Protocol for Patch-Seq of Small Interneurons

Summary Obtaining electrophysiological recordings and gene expression information from the same neuron (Patch-seq) brings forward a unique opportunity to study the transcriptional correlates of functional properties and vice versa. Here, we provide a detailed Patch-seq protocol tailored to the specialized demands of studying small interneurons. Focusing on the technically demanding process of transitioning between patch recordings and cell extraction, our protocol describes and troubleshoots steps for successfully collecting small interneurons, allowing for multi-modal Patch-seq interrogation of this crucial cell type.

a. Mix the NaCl, KCl, NaH 2 PO 4 , and NaHCO 3 stock solutions in nuclease-free H 2 O to the specified volume. See Materials and Equipment for the detailed recipe. b. Make 50 mL aliquots and store at 4 C. 10. Make 13.5 mL aliquots of 20% glucose and store at 4 C. 11. Make 1 mL aliquots of MgSO 4 and store at 4 C. 12. Make 1 mL aliquots of CaCl 2 and store at 4 C.

Prepare 13 Intracellular Solution Stocks
Timing: 1 h CRITICAL: The intracellular solution (ICS) will be in contact with the internal contents of the cell for the duration of the patch recording and will be collected with the sample. Any contaminants in this solution could potentially be sequenced and the presence of even traces of RNases will negatively impact the quality of the sequencing data generated. Use all certified nuclease-free reagents (molecular biology grade), purchased in solution form when possible, to avoid potential contamination during powder weighing (see Key Resources Table).
Note: Prepare and aliquot this solution in a separate area or room, only used for pre-amplification steps. Work under a PCR cabinet or hood.
Note: Prepare the solution in a glass volumetric flask that was previously rinsed with molecular biology grade nuclease-free H 2 O, wrapped in aluminum foil, autoclaved twice and dried by baking at 200 C.
Note: For preparing the ICS use non-DEPC treated molecular biology grade nuclease-free H 2 O (e.g., Life Technologies cat. no. AM9937; see Key Resources Table).
13. Prepare 1 M sucrose solution using RNase-free H 2 O. 14. Mix the components for 13 ICS. See Materials and Equipment for the detailed recipe. a. Weigh the K-gluconate powder and resuspend in molecular biology grade nuclease-free H 2 O in the volumetric flask.
Note: Before weighing, thoroughly clean the balance and surrounding area with RNaseZap and 70% EtOH in DEPC H 2 O. For weighing use a clean twice-autoclaved spatula and an aluminum foil weighing boat.
b. Add the required volumes of KCl, KOH, MgCl 2 , NaCl, HEPES, EGTA, Sucrose, Na 2 ATP, Na 3 GTP, and Glycogen. Prepare intermediate dilutions for the high concentration stocks to minimize pipetting errors (e.g., use a 1:10 dilution of the 5 M NaCl stock solution in order to pipette 80 mL rather than 8 mL). Mix by stirring the flask. If using a magnetic stirrer, make sure it has been previously rinsed in abundant molecular biology grade, nuclease-free H 2 O, and autoclaved twice. c. Check the pH by pipetting a few mL onto a pH strip. Adjust to pH 7.4 with KOH or HCl accordingly. d. Complete to the total volume with molecular biology grade, nuclease-free H 2 O. 15. Make 250 mL aliquots and store at À20 C until use. 16. Add 2 mL of RNase inhibitor (20 Units/mL) to one 250 mL aliquot and check the osmolarity. It should be~290 mOsm. If slightly higher, adjust the amount of glucose in the aCSF, aiming for an aCSF osmolarity that is~10 mOsm higher than that of the ICS (for example, we commonly added 13.5 mL of 20% glucose solution per 500 mL of aCSF). If the ICS osmolarity is not within the range of~270-310 mOsm, prepare a new batch of ICS (Troubleshooting 1).

Prepare Intracellular Solution Loading Tips
Timing: 2 days 19. Wrap a new box of gel loading tips (see Key Resources Table) in aluminum foil and autoclave twice. 20. Store in a separate, dedicated box, or cupboard until use.

