Combining whole-cell patch clamp and dye loading in acute brain slices with bulk RNA sequencing in embryonic to aged mice

Summary Single-cell electrophysiological recordings combined with dye loading and immunohistochemistry provide unparalleled single-cell resolution of cell physiology, morphology, location, and protein expression. When correlated with bulk RNA sequencing, these data can define cell identity and function. Here, we describe a protocol to prepare acute brain slices from embryonic and postnatal mice for whole-cell patch clamp, dye loading and post-hoc immunohistochemistry, and cell isolation for bulk RNA sequencing. While we focus on oligodendrocyte precursor cells, this protocol is applicable to other brain cells. For complete details on the use and execution of this protocol, please refer to Spitzer et al. (2019).


SUMMARY
Single-cell electrophysiological recordings combined with dye loading and immunohistochemistry provide unparalleled single-cell resolution of cell physiology, morphology, location, and protein expression. When correlated with bulk RNA sequencing, these data can define cell identity and function. Here, we describe a protocol to prepare acute brain slices from embryonic and postnatal mice for whole-cell patch clamp, dye loading and post-hoc immunohistochemistry, and cell isolation for bulk RNA sequencing. While we focus on oligodendrocyte precursor cells, this protocol is applicable to other brain cells. For complete details on the use and execution of this protocol, please refer to Spitzer et al. (2019).

BEFORE YOU BEGIN
Whole-cell patch clamp is typically used to study neuronal synaptic transmission and firing rate to understand neuronal signaling. However, the bioelectrical properties of cells can regulate cell biology beyond neuronal signaling. For instance, membrane potential is thought to regulate cell cycle progression in proliferative cells (Binggeli and Weinstein, 1986;Urrego et al., 2014). Further, neuronal activity regulates both neural stem cell (NSC) and oligodendrocyte precursor cell (OPC) proliferation and differentiation (Ká radó ttir and Kuo, 2018). Thus, studying the bioelectrical properties of a cell can contribute to our understanding of cell function in glia and progenitors. Combining standard whole-cell patch clamp with dye loading, as we describe here, allows for further morphological characterization of cells, and post-hoc immunohistochemistry can provide information on cell location and protein expression. In addition, these data can inform transcriptomic studies. In particular, if a cell population is found to be physiologically homogeneous within a condition, bulk RNA sequencing can become a simpler, faster, and more economical alternative to patch-seq (single-cell sequencing of RNA extracted following whole-cell patch clamp) or complement patchseq data (as it provides deeper sequencing, and therefore better genome coverage) to characterize cell transcriptome and correlate bioelectrical properties with cell function.
Here, we focus on OPCs, which are equally distributed in the brain throughout life, but have altered proliferation and differentiation potential, physiology, and transcriptome with age (Dawson et al., 2003;Spitzer et al., 2019;Young et al., 2013). Thus, we describe a protocol to prepare acute brain slices in embryonic, neonatal, and adult mice for whole-cell patch clamp and dye loading, perform post-hoc immunohistochemistry on patched cells, and isolate cells for bulk RNA sequencing. We 9. Prepare a 20 mL stock of 2M of D-gluconic acid in ddH 2 O.
Note: We purchase D-gluconic acid as a 51% solution in H 2 O (see key resources table). When calculating the molarity of the purchased solution, it is important to consider the density indicated by the supplier. Keep the 2M stock solution at 4 C, as it is used to make up internal solution.
10. Combine 2M CsOH and 2M D-gluconic acid (1:1) to make a 10 mL 1M Cs-gluconate stock. We use Cs-gluconate to prepare internal solution (step 17). Keep this stock solution at 4 C. We prepare 10 mL, as this is sufficient to make three internal solutions. Discard this solution after one year.
Note: We use either K-gluconate or Cs-gluconate internal solutions, depending on the experiment. Thus, we typically have all the above stock solutions. However, if exclusively using a Kgluconate based internal solution, omit steps 8-10. Conversely, if only using a Cs-gluconate based solution, omit steps 6 and 7.

Prepare drug aliquots for patch clamp
Timing: 2 h, months to days before the recording day Here, we describe how to prepare and store stock solutions for various drugs we commonly use to record from OPCs. However, these are specific to our experiments, and may not always be needed. We keep these stock solutions for several months, depending on usage.
11. Prepare a 5 mL stock of 1M glycine in ddH 2 O (dissolve 0.375 g of glycine in ddH 2 O). We use glycine in our bath solution when recording NMDA receptor currents, as it is a receptor coagonist. Prepare 0.1 mL aliquots (best for 1L of recording solution) and store at À20 C for up to one year. Avoid freeze-thaw. 12. Prepare a 3 mL stock of 50 mM of strychnine hydrochloride in ddH 2 O (dissolve 0.056 g of strychnine hydrochloride in ddH 2 O). We use strychnine in our bath solution when recording NMDA receptor currents, to block glycine receptors as we add glycine as an NMDA receptor coagonist. Prepare 0.1 mL aliquots (for 1L of recording solution) and store at À20 C for up to one year. Avoid freeze-thaw.
CRITICAL: Strychnine hydrochloride is acutely toxic, and ingestion or inhalation can be fatal. When making up 50 mM strychnine hydrochloride stock solution from powder, wear gloves, a mask, and only handle in a fume hood until dissolved. We do not make up more than 3 mL at a time. Always wear gloves when handling solutions containing strychnine.
13. Prepare a 10 mL stock of 1M barium chloride in ddH 2 O (dissolve 2.443 g of barium chloride in ddH 2 O). We sometimes use barium chloride when we wish to block potassium conductance; this is particularly important when recording from glial cells in the adult gray matter. This solution does not need to be aliquoted, and should be kept at 4 C for up to one year.
CRITICAL: Barium chloride is toxic. When making up stock solution from powder, wear gloves, a mask, and only handle in a fume hood. Gloves should be worn when manipulating solutions containing barium.
generally the amount we use for one day of recording) and store at À20 C until the day of use. Avoid freeze-thaw, and store for up to one year. 15. Prepare a $5.66 mL stock of 60 mM NMDA in ddH 2 O (we make up the entire 50 mg vial to avoid weighing out small quantities) in ddH 2 O. Prepare 0.1 mL aliquots (for 100 mL of 60 mM NMDA, which is generally the amount we use for one day of recording) and store at À20 C until the day of use. Avoid freeze-thaw, and store for up to one year.

