General procedures 130718 - BrownBlogs · AizenmanLab!Procedures.!Rev.!!Jul18!2013! Page1!of!37!!...

Aizenman Lab Procedures. Rev. Jul 18 2013

of 37

Aizenman Lab Procedures Arranged by Arseny S. Khakhalin

Introduction Experimental science is based on superstitions. Nothing is proven, nothing is evidence-‐based; we follow old protocols, not knowing which parts of them are essential, and which are a bogus, in a hope that the experiments would work again. A lot of these "protocols" are transmitted by a word of mouth, as a local lore of the place. And of course, every scientist has their own unique approaches, rituals and rites. Some of these may be traced back to their scientific heritage, to the ungrounded advices they got from their former masters in other labs. And some of these unique quirks are introduced de novo, in a desperate attempt to improve the protocols, following the intuition, and the half-‐baked knowledge of the books.

And still, generally, it is right and worthy to follow the protocols, even when they got somewhat cryptic and obscure. The benefit here is that your results stay compatible with that of your colleagues and predecessors. By altering the protocols without guidance and control you introduce the noise, which may lead you astray and render your results non-‐interpretable.

This document summarizes some of these folk tales, rituals and traditions follow in the Aizenman lab, to help the newcomers on their way of becoming successful neuroscientists.

Table of Contents

Introduction ....................................................................................................................................................... 1

Solution Recipes ................................................................................................................................................ 2

Device Manuals ............................................................................................................................................... 10

Electrophysiology protocols ............................................................................................................................. 13

Staining protocols ............................................................................................................................................ 26

Behavioral protocols ........................................................................................................................................ 27

Miscellaneous Pieces of Practical Wisdom ....................................................................................................... 37


of 37

Solution Recipes

External (ACSF)

Recipe All components of the External Solution (ACSF) fall into 2 groups. The majority of components are stable, and so you can add them to deionized water, and store the solution in the fridge (at 2°C). Some of the components would however precipitate if you try to store them for long, so you have to add them to the solution immediately before the experiment. First, here’s the stable stuff:

Substance Concentration, mM g/l NaCl 115 6.7206 KCl 4 0.298 HEPES 5 1.191 Glycine 10 μM 1ml of 10 mM stock Glucose 10 1.802 In case of Pyruvate-‐based solution, the last row (Glucose) is to be replaced with a 50/50 mixture:

Glucose 5 0.901 Pyruvate 5 0.550 And second, immediately before the experiment you also add proportional amount of the following unstable components:

CaCl2 3 3ml of 1M stock MgCl2 3 3ml of 1M stock pH = 7.2; Osmolarity = 250

Alternative ionic concentrations:

• To hyperpolarize all cells slightly (may decrease total N spikes, but improves spike-‐timing): 6 Mg, 3 Ca. • To depolarize the cells (hopefully to somewhat deactivate Na channels): 6 K, 1.5 Mg.

How to prepare the solution Here’s some general tactics for the solution preparation:

1. Take deionized water, about ~80-‐85% of the final solution volume, and mix the stable reagents in; 2. Bring pH to the target by carefully adding suitable bases (or acids). For the external a suitable pare is obviously

NaOH/HCl, as external is Na-‐based, while for the typical internal you would use KOH/KCl instead. 3. Bring osmolarity to the target by adding di water. You’d better do it in 2-‐3 iterations, in order not to overshoot

(as all measurements are noisy, including that from the osmometer). To be on a safe side, when the calculation tells you that X ml of distilled water is to be added to bring osmolarity to the target, add only 80% of this volume. This way you’ll approach the target slowly, and the ultimate result will be more precise.

Detailed instructions for the pHmeter and osmometer are provided below. With amounts shown in the tables above the final solution volume may be slightly (2-‐5%) less than the target, but the osmolarity should be correct.


of 37

Use a 1 liter beaker with a magnetic stirrer working at ~300 rpm ongoing. Use big hexagonal weighboats for NaCl, and square pieces of waxed paper for everything else. To transfer the stock solutions, choose a proper pipetter; the color of the pipetter button corresponds to the color of conical tips to use. When pressing the button, the 1st stop measures the volume you need, but if you press harder there will be the 2nd stop that empties the pipette entirely, and if you press even harder – there will be yet the 3d stop that ejects the tip. Beware that the scales take long time to switch on for the 1st time, and they can’t zero the TARE until the weight is really stable (e.g. when the window is open they cannot do that, as the wind from the ventilation system seems to be strong enough to change the “weight” at the scales; they also seem to be sensitive to your steps if you are jumping around). So be patient. Note also that while all the beakers have volume marks on them, these marks are actually ridiculously incorrect. All volumes are to be measured with the measuring cylinders, and only with them.

After the osmolarity is measured, and the stock is brought to its final volume, it should be filtered with a vacuum-‐driven huge orange disposable filter right into a bottle where it will be stored. Label the bottle, indicating the owner, the date, and the solution type.

As mentioned above, immediately before the experiment, Ca and Mg are to be added to the ACSF from stock solutions. Make a point of always adding Ca and Mg in the same order, so that even if you are sleepy or get distracted, and forget what you were doing midway in the process, you would always know which of the salts you have added already, and which is still missing. Dr. Aizenman always adds MgCl2 first, and CaCl2 second; some other people always do Ca first and Mg second (which is also incidentally the alphabetical order), but the point is in always doing it the same way. No need for guessing or hectic recollection. If you are sure that you have added some ion already, you'll always know which one it was.

Internal Solution

Recipe For the internal, these substances1 are to be added to the stock from the beginning:

What Molarity, mM g / 100 ml of solution K-‐gluconate 100 2.3430 KCl 8 0.0596 NaCl 5 0.0292 MgCl2 ·∙ 6H2O 1.5 0.0305 HEPES 20 0.4766 EGTA 10 0.3804 Target pH = 7.2; Osm = 250 (would be ~255 after ATP+GTP are added)

And these 2 substances, being unstable, are prepared and added separately, as it is described below:

ATP 2 11 mg / 10 ml GTP 0.3 1.57 mg / 10 ml

1 K. G. Pratt, W. Dong, and C. D. Aizenman, "Development and Spike Timing-‐Dependent Plasticity of Recurrent Excitation in the Xenopus Optic Tectum," Nat Neurosci 11, no. 4 (2008).


of 37

Procedure 1. We first try to make 100 ml or internal solution without ATP and GTP. Start with ~70 ml of water, and add all the

required drugs. 2. Drive the solution pH to the target (7.2), using KOH as a base (HCl in case of overshoot). 3. Measure osmolarity, and add water to make the stock 250 ± 5 Osm. We should usually end up with 90+ ml of

solution. 4. Aliquot in 10 tubes, 10 ml each, using a syringe with a disk-‐shaped filter tip (draw without a filter, expel through

the filter). All tubes but one are to be frozen at −20°C. 5. Measure ATP and GTP in amounts due for 10 ml of solution, and add them to the single tube that would go into

further processing. Use small hexagonal weighload, and when the desired weight is achieved, take 200-‐300 μl of solution from a tube with a micropipette, and move it to the weighload; dissolve the grains, and move the solution back. To dissolve the grains quicker, you may try to suck and expel the liquid through the tip several times. If you have problems with measuring the required weight precisely, you may also weigh slightly more, then move some volume as described, and then move back only part of the volume, proportional to the ratio of desired weight to the actual weight, thus ensuring correct concentration in the solution.

6. Aliquot this full solution into 20 2.5 ml conical tubes, .5 ml in each. 19 of them go into deep freeze at −80°C, while one goes into experiment. The fridge is located in the lab nearby.

7. To unfreeze a full stock .5 ml tube, just take it from the −80°C fridge, and let it thaw. To unfreeze the 10 ml of ATP/GTP-‐less stock, keep it in hot water for ~20 min, applying vortex shake from time to time.

Predicted membrane potentials for these ACSF and Internal EK -‐102 mV ENa 72 mV ECl -‐53 mV Junction Potential 14 mV

Cs Internal Per 100 ml:

CsMethaneSulfonate 80 mM * 1825 mg MgCl2 5 mM 101.7 TEA 20 mM 331.4 EGTA 10 mM 380.4 HEPES 20 mM 476.6 ATP 2 mM 11 mg / 10 ml GTP 0.3 mM 1.57 mg / 10 ml To be brought to pH 7.2 by CsOH.

* Supposedly after you add CsOH, the concentration of Cs reaches about 90 mM.


of 37

Internal Solution Flowchart

70 ml H2O Reagents

70+ ml solution

~90 ml solution

10 ml aliquotes

pH + Osm

ATP & GDP for 10 ml

10 ml full solution

x 10

9 go into −20°C

20 min in hot water

.5 ml aliquotes x 20

19 go into −80°C

Experiment!

10 min in the pocket

Filter

Filter


of 37

Additional Components

Picrotoxin PTX is dissolved in DMSO, and the final concentration in the external solution is 0.1 mM. So that you could add it in the middle of an experiment, as a single drop per total amount of 10 ml in the chamber, you should prepare a 1000 stock solution, and add 10 µl of it. The molecular weight is 602.6 g/mol, which is prepared as 30 mg per 500 µl of solution, which are then aliquoted in equal nearly-‐10 µl portions in small Eppendorf tubes. To be stored in the freezer. To be centrifuged before use (DMSO is not willing to stay at the bottom of the tube on its own, so it worth forcing it going there). LD50 = 15 mg/kg (rat), so a deadly dose would be about 1200 full chambers, or 25 pre-‐aliquoted stock vials.

Weight Volume Concentration 603 g (mol. weight) 1 l 1 M 0.603 mg (chamber) 10 ml 0.1 mM (target) 30 mg 500 µl 100 mM (1000 stock) 0.603 10 µl (transfer dose) 100 mM

Tubocurarin (TBX) TBX is dissolved in water, and the final concentration is in the range of 0.05-‐0.2 mM (Heng used 0.1 mM in his experiments). In some cases the tadpoles continue to move even after prolonged exposure to the drug. We prepare 1 ml of 100 stock, and then store it in a 1.5 ml Eppendorf tube. The stock solution can be stored in a fridge (not necessarily in a freezer). LD50 = 18 mg/kg (cat), so a deadly dose would be about 1300 chambers (or 130 stock vials).