Electrophysiology Setup
Successful whole-cell patch-clamp recordings from acutely sliced brain tissue require a suitably equipped and well-maintained patch-clamp setup or ''rig.'' For an extensive discussion of what makes up a good rig we refer the reader elsewhere (Molleman, 2003). However, the following is recommended as a minimum: An upright microscope containing its own or external light source(s) and appropriate optical filters. Sufficient magnification to see individual cell bodies in detail; at least a 403 or 633 water immersion objective is recommended. An electronic amplifier and digitizer for recording electrical signals from neurons. This should be connected to a means of reading and storing the recorded information such as a computer and/or oscilloscope. An air table and air supply to prevent any vibration interference and improve mechanical stability during recording.
A perfusion system, e.g., peristaltic pump and associated tubing, to circulate external solution through the recording chamber throughout the experiment. A reliable local supply of carbogen to oxygenate solutions. A means of heating the perfusion solutions to physiologically-relevant temperature, e.g., a temperature-controlled recording chamber and/or in-line solution heater. A camera and associated imaging software, for real time visualization of the experiment as well as image capture for later analysis.
Note: Adequate visualization of the cell and patch pipette will become essential during cell extraction steps.
A stainless-steel or platinum ''harp'' to hold the slice in place.

Cell Collection Assembly
To efficiently expel the contents of the patch pipette into the collection tube, attach a mouth pipetting assembly to the top of the glass capillary. Cells deposited in 0.2 mL PCR tubes are flash frozen on a metal plate sitting on dry ice ( Figure 1).
Note: Using mouth pipetting devices for patching, for cell collection, and for expelling the cell contents into the collection tube allows for finer control of negative and positive pressure than the use of manual syringes. To minimize the potential for contamination we used long tubing and filters (see Figure 1A). In addition, the tubing and the adaptor where the glass capillary is inserted were cleaned daily, with RNA-zap and DNAaway, rinsed in RNase-free 70% EtOH and irradiated under UV light for a minimum of 30 min.

STEP-BY-STEP METHOD DETAILS
On the day of recording/sample collection you should have all solution stocks and aliquots already prepared and all materials cleaned and stored in single-use bundles.
Use clean gloves during all steps of all procedures. Change gloves often, especially after cleaning surfaces and/or components and, crucially, before loading ICS into the recording pipette and before retrieving the pipette for transferring the sample into the collection tube.
Maintaining strict cleaning and de-contamination standards is crucial for the successful collection of high-quality Patch-seq samples.

OPEN ACCESS
Note: it is best to split the daily solution preparation and slicing between two people. This coordination facilitates the timely preparation of material, aids in avoiding cross-contamination (e.g., the internal solution is prepared and handled by a person that has not been in contact with mouse tissue on the day) and minimizes the time elapsed from slicing to recording and sample collection.

Daily Pre-recording Electrophysiology Rig Cleaning and Preparation of Materials
Timing: 30 min CRITICAL: keeping the electrode wire and electrode holder exceptionally clean is absolutely necessary for collecting good quality samples.

Pull recording pipettes.
a. Wipe all surfaces of the pipette puller thoroughly with RNaseZap and rinse with 70% EtOH in DEPC H 2 O. b. Pull glass capillaries to the desired size and shape specifications and place in a clean box with a lid.
Note: For standard borosilicate glass (OD 1.5 mm, ID 0.84 mm, filamented) we used a 2-step pull on an upright puller (Narishige PC-10, load weight: 250g, heat for each step: step 1, 59.6; step 2, 44.9. pull length, 6 mm), which generated pipettes with~5-6 MU resistance and gave identical results for autoclaved versus non-autoclaved glass. Be aware, however, that these parameters will need careful adjustment for individual experimental requirements. Unless we encountered any serious and/or unforeseen issues with pipette resistance or puller settings, we usually pulled all the pipettes used during the whole recording session before starting.
CRITICAL: keep the pulled capillaries in the plate with the lid on. Avoid contact with air as much as possible. Always handle capillaries with new, clean gloves.
6. Replace the filter on the patching mouth suction tube (Troubleshooting 3). 7. Wipe Gilson (or other brand) pipettes that will be used for loading the ICS into the recording pipettes with RNaseZap and rinse with 70% EtOH in DEPC H 2 O. CRITICAL: The collection tubes MUST be compatible with downstream sample processing, as all cDNA synthesis and amplification steps should be carried out in the same tubes, without transferring the sample.

Prepare 13 Solutions for the Day
Note: These steps must be performed in a separate pre-amplification area, in a PCR cabinet or hood.