Prepare internal solution
Timing: 1 day, months to days before the recording day 16. Combine BAPTA, Mg x ATP and Na x GTP in 5 mL ddH 2 O. 17. Add Cs-gluconate (or K-gluconate), NaCl, CaCl 2 , and HEPES. See materials and equipment for the recipe.
CRITICAL: BAPTA will only fully dissolve when the solution is near pH 7.3. Use CsOH (or KOH if using a K-gluconate based internal) to bring the solution to pH 7.3.
18. Add Lucifer Yellow and allow it to dissolve or sonicate for a few seconds. 19. Measure pH and adjust to 7.2-7.4 with CsOH (or KOH if using a K-gluconate-based internal).
Note: If the pH is higher than 7.4, the internal solution should be discarded and a new solution should be prepared. Do not use HCl to adjust pH, as this will alter the Clconcentration in the internal solution.
20. Add ddH 2 O to adjust to the final volume. 21. Measure osmolarity. Typically, osmolarity for this internal will be 290-300 mOsm.
Note: If osmolarity does not fall between 290-310 mOsm, the solution should be discarded and a new solution should be prepared.
22. Aliquot the internal solution into 0.1 mL-0.5 mL aliquots (depending on how much is needed for one experimental day) and store at À20 C until the day of use. Discard aliquots after thawing them.
CRITICAL: While preparing the internal solution, maintain it on ice as much as possible, as ATP and GTP are temperature sensitive.

OPEN ACCESS
Note: If using a perfusion system where solutions do not continuously flow through (for instance, a gravity-fed perfusion with multiple drug lines), we recommend using a HEPES-buffered aCSF for recordings, as bicarbonate-buffered aCSF needs continuous bubbling with 95% O 2 / 5% CO 2 to maintain the pH between 7.2-7.4, and changes in solution pH will affect slice health and make results uninterpretable. If using bicarbonate-buffered aCSF for recordings, omit step 24.
Note: 103 aCSF can be kept at 4 C for approximately one month. If precipitate is found in the solution, discard and prepare a new 103 stock.

Prepare 5% agarose
Timing: 10 min, months to days before the recording day We use 5% agarose when preparing slices from neonatal mice (up to P20) (Figure 1), to stabilize the brain, or in adult mice when the slicing angle requires a support to maintain the brain stable.
25. Dissolve low melting point agarose (2.5 g) in water (50 mL). a. Bring the agarose to boiling point using a hot plate. 26. Pour 50 mL of melted agarose into a 100 mm diameter petri dish (this should fill about three quarters of the height of the petri-dish). 27. Allow the agarose to set at 4 C overnight (10-20 h).
Note: Keep the agarose in the fridge after it is made. Make sure to seal the agarose-containing petri dish with parafilm to prevent it from drying out. We recommend only making one or two dishes at the time, as they can get mouldy if kept for more than three months. Discard immediately if any growth is observed.

Prepare 4% paraformaldehyde (PFA) aliquots
Timing: 2 h, months to days before the recording day We typically prepare PFA in large quantities (1-2L) for transcardiac perfusions, and set aside approximately 250 mL to prepare aliquots to fix patched slices. Thus, if preparing PFA solely to fix patched slices, we recommend preparing 250 mL of solution. This should be sufficient for $50 patch clamp experiments.

Figure 1. Preparing acute brain slices
Diagram showing a side view of slicing through embryonic (left), neonatal (middle) and adult (right) brains to prepare coronal acute forebrain slices. For embryos, we dissect the brains and embed them in agarose before slicing through the block. To prepare neonatal slices, we suggest adding a block of agarose to support the brain. For adult slices, agarose is not necessary, as the brain is stiff enough to slice at low speed, but can be used if preferred.
Note: Pre-heating the PBS in a water-bath set to 50 C can help speed up this step if needed.
29. Add the PFA powder (10 g) to the PBS (250 mL) and heat to 60 C.
CRITICAL: PFA is harmful and carcinogenic. Handle with care. Wear gloves, and prepare the solution and aliquots in a fume hood.
30. Remove the solution from the hot plate, and add a few drops of NaOH until the PFA is fully dissolved and the solution is clear. 31. Manually pH to 7.2-7.4 using NaOH (or HCl).
CRITICAL: HCl is highly corrosive. Wear gloves, and only use in a fume hood.
32. Prepare 5 mL aliquots (sufficient to fix 10 slices) and store at À20 C for up to a year.
Note: Avoid freeze-thaw. Once thawed, each aliquot should be used within the day, and discarded at the end of the day.

Prepare stock solutions for immunohistochemistry
Timing: 30 min, months to days before the recording day 33. Prepare a 500 mL stock of 5% NaN 3 PBS (dissolve 25 g of NaN 3 in PBS). This solution can be diluted to prepare the 0.05% NaN 3 PBS in which we store brain slices after fixation. Store this solution at 4 C until the day of use.
CRITICAL: NaN 3 is highly toxic and competes for the O 2 binding site on haemoglobin. Wear gloves and manipulate in a fume hood. Manipulate solutions containing NaN 3 with gloves.
34. Prepare 100 mL of 1 ng/mL DAPI in 0.05% NaN 3 PBS. Store this solution at 4 C until the day of use.
Prepare stock solutions for OPC isolation with magnetic associated cell sorting (MACS) Timing: 1 h, up to a month before the experimental day 35. Prepare a stock of 4 mg/mL DNase in ddH 2 O. Prepare 0.55 mL aliquots and store at À20 C until the day of use.
Note: Avoid freeze-thaw. Once thawed, DNase aliquots can be kept at 4 C for up to one month.
36. Prepare ovomuccoid (see materials and equipment for the recipe). a. Combine DNase, BSA and trypsin inhibitor in DMEM.
Note: Ovomuccoid can be kept at 4 C for up to one month.
Prepare 13 bicarbonate-buffered aCSF for slicing and resting Optional: Omit step 43 if using bicarbonate-buffered aCSF for recording, as pH will already have been set by bubbling.
CRITICAL: MgCl 2 should be omitted when focusing on NMDA receptor currents, as Mg 2+ blocks NMDA receptors at holding potential (À74 mV).
CRITICAL: If recording NMDA receptor currents, add glycine (an NMDA receptor coagonist) and strychnine (to block glycine receptors) to the recording aCSF. See steps 11 and 12, and materials and equipment for details.
Note: Recording aCSF can be kept for up to one week at 4 C.
Chloride a silver wire electrode and ground electrode Timing: 1 day, before the recording day 44. Place a silver wire and the ground electrode pellet in bleach overnight (10-20 h). 45. Rinse the silver wire electrode and ground pellet in distilled water before placing them in the electrode holder and the recording bath, respectively.
CRITICAL: The recording electrode should be chlorided regularly, depending on the frequency of experiments. The electrode should be chlorided if it is not a uniform dull grey and shiny spots are showing, indicating that the chloride coating is wearing off. Similarly, the ground electrode should be checked regularly and chlorided when necessary.