Weight Volume Concentration 681 g (mol. weight) 1 l 1 M 0.681 mg (chamber) 10 ml 0.1 mM (target) 6.81 mg 1 ml 10 mM (100x stock) 0.681 100 µl (transfer dose) 10 mM alternatively: 13.62 mg 1 ml 20 mM (200x stock)

Bungarotoxin (BgTX) It is a peptide, and thus needs to be stored in a freezer. Two modes of using it are described in literature: one is to keep about 100 nM (really low) concentration in the chamber; the other (and more frequently used) is to briefly (for ~5 min) dip the animal into the ~0.1 mM (1mg/ml) solution. LD50 is about 0.12 mg/kg, so one would die after drinking about 1200 chambers2.

Weight Volume Concentration 7984 g (mol. weight) 1 l 1 M 1 mg 1253 µl 0.1 mM (stock) 10 µl (transfer doze) 10 ml (chamber) 1000 nM (final)

2 Protection against alpha-‐bungarotoxin poisoning by immunization with synthetic toxin peptides. Dolimbek BZ, Atassi MZ. Mol Immunol. 1996 May-‐Jun;33(7-‐8):681-‐9.


of 37

Pancuronium In their experiments, Kurt Haas and his team dip tadpoles for 7 min in the 4 mM solution of Pancuronium3. When pancuronium is added to the bath, it seems to screw up (inhibit) the glutamatergic transmission, which seems to be the reason for this "topical" (pre-‐treatment) approach to be described in the literature.

TEA TEA (tetraethylammonium) blocks K+ channels, and should be added in 1-‐2 mM concentrations4 to the external solution. In Zebrafish in some publications up to 5 mM concentrations were used5, but for depolarizing neurons rather than simply blocking K+ currents. Note that in some studies they also put TEA inside the pipette6 (to block positive shoulder of the rectifying ions?). TEA+Cl− is dissolved in water; Tocris says that 100 mM is the official maximum solubility, but they seem to be lying, as the substance is so hydroscopic that it needs to be sealed with Parafilm (and still turns to stone despite being sealed), and also seems to dissolve really quickly without any effort. LD50 = 2630 mg/kg (rat), so a lethal dose would be about 60 000 chambers or 10 000 vials.

Weight Volume Concentration 165.7 g (mol. weight) 1 l 1 M 3.314 mg (chamber) 10 ml 2 mM (target) 16.57 mg 1 ml 100 mM (stock) 3.314 mg 200 µl (transfer dose) 100 mM

APV The desired concentration of D-‐APV is 50 µM (some people use 100 µM, but that’s if they use a racemic mixture of both potent D and inactive L forms7). Dissolves in dH2O to concentrations up to 100 mM, but hardly more than that, and requires excessive shaking. NMDA antagonists are known to induce a lot weird effects from nausea to acute psychosis an schizophrenia-‐like symptoms. Lethal doses are not known, and nobody seemed to describe the psychotropic effects in humans, so it’s hard to guess how paranoid one should be about the dangers of this drug. Judging from the concentration that efficiently blocks activity in the prep though, a dangerous zone should probably lie somewhere in the range of hundreds of chambers (or dozens of vials).

APV is usually bought in small jars that contain 10 mg of the substance. Physically 10 mg is represented as several tiny white crumbs at the very bottom of the jar; visually it is hard to believe there’s anything there (so be careful, don’t assume the jar is empty). As the amount of matter is really small, we don’t attempt to weigh it, but rather calculate how much deionized water to add to it to make a stock 3 Podgorski K, Dunfield D, Haas K. Functional clustering drives encoding improvement in a developing brain network during awake visual learning. PLoS Biol. 2012 Jan;10(1) 4 N. Iwatsuki and O. H. Petersen, "Action of Tetraethylammonium on Calcium-‐Activated Potassium Channels in Pig Pancreatic Acinar Cells Studied by Patch-‐Clamp Single-‐Channel and Whole-‐Cell Current Recording," J Membr Biol 86, no. 2 (1985). 5 R. R. Buss, C. W. Bourque, and P. Drapeau, "Membrane Properties Related to the Firing Behavior of Zebrafish Motoneurons," J Neurophysiol 89, no. 2 (2003). 6 M. Zhang et al., "Functional Elimination of Excitatory Feedforward Inputs Underlies Developmental Refinement of Visual Receptive Fields in Zebrafish," J Neurosci 31, no. 14 (2011). 7 C. J. Akerman and H. T. Cline, "Depolarizing Gabaergic Conductances Regulate the Balance of Excitation to Inhibition in the Developing Retinotectal Circuit in Vivo," J Neurosci 26, no. 19 (2006).


of 37

solution of proper concentration (100 mM in this case), and then aliquot it Eppendorf tubes, and store in the −20°C freezer. For APV it means 507 µl of water; ~25 vials, 20 µl each, containing drug for 4 experiments (10 ml of external each). Frozen aliquots are good for about a year, and should not be thawed and frozen again more than 2 times (so either prepare more external, or do more than 1 experiment per day, or re-‐aliquot them into smaller portions every 4 experiments, or throw excess away). Dry substance can be stored at room temperature, and is good for about a year (even though Tocrics promises it to be good for only half a year).

Weight Volume Concentration 197.13 g (mol. weight) 1 l 1 M 0.099 mg (chamber) 10 ml 0.05 mM (target) 10 mg (batch) 507 µl 100 mM (stock) 0.099 mg 5 µl (transfer dose) 100 mM

NBQX The desired concentration is 20 µM8. Old-‐style NBQX is not polar, and so does not dissolve in water, but only in DMSO. Modern version of NBQX however is a Na salt, and it dissolves perfectly, at least up to 100 mM (and probably higher). Lethal doses are not known, but in rats behavioral changes started to be manifested at 20 mg/kg, and strong ataxia developed at 60 mg/kg9, which means that one would have to expect certain effects after drinking about 20 000 chambers (7000 vials).

The rules for storage and aliquoting are same as for AP5, only substance is dark brownish, should be stored at −20°C even in dry form, and all the numbers are of course different.

Weight Volume Concentration 398.26 g (mol. weight) 1 l 1 M 0.080 mg (chamber) 10 ml 0.02 mM (target) 10 mg (batch) 1255 µl 20 mM (stock) 0. 080 mg 10 µl (transfer dose) 20 mM

Gramicidin

Gramicidin is to be dissolved in dimethylsulfoxide (DMSO) to produce a stock solution of 5 mg/ml, which should then be stored in a cool place. 4 µl of this stock are to be added to 1 ml of prefiltered intracellular solution every 3 h, and sonicated for 30 s to produce a final gramicidin concentration of 20 µg/ml10 (as we usually use 0.5 ml aliquots for the internal, the transfer dose would be 2 µl). It seems to

8 M. R. Bell et al., "A Neuroprotective Role for Polyamines in a Xenopus Tadpole Model of Epilepsy," Nat Neurosci 14, no. 4 (2011). 9 Filliat P, Pernot-‐Marino I, Baubichon D, Lallement G. Behavioral effects of NBQX, a competitive antagonist of the AMPA receptors. Pharmacol Biochem Behav. 1998 Apr;59(4):1087-‐92 10 Akerman and Cline, "Depolarizing Gabaergic Conductances Regulate the Balance of Excitation to Inhibition in the Developing Retinotectal Circuit in Vivo."


of 37

be a good idea to store the internal (the whisker syringe) on ice when gramicidin is used, as it does not seem to be stable in the solution.

If purchased from Sigma, it is shipped in small jars of 500 mg that should be stored at 2°C, and is good for 3 years. It is usually quite pure (>90%). The solution in ethanol is said to be stable at 2°C for 30 days11, but as we know from literature that it becomes inactive in the internal solution after about 3-‐4 hours, we’d better keep it frozen. Before the experiment, measure some 2-‐5 mg of the substance into a small Eppendorf tube; calculate proper amount of DMSO; add it and shake (with a shaker); aliquot in ~50 µl and freeze at −20°C. Then use each aliquot for about a week, keeping it at 2°C (?). When working with the DMSO, use latex gloves. See below for the recording protocol.

Weight Volume 20 µg 1 ml (target) 5 mg 1 ml (stock) 10 µg (transfer dose) 2 µl (per 0.5 ml of sol.)

GABA

Working with GABA is funny, because it is a selective agonist, and so the working concentrations are those of µM (from 25 to 200 µM according to literature), but at the same time to apply it locally one would fill pipettes with it, and so the volumes used are really small (0.5 ml per experiment). It gives the process of preparing it a certain homeopathic tint: to get a load of aliquots one would take a tiny dust-‐like particle of GABA, dissolve it in about 1 ml of solution, then take a few µl of this “stock solution” (a single drop, barely visible) to further dissolve it in ~10 ml of external. In our lab we used 100 µM solution of GABA (Colin Akerman used 50 µM12).

GABA is cheap, comes in little jars, can be stored for years at room temperature, and is not too hygroscopic. While preparing the solution for local application, it is advisable to add a speck of some fluorescent dye (such as Lucifer yellow) in it: this will make it possible to use fluorescence to calibrate the Picospritser for puffs of proper strength, and also it will make preparation of the solution somewhat easier (more sane subjectively), as at least the substance will have some yellowish color.

Weight Volume 103 µg 10 ml (pre-‐aliquots) 5 µg 0.5 ml (dose)

11 Sigma Product Information for G5002 12 Akerman and Cline, "Depolarizing Gabaergic Conductances Regulate the Balance of Excitation to Inhibition in the Developing Retinotectal Circuit in Vivo."


of 37

Device Manuals

Scales 1. Switch on (pressing the I/Ȯ symbol) 2. Wait really long, with doors of the weighting chamber closed, trying not to jump near the table, and

not to shake the table. The scales are trying to set a zero point, and the more stable they feel – the quicker it happens.