OPEN ACCESS
Note: The RLT plus lysis buffer precipitates readily, so these aliquots need to be prepared on the day of use.
Note: the use of a strong lysis buffer is a requirement for accessing the mRNA content inside the nucleus. Its use in this protocol is made possible by the incorporation of mRNA capture steps into the sample processing pipeline (step 23).
13. Wipe an ice bucket with RNaseZap three times and rinse with 70% EtOH in DEPC H 2 O. Load with dry ice and place the clean, dry metal plate on top. Place the lid on.

Patch-Clamp Recordings
Timing: 2 h The following is a brief description of the process of obtaining whole-cell recordings, focusing on the need to minimize sources of contamination during Patch-seq. More detailed accounts of the technical aspects of whole-cell recordings are available elsewhere (Molleman, 2003;Ogden, 1994;Segev et al., 2016).
The recording protocol needs to be previously optimized to ensure the required functional parameters are obtained in the shortest period of time, to minimize mRNA degradation. As a general rule, do not spend more than 10 min on this step after membrane rupture.
14. Prepare acute slices for electrophysiological recordings. a. With appropriate ethical approval, obtain acute slices of the CNS region under study. We use a standard acute slicing protocol (Galliano et al., 2018). For the purposes of Patch-seq the only specific modification is the preparation of clean aCSF (see Materials and Equipment). Dissection and slicing occur in ice-cold oxygenated sucrose solution. Slices are then kept in clean, oxygenated aCSF at 32 C-34 C for 30 min, before being maintained in oxygenated aCSF at 22 C-24 C for the remainder of the experiment. Individual protocols should be optimized for the requirements of particular animal models, tissues, equipment, or experiments. b. Perfuse the aCSF extracellular recording solution (containing the channel blockers) through the system and wait for the recording chamber to equilibrate to the correct temperature. c. Transfer a brain slice from your recovery chamber to the recording chamber and fix in place with a harp. 15. Load 2 mL of ICS (with RNase inhibitor) into a pulled glass capillary using a gel loading tip.
Note: If the gel loading tip does not have a filter place a P10 filter tip between the pipette and the gel loading tip. We found that the Eppendorf gel loading tips (Microloaderä, cat. no. 5242956003) and the ClearLine P10 filter tips (cat. no. S1120-3810) are perfectly compatible.
Note: gel loading tips affect the accuracy of volume measurements. If you need to pipette exactly 2 mL, because the samples will be processed by an automated pipetting station, you can accurately measure the 2 mL using a standard P10 filter tip, place the drop on the back of a microcentrifuge tube lid and aspirate it with the gel loading tip.
Note: Obtaining high-quality whole-cell somatic recordings from small neurons usually requires the use of small-tipped, relatively high-resistance pipettes. Typically, we record from olfactory bulb dopaminergic neurons using pipettes with resistance of~4 MU. For Patchseq recordings, we had to adjust this to~5-6 MU (see Cell extraction). This made obtaining ll OPEN ACCESS STAR Protocols 1, 100146, December 18, 2020 good quality recordings a bit more cumbersome due to higher access resistance, but ensured that the pipette tip was just the right size for cell collection.
a. Keeping positive pressure on the pipette, lower the electrode into proximity with your chosen cell (Figure 2, step 1). b. Zero the voltage offset on the amplifier. c. Move the pipette into contact with your cell and remove the positive pressure.
Note: Traditionally this involves lowering the pipette vertically or obliquely toward the surface of the cell until a dimple in the cell membrane is observed. With very small cells the positive pressure from the pipette tip may be enough to significantly distort or even blow the soma away upon approach. Carefully adjusting the level of positive pressure, approaching the cell from the side or at an angle, and/or swiftly switching off the positive pressure when close to the cell soma can help successful seal formation. d. Once a giga-Ohm seal is formed, use negative pressure to break into the cell and establish a whole-cell patch configuration (Figure 2, step 2). e. Perform the desired electrophysiological experiments (Figure 2, step 3). f. Evaluate recording parameters online (e.g., leak current, resting potential, etc.). If recordings meet quality thresholds, proceed with sample collection.
Note: Criteria will vary according to individual experimental demands, but should be kept consistent across a study; with the small cells and small pipettes in play here, ours were access resistance %33 MU with no change >25% across a recording, and holding current %100 pA.