Prepare solutions for OPC isolation with MACS
Timing: 1 h 30, on the experimental day 46. Prepare 0.5% Bovine Serum Albumin (BSA) in PBS. a. Prepare $50 mL of BSA per brain. b. Filter the buffer with a vacuum filter, and keep the vacuum on until all bubbles have disappeared, as bubbles can block the columns. c. Store at 4 C until needed.
Note: While BSA can sometimes be difficult to dissolve, 0.5% BSA should dissolve promptly, without vortexing.
Note: While we typically prepare BSA on the experimental day, any remaining solution can be kept at 4 C for future use, for up to one month.
47. Prepare 30 mg/mL L-cysteine in DMEM. Keep at 4 C until needed. Discard any remaining Lcysteine after use. a. Prepare 0.2 mL L-cysteine per brain. 48. Prepare papain dissociation medium (see materials and equipment for the recipe). a. Prepare 1 or 2 mL per brain (see step-by step method details step 45). b. Combine papain and DNase in DMEM. c. Omit L-cysteine, which is only added 10 min prior to the experiment (see step-by step method details step 45a(i)).
Note: Papain dissociation medium should be prepared the day of the experiment, and kept at 4 C until immediately prior to the experiment. Any remaining medium should be discarded. Other Dissection scissors (large and small) n/a n/a Glass beakers n/a n/a Glass petri dish n/a n/a Forceps (curved and straight) n/a n/a Spatula (large and small) n/a n/a Note: The expected osmolarity of this solution is 315-330 mOsm.

KEY RESOURCES
CRITICAL: 13 slicing and resting solution can be made up weekly and stored at 4 C. pH should be equilibrated at the start of each experimental day with at least 30 min of bubbling with 95% O 2 / 5% CO 2 .
Optional: Bicarbonate-based aCSF can also be used for recordings. For that purpose, prepare a separate 13 stock and omit kynurenic acid. Make sure to add the correct MgCl 2 concentration for the experimental design.
CRITICAL: 103 HEPES-buffered aCSF can be used for up to a month and should be stored at 4 C. Discard the stock solution if any precipitate is observed.
Note: The osmolarity of HEPES-buffered recording aCSF is expected to be 315-330 mOsm.
CRITICAL: 13 recording solution can be stored at 4 C and used for up to a week.
Optional: If recording NMDA receptor currents, add the following to the recording aCSF immediately prior to recording: CRITICAL: Strychnine hydrochloride is acutely toxic, and ingestion or inhalation can be fatal. Gloves should be worn when manipulating any solutions containing strychnine.
Optional: If recording in adult grey matter, add the following to the recording aCSF immediately prior to recording: CRITICAL: Barium chloride dihydrate is toxic. When making up stock solution from powder, wear gloves, a mask, and only handle in a fume hood. Gloves should be worn when manipulating solutions containing barium.
Note: We add 25 mg Lucifer Yellow (one vial) into 25 mL of internal solution, as this is the simplest method, and an exact final concentration is not critical.
Note: The expected osmolarity for this internal solution is 290-300 mOsm.
CRITICAL: Internal solution should be aliquoted and stored at À20 C for up to one year. The aliquot size should correspond to the amount of internal solution needed for one day of recording (typically between 0.1 mL and 0.5 mL). Once thawed, internal solution should be kept on ice and protected from light, and should be discarded at the end of the day.
Alternatives: Neurobiotin, Alexa Fluor dyes or fluorescent-conjugated dextran can be used as alternatives to Lucifer Yellow.
Alternatives: Cs-or K-methanesulfonate can be used instead of Cs-or K-gluconate in this recipe. Cs-or K-chloride can also be used, but a high intracellular chloride concentration will affect the chloride reversal potential, and may result in inhibitory inputs depolarizing the cells.

STEP-BY-STEP METHOD DETAILS
Preparing acute brain slices

Timing: 2 h
Here, we describe how to prepare acute brain slices from embryonic (E13-E20) or postnatal brains (RP0) for patch clamp recording. We use both male and female animals.
1. Before you begin, prepare for the dissection and slicing. a. Bubble the slicing aCSF with 95% O 2 / 5% CO 2 for at least 30 min (to set pH to 7.2-7.4), on ice. b. Calibrate the vibratome blade according to the manufacturer's instructions. c. Place the slicing chamber in the freezer for at least 5 min. i. Once it is cooled, place the slicing chamber on the vibratome, and fill the outer chamber with ice. d. Once you have bubbled the slicing aCSF for 30 min, fill the chamber with ice-cold slicing aCSF, and continuously bubble the slicing solution with 95% O 2 / 5% CO 2 until the last slice is cut (step 6), so that the pH of the aCSF is maintained throughout the procedure. e. Cool two beakers (one 10 mL and one 50 mL) and a glass petri dish on ice. f. Fill the resting chamber with slicing aCSF and continuously bubble with 95% O 2 / 5% CO 2 until the end of the recording day. g. Fill the two beakers with bubbled slicing aCSF (after the 30 min of bubbling).
Optional: When slicing from young postnatal brains (up to P20), we suggest using a block of agarose to support the brain while slicing. This is also useful when slicing from brains regions where the required slicing angle is not readily maintained (such as the cerebellum), but is usually not needed when cutting coronal forebrain slices from adult mice, as the brain is stiff enough to remain stable when slicing at low speed. If you choose to use agarose, cut out a cube of 5% agarose and glue it on the vibratome slicing platform. Make sure that the brain will rest against a flat surface (Figure 1). Optional: If preparing embryonic slices, dissolve 1 g low melting point agarose in 50 mL of 13 recording aCSF by bringing the agarose to boiling point using a hot plate. Then, cool down the 2% agarose to avoid damaging the tissue when embedding, but do not allow it to set. This can be done while the slicing aCSF is bubbling (step 1a).
2. Humanely sacrifice an animal in accordance with local regulations.
Note: When preparing embryonic slices, sacrifice the pregnant female first, then dissect the embryos.
3. Remove the head and immerse it in the 50 mL glass beaker containing bubbled slicing aCSF. 4. Dissect the brain. a. Cut the skin along the midline, exposing the skull, and use it to hold the head by folding it under the head. b. Make a cut between the eyes, make lateral cuts in the skull anterior to posterior, and remove the skull. Alternatively, make a cut near the midline above the hindbrain, peel off the skull above the brainstem and cerebellum, and then make lateral cuts in the skull above the forebrain, posterior to anterior, and peel off the skull. c. Carefully extract the brain using a small spatula. For best results, dip the spatula in slicing aCSF first. d. Immerse the brain in the 10 mL glass beaker containing bubbled slicing aCSF. e. Fill the glass petri dish with bubbled aCSF and transfer the brain into the petri dish using the large spatula. Keep the petri dish on ice.
Optional: For postnatal brains, dissect the meninges, and then dissect the region of interest (for instance, the forebrain).
Optional: For embryonic slices, fill one well in a 4-well plate halfway with the melted agarose. Place a dissected brain in the well and add more agarose to fill the well. Ensure that the brain is correctly positioned for the required slice orientation (coronal/sagittal/transverse). Repeat this procedure for each brain. Place the plate on ice or in the fridge to allow the agarose to set rapidly. We recommend preparing embryonic slices with two people: one person should dissect while the second person should embed the brains.
5. Glue the brain to the vibratome slicing platform in the correct orientation for the required slices (Figure 1). Consult a brain atlas to determine the right slicing orientation to keep axonal fibers intact in the region of interest.
Optional: For embryonic slices, once the agarose has set, cut a block around the embedded brain and glue this block to the vibratome slicing platform (Figure 1).
6. Cut 225 mm-325 mm thick slices. Cut enough slices for a full day of recording.
Note: Slice thickness is a compromise between axonal integrity (thicker slices allow for better axonal preservation) and cell visibility, especially when patching small cells and using fluorescent reporter lines. For instance, when recording from OPCs in postnatal brains, we typically use 225 mm thick slices. For embryonic slices, we recommend cutting 325 mm thick slices, as embryonic slices are fragile, and slice integrity is not always maintained in 225 mm thick slices. Embryonic slices are more transparent, as they are not myelinated, and thus, cell visibility is better, even with thicker slices.
7. Transfer each slice into the resting chamber using the glass Pasteur pipette. 8. Allow the slices to rest for 1 h before recording.