3. Do the scaling. TARE button is very useful. Close the door and wait the thing to stabilize each time.

pHmeter 1. Switch on 2. Move the sensor (glass rod with wires in it) out of the little jar with yellow buffer 3. Open a side hole on the upper part of the sensor by rotating a purple ring clockwise 4. Rinse the sensor tip with dH2O so that the yellow buffer doesn’t spoil our solution 5. Immerse the sensor into our solution that is being mixed on the magnetic stirrer. Use the strange springy arm to

hold it in place. The arm doesn't hold very strong, so be careful. 6. Wait while pHmeter writes “Stable” 7. Add a drop of 10N base (NaOH for external, KOH for internal). Use plastic pipette with a sharp tip 8. Repeat 6-‐7 until you are ~1 pH below the target, then switch to 1N base and continue 9. If an overshoot – correct with HCl, and repeat 8 if necessary 10. When all set, perform steps 5-‐1 backwards, returning pHmeter to its initial state

Osmometer 1. Move black lever at the right side towards you 2. Carefully pull the black thing out, so that a shiny plate is seen 3. Take a small paper disk from a box with a forceps, and place it carefully on a shiny plate, centered 4. Take a black pipette, put a tip on it (it has its own tips) 5. Move a drop of solution (as much as the pipette takes actually) to the disk, placing pipette in a notch, ensuring

that the disk is still centered 6. Without hesitation carefully push the thing into the apparatus 7. Move the lever back from you 8. The apparatus starts to countdown. After the countdown it will say the osmolarity. Meanwhile you can remove

a tip from the pipette by lifting (rather than pressing) a radial lug up 9. After the measurement, open the thing again, remove the disk, carefully and gently wipe the plate, add a drop

of distilled water, wipe it again so that it is dry, push the thing back, and close the right lever. The apparatus will start the countdown again, and if the plate is clean – it will go into standby (=READY) mode. If it measures anything – then the plate was not clean and dry, and needs to be re-‐wiped.

10. From time to time the osmometer may need calibration. Solutions for calibration are available, and naturally that one should be used, which has the closest Osm to our target Osm.


of 37

Picospritser & local application The thing can be triggered either manually (there’s a button on the left), or with a remote control (that does not quite work, and sometimes either does not provide a puff, or provides 2-‐3 puffs in a burst), or via the TTL pulse (advisable). To make the valve work, first “enable” the respective channel.

The application duration can be changed from several ms to several seconds. Although theoretically durations of less than 3 are possible, it does not seem to open a valve when the durations are shorter. The pressure can be changed with a black knob on the right side of the panel: rotating it counterclockwise makes the pressure drop at some point.

The holders we use are designed for 0.78 mm outer width capillaries (almost twice thinner than our standard 1.65 mm recording capillaries). The same program can be used to pull the capillaries into pipettes though, and the opening sizes seem to be pretty comparable (broken #77 program pipettes are used for electroporation, while #45 pipettes are perfect for local application). When installing or removing an electrode from a holder, be careful not to break it, as if you do – there will be no way to push the glass out of the holder, and it is likely to become unusable. To fill the pipette make a special super-‐thin barely-‐usable whisker-‐syringe.

For local application of GABA Colin Akerman used pressure of 5-‐10 psi and 20-‐50 ms pulses, while I find that 30 psi with 3 ms pulse duration and #45 pipette work equally well, and provide a nice time-‐locked response. The pipette opening should be brought pretty close to the cell (about 2-‐3 cell bodies away). It is better to provide the puffs periodically, as in this case diffusion will wash drugs from the pipette tip to the same expect each time, making concentrations stable across applications.


of 37

Master-‐8 General principles of operation:

• To learn current program: ALL • To switch to a program 1: ALL, 1, ENTER. (up to 8 programs) • To switch a “manual channel” ON and OFF: use the trigger below the light. • To send a pulse to channel 2: press 2. • To regret: CLEAR DISPLAY. To deeply regret: RESET.

How to Program it:

• To switch channel 3 into… o Manual ON/OFF mode: DC, 3, ENTER. o Independently run in a cycle mode: FREE, 3, ENTER. o Be triggered by channel 1: TRIGGER, 3, ENTER. CONCT, 1, 3, ENTER. o Be triggered by ch. 1, delivering trains of pulses: TRAIN, 3, ENTER. CONCT, 1, 3, ENTER.

• To change channel 1 pulse… o Interval of cycling to 2 s: INTER, 1, 2, ENTER, 0, ENTER. (Here “0” stands for seconds

(1e0). Naturally, this setup is not applicable to triggered channels). o Duration to 9.5 ms: DURA, 1, 9.5, ENTER, 3, ENTER. (Here “3” stands for ms (1e-‐3)) o Trigger-‐to-‐pulse delay to 100 ms: DELAY, 1, 1, ENTER, 1, ENTER. (“1” for 1e-‐1). o Pulses-‐per-‐train (if applicable) to 5: M, 1, 5, ENTER, 0, ENTER.

• To disconnect channel 3 from channel 1: CONCT, CONCT, 1, 3, ENTER. To disconnect all in and out connections from channel 1: CONCT, CONCT, 1, ENTER.

• To stop channel 1: OFF, 1, ENTER. To stop all channels: OFF, ALL, ENTER. • To clear

o Current program: OFF, ALL, ALL, ENTER. o All memory: OFF, ALL, ALL, ALL, ENTER. (Never do that in practice!)

Fragments of Arcane Knowledge:

• All time periods should be < 3999s. Min values are 40us for Duration, 100 us for Delay and 60 us for Interval. Also if Interval is <= Duration, it returns an error.

• TTL pulses used in equipment have 0 V for “0”, and 5 V (more than about 2.5 V) for “1”. • Ideal settings for the Robotic Arm: D=1e-‐4; knob looking up. • “Triggered channels” (that are not connected internally?) are triggered by pulses that come to

“EXT 2” input. Channels 1 and 2 may be GATED, and deliver pulses continuously while EXT1 or EXT2 voltage meets some criteria. Read the “manual” if necessary.


of 37

Electrophysiology protocols

Generic protocol 1. Startup

a. Take a small Petri dish, and fill it with MS-‐222 anesthetics. Catch a tadpole with a wide plastic pipette, and place it in this Petri dish. It takes ~5 min for brief anesthesia, and about ~15 min for deep anesthesia (which you would prefer for in vivo preparations). Don’t leave tadpoles in MS-‐222 for hours, as it becomes toxic when exposed to light.

b. Dismount the chamber; rinse it with dH2O. Check that all pins are in place (2-‐3 longer tips for the tadpole operation table; 2-‐3 short tips for a full brain preparation; 3 medium-‐sized pins for in vivo experiments).

c. Fill a tube with ~10-‐15 ml of external solution. Add MgCl2 and CaCl2 from stock solutions (if you have 10 ml of external, with our pippeters you need to add 15 µl of each stock twice). Shake it well.

d. Fill the chamber with the external. Put it under the dissection binocular microscope. Move the tadpole there with a plastic pipette (try not to transfer too much MS-‐222 into the bath). Make the preparation (see below). If you work on full brain prep, don’t forget to unpin and remove the carcass.

e. Place the chamber in the rig. Attach the ground wire. f. Switch on the air table, microscope light, manipulators and stimulation box. g. Switch on the computer, rig and amplifiers; launch programs h. Ventricular membrane removal: Take a microelectrode pipette (old one, from the previous day),

and break the very tip of it by gently touching a kimwipe. Alternatively you can micro-‐brake the tip with a forceps under the binocular microscope in a “controlled” way. Install this pipette into the pipette holder (don’t fill it with any solution), and lower it to the cells layer (as described below), as if for a recording. Finally, by moving it with micro manipulators, and applying measured suction, remove (rip and suck) top layer of cells (described in detail below). Leave the brain at rest for about 10 min, to let the cells heal a bit after you disturbed the tissue with your membrane removal.

i. Pull 6-‐8 electrodes (program 45). Fire-‐polish the backs. j. Prepare the whisker-‐syringe, conical tip syringe, and a small blue tip filter. k. Take a small conical tube of internal (unless you record cell-‐attached, with an external in your

pipette) from a deep fridge, and let it thaw. You may either put it in the pocket (if you plan to start the recording in a few minutes), or just leave it on the table (if you plan to start it in about 30 minutes).

l. If you work with external in your pipette (cell-‐attached or focal potentials recordings) – draw about 1 ml of external with a syringe using a conical tip; replace the tip with a small blue filer, and drop by drop transfer it to the whisker-‐syringe.

m. Take internal from a tube with a conical tip syringe, and transfer it into a whisker-‐syringe, at the same time filtering it (with a small blue tip filter).

n. Place the stimulating electrode.


of 37

2. Recording cycle (for the standard whole cell protocol) a. Fill a pipette with internal up to about 1/3 of pipette length (using the whisker-‐syringe). Avoid

introducing large bubbles. Remove small bubbles in the tip by gently knocking the pipette with a sharpie, while holding the tip down, and staring on it really hard (it helps to scare the bubbles away). With a rolled thin scrap of kimwipe remove the solution from the rare end of the electrode.

b. If necessary, place the intracellular-‐filled whisker-‐syringe in a cool zone if necessary (on ice). c. Install the pipette (by loosening the holder, and then screwing it back tight). If the orange rubber

sealing thing in the holder is new, it can be so tight that you'd have to unscrew the holder completely, manually put the orange thing on the pipette, and then assemble everything back. Note that the diameter of pipettes within the batch may vary slightly, and every now and then you may have a pipette that just would not fit into the holder. Don't break the thing – just admit that it is a wrong pipette to use (and maybe warn others that this particular batch is weird).

d. Apply slight positive pressure (~0.02 ml on a standard 1ml syringe). Don't overdo with positive pressure: you don't want internal to spill all over the cells, as it depolarizes and disturbs them. Also the more flow you have, the sooner your tip will be clogged with some dirt from the solution.

e. Immerse the electrode in the chamber, lower it down a bit, and find the tip under low magnification. You may like to refocus much higher than the brain in order to find the tip. You can also use the meniscus (a black shade that appears at the sides of your visual field when the electrodes touches the surface and all the surface got bent by the capillary forces / surface tention). By following the shade you can roughly guess the pipette location. It also happens to wiggle the pipette a teeny bit left and right (red knob), and see if you can catch any motion with your eyes.

f. Set the electrode to Voltage Clamp at 0 mV (or just have the "clamp" checkbox unchecked, while formally in VC mode). Don't try to clamp at any other potential yet, because the current through the pipette would be huge, and it would de-‐chlorinate the silver wire. Press the "Pipette Offset" button.