Cell Extraction
Timing: 10 min To maximize single-cell sample purity this procedure must be carried out entirely under good electrical seal (usually >1 GU). Ideally, it should not take longer than 10 min, to minimize mRNA degradation.
Note: The specific technique described here, including the ''corking'' approach, requires some optimization with regards to pipette size, negative pressure applied, and movement Corresponding representative seal test protocols are shown for each step.
Step 1: A clean, unobstructed pipette is lowered into the vicinity of the target cell.
Step 2: A giga-ohm seal is formed between the pipette and cell.
Step 3: The cell membrane is ruptured yielding electrical access to the inside of the cell. At this point the membrane properties of the cell can be interrogated (e.g., inset trace, showing cell response to 500 ms of depolarizing current injected in current-clamp mode).
Step 4: Negative pressure is applied, drawing the cell contents into the pipette and forming a strong seal onto the cell nucleus.
Step 5: The pipette is slowly retracted from the tissue bringing the cell nucleus with it. Seal test recordings shown here were obtained by holding the cell at À60 mV and stepping the membrane voltage sequentially to À50 mV and then À70 mV for 10 ms before returning to the holding potential. Scale bar, 10 ms and 1 nA for #1; 10 ms and 200 pA for #2-5; 100 ms and 25 mV for inset in #3.

OPEN ACCESS
of the pipette away from the tissue. Although it might take a little practice, following the steps below is crucial for ensuring successful, clean collection of the recorded interneuron's intracellular contents. In particular, collection of the cell nucleus significantly increases the cDNA yield (Cadwell et al., 2016;Troubleshooting 4, 5, 6, 7, and 8). Ensuring a good ''cork'' of the nucleus in the pipette tip allows the nucleus to be collected with a small-tipped pipette without losing the high-resistance seal which optimizes sample purity.
17. Collect the cell content, including the nucleus, using the recording pipette (Figure 2, steps 4 and 5). a. Apply negative pressure until the cell has reduced significantly in size. This may take up to several minutes, and with small-tipped pipettes it may require constant suction far stronger than that required to initially obtain a gigaseal. If electrical seal begins to fail, release pressure and allow the cell to recover. b. When most of the cell contents have been suctioned into the pipette, begin to manipulate the pipette into proximity with the nucleus. Identifying the nucleus' location is helped by its generally darker appearance and its more solid presence; it does not move or distort as readily as the plasma membrane in response to pipette manipulations. Continue to apply negative pressure and monitor electrical resistance to observe a change in access resistance when the seal is formed on the nucleus. c. Slowly begin to move the pipette away from the tissue. Raise the pipette vertically from the slice in stages, allowing time at each step for the electrical seal to stabilize. If the previous steps were successful the nucleus should be seen accompanying the pipette tip (''corking''), if this is not the case, lower the pipette again in line with the cell body and repeat steps a-b (Troubleshooting 4).
CRITICAL: Due to the small size of olfactory bulb dopaminergic neurons, in order to obtain a significant amount of mRNA for proceeding with cDNA synthesis and sequencing library preparations, we needed to ensure that we collected the cell nucleus.
Note: we optimized this step for consistently obtaining a nucleus ''cork'' during cell extraction. We achieved this by using smaller tip pipettes (~5-6 MU) for recording so the nucleus was slightly bigger than the tip and so would generally not be sucked into the pipette.
d. Finally, as the pipette is extracted from the tissue along with the cell body/nucleus, it may still be connected to the tissue via intact but stretched neuronal processes. Significant care is required here to prevent rupture of the cell and loss of electrical seal due to the tearing of these processes. Move the pipette slowly away from the tissue until the nucleus/soma can be confidently seen to be separated from the slice. Only after this is confirmed is it safe to remove the pipette from the solution. (Figure 2; Troubleshooting 4, 5, 6, 7, and 8). e. Check that there is no debris attached to the outside of the pipette (Troubleshooting 9). 18. Transfer the cell to the collection tube.
a. Remove the glass capillary, or detach the entire pipette holder assembly from the headstage (this will depend on the holder specifications).
Note: Removing the glass capillary from the pipette holder does not affect the small volume of liquid at the tip. It is safe to remove it from the holder.
b. Quickly bring the tip of the recording pipette to the bottom of a PCR collection tube, containing 2 mL of lysis buffer.