OPEN ACCESS
CRITICAL: Time is crucial when preparing acute brain slices. If the delay between sacrificing an animal and placing the brain in the ice-cold aCSF-filled vibratome chamber is too long, the slices may not be healthy enough, which will make patch clamp recordings particularly difficult. Troubleshooting 1.
Whole-cell patch clamp recording in reporter mice with dye loading Timing: 8-10 h Here, we describe how to record from OPCs in acute brain slices from fluorescent reporter mice such as NG2-EYFP, NG2-dsRed, or PdgfraCre ERT2 :tdTomato using whole-cell patch clamp. OPCs can be identified in acute brain slices without any fluorescent label, but this requires a trained eye, and posthoc immunohistochemistry to confirm cell identity (for more detail on identifying oligodendrocyte lineage cells for patch clamp, see Agathou and Ká radó ttir (2019)). The following steps can be used for recording non-neuronal cells in acute brain slices with or without the use of reporter mice.
9. Prepare the pipettes and patch clamp set up for the day. a. Fire-polish both ends of glass capillaries using a Bunsen burner (unless you are buying in firepolished glass capillaries). b. Use the fire-polished capillaries to pull glass electrodes using an electrode puller. For OPCs, we recommend pulling electrodes with a 5-7 MU resistance.
Note: It is possible to use electrodes with resistances as low at 4.5 MU to patch OPCs, but this is very challenging, as they tend to have almost the same diameter as most OPCs. A resistance of 5-7 MU is a good compromise between ease of seal formation, ease of cell opening, and series resistance during the recording. Troubleshooting 2 and 4. c. Thaw an aliquot of internal solution and keep it on ice and protected from light.
Note: To record from OPCs, we typically use Cs-gluconate or K-gluconate based internal solutions. Cs-gluconate allows for more stable recordings than K-gluconate, as it blocks some K + channels (Hille, 2001). We tend to use Cs-gluconate when recording NMDA currents as these currents can be very small, and could be masked by K + channel noise. Cs-gluconate internal is also recommended to record voltage-gated Na + currents. When recording K + currents, a Kgluconate based internal solution is more appropriate. d. Thaw an aliquot of 4% PFA and keep it at room temperature (19 C-23 C) for fixing slices after recording.
CRITICAL: PFA is harmful. Handle with care. Wear gloves and use in a fume hood or under an air extraction vent. e. Turn on all the electrical equipment. f. Oxygenate the recording aCSF with 100% O 2 (when using a HEPES-buffered recording solution; for a bicarbonate-buffered aCSF, bubble with 95% O 2 / 5% CO 2 ) and start running the solutions through the gravity-fed perfusion. Ensure that the lines are free of air bubbles. Troubleshooting 6. Bubble the solutions throughout the recordings.
10. Transfer a slice from the resting chamber into the recording chamber using a glass Pasteur pipette. a. Use a harp (a thin U-shaped non-conductive metal strung with nylon, for instance) to hold the slice in position at the bottom of the recording chamber. 11. Use the 4/53 objective to locate the slice under the microscope, and identify the region of interest ( Figure 2).

OPEN ACCESS
a. Use the fluorescence to identify reporter-expressing OPCs. b. Switch back to brightfield to identify a good OPC to attempt patch clamp.
Note: Under IR-DIC, good OPCs appear smooth, with a soft and clear outline, while shiny cells with a hard outline should be discarded as they are difficult to seal. Round cells with a visible nucleus should be avoided, as they are dying cells. Cell location (proximity to other cells, depth in the slice) cannot be assessed when using fluorescence only. Thus, it is best to alternate between fluorescence and brightfield until a suitable cell is identified (healthy, with good access and good slice depth).
Note: When using fluorescent reporters under the NG2 promoter, we avoid cells adjacent to blood vessels, as pericytes also express NG2. Pericytes are covered by vascular basement membrane and are therefore much more difficult to seal than OPCs, requiring smaller pipette tips and/or pre-incubation of slices in collagenase, as well as a longer time to seal (Mishra et al., 2014;Peppiatt et al., 2006). Nonetheless, the current-voltage relationship (IV) differs between OPCs and pericytes, with pericytes displaying a different capacitive current and potassium conductance (Sakagami et al., 1999). It is therefore particularly important to apply a voltage step protocol (see step 27) at the beginning of each recording, as each cell type have their specific ion channel expression and IV. We recommend consulting several IVs for the cell type of interest before starting patch clamp experiments. c. Mark the location of the cell on the screen to easily locate it once the electrode is in the bath.
13. Fill a glass pipette with internal solution and fit it in the electrode holder on the micromanipulator headstage so that the chlorided electrode is in contact with the internal solution. 14. Raise the objective so that it is still in contact with the bath solution, but there is enough space to lower the electrode into the bath. 15. Lower the electrode into the bath, and begin applying a G5 mV (or 10 mV) 15 ms voltage pulse every 30 ms. Display the resulting current on your computer monitor using Clampex (or another appropriate software). Continue these pulses until you have achieved whole-cell configuration (Figure 3). 16. Adjust the pipette offset current to zero on the amplifier. 17. Apply positive pressure to prevent debris from collecting at the tip of the electrode. We use a lab-made system where we apply positive pressure by blowing into a 1 mL syringe connected to the headstage by tubing.
Note: This initial positive pressure can be achieved by blowing into the syringe and locking the valve connecting the syringe to the tubing.
18. Find the electrode under the microscope, and lower the electrode and objective until you are just above the cell of interest. 19. Increase the positive pressure by continuously blowing into the syringe as you approach the cell. 20. When the electrode contacts the cell, a dimple should appear and resistance should slightly increase while current decreases. Switch from positive pressure to negative pressure (achieved by suction into the syringe). 21. Set the holding potential to À60 mV on the amplifier.
Note: We measured a junction potential of À14 mV between our internal solutions and recording aCSF. Corrected for junction potential, a holding potential of À60 mV on the B A C Figure 3. Sealing and opening a cell (A) When the electrode is in the bath, applying a À5 mV voltage pulse leads to current flow across the pipette resistance (R pipette ) only. This current can be calculated using Ohm's law, V=IR.
(B) Once a giga-ohm seal is achieved, the same voltage step will elicit current flow across the pipette resistance and the seal resistance (R access ). We use Ohm's law to calculate this current, but it is near 0, as R access is very large. In most cells, the formation of a giga-ohm seal is not instantaneous. It can be monitored by tracking the change in current (and resistance) with (A) as a starting point and (B) as the end point. (C) Once the cell is open, current flows across the pipette resistance, access resistance (adding up to the series resistance, R series ), the membrane resistance (R membrane ) and the membrane capacitance (C m ). Analysis of the capacitive current elicited by a À5 mV step allows the calculation of R series , R membrane and C m (see Figure 8).