g. Start the resistance test, compare the resistance vs. expected value for this type of electrodes (our ideal is about ~8-‐12 MΩ). Inspect the tip visually under low magnification, to check if it’s not clogged or broken. Quite frequently there are bubbles in the tip, in which case you'll have to blow them through by applying ~1-‐3 full syringes of positive pressure, each time locking the air that was already pumped in with a 3-‐way-‐valve. 2-‐3 syringes are always enough to get rid of the bubbles (if 3 syringes don’t help – the tip is either clogged with something stiff, or there is some other problem with the rig, such as a leak in the pressure system). Remove the excessive positive pressure immediately after the bubbles disappear by opening a 3-‐way valve (to avoid strong flow of solution), and re-‐apply the small positive pressure again, as you did it initially, to prevent the tip from clogging. Write down your electrode resistance. Mind that blowing bubbles through should happen high above the brain, otherwise you'll destroy it, as if by a nuclear explosion.

h. Go down towards the brain under low magnification, stopping a bit before the surface (at the level at which the geography of the brain can be guessed already, but still stays much out of focus). Always refocus down first, and only then bring the electrode to the focal plane. This way, even if you forget to change the manipulator speed, or misjudge distance to the brain, you won't crash into the brain, destroying it. Switch the manipulator box to slow speed (“fine turtle” rather than “coarse rabbit” mode; the switch is on the back side of the manipulator control box), and switch the microscope to higher magnification. Refocus to the tip (usually we make it so you had to refocus


of 37

down a bit), and go further down towards the cells. Use only fine refocusing knob when under high magnification.

i. Do “Capacitance compensation”. Note that the “Capacitance compensation” depends on the depth at which the pipette is immersed into the solution, so you should do it when the tip is relatively close to the brain surface. The step on the resistance test should look nice and flat.

j. Find a good cell; target it. Refocus above it; place the electrode above the target. Focus down to the cell; slowly bring the electrode down, touch the membrane at the point roughly between the center and the edge of the cell, or maybe even closer to the center. You should see an invasion (a dark "dip") as the electrode approaches the cell, as there should be a flow of liquid coming out of the tip, and it should bend the membrane. Quickly (but calmly) remove positive pressure (you don’t want to spill internal above the cells, as it would depolarize them, and make them unhappy). See the invasion disappear, and the resistance increase at least slightly. Clamp the pipette at a negative potential (about −50 or −60 mV). Gently (!) apply the tiniest bit of negative pressure if necessary. For a clean electrode it should immediately create a Gigaseal (1-‐8GΩ). Sometimes the negative suction is not even necessary: as the pressure is released, and the dip disappears, the membrane just moves towards the pipette tip, and seals it over. Wait several seconds, for the membrane to move around the pipette; for the seal to improve, and for the cell to calm down. Gently adjust the electrode tip position if necessary, so that the membrane is not stretched. Keeping the electrode at a negative potential helps to “calm down” the cell after you mechanically disturbed it.

k. Go Whole-‐Cell by applying a decisive, but short and gentle suction. Use your finger/thumb to close the hole on the control valve, and be ready to open it again to remove the suction. Don't "lock" the valve: you don't want to be that brutal to the cell. Open it, and just close the whole with some part your other hand. Some people (Carlos) like to apply negative pressure gradually until the membrane is broken, and then rely on their reaction for quickly releasing the pressure (definitely before half of the cell content is sucked into the pipette). Some other people are not fast enough to rely on their reaction speed. These slower people would rather apply some "standard pulse suction" in order to get into the cell (in which case you may have to repeat this "pulse suction" some 2-‐3 times) 13. If you prefer this approach, close the opening with your thumb, move the plunger backwards “drawing” about 0.1-‐0.2 ml in a standard pipette; and while doing it release the hole. The whole affair should last a fraction of a second, and produce a subtle “pop” sound. While doing this (in either way), keep your eyes on the seal test. When you are successful, you’ll see the resistance decrease in a jump, and capacitance current to appear on the step (normal cell parameters for tectal cells: capacitance of 10-‐15 pf; membrane resistance of 800-‐1000 MΩ, access resistance of 100-‐300 MΩ).

13 There are people in other labs who kiss the mouthpiece (syringe without a plunger) to break the membrane. While undoubtedly rich in imagery and metaphor, this method is also messy, unsafe and epidemiologically questionable, so we don't follow it. Surely if you can produce a brief pulse of gentle negative pressure with a sensual kiss, you can also do it with a syringe and a valve hole to cover. There are also some other people in some other labs who apply "buzz" (quick and strong current injection) to break the membrane (there's a special button for providing a "Buzz" on the Clampex panel). We don't do it for our cells, as they are small and tender, and usually die if you try to buzz them; but for some other (bigger, sturdier) cell types, and other electrode sizes, it may be useful.


of 37

3. Switching off a. Switch off the computer, amplifiers, microscope light, manipulators and stimulation box. Close the

air-‐table valve. b. Unpin and remove the brain (under binoculars); rinse the chamber with dH2O. If you reuse same

sylgard block over and over again, you may also consider rinsing the chamber with ethanol, and leaving it this way to dry – it would prevent bacteria from growing, and your sylgard will be less poisonous for the brain.

c. Rinse the whisker syringe and the external solution tube. d. Gently wipe (touch) the immersion objective with a very special Kimtech paper.

Checklist for Junior Patchers 1. Fill new pipette to 1/3 – 1/2 of its length 2. With a kimwipe (rolled into a little paper-‐needle) remove excess liquid from the back 3. Get rid of bubbles (through knocking the pipette gently) 4. Install the pipette 5. Apply positive pressure (<0.1 ml of air) 6. Go down until the tip touches the liquid 7. Measure pipette resistance (should be 5-‐15 MΩ; do it in "Bath" mode). 8. Focus on the tip, ensure if there's no bubble or junk inside

a. If any – blow it through (push – lock – push – lock … when junk goes through – release) 9. Focus lower, then move the pipette to the focus plane. Repeat until slightly above the brain. 10. Switch the manipulator to low speed 11. Change objective 12. Find the tip 13. Lower to just above the cell layer 14. Choose the cell 15. Cancel pipette offset & pipette capacitance 16. Patch:

a. Place the tip exactly above the point between the center and the side of the cell b. Focus at the cell (slightly above the midline) c. Move pipette down until it is near the cell and a dimple is seen d. Release the pressure e. Put the amplifier at voltage clamp at -‐60 mV f. Gently apply suction – until gigaseal is formed g. Switch to "Seal" mode. Apply measured suction until you break into the cell.

17. Measure cell parameters (in "Cell" mode), and follow the protocols.


of 37

Tight Cell-‐Attached recordings The only difference of TCA recordings vs whole cell ones is the following one: you should use external solution instead of internal, and stay in Gigaseal mode, without rupturing the membrane. After waiting a bit in VC mode at −70 mV, go Current Clamp with zero current. You may also prefer using external in the pipette rather than the internal, as it seems to make the seals somewhat better, although in this case you lose the possibility of going whole cell afterwards.

Loose Cell-‐Attached recordings All the preparations for the LCA recordings are like those for the Tight Cell Attached mode, with 2 important differences. First, you definitely need to use external in the pipette. Second, before patching an actual cell, you should make the pipette tip “dirty”, so that it stays in loose cell-‐attached, and does not make a Gigaseal. One way to do it is to find a patch where the ventricular membrane is still intact, release the positive pressure, and several times touch, or even bump into the ventricular cells. They seem to be covered with something (cilia? proteoglycans?) that prevent Gigaseal formation, and they are also strong enough not to be damaged by this bumping, but it makes a tip dirty. Alternatively, you may find a “victim cell”, and make a “fake whole-‐cell” on it, departing from the cell immediately afterwards. Regardless of the approach, in some cases you may still inadvertently get a Gigaseal on the first cell you try to patch, in which case you should just go “fake whole-‐cell” again, and then get rid of the cell again, hoping that after this little adventure the tip has become even dirtier, and thus better.

How to depart from the cell? If you just move the electrode away, especially if you do it very slowly, or very fast, you would just rip the cell out of the tissue, and it will be secured on your pipette tip. You may even like it when you do real recordings (as by doing so you would remove dead cell bodies from the brain), but you definitely don't want the cell on the tip if you just try to prepare it for LCA. You may also apply positive pressure, but if you do it without moving the tip, the cell will probably pop up, and the dead "cell shell" will still sit at the end of your tip. Disgusting. So in reality to get away from the cell you should perform a combination of these actions: you should start moving away from the cell (decisively = not too slow), and when the cell is visibly stretched – apply some modest positive pressure. Then continue moving the tip away. Every now and then it does not work, but overall you should be able to depart from cells with some ~80% success rate.

Finally with the “dirtyish” tip, follow the “standard patching procedure” with a target cell. The difference will be that instead of a Gigaseal, after removing positive pressure, and then applying negative pressure, you should expect to see only moderate and gradual increases in “patch resistance”. Continue to apply negative pressure slowly, gradually, and steadily, to suck a bit of the membrane inside. Then release, and see if the resistance goes down much. Usually in 2-‐3 attempts you can reach a resistance of ~80-‐100 MΩ, that could relax into ~100-‐120 MΩ after you release the pressure, however resistances of about 60-‐70 MΩ are usually high enough already to allow a nice LCA recording. The “maximum achievable resistance” seem to be dependent on the prep (the age of the animal, the external solution composition, and on whether there was a lot of bleeding around the brain), so if several cells in a row get torn at ~70 MΩ, just accept it as a possible “practical maximum” for this particular prep. During the negative pressure application, control the cell visually, as it will visibly shrink while being sucked into the pipette, so you may need to re-‐adjust the focal plane and the electrode position during the process (between the suctions).


of 37

Several cells in a row can usually be clamped in loose-‐cell-‐attached mode with the same electrode, and the ease of getting a reliable LCA will only increase. Gradually however the tip would catch membrane flips with ion channels on them, and these channels would make the recordings dirty. You’ll need to change the pipette if the cell ghost (blown-‐up cell) got stuck on the tip, or if a particularly big membrane flap would lead you into a “fake Gigaseal” of about 300 MΩ – 2 GΩ without actual access to the spiking cell.

Perforated patch (Gramicidin) recordings In short, perforated patch recordings resemble those performed in whole cell or tight cell-‐attached modes. The perforating peptide (gramicidin) is added to the solution inside the pipette; after the gigaseal is achieved, one needs to wait several minutes for the peptide to get into the membrane and allow voltage control over the cell. The protocol is somewhat tricky however, so it worth describing the details.

