my problem but i have got it now so i aplogise for sending that mail. I wanted to know that i created subtrees of a dendrite using both hoc code and gui but when i see my shape plot, dendrites are shown by straight lines but in some other files that i saw like sheasbyfohlmeister1999.zip they are shown by curvey lines in the shape plot . Is there anything wrong with my code? sorry for this silly question

[ted]

To make a disc electrode i thought that i should start with a point and take a current density sigma. taking the point, i can assume a disc of radius r.
for calculating the current i could integrate it taking a ring of thickness dr as the element of integration. . . .

Are you assuming that the cell lies near the axis of the electrode? If yes, do its dendritites lie approximately in a plane parallel to the plane of the electrode? If yes again, then I would guess your model cell should include an axon that exits the plane of the dendrites and extends for some distance.

I'd be surprised if there isn't already a published literature about the field produced in tissue by specific electrode geometries. If the distance between the electrode and the cell is large compared to the size of the electrode (at least 5 x diameter), you might be able to get decent approximate results by treating the electrode as a point source of current.

[shwetakgp]

for the disc elecrode you saw my model , my soma and axon are collinear to the point electrode.So how should i do it?. Do you mean to say that atleast one section should not be in a parllel plane. actually i have to do the simulation just like the experiment is going to be performed and a disc electrode has been used so i have to put a disc electrode. But if you are saying that at a large distance the effects of disc and point electrode are same then i could take electrode as a point but what distance i should take??

[shwetakgp]

dear ted,
i was able to simualte a pulse train by vector clsaa's play method .It shows a spike with the potential reaching 20 to 30 but but the problem is that before showing a spike it first decreases from resting potential to -90 something. Is there something wrong?
i read your mail about using neuromorpho but that paper from which the slamander values have been taken is not accessible to me. i am basically concerned about the dendritic tree which i am not able to create

[ted]

i was able to simualte a pulse train by vector clsaa's play method .It shows a spike with the potential reaching 20 to 30 but but the problem is that before showing a spike it first decreases from resting potential to -90 something. Is there something wrong?

Maybe. What happens if there is no stimulus?

i read your mail about using neuromorpho but that paper from which the slamander values have been taken is not accessible to me.

Are you sure? R.F. Miller and J.F. Fohlmeister have publised several articles in Journal of Neurophysiology, a journal that makes articles freely available 1 year after publication. A search of Pubmed turns up many such articles
http://www.ncbi.nlm.nih.gov/pubmed/?ter ... ter+miller
http://www.ncbi.nlm.nih.gov/pubmed/?ter ... der+retina
Sheasby, Brent W. and Jurgen F. Fohlmeister. Impulse encoding across the dendritic morphologies of retinal ganglion cells. J.Neurophysiol. 81: 1685-1698, 1999.
will also be available from that journal's web site.

[ted]

when i see my shape plot, dendrites are shown by straight lines but in some other files that i saw like sheasbyfohlmeister1999.zip they are shown by curvey lines in the shape plot

Real cells have curves. Some models of real cells don't bother to preserve curvature--they just represent each neurite as a straight branch; this is typical of model cells created by drawing branches with the CellBuilder. It won't make any difference to the electrical properties of the neuron itself, as long as lengths, diameters, and membrane and cytoplasm properties are the same as in the biological originals. However, the difference in shape will affect the response of the model to an external electrical field.

[shwetakgp]

i wanted to create a cell from shaesbyfohlmeister 1999 by GUI.I am facing few problems
1) They have not mentioned the length and diameter of any of the dendritic subtrees!!
2) in some cells values are like 36 subtrees for one dendrite,18 for other and 2 or 3 for other. its getting crowded!!!
and there is an orientation in which the dendrites are kept which i am not able to make and which can change the overall shape of the cell maybe . even if i make some of dendrites ,some of them are downwards and some upwards. i cnt match their orientation. Thus my cell will never be like that cell

[shwetakgp]

and for the stimulus this is what happens. after a spike is achieved nothing happens to its peak voltage for some values of amplitude, when i change the values of amplitude further the spike peak starts dipping i.e peak voltage starts dipping.... then on furthur increasing the amplitude the peak goeas into negative ..>90 .maybe i am not explaining it in a good way. I wanted to upload a picture of what is happening exactly but i am not able to

[shwetakgp]

dear ted,
i still dont have a clue about how to make the cell without the length of dendrites . and i still cnt find it!!!!