OPEN ACCESS
amplifier is a À74 mV holding potential. However, junction potential should be measured experimentally for each internal-external solution combination, and the holding potential on the amplifier may need to be adjusted.
22. Continue to apply negative pressure until a giga-ohm seal is achieved. Monitor the formation of a seal on Clampex, using the continuous voltage pulses and paying close attention to the resistance, which should reach at least 1 GU, and the resulting reduction in current ( Figure 3B). Troubleshooting 2 and 3.
Note: Different OPCs seal at different rates. For some cells, simply switching from positive pressure to light negative pressure is enough to obtain a giga-ohm seal. However, other cells require sustained negative pressure for 30 s-1 min before a seal is achieved. The negative pressure will have to be adapted for each cell. This can be gaged by monitoring the increase in resistance and decrease in current.
23. Once a giga-ohm seal is achieved, compensate pipette capacitance on the amplifier ( Figure 3B). Follow the instructions in the manufacturer's manual to apply compensation. 24. Apply short, hard, negative pressure pulses to open the cell. Once the cell has opened, a membrane capacitive current will be visible in response to the voltage pulses ( Figure 3C). In addition, the Lucifer Yellow-containing internal solution will be visible in the cell (Figure 2). Troubleshooting 4 and 5.
Note: As with sealing, different OPCs will require different negative pressure to open. Be cautious, as too much negative pressure will break the seal and the cell will have to be discarded. However, some cells will require many pulses before opening. OPCs in slices from aged mice are typically harder to open than OPCs in slices from neonates. Troubleshooting 4 and 5.
25. Apply a 50 ms G 5 mV voltage pulse. From this pulse, we calculate R s , membrane capacitance (C m ) and membrane resistance (R m ). See quantification and statistical analysis for more detail. 26. Compensate series resistance (R s ) (according to the instructions in the manufacturer's manual). If recording large current changes, it is critical to compensate for R s . Monitor R s throughout the recording by regularly applying a 50 ms G5 mV voltage pulse.
Note: We recommend recording from cells with an R s % 25 MU. However, if R s is larger than 25 MU, but the cell has a stable baseline and there is minimal leak current (see step 28), recording from OPCs with R s % 40 MU is still acceptable. If R s is larger than 40 MU, the cell should be discarded. Similarly, if R s changes by more than 20% throughout the recording, the cell should be discarded.
27. Apply a protocol delivering voltage pulses to step the membrane potential from À120 mV to +40 mV (using 10 mV or 20 mV steps) and record the current responses. From this protocol, you can determine the ion channel expression in the recorded cell. While we found that OPCs are heterogeneous with respect to their voltage-gated ion channel expression (Spitzer et al., 2019), this profile can be used to distinguish them from more mature oligodendrocytes if necessary (see Agathou and Ká radó ttir, 2019). See quantification and statistical analysis for more detail on quantifying voltage-gated Na + channel current responses.
Optional: When recording in the grey matter of adult animals, we recommend switching to a recording aCSF containing BaCl 2 after the first voltage step protocol. This will block potassium conductance, which has been shown to increase with age (Maldonado et al., 2013). This is particularly important if comparing bath applied agonist responses in neonates and adults; however, avoid using BaCl 2 when investigating K + currents. 28. Apply any stimulation or voltage protocols and drugs required to answer your research question. For instance, in Spitzer et al., 2019, we typically applied NMDA, then kainate, as we were studying changes in NMDA and AMPA/kainate receptors with age and brain region. See quantification and statistical analysis for more detail on quantifying receptor agonist responses. a. Once you have recorded a stable baseline, bath apply NMDA first. If the cell responds, keep NMDA on until a stable peak is reached or the response begins to recover (indicating receptor desensitization). Wash out with recording aCSF. If the cell does not respond, leave NMDA on for 4-5 min to confirm that the lack of response is due to an absence of NMDA receptors, and not receptor desensitization due to a decrease in perfusion flow rate. Troubleshooting 6. b. Once the cell has fully recovered from an NMDA response and you have acquired a stable baseline, bath apply kainate. Keep kainate on until a stable peak is reached.
Note: If the recording becomes unstable and does not recover within a few minutes, the seal may have broken, or the cell may be dying. The recording should be stopped, as the results may become uninterpretable. Similarly, if the leak current is larger than À400 pA (in Cs-gluconate; À200 pA in K-gluconate), the recording should be stopped.
29. Once the recording is done, capture an image of the Lucifer Yellow filled cell ( Figure 2). Switch back to the 4/53 objective to capture an image of the pipette location on the slice ( Figure 4A). Record this in your lab book. This will help to locate the cell when imaging post-hoc antibody staining. 30. Remove the electrode from the cell, trying to keep the cell in the slice. Depolarizing the cell to 0 mV may help.
Note: For cells with a high seal resistance (>10 GU), we recommend slowly pulling the pipette away and upwards. A thread of membrane should appear between the electrode and soma, and detach from the cell body, leaving it intact in the slice. If this is not sufficient, or for cells with lower seal resistance, positive pressure or moving the pipette sideways may help to detach the electrode from the cell body. When neither of these methods work, pulling the pipette away quickly, or gently taping the stage may help, although the dye may leak out. In addition, long recordings make it significantly more difficult to keep the cell in the slice when removing the electrode.
31. If attempting your next cell on the same slice, repeat steps 12-30. We recommend changing slices approximately every hour, as slice health will deteriorate in HEPES-buffered recording aCSF. If using drugs that are difficult to wash out, or bind irreversibly, change slices after each cell. 32. If changing slices, repeat steps 10-30.

OPEN ACCESS
33. Note the slice orientation in your lab book. It may be necessary to make a mark on symmetrical slices. Use a disposable scalpel to do so. This will be helpful for post-hoc immunohistochemistry. 34. Immediately fix the finished slice in 4% PFA in a 24-well plate for one hour at room temperature (19 C-23 C) on a rotating shaker, or overnight (10-20 h) at 4 C, on a rotating shaker. After fixing the slices, wash 3 3 30 min in PBS and store at 4 C in 0.05% NaN 3 in PBS.
Pause point: When kept in these conditions, fixed slices can be used for up to a month after patch clamp with successful recovery of Lucifer Yellow filled cells. However, for best results, we recommend performing immunohistochemistry within a week following patch clamp and slice fixation.