Take a tube of internal solution. Some people prefer using slightly weird internals (such as ones with extremely low or extremely high Cl− concentration, to be able to notice the moment when the perforated patch is broken and the recordings goes whole-‐cell14. Filter the internal into another Eppendorf tube (you have to do it before the gramicidin is added, otherwise it will be filtered out). The tubes on the table are usually quite dusty, so it worth take one from a rarely disturbed bag somewhere deep in the shelves. Add required amount of DMSO with gramicidin in it; it may cause partial precipitation of the peptide. Close and vortex (shake) the tube; then attach a piece of plastic foam to it (put the Eppendorf tube through the hole in the foam), and let it float in the sonicator; sonicate for ~2 min. Suck the contents of the tube into a whisker syringe. During the recording, keep the syringe with the gramicidin internal on ice. Prepare new solution every 3-‐4 hours (as allegedly gramicidin gradually deactivates while in solution).

As the gramicidin solution is not filtered, and as at high concentrations the peptide tends to precipitate, you cannot blow the bubbles through the pipette tip, and also you can’t apply strong positive pressure while approaching a cell, for the same reason. There are two ways to tackle that. One approach is to filter the internal with a filter that has relatively large pores, however in this case the concentration of gramicidin in the solution will get lower, and you’ll have to wait longer before getting enough access to the cell (possibly about 20-‐30 min). Alternatively, you can use a front-‐filling technique: just keep a small tube of filtered gramicidin-‐free internal, and dip the pipette tip in it for several seconds (10-‐20 s). Due to the action of capillary forces the tip will get filled with the solution (it is almost impossible to see the liquid inside the tip, but it should be there). Then the remaining of the pipette is filled with gramicidin-‐containing solution as usual. There will be some bubbles in the tip, and remember: this time you cannot blow them through, so hold the pipette in your hand vertically, its tip down, and gently knock it with a pen or a marker for 1-‐5 min, until the bubbles leave the tip. 5 minutes of knocking may sound like a pretty long time, but as a result of this time investment you'll get a pipette with a clean unclogged tip, and full of gramicidin, making the yield potentially really high.

14 Akerman, Cline 2006


of 37

Another reason not to apply positive pressure while approaching the cells is that you don’t want to spill the stuff all over their outer membranes – they will get leaky, and die if you do that. It is not advisable however to completely get rid of positive pressure, as in this case the electrode keeps sucking stuff from the solution (due to the capillary forces action), and the tip quickly becomes dirty. The best idea is to close the valve before installing an electrode into the holder: it will create some positive pressure, just enough to prevent the external solution from entering the tip, making it dirty. It will also make the patching easer: you won't probably see the dent on the membrane as you go close to it, but theoretically it will still be there, as for the normal patch-‐clam protocol; the tip however won't get clogged, as the flow of stuff from the tip will be really low.

After you patched a cell, establishing a Gigaseal, just wait for 5-‐10 min and see the access resistance drop, and the postsynaptic events to become visible. Usually at the beginning you would see a normal gigaseal (1-‐4 GΩ), which would gradually evolve to a stable perforated patch case, which looks like a “very small and resistant cell" (capacitance of ~5 pF, membrane resitance of 1-‐2 GΩ, access resistance of ~100 MΩ). After doing a recording, apply brief suction, going whole-‐cell: the capacitance should increase 2-‐3 times in a jump (up to about 15-‐20 pF), the access resistance could decrease slightly (down to 20-‐30 MΩ), while the cell membrane resistance would usually decrease gradually with gramicidin perforating it, killing the cell.

Ca imaging protocol 1. Prepare 5 measures of external solution (1 for preparation & staining, 3 for washing the dye

out, and 1 for the recording). Pour 2 measures in separate tubes (preparation & final recording), and add the paralytic agent (tubocurarine) in them (you’ll have to add 100-‐200 µl of tubocurarine stock solution per tube). Save the remaining no-‐paralytic solution in the beaker (probably sealed with Parafilm).

2. Place the tadpole in anesthetic (MS-‐222). 3. Take 1 portion of dye from the freezer; dissolve it in 10 µl of DMSO + F-‐127. Put in the sonicator

for 15 min (it won’t dissolve without that). At all stages try to protect the solution from light (wrap it in tin foil?).

4. Pour 1 portion of external solution (with paralytic) in the chamber. Make in vivo preparation. 5. Place the chamber under the microscope. Remove the membrane. 6. Take the chamber out, and add the dye to the solution. Leave for 30 min in a very dark place

(black bowl + tin foil, or something similar). 7. Place a prism in the rig, and center the image on the fiber. 8. After the tectum is (hopefully) stained, carefully pour out the solution, and rinse the chamber 3

times with the external. Be sure not to break the tadpole while adding the liquid, or while it is exposed to air. Finally, add the last 5th portion of the external (this one is to be again with the tubocurarine).

9. Place a fiber facing the eye, at about 1 eye diameter from the lens. 10. Switch the excitation light; calibrate the camera; find a good place. 11. Do the recording.


of 37

In-‐vitro (whole brain) preparation 1. Tadpole is anesthetized with MS-‐222 for 5-‐10 min

2. It is pinned to a “preparation” Silgard block roughly in the same manner as it is done for in vivo experiments (see below), only without tilting, but rather symmetrical. 3 pins are to be used for stage 48-‐49 animals (right, left, forward), while for stage 45-‐46 2 pins are enough (there’s no space to place the 3d pin anyway). The right pin (for right-‐handers) is to be really sharp, as you’ll have to poke the tadpole with it, while it is not yet attached to anything else, helping yourself only with a forceps. The other two pins can be blunter.

3. The brain is opened (see figure). First you take a new needle, and stick it under the skin above the hindbrain. Then you make an incision of the skin above the brain all the way down to the olfactory bulbs (towards the space between the nostrils). Widen the sides of the incision a bit, to have access to the brain. Then cut the upper commissures above the hindbrain, at the caudal part above the tectum, and finally between the olfactory bulbs. Be careful not to go to deep, not to damage the midline structures, especially around the Optic Chiasm, as you’ll have to dive somewhat down right after it, to cut between the olfactory bulbs. All these procedures are better done when the brain is still secured by the nerves inside the body.

An alternative to cutting right above the brain is in cutting the skin medial to the brain, flipping the piece of skin over to the side (thus opening the brain), and then carefully cutting through the membrane with melanocytes. Some people feel having more control over the prep if it is done this way.

4. After that is done you can already pick the brain with a sharp forceps by the stump of the spinal cord, and pull it away. In some cases it may be also useful to cut the nerves to the right and to the left from the brain. For younger tadpoles you’ll have also to cut below the brain, as the brain is not yet fully solid at this stage, and you can easily tore it in two parts while pulling the hindbrain out. Also if you wish to preserve a bit more of the spinal cord than just the thump, you’d better cut backwards from the point of your initial incision, opening the spinal channel.

5. The brain is then transferred to the recording part of the chamber, and pinned down (see the photo). As we don’t study olfactory so far, it is usually safe to secure it with 2 pins going through the olfactory bulbs, and to place the 3d pin somewhere in the hindbrain, in the region you don’t need for your recordings. Don't overstretch the brain when you pin it, as in this case the cells will gradually give up during the recording, and the tissue will move in respect to the electrode tip, which is extremely annoying if you want to hold the cell. It happens every now and then anyway (it also seems to happen when the osmolarity of the external solution is slightly off, causing some shrinking / swelling), but at least try to minimize the drift by pinning the brain properly.


of 37

Stimulation & Recording. The stimulation is frequently done at Optic Chiasm (marked with a red arrow on the photo). The recordings, when they are performed in the Optic Tectum, are usually limited to the central part of it (marked with an orange circle).


of 37

The preparation scheme

A. Place the pins B. Cut from the tail base up towards the nostrils C. Move the skin, making the opening wider (you should see both tecta) D. Cut the commissures (including between the olfactory bulbs, which is lower) E. Cut the nerves on both sides of the brain F. If you need the spinal cord: cut above it G. Pull the brain out


of 37

In-‐vivo preparation Pinning: You’re supposed to see the heart to the left from the otic chamber. So that brain, the otic chamber, and then the heart somewhere below. In this case one pin goes straight down, not to reduce your field of view in the future.

Pin on the right side (#2 as they are numbered above) goes first; then #1, and then finally #3. The triangle between the huge vessel, the eye and the brain is relatively free of everything (there are only gills and muscles there), so it’s a good place for a pin. Be careful not to damage the vessel, or the nerves/vessels that go to the eye though. Note also, that while you’ll be working on right tectum, you’ll be projecting to the left eye, so don’t harm or damage it.

The brain is surrounded with big and tiny vessels, so it is impossible to avoid bleeding. Bigger vessels seem to be located above the hindbrain though, so it may worth trying to open tectum only, without touching the hindbrain region. Any bleeding that occurs during the preparation needs to be immediately taken care of. Ideally all blood is to be sucked out with a special syringe that has its needle blunted; alternatively you can move whatever instrument you are using circularly at one side of the brain to create the flow, and to let this flow move the blood out. The reason for this requirement is that after the blood coagulates, it becomes almost impossible to get rid of it: it glues to the tissues around, and becomes stronger to tear than the brain tissue! Once coagulated, it is also quite

impenetrable for the recording pipette, so you have to suck it out or dilute if before it is too late. Also it looks like the neurons that got in direct contact with the blood, even diluted, have their membrane properties changed, and it becomes harder to establish a stable loose-‐cell-‐attached contact with them.

This pin goes first

Otic chamber

gut


of 37

An old unused recording pipette with its tip briefly brought close to the Bunsen flame comes really handy for in vivo preparation. After meeting fire, the pipette tip is transformed into a small perfectly smooth sphere sitting on a glass stem. Bringing pipettes closer to the flame, or prolonging the meltdown one can crate spheres of different diameter. These spheres can then be used to manipulate soft brain tissues: to open the brain (after the commissures are cut with a syringes), push tectal lobes apart, smoothen the surface etc. A blunt forceps, even a polished one, is still very rough, and damages the tissue, tearing away upper layer of cells each time it comes in contact with the tectum. In some cases it may be a desired feature (in can actually help with membrane removal), but in the majority of cases it is safer to use glass sphere for all manipulations.