OPEN ACCESS
STAR Protocols 1, 100146, December 18, 2020 CRITICAL: It is essential that the tip of the capillary does not touch anything before contacting the lysis buffer. Special attention must be paid to avoid touching the walls of the collection tube (Troubleshooting 10).
c. Gently break the tip of the capillary against the bottom of the collection tube. d. Apply positive pressure to expel the whole of the pipette content into the lysis buffer. Avoid making bubbles.
Note: If in the previous step the glass capillary was removed from the pipette holder, a mouth pipetting assembly can be quickly placed on the top of the glass capillary to apply positive pressure ( Figure 1A).
19. Quickly place the tube on the metal plate sitting on dry ice to flash-freeze the sample ( Figure 1B). 20. Assign a sample collection score to the cell, based on successful collection of the nucleus, extraction time, quality of seal, presence of debris on the outside of the pipette and successful depositing into lysis buffer (Figure 3; Troubleshooting 4, 5, 6, 7, 8, 9, and 10). 21. Store samples at À80 C until processing. 22. Collect negative controls during each recording/collection session. a. Load a pipette with ICS and lower to the tissue as if aiming to perform a recording. b. After making contact with the tissue, retract the pipette and transfer the contents to a collection tube. c. Flash-freeze the sample and store at À80 C until further processing.
Pause Point: Frozen samples can be stored at À80 C for up to 6 months before further processing.
mRNA Capture and scRNA-Seq

Timing: 1-2 weeks
The focus of this protocol is on Patch-seq sample collection: the steps required to obtain high-quality material for sequencing after whole-cell recordings in small interneurons. We do not aim to Perfectly extracted cells must meet all the criteria outlined: successfully extracted nucleus, seal maintained throughout the whole extraction process, no debris observed on the outside of the pipette, no contact of the tip of the pipette with any surface before it reaches the lysis buffer in the collection tube, and less than 10 min duration for the entire extraction process. Assign a score to each cell and only process for sequencing cells with ''perfect'' or ''very good'' scores. Cells with ''good'' scores can also be sequenced if, for example, sample numbers are low or there is particularly interesting physiology.
provide a comprehensive description of the downstream procedures needed to generate and analyse scRNA-seq or other -omic data. Detailed protocols for this are published elsewhere (Macaulay et al., 2016;Picelli et al., 2014;Troubleshooting 11, 12, and -13). Instead, here we briefly outline a possible workflow based on our own experiments.
Note: Although this adds a significant number of steps to sample processing, it presents a number of advantages. First, this step is necessary for removing any traces of the RLT plus lysis buffer. Second, it helps in controlling the total volumes of subsequent reactions. Third, it generates a separate gDNA fraction that is available for pursuing questions at the genomic level.
Note: before starting the G&T procedure, add 1 mL of spike-in RNA (typically, 1:1-2 million dilution, but this will depend on the cell type) to each sample. These artificial RNAs will serve to control for technical artifacts during cDNA synthesis, library preparation and sequencing (Jiang et al., 2011). 24. Perform cDNA synthesis and amplification following the Smart-Seq2 protocol (Picelli et al., 2014). 25. Quantify and quality control the cDNA using standard procedures (e.g., Qubit and bioanalyzer) and by running qPCRs for relevant marker genes.
Note: extensive cDNA quality control (Qubit + bioanalyzer + qPCRs; Figure 5) is highly recommended during the initial troubleshooting stages, when Patch-seq workflows are being established anew. This will provide the necessary feedback for outlining the cell collection scoring criteria.
26. Prepare sequencing libraries by tagmentation using the Illumina Nextera XT kit and introduce sample indexes. 27. Quantify and quality control the pooled libraries using standard procedures (e.g., Qubit, bioanalyzer, and qPCR quantification). 28. Sequence the libraries on the appropriate Illumina platform (e.g., HiSeq2500). Typically, samples are pooled to obtain~2 million reads per cell. For cell types with complex transcriptomes (e.g., neurons) this sequencing depth will ensure quantitative gene expression data for >5,000 genes, comprising highly, averagely and lowly expressed genes, and ensuring these include genes coding for ion channels and transporters which are of particular interest for experimental questions addressed using Patch-seq protocols. 29. Demultiplex, QC, map and quantify the sequencing reads following standard protocols (Vieth et al., 2019). 30. Perform sample-level QC to identify successfully sequenced samples and eliminate low-quality cells (see examples in Figure 5). Continue with your data analysis pipeline (Luecken and Theis, 2019;Lun et al., 2016).