Timing: 2-3 days
Here, we briefly describe how to use immunohistochemistry to confirm the identity of patched cells, or obtain more information on the cells, such as location (with layer specific markers for instance) or proliferative state. As this has previously been described in great detail (please consult Karadottir and Attwell (2006) for an in-depth protocol), we simply summarize the key steps.
35. Wash the slices for 15 min in PBS. 36. Incubate the slices in blocking solution (10% goat or donkey serum and 0.5% Triton-X 100, in PBS) at room temperature (19 C-23 C) on a rotating shaker for 4 h. 37. Incubate the slices with primary antibodies in PBS overnight (10-20 h), at room temperature (19 C-23 C) on a rotating shaker.
Note: To confirm OPC identity, we recommend staining with an antibody against your reporter (this is not always necessary, as the endogenous signal may still be detected postfix; however, EYFP fluorescence is quenched by PFA fixation, so using an antibody is important to detect the cells in EYFP reporter mice), or anti-NG2 or anti-Pdgfra, two OPC markers ( Figure 4B). In addition, staining with anti-Olig2 (an oligodendrocyte lineage marker) can help confirm that the cell is an OPC rather than a pericyte (as pericytes are also NG2+). Appropriate markers should be used if performing this protocol with a different cell type.
Note: To confirm cell location, a range of markers can be used, depending on the region of interest. For instance, in Spitzer et al., 2019, we located cells within specific cortical sections.
To do so, we used Cux1, which labels cells in cortical layers 2/3, and CTIP2, which labels cells in cortical layers 5/6 ( Figure 4B).

Wash 3 3 30 min in PBS.
39. Incubate the slices with secondary antibodies in PBS at room temperature (19 C-23 C) on a rotating shaker for 5 h.
Pause point: Alternatively, the secondary antibody incubation can be done overnight (10-20 h) at 4 C on a rotating shaker. Note: To locate Lucifer Yellow-filled cells in brain slices, we recommend using a low magnification objective (203) and scanning the slice through the eyepiece. Recording the cell location after patch clamp should help to confine the search region.

Isolation of OPCs for bulk RNA sequencing
Timing: 5 h Here, we describe how to isolate OPCs from wild-type C57BL/6 mice at different timepoints using Magnetic Associated Cell Sorting (MACS). We briefly list our pipeline for RNA sequencing following cell isolation, but do not detail it, as this is not the focus of this protocol (and thus, is not included in the timing for this step).
Pause point: This section of the protocol is independent from the preparation of acute brain slices, whole-cell patch clamp and post-hoc immunohistochemistry sections. Thus, it can be performed at any given time.
45. Before you begin, prepare for the dissection: a. Prepare a 7 mL universal tube per sample with 2 mL (adult brain) or 1 mL (neonatal or embryonic brains) papain dissociation medium. i. Add L-cysteine to the papain dissociation medium (see materials and equipment) and incubate for 10 min at 37 C. 46. Humanely sacrifice an animal in accordance with local regulations. 47. Dissect the brain (as described in Preparing acute brain slices) and place the whole brain, or the dissected region of interest in a large weighing boat. 48. Quickly cut up the brain in small pieces using a scalpel. 49. Add the tissue pieces to papain dissociation medium.
a. Pool 2-3 embryonic or neonatal brains in 1 mL papain dissociation medium. b. For adult samples, dissociate in 2 mL papain dissociation medium. 50. Use scissors to cut the tissue into smaller pieces. 51. Incubate for 1 h at 37 C. 52. During the incubation, prepare for the myelin removal: a. Cool the ovomuccoid, PBS and BSA on ice. b. Prepare 7 3 15 mL tubes (if using 3 columns per sample) per sample and 1 3 50 mL tube per sample. c. Prepare the LS MACS columns, MidiMACS Separators, and MACS Multistand. 53. Inhibit digestion with 1 mL ovomuccoid. 54. Resuspend the tissue and transfer to a 15 mL tube. 55. Complete to 8 mL with ovomuccoid. 56. Centrifuge for 10 min at 300 g. 57. Wash in 7 mL ice-cold PBS. 58. Proceed with the myelin removal using Myltenyi Biotec Myelin Removal Beads II following the manufacturer's instructions for magnetic labeling and depletion with LS columns: https:// www.miltenyibiotec.com/GB-en/products/myelin-removal-beads-ii-human-mouse-rat.html#gref Troubleshooting 7. a. We recommend filtering the samples before adding the Myelin Removal Beads. For best results, use a 70 mm cell strainer (as an additional step between steps 2 and 3 of the magnetic labeling in the Myltenyi protocol). b. Ensure that you wait for the columns to be empty before applying the wash (step 4 of the depletion with LS columns in the Myltenyi protocol). c. At the end, combine the flow-through from all columns used for the same sample.

OPEN ACCESS
CRITICAL: Perform the myelin removal on ice, and in a microbiological safety cabinet to preserve RNA.
Optional: For embryonic and neonatal samples, omit the myelin removal step, as it is not necessary, and may cause tissue loss ( Figure 5). 59. Proceed with the OPC isolation using Myltenyi Biotec CD140 MicroBeads following the manufacturer's instructions for OPC isolation: https://www.miltenyibiotec.com/GB-en/products/ cd140a-pdgfra-microbead-kit-mouse.html?countryRedirected=1#gref Troubleshooting 7. a. Be sure to follow the instructions for magnetic labeling and magnetic separation with MS columns (rather than depletion). b. During the first centrifugation step (step 2 of the magnetic labeling in the Myltenyi protocol), prepare the MS columns, the MiniMACS Separators, and the MACS MultiStand, as well as 2 3 15 mL tubes per sample. c. Be sure to remove the columns from the MiniMACS Separators before plunging them to collect the isolated OPCs.
CRITICAL: Perform the OPC sort on ice, and in a microbiological safety cabinet to preserve RNA.
60. Proceed with RNA extraction using Qiagen's RNeasy Micro Kit according to the manufacturer's instructions: https://www.qiagen.com/gb/products/discovery-and-translational-research/dna-rna-purification/ rna-purification/total-rna/rneasy-micro-kit/?clear=true#orderinginformation a. Be sure to include DNase when performing step 5 in the Qiagen protocol. Failure to do so leads to genomic DNA contamination ( Figure 6B). b. Be sure to add b-mercaptoethanol to buffer RLT, as cells isolated from tissues will be rich in RNAses (as described in the first note of the Qiagen Protocol; see Figure 6C). Pause point: RNA can be kept at À80 C until all the samples have been processed.