Membrane removal The main difference between areas of the brain covered with a membrane, and those not covered, is that under the membrane you can’t see neuron cell bodies clearly, as separate roundish blobs. The other peculiar feature of membrane cells is that small bubbles are abundant in them (this feature is helpful once the majority of membrane is removed, only some solitary cells are still left in their place). Sometimes bubbles happen to sit among the neurons as well though, especially in younger brains, so don’t try just to catch all the bubbles – that’s not a universal trait. To compensate for this nuisance, membrane removal is generally easier in younger than it is in older animals, as the epithelium cells are not sealed so strong with each other in stage 45-‐46 tadpoles.

If you used a blunt forceps to open a brain, there might be membrane-‐free areas on tectum surface when you place it under the microscope already. In this case it worth starting from this membrane-‐free spot, peeling the membrane to the sides of it, cell after cell, gradually expanding the exposed area. Be careful though, as even those membrane cells that are only partly surrounded by other membrane cells, still tend to be more resistive to suction than neurons beneath them, so there’s always a risk to suck too many neurons compared with the membrane cells.

If no membrane-‐free spot is available, you can poke the membrane with your pipette tip, and then move it forward, to let it go under the upper layer of cells, stretching the membrane. As pipette has a conical shape, this manipulation would tear the membrane above the pipette. All you’ll have to do afterwards is to suck out the flapping scraps on both sides. If afterwards you turn the chamber slightly, you’ll get access to a dozen or more cells on one side of the pit.


of 37

Mauthner Cell recordings We used approximately 1 month old animals (stage 49 or “49+”), avoding the bowls where most of the tadpoles had already died. The closer the tadpole is to metamorphosis, the larger the Mauthner cells (Moulton, Jourand, and Fox, 1968). At the same time from behavioral data on startle responses in XL tadpoles is it known that they are most responsive to stimulation (at least to acoustic one) at around stage 47.

Dissection: 1. Start as usual. 2. In order to flatten the hindbrain, it is necessary to cut the overlying cells in the caudal hindbrain

at the beginning of the spinal cord. The more of these cells that are cut, the easier recording from PHP cells will be (especially if recording from both sides of the brain).

3. Remove the layer of melanocytes from the spinal cord as far as it is necessary (depending on whether recording or stimulating in the SC).

4. Pin each olfactory bulb and the spinal cord. Instead of the SC pin, it is also possible to pin the hindbrain at a caudal and lateral point.

Hindbrain orientation: There are 8 hindbrain segments (rhombomeres). M-‐cell and most PHP cells are located in the fourth segment, which is at the widest point of the flattened hindbrain. The clearest sign of its location is the decussation of two heavily myelinated axons at the midline. If this is not visible, one may try to count the segments from the tectum.

Some clues: • Each segment has a superficial layer of very small cells that is absent (or located deeper)

between segments. • The first segment is the largest, the second is much smaller, and segments 3-‐7 are similar in

size. • Sometimes the segments may appear darker due to the increased thickness of the brain at

those points. • It is also occasionally possible to see the insertion site of the VIIIth nerve, which is in the fourth

segment.

The M cell is located laterally in the un-‐filleted brain. In the filleted brain, it is located just lateral of the halfway point between the midline and the edge of the flattened hindbrain. If counting segments by the small cells, the M cell is found below the caudal edge of the small cell area in the fourth segment. Most visually responsive cells are superficial and medial from the MC.

Field potential recordings, method 1: Remove the superficial cell layer over the MC and insert a program-‐45 electrode (resistance 9-‐12 MΩ) into the area. Use a small amount of positive pressure and insert the electrode about as deep as the axon decussation level. Look for a location with response around 0.5 mV (to either tectal or antidromic stimulation).


of 37

FP recordings, method 2: Break a program-‐45 electrode and use negative pressure to draw in cells around the MC without clearing the superficial cells. This method is quick and effective (responses are much larger) but also kills cells in the area.

For patching: clear the superficial cells over a large area in the fourth rhombomere. Although visually-‐responsive cells are found at every level, large cells with interesting properties (putative collateral PHP neurons) are found at the level of the MC. Note: there are also glia and many dendrites/axons in this area (putative axon cap-‐like structure). Getting a tight seal on a cell may require maneuvering of the electrode and positive pressure to avoid these obstacles.

Staining protocols

Fixatives & the fixing procedure 1. Put Nitrile gloves on. Prepare goggles & a dust mask. 2. For 50 ml of 4% PFH solution (just enough to fit in one 50 ml tube, which is very convenient) –

put 25 ml of dH2O in a beaker on the heater in the fume hood. Stir. Turn the heater to ~150°C. 3. Put on goggles & a mask. Weigh 2g of paraformaldehyde (nasty dust) in a weighload inside the

fume hood. Add the PFH to water. Discard the contaminated weighload in the plastic bag, which is inside the fumehood, on the right. You can take the dust mask off now.

4. Add 2 drops of NaOH to the beaker. with a plastic pipette. Heat to 65°C (but not much higher). Wait for PFH to dissolve.

5. When the solution is clear, let it cool to room temperature. Add 5 ml of 10x Sodium Phosphate Buffer. By adding more NaOH bring the pH to 7.4 (use HCl in case of an overshoot).

6. Adjust volume to 50 ml. You may filter it if the objects to fix are going to be extremely sensitive. 7. Pour in a 50ml tube, cover it with tin foil, and place it in the Fridge. The solution is good for 1

month.

Dextranes staining15 1. Put some albinos in MS-‐222. 2. Take them out, put on a Sylgard / Kimwhipe / Petri dish. Make the necessary cuts. 3. Remove the excessive liquid with a pipette, then with a Kimwipe. Let the tadpoles dry slightly,

but so that they don’t die. Ideally the incision should be wet, everything immediately surrounding the incision should be dry, and then the remaining parts of the tadpole should be wet again, but they are just too small to achieve this ideal state, so maybe just catch the moment when they are dry already, but yet not for long.

4. Take a small jar with dextrane crystals (they are normally stored in the −20°C freezer, and should be protected from light with a tinfoil). Take a bit out of there (with a syringe needle, or with a glass tube – with something thin). Apply the crystals to the incision place. Let them be dissolved.

15 Modified from: H. Straka, R. Baker, and E. Gilland, "Rhombomeric Organization of Vestibular Pathways in Larval Frogs," J Comp Neurol 437, no. 1 (2001).


of 37

5. Cover the place with a Petry dish lid, and let it stay for 5-‐15 minutes. You may place some droplets of water around the tadpoles, but not touching them, in a hope that it would evaporate under the lid and bring the relative humidity close to 100%.

6. Wash out the excess of dye with Steinberg. Then place tadpoles in MS-‐222 (in a small Petri dish), and let them be incubated at 18°C overnight.

7. Next morning – fix the tadpoles for 2-‐6 h in 4% PFH. Then store them in the buffer.

Lipophylic staining of full brains 1. Extract some brains in ACSF 2. Move the brains to a small Petri dish of PFH fixative with a pipette. Fix overnight at 4°C. 3. Remove the fixative from the Petri dish – first with a pipette, then with a Kimwipe. Catch the

moment when the brains are not too wet, but not yet too dry, and stain them. 4. The staining is done in the following manner: you put some ethanol-‐diluted DiI (or whatever) in

a whisker-‐syringe. You put a small droplet near the brain. It starts to evaporate immediately. But you are faster than the vapors, and you “connect” the droplet with the spinal cord with some kind of a tool (like fire-‐sealed electrode for example). Thus the droplet doesn’t have enough time to cover the brain entirely, but just stains the tip of the spinal cord.

5. When staining is done, you add buffer, and incubate for 3-‐4 days at room temperature.

Behavioral protocols

Seizure Experiment 1. Get control and treated tadpoles from the incubator (stage 47 works the best). Place each

group of tadpoles in a bowl filled with fresh Steinberg’s. Leave them for one hour on the bench top, so that whatever treated tadpoles were treated with could washout from the body.

2. Set up the Styrofoam box underneath the camera. Set up the six-‐well plate, with 5, 7.5 or 10 mL of PTZ solution in each well. You'll have to wait longer with lower concentrations of PTZ, but in case of weak effects they may be clearer (otherwise everything happens too quickly to notice a difference). Switch the Styrofoam light on.

3. When the tadpoles are ready (in 1 hour after stage 1), start recording the six-‐well plate in EthoVision. We start recording before transferring the tadpoles to PTZ, as want to see the entirety of the tadpoles’ behaviour.

4. Transfer three control tadpoles on one row of wells, and three treated tadpoles into the other row. Record for 20 min.

5. Analysis: Export raw data from Ethovision as an Excel spreadsheet. Use the MATLAB file “seizures.m” to process the files. This program returns frequency of seizures, time until first seizure, and average seizure length.


of 37

Habituation Experiment 1. Get several control and treated tadpoles from the incubator (here we usually work with s49).

Place both groups of tadpoles in separate fresh containers of Steinberg’s, for all substances they were raised in to wash out.

2. Set up the apparatus in the meantime. a. If you use the "Robotic arm", position the styrofoam box and the “hitting arm” as

shown on the photo below. Position a six-‐well plate at the upper right corner of the Styrofoam box.

b. If you use the Sound-‐Pulse Precision Point-‐Injection Device (SPPPID), just put the

SPPPID on top of the Styrofoam box. Switch the Styrofoam light on. 3. Prepare the program

a. For the Robotic arm: Verify that Master-‐8 is following the correct protocol. Unless the program has been deleted from Master-‐8, it should be Program 7 on Channel 6. The interval is set to 5 seconds and the duration reading is 4—3 (which supposedly means 4 ms).

b. For the SPPPID: start the "clicker.m" program in Matlab. Connect the SPPPID to the Matlab-‐running computer with an appropriate cable (long audio cable).

4. Fill the wells with 6.5 mL of Steinberg’s each. 5. After one hour of washing-‐out, transfer tadpoles into the wells (treated on one side, and

controls on the other). Note that you should alternate the sides for the plate where the treated and control tadpoles go on a trial-‐to-‐trial basis, as whatever you do the stimulation in different wells will be different, and you don't want to introduce a bias in your measurements. If you can rotate the plate (for the "Robotic arm"), mark the control and treated wells, and perform the rotation before each trial. If the plate is secured in the apparatus (SPPPID), introduce some kind of a marker (a piece of scotch tape would do) and change it trial to trial, so that you could always tell from the video which raw was treated. But also of course carefully write it down in the notebook.


of 37

6. Start recording in EthoVision (“Acquisition 2012 startles 6well” file can be used as a template); then start the clicking (in either Master-‐8, or Matlab GUI). Run the hits for 2 min. Then stop the hits and stop the EthoVision trial.