EXPECTED OUTCOMES
Adapting patch recordings of very small interneurons for compatibility with sample collection for Patch-seq presents a number of challenges that impact the recording itself, the sample collection procedure, and the number of successfully collected cells per session. Because we performed recordings with smaller than usual pipettes to improve the cell collection step, the rate of successfully patched cells was lower than usual. This elevated difficulty in obtaining patch recordings, coupled with the need to study specific fluorescently-labeled neurons and the practical difficulties inherent in the cell extraction process for small interneurons, meant that we typically collected around 4 good quality cells per day, among roughly equal numbers of failed attempts (Figure 4).

OPEN ACCESS
Using a number of electrophysiological parameters and our cell extraction scoring system, we could decide, at the time of collection, whether each individual sample was of good enough quality to warrant moving ahead to cDNA synthesis and sequencing ( Figure 3). This presents an advantage against other single-cell collection methods (e.g., FACS) for which there is no experimental certainty on the quality of each individual sample prior to cDNA synthesis and sequencing. Overall, collected samples that were subsequently sequenced yielded excellent quality sequencing data, successfully quantifying thousands of genes per cell (>80% of sequenced cells passed all sequencing QC metrics; see ''Good quality cell'' example in Figure 5).
Our sample collection and processing workflow included a series of steps to pull-down the mRNA, following the G&T protocol. This allowed for flexibility in the volume of ICS used for recording and the lysis buffer used for sample collection. More importantly, the protocol produces a separate fraction, containing the genomic DNA, that is now available to pursue new experimental questions.
Carefully choosing the recording protocol, following the optimized cell collection steps outlined above, and sequencing at enough depth to obtain accurate expression information for relevant genes (e.g., ion channels and membrane transporters) will allow for accurate analysis of the relationships between functional parameters and gene expression patterns.

LIMITATIONS
The collection of samples for Patch-seq is extremely low throughput, especially in comparison with other scRNA-seq approaches. In addition to the Patch-seq-ed experimental samples, we strongly recommend performing scRNA-seq, in parallel, on the same neuronal type/stage using a higher throughput sample collection method (e.g., FACS or manual sorting). This will greatly enhance data analysis strategies and improve statistical power. Also, performing a whole tissue scRNA-seq experiment (e.g., 10X genomics) to analyze gene expression in all constituent cell types of the brain region in question, including those targeted for Patch-seq, will allow for sample specific quality and contamination control (Tripathy et al., 2018).
A number of contamination and quality control measurements make the patching procedures more laborious than a stand-alone electrophysiology protocol, potentially reducing the number of successfully collected samples per session. First, because during electrophysiological recordings the intracellular environment is exposed to the ICS and the recording electrode, these two components are crucial to obtaining good quality samples. Keeping them absolutely free of contaminants, including nucleases, is laborious but a requirement for collecting useful samples. Second, during cell extraction extreme caution must be observed in maintaining an electrical seal. This guarantees that the intracellular environment never comes into contact with the circulating aCSF. Because this external solution is bathing a tissue slice with dying cells and debris from already dead cells, it is a major source of cross-contamination. Sticking to the ''extraction under seal'' requirement greatly reduces the number of successfully collected cells per session, especially during early training stages. For our sample collection process, because we were patching with smaller than usual pipettes, maintaining the seal was made all the more complex and benefited from this experiments being performed by a highly skillful and experienced electrophysiologist. With practice, this step becomes less of a bottleneck. Due to the small size of olfactory bulb dopaminergic neurons, in order to obtain a significant amount of mRNA for proceeding with cDNA synthesis and sequencing library preparations we needed to ensure that we collected the cell nucleus. Efficient access to nuclear mRNA requires a stronger lysis buffer, like the RLT plus buffer used here. This presents the limitation of seriously affecting downstream reactions. cDNA synthesis cannot directly be performed on a cell collected on RLT plus buffer, as the denaturation agents present will inactivate the retrotranscriptase. To solve this issue, we introduced extra sample processing steps for capturing and washing the mRNA, following the G&T protocol (Macaulay et al., 2016). The downsides of this are the added cost of reagents and that the lengthier protocol meant that samples could not be processed on the day of collection. We stored all our collected samples at À80 C and processed them in batches. Storage in this way was not associated with any change in sequencing quality up to our longest storage period of almost 7 months (Pearson correlation with days in storage for ''good'' cells; total reads per cell, R = À0.07, p = 0.57, n = 68; total genes per cell, R = 0.15, p = 0.21, n = 68).
The G&T protocol, though lengthier and more costly, presented a much welcomed advantage. Because the mRNA is pulled down with magnetic beads and the starting volume consequently does not affect downstream reactions, we could perform electrophysiological recordings using a significantly higher volume (2 mL) of ICS than those previously reported for Patch-seq experiments (0.1-0.9 mL; Cadwell et al., 2017;Fuzik et al., 2016;Troubleshooting 14). Finally, in our hands, and in line with previous publications (Macaulay et al., 2016) the incorporation of the G&T steps to our sample processing pipeline did not affect the quality and/or quantity of scRNA-seq data obtained. Overall, the incorporation of G&T steps presented more advantages than limitations. However, these steps for mRNA capture and wash are not an absolute requirement for the preparation of scRNA-seq libraries. If the G&T protocol is not possible or preferred, samples should be collected in a milder lysis buffer (e.g., 0.2% Triton X-100) and the volume of ICS used should be lowered so as to not hamper the cDNA synthesis reaction.