61
. Run a quality control check on the RNA a. Quantify the RNA in your samples, using a Qubit Fluorometer as this will determine which kit is needed to prepare the cDNA library. For instance, after MACS sorting OPCs, we found that our RNA concentration fell in the 250 pg-10 ng range, and therefore used the kit described in step 62. b. Measure the RNA Integrity Number (RIN; see Schroeder et al., 2006) using an Agilent Bioanalyzer. Typically, RIN values above 7 can be used for analysis (Sheng et al., 2017) (Figure 6). 62. Prepare the RNA-seq library using Takara Clontech's SMARTer Stranded Total RNA-seq Kit v2 -Pico Input Mammalian according to the manufacturer's instructions: https://www.takarabio. com/learning-centers/next-generation-sequencing/technical-notes/rna-seq/stranded-librariesfrom-picogram-input-total-rna-(v2) 63. Sequence the libraries on an Illumina sequencer (or another appropriate instrument). 64. Assess read quality and trim the reads using standard protocols (for instance, Trim Galore! (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore). 65. Align the reads to a reference mouse genome (such as e GRCm38/mm10) using standard protocols (for instance, Tophat (version 2.1.1, https://ccb.jhu.edu/software/tophat/index.shtml). 66. Proceed with normalization and differential expression analysis with appropriate software and protocols (for instance, the R Bioconductor DESeq2 package (Love et al., 2014; https:// bioconductor.org/packages/release/bioc/html/DESeq2.html). 67. Perform any other required analyses, such as GO term analysis with appropriate software and protocols (for instance DAVID (https://david.ncifcrf.gov/) and REVIGO (Supek et al., 2011; http://revigo.irb.hr)).

OPEN ACCESS
Note: Unless you have a strong background in bioinformatics, we recommend collaborating with a bioinformatician to analyse RNA sequencing data.

EXPECTED OUTCOMES
With this protocol, it is possible to determine passive membrane properties, and voltage-gated ion channel and neurotransmitter receptor expression in cells in acute brain slices at different ages , and in different brain regions (Spitzer et al., 2019). In addition, dye loading during whole-cell patch clamp allows for morphological characterization, cell location determination, and investigation of protein expression ( Figure 5).When cells are physiologically homogeneous within a condition (such as within a timepoint or brain region), bulk RNA sequencing provides additional information on cell function ( Figures 7D and 7E). Together, these data can determine cell identity or state and cell function ( Figure

QUANTIFICATION AND STATISTICAL ANALYSIS
Calculating R s , C m , and R m Here we describe how to measure series resistance (R s ) and passive membrane properties including cell capacitance (C m ) and membrane resistance (R m ) from capacitive transients elicited by a À5 mV pulse ( Figure 8B). In whole-cell voltage clamp of a compact cell, the electrical circuit formed comprises a membrane capacitance, a membrane resistance, and the series resistance (comprised of the pipette resistance and the access resistance), in series with the membrane capacitance and resistance ( Figure 8A).
1. To measure R s , we take advantage of the fact that when a voltage step (V step ) is initiated (t=0), the charge across the membrane capacitor and resistor is null ( Figure 8C). Thus, the current flow I, is and R s is (C) At the onset of the voltage pulse, t=0, the current only flows through R s , allowing us to calculate it using Ohm's law. However, the current at t=0 is not recorded due to pipette capacitance; thus, we fit the current response to a single exponential (in red), to extrapolate the current at the onset of the voltage pulse. Once the current has reached a steady state, we can calculate R m .
2. To measure R m , we take advantage of the fact that at steady state (t=N), the membrane capacitor is fully charged and current only flows through the resistors. Thus, the current flow I, is Isolating R m and substituting R s from the equation above, we find R m = V step , I t0 À I tN I tN ,I t0 3. To find I t0 and I tN , we use the fact that the current flow through this circuit can be described by the following single exponential (Tessier-Lavigne et al., 1988): Thus, we fit the recorded capacitive transient using a single exponential, IðNÞ + A,e Àt=t and extrapolate the curve to the onset of the voltage pulse ( Figure 8C), allowing us to determine I t0 and I tN , and therefore to calculate R s and R m .
4. From this single exponential fit, we also determine t, the time constant of current decay. t is also described as t = C m , R s ,R m R s + R m 5. Rearranging this equation, we can calculate C m : We perform this fitting and calculate R s , R m and C m with a lab written Matlab script (available upon request), using the equations described above.
Measuring the peak voltage-gated Na + current Here we describe how to measure current responses from the activation of voltage-gated Na + channels during a voltage-step protocol. We use a lab written Matlab script (available upon request) to perform the steps described below.
6. Subtract the capacitive and leak currents. a. We perform this subtraction mathematically. To do so, we assume that the capacitive and leak currents follow Ohm's law, and are proportional to the amplitude of the voltage pulse applied. We use the current resulting from the À100 mV pulse as our template leak current, as there are no voltage-gated transients at this potential during the time frame in which voltage-gated Na + channels activate, and scale it to each voltage pulse using a multiplier, mul: such that, for a holding potential V hold of À60 mV (uncorrected for junction potential), ll OPEN ACCESS STAR Protocols 2, 100439, June 18, 2021 b. We then subtract these scaled currents from our recorded currents, using the following equation: I subðiÞ = I i + mul i ,ðI 100 À I baseline Þ À I baseline and obtain capacitance and leak subtracted current responses (Figure 9). 7. From these subtracted currents, we measure the peak voltage-gated Na + channel current as the peak inward current in the first 7 ms following the onset of the voltage pulse ( Figure 9).

Measuring NMDA and kainate current responses
Here we describe how to measure current responses to NMDA and kainate application ( Figure 10).
8. Open the recording in Clampfit. 9. Measure the baseline current, I baseline , at t baseline .
Note: For noisy recordings, we suggest filtering the recording before measuring current responses using a lowpass Bessel filter. This is particularly useful for NMDA-induced responses, as they are typically small (2-10 pA), and can fall within the noise level in unfiltered recordings.
10. Measure the peak current, I pk , at t pk . 11. Measure the recovery current, I wash , at t wash . This is measured once the current has recovered to a stable baseline ( Figure 10). 12. To account for the potential baseline drift during the drug application, calculate the current response, as ( Figure 10): I response = ðI baseline À I pk Þ À I drift where I drift ðI baseline À I wash Þ = À t pk À t baseline Á ðt wash À t baseline Þ rearranging and substituting I drift in the first equation, we find I response = À I baseline À I pk Á À ðI baseline À I wash Þ, À t pk À t baseline Á ðt wash À t baseline Þ 13. To calculate receptor density, divide the current response by the membrane capacitance, which is a proxy for cell size.
Optional: If all cells in each condition have the same capacitance, the total current can be compared.