7. Give tadpoles 5 min of rest before the next round of hits. 8. Repeat the protocol (2 min of hits, then 5 min rest) until the tadpoles have had 5 rounds of hits.

After the 5th round of hits, give the tadpoles 15 min of rest. For each 2-‐min round of hits, start a new EthoVision trial. Note that because you have 5 minutes pauses between 2 minutes sessions you can run 2 batches of tadpoles at the same time, intermittently.

9. Finally, after last 15 min of rest, give the tadpoles one more final round of hits (2 min). 10. Analysis: Export raw data as an Excel spreadsheet. Use MATLAB file “habituation_6well_a.m”

and input the correct Trial #. This program returns the average speed of responses. The program tries to find the moments the hit was delivered automatically, based on the tadpole behavior (so do check the plots it produced! If the tadpoles did not react at all, the program won't be able to find the signal in the noise, and will measure "escape speed" at some random points of time).

11. Take the output from this program and put it into the Excel document “habituationformatlab.xlsx” (located in: C:\Users\AizenmanLab\Documents\Jenny).

12. Then, use MATLAB file “habituation_6well_part2.m”. This program reads the habituationformatlab.xlsx document (make sure that you type in the relevant sheet number for the trial you want to analyze). This program returns rapid habituation, short habituation, long habituation, short recovery, and long recovery.

PPI Experiment 13. Get s49 control and treated tadpoles from the incubator. Place both groups of tadpoles in

(separate) fresh containers of Steinberg’s. Basically do everything as in protocols described above.

14. Set up the SPPPID contraption under the camera. Make sure the speakers are turned to their maximum volume. Make sure the computer is set to maximum headphone volume. Fill each well on the six-‐well plate with ~6.5 mL of Steinberg’s.

15. On EthoVision (“Acquisition 2012 ppi” file), set up six arenas (one for each of the six wells on the plate). Open the MATLAB file “clicker.m” and run it by pressing F5. Set parameters accordingly. Set Ratio=0.05 and ISI=100 ms. Leave other parameters as their default values. Importantly, the time between hits is 20 s.

16. After one hour has passed, place three control tadpoles and three treated tadpoles into the six-‐well plate. Start recording from EthoVision. Start the MATLAB program (i.e. start hits).

17. Stop the EthoVision trial and stop the hits after 5 minutes (approximately 15 hits) have elapsed. 18. Switch out the tadpoles for a new trial. Each animal participates in two trials. Transfer tadpoles

from the experimental apparatus to a different six-‐well plate, in order to keep track of which tadpole is which. Each tadpole should have at least 15 min between trials. Also, the tadpoles should switch wells for their second trial (e.g. if control tadpoles were in the first three wells for Trial 1, then they should be the last three wells for Trial 2).


of 37

19. Analysis: Export raw data as an Excel spreadsheet. Use MATLAB file “ppi_6well.m” and input the relevant Trial # and whether the trial’s first hit was a single hit (singleFirst = 1) or a double hit (singleFirst = 0). This program returns average speed of response to single hits, average speed of response to double hits, and percent PPI.

Miscellaneous EthoVision Tips 20. Make a new arena setting for every new day of experiments, and name the arena setting as

such. The easiest way to make new arena settings is by grabbing a background image right before the experiment starts.

21. For the above protocols, don’t mess with the detection settings. They work fairly well, even on separate days of experiments.

22. For exporting trials, it is useful to have a folder especially set up as an export location. This is because EthoVision cannot export only a few trials (e.g. the 3 trials you did one day); instead, EV will export all the trials that are in that file, even if you have analyzed them previously. So, export all trials to a separate folder and then copy-‐paste the trials that you haven’t analyzed yet to a main folder.

23. If you try to acquire via EthoVision and it tells you that there’s a video source error, then restart the computer (don’t do anything to the camera). This has resolved the issue every time in the past.

Tadpole-‐Dot Collisions

Startup 24. The tadpoles get up at 7 am, and go to bed at 7 pm. It is advisable to do the experiments in the

2nd quarter of their day: from about 10 am to about 3 pm. After that they become drowsy. 25. Turn the Right Computer on:

a. Ensure that the video cable is connected to the CRT monitor. b. Switch the computer on. If the big CRT monitor switches on as well (the LED turns

green) switch it OFF immediately. Otherwise the computer will make it the lead monitor, and you'll have to switch the whole system off and start all over again.

c. After the computer is up and running (with the vertical monitor as the main one) switch on the CRT. It should show a Native American. (If it doesn't show a Native American, but goes black, you'll have to do all that Right-‐Clicking on the desktop, going to "Settings", and then "Extending" the desktop to the CRT.)

26. Start Matlab (Orange wavy icon on the "quick launch" toolbar) 27. Go to the Matlab main window (console), and type "dishRig2". That's the name of our program.

The CRT should turn white, and show a beautiful black military ring with numbers around it. 28. Move the CRT screen and/or the camera (by adjusting the lever) so that the ring is fully with

the camera field of view, along with all the useful messages written around it. Make sure that the camera sees the messages as more-‐or-‐less horizontal (this way they fit better, and are more readable).


of 37

29. By rotating the rings on the camera, ensure that the screen and everything that is on it are in focus (lower ring), and also that the aperture is right (upper ring). The target is to make the pinkish retrace beam artifact that slowly moves across the monitor to disappear (or almost disappear), while the letters / numbers, and the tadpole, are still be visible.

30. Turn on the Left Computer. Start EthoVision (an orangish icon with a square and a camera, or something in this vein). Open your experiment file, go to acquisition <details are missing here>, and ensure you see something on the screen (the current view that the camera sees, whatever it is at this stage).

31. Carefully put the black cloth round the CRT. Ensure that: a. It doesn't occlude the scene for the camera. Put a cardboard ring inside the tent, so

that it wouldn't occlude the view. b. the tadpole doesn't see you, and cannot guess your intentions (the tent should be all

around the dish, without gaps). Still you should be able to close and open it as needed. c. it does not cover the vents on the former bottom side of the monitor, as you don't

want the thing to overheat. Use pegs, sticks and random pieces of cardboard to keep the cloth from closing the vents.

d. The picture is still in focus, and the aperture is still the same as when you set it up. (It is quite easy to upset both of them while attaching the cloth tent to the camera).

32. Take some cold Steinberg's from the 18°C incubator, fill the Petri dish with it, and put the dish exactly onto the black projected ring. Put your head into the tent and position it precisely.

33. Take a bowl of tadpoles from the incubator. Take another bowl and fill it with Steinberg (you will place "used" tadpoles there). Find a wide-‐opening tadpole-‐friendly (clean) transfer pipette.

Experiment 1. On the RIGHT computer (using the remote keyboard) set the dot size and dot speed are as you

like them to be. Both values are displayed on the CRT monitor (to the right and bottom from the ring), as well as on the Right computer monitor (only when you change them). Press "h" if you forgot the functional keys for the "dishRig2" program.

2. Take one tadpole and carefully place it to the dish. Cover the black cloth back. A tadpole may freeze for several seconds (as it was moved from bright light into dark surrounding), but in a while it should start swimming in circles in the Petri dish.

3. On the RIGHT computer press "n" (new tadpole). The system will change T value (tadpole number) on the screen, and also will zero the stopwatch (in minutes) for the time this tadpole spent in the chamber.

a. Note: if you press "h" at any point, you'll see a help screen with various crazy commands that you can give.

4. On the LEFT computer, start the recording <details are missing here>. Start a new trial (new video file) for each tadpole, and for each dot size, if you are using more than one dot size per tadpole (although we should discuss if it is advisable).

5. On the RIGHT computer, target the dot with PgUp / PgDown keys (a notch would move around the ring, indicating the target). When you think you can hit the tadpole, press "z". The system will send a dot, and also will start countdown in seconds till the time you can send the next dot


of 37

(currently it is set at 20 s). The countdown values are written to the left from the ring. When it is time for you to send next dot, it will also produce a sound.

6. Repeat the previous step until either the tadpole stops swimming (although you may try to knock the CRT monitor slightly, to make it swimming again), or 15 minutes (?) passes. In either case – change the tadpole after that.

7. Every several tadpoles (3?) you might want to change the water in the Petri dish. 8. Try not to move the Petri dish (and the CRT monitor) even slightly during the set of

experiments! Especially during the acquisition, but also between acquisitions. Be very careful about that, otherwise you'll have to adjust arenas during tracking each time you moved the monitor, and it is extremely boring and would make the EthoVision file messy.

9. After the full set of experiment – carefully switch off the majority of whatever you switched on. Place the "used" tadpoles back into the same bowl where you took them from (this actually needs to be discussed – I guess it depends on how soon somebody would use them again for any experiments).

Tracking in Ethovision Tracking can be done either offline (after you recorded the videos), or on-‐the-‐fly, or actually both (draft tracking during the acquisition, and some re-‐tracking later if necessary). The procedure below is more applicable to offline tracking, but on-‐the-‐fly tracking is similar.

1. Go to "Arena" section, duplicate one of existing arenas, and modify (move) it, insuring that it covers full Petri dish bottom, but not its sides (as otherwise the system will track the reflections of the dot in the sides of the dish). Name this new arena in the same clever way, so that you don't get lost, and can re-‐use the arenas later if you need to re-‐track the videos. One approach would be to name them in the same way you named your trials.

2. Depending on the size of the dot you'll be tracking, choose corresponding "Detection settings" (I here assume that they are already pre-‐created and named properly).

3. To track a video file for the first time: go to "Acquisition", press a "sunny star" button (or whatever it is) to create a new trial; select a video file for your trial; ensure that "Speed is defined by tracking progress" checkbox is checked, and start the tracking (by pressing a "green sphere" button). – This point is slightly different if you are re-‐tracking an old trial, but I'm not sure how exactly it should happen, so thing aspect of the process is not covered intentionally.

4. Look at the screen while it is tracking. If it does a bad job – change something. If it is OK – just use this opportunity to prepare yourself for future manual clean-‐up. If you notice something obvious (mislabeling, missed collision, wrongly processed occlusion or something of this kind) – make a note (roughly indicating the time when it happened), so that you can return to that later.