TROUBLESHOOTING Problem 1
The osmolarity of the ICS is very high, and stable recordings are impossible (Before You Begin, step 16).

Potential Solution
The addition of RNase inhibitor and glycogen to the ICS can significantly alter the osmolarity of standard recording solutions. Small excesses in ICS osmolarity (<20 mOsm) can be offset by adding additional glucose to the aCSF. Larger differences can be troubleshot by making ''mock'' solutions first that do not contain expensive nucleotides, to establish the effect of glycogen and RNase inhibitor on ICS osmolarity.

Problem 2
A voltage offset is present which is not removed when pipette is ''zeroed'' (step 16b).

Potential Solution
Because the pipette holder is disassembled and thoroughly cleaned every day, occasionally a drop of liquid can remain lodged in the smaller nooks and crannies. Ensure that every part of the holder is perfectly dry before re-assembling.

Problem 3
Loss of pressure or poor pressure control through the pipette holder (step 16c).

Potential Solution
Ensure the tubing between the pressure system and head stage/pipette holder is clean and unobstructed. Repeated strong pressure can affect the integrity of the filter. For this reason, and as a contamination control measure, change the filter daily.

Problem 4
Cell extraction fails. The nucleus was clearly left behind when removing the pipette (step 17).

Potential Solution
During the extraction procedure, move the pipette closer to the nucleus and apply negative pressure until a seal is formed on the nucleus. This is observed as a change in access resistance (Figure 2). Continue applying negative pressure, pulling the nucleus away from the rest of the cell. This will leave all projections behind, but will extract the nucleus under seal (i.e., nucleated patch), which contains most of the cell's mRNA. When the cell is successfully out of the tissue, it should still be under seal. In DIC, the nucleus is easily spotted when it was successfully pulled out of the tissue and it is ''corking'' the pipette (Figure 4, top right). If the nucleus was nonetheless left behind, and not clearly visible ''corking'' or inside the tip of the pipette (Figure 4, bottom right), discard the sample.

Problem 5
Cell extraction fails. Electrical seal clearly lost during cell extraction (step 17).

Potential Solution
This problem has no solution for that specific sample. The cell must be discarded. During the next cell collection attempt, apply milder negative pressure and wait for the cell to accommodate. Different interneuron types may require slightly different extraction techniques in terms of pipette tip size, negative pressure strength, and/or speed of pipette retraction. Optimizing the extraction technique in advance of commencing the experiment is highly recommended.

Problem 6
Cell extraction fails. The cell was clearly left behind when removing the pipette (step 17).

Potential Solution
This is a major problem with cells with somas that are solidly embedded within local neuropil, as is the case for olfactory bulb dopaminergic neurons. Here, it pays off to take slightly longer to pull the cell out of the tissue. Retract the pipette only slightly, wait for the seal to stabilize and keep an eye on the cell making sure there are no projections that are stretching and still attached to the tissue. With time, those connections will break and seal off and an intact sample, under seal, will be collected. If the problem persists, try modifying the pipette tip size. If the pipette tip is small, sufficient traction of the cell nucleus is difficult; too large and the nucleus will not become lodged in the tip and the seal may be lost.