LIMITATIONS
We described a protocol to measure ion channel currents and neurotransmitter receptor currents in cells from different timepoints and/or brain regions. However, the external and internal solution recipes we describe here are specific to our purposes, and different components may be needed for other purposes. For instance, we use BAPTA as a Ca 2+ chelator in our internal solution, as we found that it helped with stability and response rundown during long recordings; however, if interested in G-protein coupled receptors, you may wish to use EGTA, a less potent chelator.
Here, we describe a whole-cell patch clamp protocol, and not perforated patch. When interested in downstream intracellular signaling, using perforated patch may be more suitable, as the native intracellular milieu is more conserved than in whole-cell patch clamp, where it is mostly replaced by the internal solution.
We perform our experiments at room temperature (19 C-23 C), and in HEPES-buffered aCSF, rather than physiological temperature. Thus, receptor kinetics are difficult to interpret. Recording in bicarbonate-buffered aCSF at physiological temperature would be more appropriate if focusing on receptor kinetics, but the perfusion flow must be constant so that pH is maintained with 95% O 2 / 5% CO 2 bubbling. In addition, we typically bath apply receptor agonists such as NMDA and kainate. As bath application requires some time for the whole bath to reach concentration, it should not be used when attempting to study receptor kinetics, as receptor desensitization does occur. Following this, the flow rate should be constant between experiments. Alternatives to bath application include fast perfusion systems, where a pipette placed in the bath near the cell is used for drug delivery, or puffer systems, where a pulse of positive pressure is used to eject a small volume of drug onto a cell.
Cell recovery after dye loading can be challenging. With practice, approximately 60% of cells remain in the slice following a recording; however, this proportion decreases with longer recordings. In addition, not all recovered cells can be found when imaging a post-hoc staining. Thus, while dye loading provides morphology, location, and protein expression data, it is difficult to obtain these data for every recorded cell. Figure 9. Analyzing voltage-gated Na + channel currents We measure voltage-gated Na + channel currents by recording the current response to a voltage step protocol ranging from À120 mV to +40 mV (not corrected for junction potential). The left panel shows the voltage steps and a representative current response. To isolate Na + currents, we first perform leak and capacitive current subtraction using the current at the À100 mV step (in red) as our template leak current and scaling it to the other voltage steps. From the subtracted current responses (right panel), we measure the peak Na + current as the peak inward current (blue line) during the first 7 ms of the voltage steps. While electrophysiology remains the preferred method to study voltage-gated ion channel and neurotransmitter receptor function, it can be difficult to link physiological properties to cell function.
Combining multiple approaches, such as electrophysiology, cell morphology, immunohistochemistry, and bulk RNA sequencing, as we have described above, can be useful to assess whether cell subtypes, or states, characterized by a specific electrophysiological profile are more likely to perform specific functions. However, bulk RNA sequencing is only appropriate for this purpose when cells are relatively homogeneous within a condition. If the cells display heterogeneity with respect to their physiological properties, patch-seq (single-cell sequencing of RNA extracted from a patched-cell) is likely to be more appropriate, although bulk sequencing can be complementary to patch-seq in this case.
Finally, MACS may not be an appropriate sorting method for all cell types. For instance, microglia are activated by MACS sorting, and thus, for these cells, Fluorescence Associated Cell Sorting (FACS) is more appropriate. We refer the reader to the following protocol for FACS (Barbar et al., 2020).

TROUBLESHOOTING
Problem 1: The acute brain slices are unhealthy Most of the neurons on the brain slice are dead or dying (steps 4 and 5).

Potential solution
The dissection may not have been fast enough. Try to increase your dissection speed. Ensure that you do not damage the brain extensively while dissecting and slicing, as this may increase cell death.
Check your slicing and resting aCSF and make sure that you have included glucose in the solution (a sign of glucose omission is that blood vessels in the slice become very prominent). Measure the osmolarity of the slicing and resting aCSF, which is expected to be 315-330 mOsm.
Problem 2: Difficulty to seal cells The resistance increases when you begin applying negative pressure, but quickly decreases again, or you suck up the cell (step 22). Figure 10. Analyzing kainate or NMDA receptor currents Representative response to the bath application of 30 mm kainate. To measure the peak response to kainate (or NMDA), we subtract the baseline current (I baseline ) from the peak current (I peak ). However, we also account for a baseline drift (I drift ), as this can occur with during long drug applications.

Potential solution
Try pulling pipettes with a slightly higher resistance. OPCs are very small cells, and it can be challenging to start patching them with lower resistance pipettes. When training to patch OPCs, start with pipettes around 7 MU, and make your way down to 5-6 MU.
Problem 3: Difficulty to seal cells The cell does not seal when you apply negative pressure (step 22).

Potential solution
If this occurs occasionally, you may have selected an unhealthy cell. Try selecting less shiny or grainy cells.
If this occurs systematically, there may be a leak between the 1 mL syringe where you apply pressure and the electrode holder. Check that you can apply positive pressure; if this doesn't work, check for cracks in the syringe, valve, tubing or holder to identify a potential leak, and replace the faulty component.
If you can still apply positive pressure, a leak is unlikely to be the answer. Check slice health. If your slice is unhealthy, discard it and change to another slice.
If your slice is healthy, your solutions may be the problem. Try patching the cells with another internal solution if you have one available.
If your internal solution is fine (changing internal does not help with patching), check the osmolarity of your external solutions.
If the cell is taking a long time to seal, make yourself a cup of tea and it may seal in the meantime.
Problem 4: Difficulty opening the cell The cell does not open once you have achieved a giga-ohm seal (step 24).

Potential solution
Try pulling pipettes with a slightly lower resistance.
Too much negative pressure applied during sealing, apply less (or try stopping applying positive pressure when close to the cell and allow the cell to seal itself).
Problem 5: Difficulty opening the cell The seal is lost as you open the cell (step 24).

Potential solution
Try using less negative pressure to open the cell.
If this happens systematically, it may indicate a problem with slice health, or your solutions. Check slice health. If your slice is unhealthy, discard it and change to another slice.
If your slice is healthy, your solutions may be the problem. Try patching the cells with another internal solution if you have one available.
If your internal solution is fine (changing internal does not help with patching), check the osmolarity of your external solutions.
Problem 6: The flow through the gravity-fed perfusion stops The flow rate is significantly decreased or stops (step 28).

Potential solution
There may be bubbles in the line. Make sure that you prime the lines before you start your patching day (run the solution through until bubbles are gone). Clear any bubbles that appear during the day.
There may be bacteria growing in the lines, as all solutions flowing through the lines contain glucose. Flush the lines with bleach and wash with distilled water.
Problem 7: MACS columns are blocked during myelin retrieval or OPC sort The tissue suspension does not flow through the columns (steps 58 and 59).

Potential solution
Add DNase (0.04 mg/ml) to the 0.5% BSA MACS buffer.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Ragnhildur Thó ra Ká radó ttir (rk385@cam.ac.uk).

Materials availability
This study did not generate new reagents.

Data and code availability
This study did not generate new data nor code, but Matlab scripts to measure R s , C m and R m and voltage-gated Na + currents are available upon request.