Fluorescent pulse-chase labeling to monitor long-term mitochondrial degradation in primary hippocampal neurons

Summary The accumulation of dysfunctional mitochondria is a hallmark of neurodegenerative diseases, yet the dynamics of mitochondrial turnover in neurons are unclear. Here, we describe a protocol to monitor the degradation of spectrally distinct, “aged” mitochondrial populations. We describe the preparation and transfection of primary rat hippocampal neuron cultures. We detail a mitochondrial-damaging assay, a SNAP pulse-chase labeling paradigm, and live imaging to visualize the mitochondrial network. Finally, we provide steps to quantify mitochondrial turnover via lysosomal fusion. For complete details on the use and execution of this protocol, please refer to Evans and Holzbaur (2020a).

7. Prepare AM and Maintenance Media (MM) for primary hippocampal cultures. 8. Prewarm 5 mL of HBSS and 35 mL of AM in a 37 C water bath. 9. Place a 15 mL conical containing 10 mL of HBSS on ice. 10. Sterilize scissors, forceps, and dissection area with 70% ethanol.

MATERIALS AND EQUIPMENT
Preparation procedure: Dissolve boric acid and sodium tetraborate in 400 mL of ddH 2 O. Filter buffer using a 0.22 mm filter unit to sterilize and if necessary, adjust pH to 8.7-9.0. Stock solution can be stored for $4 months at 4 C.
CRITICAL: Boric Acid and Sodium Tetraborate are Particularly Hazardous chemicals. Store in a cool, dry, well-ventilated area away from incompatible materials. Wear proper PPE when working with chemicals and know institutional approved procedures for cleaning spills. Fiji (Schindelin, 2012) https://imagej.net/software/fiji/ TurboReg (Thevenaz, 1998)  Poly-L-Lysine

Reagent Final concentration Amount
Preparation procedure: Dissolve 100 mg of PLL in 50 mL of borate buffer. Make 1 mL aliquots and store at À80 C for up to $3 months. For each experiment, thaw aliquot(s) and dilute to 0.5 mg/mL with sterile ddH 2 O.
Preparation procedure: Filter medium using a 0.22 mm filter unit to sterilize. If making 1M HEPES from powder, adjust pH to 7.4. Solution is stored at 4 C for $2 months.
Preparation procedure: Filter the medium using a 0.22 mm filter unit to sterilize. Store at 4 C for 2-3 weeks. Horse serum is inactivated by heating to 56 C for 30 min.
Preparation procedure: Filter the medium using a 0.22 mm filter unit to sterilize and then add B-27 supplement. MM can be stored at 4 C for 3-4 weeks. Preparation procedure: Dissolve AraC in ddH 2 O. Aliquot into 0.5 mL and store at À20 C for up to 4 months.

Reagent
Preparation procedure: Dissolve SNAP or Halo substrate in DMSO and vortex for 10 min. Prepare 2 mL aliquots in amber tubes and store in the dark at À20 C for 6 months. Ligands come as 50 or 100 nmol and should be diluted to a final concentration of 1 mM with DMSO.
Working solution can be stored in the dark at 4 C for 2 months. Stock should be prepared in amber tubes. Harvesting embryos from an E18 timed-pregnant rat should be performed quickly to maintain highquality tissue. It is important to use fresh forceps and scissors between steps to prevent contamination.
1. Euthanize an E18 timed-pregnant rat using methods outlined in an institution-approved protocol, such as CO 2 euthanasia or decapitation. 2. Spray the ventral area of the female rat with 70% ethanol and make a transverse incision between the dermal layer and abdominal fascia. 3. Using the scissors and forceps, separate the layers and cut to expose the abdominal fascia (Figure 1A). (E) Incision to the mid-sagittal plane of the skull to expose the brain. Scale bar, 0.5 cm.
(F) E18 brain with intact hemispheres; brain stem was removed. Scale bar, 0.25 cm. (G) Right and left hemispheres of the brain. The meninges are removed from the hemisphere on the left but not the right. Scale bar, 0.25 cm.
(H) Interior of the brain with the hippocampus outlined in black. Scale bar, 0.25 cm.

OPEN ACCESS
4. Using a fresh pair of scissors and forceps, make an incision to open the abdomen and cut to expose the abdominal cavity ( Figure 1B). 5. Using a fresh pair of scissors and forceps, remove the uterus and place it in a 15 cm dish containing ice-cold HBSS ( Figure 1C).
Note: Use fresh pairs of scissors and forceps to prevent contamination (troubleshooting 1).
6. Remove the embryos from the uterus and place them into a 10 cm dish that contains ice-cold HBSS. 7. Separate and place the heads into a new 10 cm dish that has ice-cold HBSS ( Figure 1D). Keep the plate on ice.
Note: Placing the specimen into new HBSS helps wash away residue and blood.
Note: Make an even and straight cut when separating the head. This step is vital as the head must balance on the cut surface.