Track Editing 1. Use both the keyboard (Ctrl-‐left and right arrows and other shortcuts, as described in the

Official Manual), and the mouse. See the EthoVision manual: sections "Control the playback" and "Track Editing" for the lists of keyboard shortcuts. Make the program draw 1 second trace behind the dots – it helps to see what is happening.


of 37

2. Go through the recording, looking for 1) Collisions that aren't tracked properly, 2) Non-‐collisions that were tracked as if they were collisions: those cases when the system did not track the objects properly, and the tracks were placed close enough to each other to look like collisions. Assume that all the subsequent stages of processing will be blind (only tracks, no video) and fully automated.

a. For each collision (each case when the tadpole and the dot were closer than about one third dish radius from each other, even if there were no escape response) clean about 1 s before and 1 s after their encounter.

b. Look through the whole recording, and if the system wrongly tracked the objects (for example placing the "Tadpole" on the rim of the "Dot"), creating a "Fake collision", just delete these erroneous areas. The tracks should not contain any "close encounters" between the "Tracked tadpole" and the "Tracked dot" that are not real collisions.

c. The quality of the tracks everywhere else mostly irrelevant. But if you see a bad case of mistracking, it's better at least to delete it.

After you're done with editing, export all the tracks to some sane location. Don't forget to name the folder properly, so that we can identify these tracks later on.


of 37

Tailflicks

Experiment 1. Prepare 2% low-‐melting point agarose gel (0.20g low-‐

melting agarose in 10ml Steinberg’s) in a small beaker, and heat it at 200-‐300°C on the stove until boiling (microwave is much quicker, but dangerous, as once the boiling point is reached gel runs out from the beaker almost instantaneously).

2. Anesthetize some Stage 48-‐49 tadpoles in MS-‐222 for about 5 minutes.

3. Once the gel cools to a mild temperature (feels warm, and not hot, but is still liquid), transfer tadpoles into Steinberg’s (to rinse off MS-‐222) and then individually onto small Silgard platforms. Remove excess solution using a transfer pipette but don’t dry them out completely (e.g. using Kimwipes). If Silgard platforms are carved, it works better, but flat platforms are also OK. The temperature of the gel is the first critical point of the whole process.

4. Cover each tadpole with the gel, taking care to keep the tadpoles on the platform. 5. Once the gel solidifies a bit, add a second layer to make sure all parts of the tadpole are

covered. And then maybe a third one. 6. After 5 minutes of drying, choose an eye to clear from the

gel. If the animal is slanted (and ideally it should be!) choose the eye that points upwards. Using a razor, take the gel off from the tail, from the chosen eye, and from around the mouth. You may also try to clear under the mouth. Leave a lot of gel on the right side, and in the left bottom corner, to keep the tadpole secured. The gel carving should be done quickly, as tadpole is drying and suffocating all this time. This is the 2nd most critical point of the whole process.

7. Finally, fill each container with Steinberg’s. Usually, the tadpoles start swimming immediately. 8. Turn on Master-‐8, computer, and Matlab. Plug in IR light, and connect the green light to the

battery. Turn on Motion Studio (click CamerasOK open). Set Master-‐8 to program 1 (All, 1, enter). Turn on LCD and LED switch.

9. I place all swimming tadpoles into the box immediately, to allow them time to adjust. Generally I leave the flap open so there is as much light as possible inside the box.

10. Place one container on the glass, hold it in place using wax, and aim the fiber towards the cleared eye. The camera/motion studio can be used to visualize in two dimensions, and I look


of 37

through the side of the container to set the third dimension. (When the LCD is off, red light is reflected off the eye once the fiber is in the correct position).

11. Adjust the Motion Studio window (and/or physically move the camera) to capture the tadpole. 12. The position of the fiber and the responsiveness of the tadpole can be verified by testing crash

or grid (with q = 1). Set Motion Studio to “live” to watch for tail flick. 13. To start, set q = 0 on Matlab and enter the order of stimuli. I generally chose at random but

tried not to use the same sequence twice. The stimuli are C (crash), F (flash), G (grid), B (back-‐crash), 0 (control), and R (ramp). On Motion Studio, set the image folder name (the date) or

other settings by clicking the button. Then click record. Hit enter on Matlab.

Default settings: Matlab will run through 200 stimuli in the order selected. Motion Studio will save them as FolderName_001 through FolderName_200. [I use the date, in MMDDYY format, for the folder name]. The interval between the onset of two stimuli is 45 seconds. The stimuli are 2 seconds long16. Program 1 on Master-‐8 is set to trigger the camera 0.5 seconds after channel 1 is activated and to trigger Matlab 1 second after channel 1 is activated. The camera records at 100 Hz.

Processing 1. Open the file in ImageJ17 2. Choose “Use Virtual Stack” and “Convert to Grayscale,” then click OK.

3. In Image J window, double-‐click the button and choose “Auto-‐Next Slice” (in addition to “Auto-‐Measure” that should be selected by default)

4. Enlarge the video screen and click on the tip of the tail. There should be 300 frames per video for the settings we used after 10/16/11, so repeat 300 times. The program will automatically record positions for the 300 frames. At the very end, also click the center of the gut (to get a reference opint).

5. Select all and copy the positions. 6. Paste the positions to the Excel raw data file. ImageJ produces positions in this format:

# Label Area Mean Min Max X Y Slice The “Label” contains the time of each frame. "#" should equal “Slice” if no frames are skipped or clicked twice. The gut point will be #301 but slice 300. X and Y are the key variables.

7. Recalculate XY coordinates into tail angle.

In rare cases, the angle of the tail will be greater than 90 or less than -‐90. Be careful with the formula.

Repeat this for each video. Organize the measurements by stimulus type in the Excel file "DK Data" on the Motion Studio computer. The first tab in this file contains the key. The naming convention in 16 Actually, the stimulus onset varies but is usually 40-‐70ms late (converter/computer delay). The end of the stimulus (light on) is usually faster and less than 40ms late (or on time). Therefore, the stimuli are a bit shorter than 2s when working with w = 0. The delay time was tested on 8/8/11 (video “Test_014”) and on 9/19/11 (videos “Test_017” through “Test_019”). Some of the results are in the Excel file “Delay calculations.” There is no camera delay (tests not saved, 6/3/11) but the first image is at t=1ms. 17 All files recorded after 10/16/11 open directly in ImageJ. Earlier files were (or must be) first converted to a compatible format using the programs VLC Media Player [profile: video-‐H.264; AAC(MP4)], Avidemax [video: YV12 (raw)], and Virtual Dub [Save As: old format AVI], in that order.


of 37

original "DK Data" files is different: the name of each tab is YYMMDDNNN, where NNN is the three-‐digit number of the tadpole for the day (the first tadpole of the day is 001, for example). This format change is for Matlab compatibility.

Preparing random video files for ImageJ Image J can only read very simple video files, and there are thousands of codecs out there, so there's absolutely zero chances that it would read the files you need. In the worst case (you can play your files, but none of the programs you have opens them) you'll have to follow this procedure:

1. Open the file in VLC Media Player (that presumably plays the files). Choose "Save / Convert" from the File menu, and save it as a "Video-‐H.264 + AAC (MP4)" profile.

2. Open this file with AvidDemux, and save it as video YV12 (raw). If you can crop it – it is highly advisable, as ImageJ would try to load your video into its limited memory, and is rather likely to fail. So you'd better crop it. You can only use dimensions that are multiples of 5 px.

3. Open the saved file in VirtualDub. Save is as an "Old Format AVI" (there's a special entry for that in the File Menu".

4. Now you can open it in ImageJ (again, if it would have enough memory).

Presumably, there's a way to tell ImageJ that it should grab a lot of virtual memory at startup, but I don't know how to do that. If you face this problem – try reading ImageJ specifications.


of 37

Miscellaneous Pieces of Practical Wisdom If you start a recording, but there's no noise and no stimulation artifacts – the Clampex is probably in the "Demo mode". Maybe you started it when the digitizer was off. You'll have to find the digitizer again – there's a special dialog for that (look somewhere in the "Options" menu).

If the noise is high, and you can't get the voltage clamped – you might have forgotten to put the ground electrode in the dish, or the ground wire is broken. Replace it.

If the baseline moves all the time (especially after you just changed the holding voltage), the wire needs to be re-‐chlorided.

If you think you use same pair of solutions you used yesterday, but the pipette offset differs a lot from the one you used to have yesterday, either your solutions are wrong, or the wire needs to be re-‐chlorided.

If you have some strong noise – maybe you ground got broken. Or maybe somebody is making a prep right now, with this beautiful powerful light source being on, as it creates some tremendous amount of electromagnetic noise.

If the brain moves slowly, and you cannot hold the cells, having to adjust the position of the tip all the time, you may have overstretched the brain while pinning it down to silgard. Or maybe the osmolarity of your solution is all wrong.

If your brain becomes damaged and screwed, getting wound up on your pin when you try to pin it, it just means that the pin is bad. The forceps left some scratches on it, and now it acts as a screw or a drill, rotating the tissue it is being stick into. Just make a new pin. To make one, take a long sharp pin, put it into Silgard, hold the upper part with a forceps, and cut with scissors at a desired length.

By the way, only about one quarter of pins that are shipped to us are actually sharp. The majority of pins in these boxes are blunt from both ends, so you may want to spend some time selecting good pins for your prep.

When starting the experiment, never put the stimulating electrode down into the tissue before switching all the electronic boxes on. When you switch them on (especially if you do it in incorrect sequence; correct generally being from the center towards the periphery for stimulating branch, and from the periphery to the center for recording branch), you may pass strong currents through the electrode. If you place it in the tissue, and then switch everything on, you would fry the tissue, and you'll see electrolysis bubbles near the tip of the electrode.

Manipulators make noise when they move. This noise may resemble spikes, so don't be confused.

In the "Ca image rig", the full turn (100 units) of the fine focus knob seems to be equal to 117 µm (based on measurements of a piece of plastic from a Petri dish lid).

General procedures 130718 - BrownBlogs · AizenmanLab!Procedures.!Rev.!!Jul18!2013! Page1!of!37!!...

Documents

Transcript of General procedures 130718 - BrownBlogs · AizenmanLab!Procedures.!Rev.!!Jul18!2013! Page1!of!37!!...