Using uLEDs to perform a micro-lesion for probe localization?

[mattm...@gmail.com]

We would like to improve our ability to localize reocrding sites in our sectioned brain tissue. We found DiI to not be precise enough, and doing electrical lesions is challenging because unplugging the recording headstage in our acute head-fixed recording risks moving the probe. So, we wondered, does it seem feasible to use heat from the uLEDs to create a lesion? Has anyone tried this? Grateful for thoughts/suggestions! -- Matt

[Liset M de la Prida]

Hi Matt

We don't use to mark them because in the dorsal hippocampus localization is validated by recording. If you are recording from elsewhere, you can paint them at the back with fluorescent DiI. We do this regularly for other probes such as Neuropixels and it works well. If you are good with histology, you can even localized the tracks without marking them, which is 'very easy' if you implant them chornically. Also doable if you are going acutely.

We do not recommend using current to generate lesions with these electrodes, because the charge transfer will not be enough and you can damage them. 

If you have any additional comment let us know.  I guess you are aware on the Plexon webinar https://plexon.com/news-and-events/neurolight-webinar-series/

Regards

Liset

http://hippo-circuitlab.es/

[Dan English]

The current required to produce heat with the LEDs will burn them out far before creating a lesion. If the DiI/tracks are not enough, and you really want to label where the uLED was stimulating, consider a photo-convertible fluorescent protein such as Eos or CaMPARI

[mattm...@gmail.com]

Thank you both for the informative replies, Liset and Dan! A few more thoughts:

If you are recording from elsewhere, you can paint them at the back with fluorescent DiI

Yes, we are recording acutely in some tiny brainstem nuclei, so we lack the nice electrophysiological signatures the hippocampus so conveniently gives us! Maybe we need to spend more time calibrating the right DiI amount, but so far we found that if we apply enough to be able to see the (hypothesized) tip of the probe, the Di diffuses laterally several 100um, which is too much for the precision we need. If we apply less Di, then it seems like there isn't enough to make it all the way down to the recording site. We then see a fluorescent track but only part of the way down, as if the Di wears as the probe is lowered.

So we figured we have to use a different approach like a micro-lesion.

We do not recommend using current to generate lesions with these electrodes, because the charge transfer will not be enough and you can damage them. 

We were encouraged to see this NeuroNexus microlesioning guide, did that not work for you, Liset? We are working on making a flexible cable that would allow us to unplug the recording headstage to gain access for the microlesion procedure without moving the probe from the exertion forces.

The current required to produce heat with the LEDs will burn them out far before creating a lesion. 

Too bad, thanks for confirming.

If the DiI/tracks are not enough, and you really want to label where the uLED was stimulating, consider a photo-convertible fluorescent protein such as Eos or CaMPARI

Yes, that's a promising idea. But of course we need the uLEDs for optotagging too... maybe the next generation dual-wavelength uLEDs will take care of that!

Thanks again, and if any other avenues come to mind I'd be very grateful if you could let me know!

[Liset M de la Prida]

With the DiI you will not be able to identify where the tip was precisely. Only the track. 

It would be great that microlesions work, but you typically need to pass a high current to produce a visible mark which is not good for the probe. I suggest discussing this with people at Neurolight Tech to evaluate the limits. I guess this is very challenging.  

Thank you for sharing the NN lesioning guide. Personally if you are able to produce them I would consider running a more patterned array for unambiguous identification.

For CaMPARI, etc... I am not so optimistic. First, you need precise wavelenghts that may not be targeted by the uLED emission spectra. Second, because it is based on inducing calcium activity the reporter is not necessarily reflecting the light spot but the activity cascade, which could be even activated from dendritic locations. Of course, it would be great to give it a try.  Some people have tried to use cfos IH but it was not so straitforward neither.

The problem is still calling for solutions.

[Christos Lisgaras]

Hi everyone, 

Any suggestions on Dil's concentration? I have heard people diluting 0.5% Dil in DMSO or ethanol, but wondered what is your experience. Thank you in advance!

Christos

Scharfman Lab

[Liset M de la Prida]

Hi Christos

We do not dilute DiI. Basically use a some micropipette to handle a drop of DiI. Gently bring the drop to the back of each shank and drag the drop as if you were painting. Be very careful not to hit the shank. We do this under a scope. The probe is fixed to a manipulator or something. Use your good hand to hold the drop and the other hand to suport your good hand to avoid shaking. 

You only need one or two passes with the drop to have your probe painted (learn to note the reflex). Let it dry (like 5 min) and then you are ready. 

Liset

[Christos Lisgaras]

Thank you, Liset! Really helpful!

Christos 

[mvoeroes]

Hi Matt,

I think you can try two different solutions:

1. Solder a custom connector with male 36-pin Omnetics on one side and 2 female 36-pin Omnetics on other side. One female connector can be used to plug in your headstage and the other one for micro lesion after your experiments. The down side of this approach that it can introduce noise, so you need to have a good Faraday cage/EMI protection around your setup, and it takes forever to solder this adapter. I have used such a thing in the past, so it is definitely doable.

2. I have talked with Nate about this option and they are willing to do some custom modifications (see in the attached picture from Cambridge Neurotech). The idea is that they move out n (number of omnetics pin, hopefully on different shanks, but you can define) number of pin and solder cables to them before they cover it with epoxy. During recording, you should gnd them, not to introduce noise and at the end of the recording you can apply current through them. Unfortunately, you will loose recording channels.

Best,

Misi