[ted]

for the disc elecrode you saw my model , my soma and axon are collinear to the point electrode.

An extracellular field won't stimulate a cell if the extracellular potential is nearly identical over the surface of the cell. If a cell's dendritic branches all lie in a plane that is parallel to the plane of a disc electrode, and the axis of the disc passes through the "center" of the cell's dendritic field, the extracellular potential gradient will be nearly orthogonal to each branch of the cell and the stimulus will be relatively ineffective. An electrode on the "front" of the eye (cornea or sclera) will be nearly parallel to ganglion cells at the opposite pole ("back") of the eye, so any stimulus those cells experience will result mostly from the extension of their axons past the optic disk because that will be the part of the cell that spans a wide range of extracellular potentials.

[ted]

i wanted to create a cell from shaesbyfohlmeister 1999 by GUI.

The GUI's tools don't give you the ability to specify the detailed trajectory of any branch, and that means you can't just "draw" the shape of a cell and expect it to look like a real cell or respond to an extracellular field in the same way a real cell does. Besides, the GUI is very poorly suited for constructing topologies that involve more than about a half a dozen branches--it's just way too tedious. You're stuck with having to use detailed morphometric data like a Neurolucida or Eutectic reconstruction of a cell.

The Sheasby & Fohlmeister files in ModelDB create a wonderful user interface, but if you want to use any one of those cells you're going to have to unpack its morphology from all that elegant code. And you're going to have to make sure that it has exactly the right biophysical properties. And you're going to have to do something about the axon, initial segment, and something the authors call narrowr, which are specified in terms of L and diam--because the authors probably used "average" or "typical" values to specify the properties of these sections, and most likely didn't have actual x,y,z,diam measurements from any particular cell's axon. Therefore the axon specification contains no useful information about orientation--and as I indicated in a previous message today, the efficacy of extracellular stimulation depends strongly on the orientation of neurites relative to the extracellular field.

That's why I suggested looking at morphologies in NeuroMorpho, where you might be lucky enough to find one or more ganglion cells that include axonal data.

[shwetakgp]

dear ted,
if i have a the value of area of the dendritic field . Is it okey if i draw a circular section of area equal to dendritic field area connected to the soma at one side?????? or a circular segment equal the area of soma plus dendritic field area??. I could consider the all the dendrites are within that area and whatever effect all the dendrites together will give a similar effect could be given by the circular segment.please reply asap

i did run my code using soma plus dendritic field area but when i give a stimulus of suppose duration= 1 the potential of soma goes to -100 or even lower after achieving a spike i am sending my code and an image showing what my issue is...please reply asap..thank you

[ted]

dear ted,
if i have a the value of area of the dendritic field . Is it okey if i draw a circular section of area equal to dendritic field area connected to the soma at one side?????? or a circular segment equal the area of soma plus dendritic field area??

I don't think so. Extracellular stimulation isn't just about surface area, it's also about branch length, diameter, and orientation relative to the field.

[shwetakgp]

dear ted,
the values of conductances of the ion channels which are already present in the cell builder, are they the default values of the conductance??? if i use those values do i have to mention any reference ... if yes then what...in every paper i get the value of conductance of na,k and leakage channel but the cell builder wants maximum conductance. what should be the value to max conductance if conductance is say 70 ms/ohm2
and if i make intial, narrower and axon in cell builder joined side by side. would that work if not then how to make them?
and my response of axon is that the i threshold is same as soma but the peak at soma is 3 to 4 seconds before axon..is it okey??