Problem 7
Cell extraction fails. No noticeable amount of cytoplasm is observed going into the pipette and there is no detectable change in cell size (step 17).
Potential Solution This is a major problem for small, compact neurons such as olfactory bulb dopaminergic neurons. After finishing the recordings, and without moving the pipette, apply slight negative pressure to suck the small amount of cytoplasm into the pipette. We usually observed some red fluorescence inside the pipette at this step, without actually noticing any change in the size of the cell soma. At this point, this is not an extraction failure. These interneurons are so small that their somas are virtually all nucleus, hence the lack of size shrinkage when suctioning the cell content. As described in the step-by-step section, the approach to obtain enough sequencing material is to ensure the collection of the nucleus.

Problem 8
Seal collapses at a very late stage of extraction (step 17).

OPEN ACCESS
STAR Protocols 1, 100146, December 18, 2020 Potential Solution Even after perfect extraction, a sudden collapse of the seal can occur, often accompanied by a change in the visual appearance of the cell from a smooth ball to a more translucent and defined nucleus shape. If this happens when the pipette is far away from the slice, and is only partial, with significant nucleus corking maintained, the cell might still be successfully included in future analyses.
If not, discard the sample.

Problem 9
Cell collection fails. Tissue debris is observed on the outside surface of the pipette (Figure 4, bottom left) (step 17).

Potential Solution
After the cell has been successfully extracted from the tissue, under electrical seal, inspect the pipette, by moving the focus up and down, to ensure there is no debris attached. If debris is observed, then slowly move the pipette around to see if it detaches. Alternatively, as a last resort, because it can dislodge the nucleus and/or rupture the seal, move the pipette outside of the solution and re-submerge it. Visually inspect the pipette to confirm debris was removed and ensure that electrical seal was maintained throughout. If either of these fails, discard the sample.
If debris is regularly observed, make sure the aCSF was prepared with high-quality water (see Key Resources Table) and that the tubing is clean. If necessary, prepare new aCSF stocks and change all the tubing for the perfusion system. Also consider targeting cells that are closer to the slice surface.

Problem 10
Cell collection fails. The tip of the pipette touched another surface (e.g., upper side wall of collection tube; Figure 3) (step 18).

Potential Solution
Discard the sample. Also, drink less coffee.

Problem 11
Low (or no) cDNA obtained for an individual sample (steps 24 and 25).

Potential Solution
Sample collection and/or storage failed. Only process and sequence samples that meet all the sample collection criteria: (e.g., seal maintained throughout the extraction, cell quickly placed into tube and frozen; Figure 3).

Problem 12
Low (or no) cDNA obtained for an individual sample (steps 24 and 25).

Potential Solution
Double-check all solution preparation and sample collection steps to ensure all contamination control and nuclease-free requirements are met. Prepare new solution stocks and single-use aliquots.
Ensure the pre-amplification area and PCR cabinet are free of contaminants. Re-clean all equipment and benchtop surfaces. Change gloves more frequently, especially before loading the ICS into the recording pipette and during sample collection. Always run positive control samples in parallel.

Problem 13
Low (or no) cDNA, low (or no) reads/genes obtained for an individual sample (steps 24 and 25).

OPEN ACCESS
Potential Solution mRNA capture, cDNA synthesis and/or library preparation may have failed (see ''Failed cell'' example in Figure 5). The preparation of scRNA-seq relies on the capture and amplification of a very small amount of starting material. There are multiple reasons why any of these steps may fail, resulting in sample loss. These troubleshooting aspects are beyond the scope of the present protocol. Readers should refer to relevant publications (Macaulay et al., 2016;Picelli et al., 2014), or contact the authors for in-depth discussions.

Problem 14
Difficulties patching with lower than usual volume of ICS (step 16).

Potential Solution
Load the patching pipette with 2 mL of ICS and bring the volume back down for the cDNA synthesis reaction by first performing an mRNA capture step.

Lead Contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Matthew S. Grubb (matthew.grubb@kcl.ac.uk).

Materials Availability
No new mouse lines or materials were generated in this study.

Data and Code Availability
Sequencing data corresponding to the example cells depicted in Figure 5 is available at NCBI GEO (GSE151709). Data analysis code is available upon request.