Hippocampal dissection
Timing: 60 min This section details the careful removal of hippocampi from dissected E18 rat embryo brains. The dissection should be performed efficiently under a stereo or dissecting microscope. Minimizing time spent dissecting the hippocampi will produce higher yields of viable cells.
8. Place cold HBSS into the lid of a 10 cm dish and add the embryo heads. 9. With the head in the prone position, make an incision into the mid-sagittal plane of the head from posterior to anterior using vannas scissors ( Figure 1E). 10. Peel back the cranium. Using a closed pair of vannas scissors, gently scoop under the brain in an anterior to posterior direction to remove it from the skull. Cut the brain stem to separate the brain ( Figure 1F). 11. Make a sagittal cut to separate the right and left hemispheres. 12. Remove the meninges from the brain's surface using two fine-pointed forceps ( Figure 1G). The meninges contain blood vessels, making them appear red in color. It is easiest to remove starting from the exterior, working towards the brain's interior. 13. With the interior hemisphere facing up, remove the hippocampus using vannas scissors. The hippocampus is located below the cortex and appears dark and moon-shaped ( Figures 1H  and 1I). 14. Use a P1000 pipette to transfer hippocampi to the 15 mL conical containing ice-cold HBSS.
Note: It is fastest to work in batches of 4-6 brains, completing the same step for each brain before proceeding to the next step (troubleshooting 2).
Note: It is essential to use ice-cold HBSS. The tissue becomes tacky, fragile, and challenging to work with when warm.

Hippocampal tissue dissociation and plating
Timing: 6 h This section contains steps for the trypsinization and trituration of hippocampi. Once the tissue is dissociated, cells are plated on the glass surface of PLL-coated imaging dishes. These steps should be performed in a Biosafety cabinet. 15. Remove HBSS from the 15 mL conical with a serological pipet. Try not to disturb the hippocampi. 16. Add 4.5 mL of warm HBSS and 0.5 mL of 2.5% trypsin. Invert the conical once and incubate in a 37 C water bath for 10 min. 17. Stop digestion by removing the trypsin solution using a 10 mL serological pipet. 18. Add 10 mL of warmed AM, allow hippocampi to settle (approximately 30 s), and remove media using a 10 mL serological pipet. 19. Repeat step 18 two more times. 20. Add 1 mL of AM and triturate tissue with a fire-polished Pasteur pipette. Try not to introduce air bubbles. Proceed until the media is cloudy and tissue pieces are no longer visible ($10-20 times).
Note: Fire-polished Pasteur pipettes have smoother edges and a narrowed tip which help break up tissue more efficiently while reducing cell damage.
21. Dilute 10 mL of cell suspension with 10 mL of 0.4% Trypan Blue Stain. Determine the cell viability and count using an automated counter.

Alternatives:
The cell count can also be determined using a hemacytometer.
CRITICAL: Trypan blue is a carcinogen and may cause cancer. Wear appropriate PPE when working with the solution.
22. Plate 1.25 3 10 5 cells onto the pre-coated/washed glass surface of 35 mm MatTek dishes containing warmed AM. Add dropwise, covering the entire glass surface. 23. Keep cultures at 5% CO 2 and 37 C for $2-5 h to provide adequate time for cells to settle and attach to the glass surface. 24. Ensure cells have adhered through visual inspection using a light microscope. Upon confirmation, replace media with MM and maintain cultures at 37 C and 5% CO 2 .

Maintaining hippocampal cultures
Timing: 6 d Hippocampal cultures are maintained by performing media exchanges every few days. In addition, AraC is added to reduce non-neuronal cell growth the day after plating. These steps should be performed in the Biosafety cabinet.
25. The day after plating, or 1 day in vitro (DIV), add 1 mM AraC to cultures to prevent the proliferation of non-neuronal cells. It is best to dilute AraC in a small amount of MM ($200 mL) and add it dropwise to the dish. Gently mix to ensure there is an even distribution of AraC.
Note: Adding AraC the day after plating reduces the number of non-neuronal cells without affecting neuronal growth or survival. If performing studies on aged cultures, it is best to continue adding AraC during feedings.
26. Every 3-4 days, replace 1 / 4 of culture media with equilibrated MM (i.e., media that has been warmed to 37 C and equilibrated to 5% CO 2 for a minimum of 30 min).
Note: Neurons are sensitive. Minimize stress by working quickly when handling cells outside the CO 2 incubator (troubleshooting 2). Note: According to the manufacturer, DNA-Lipofectamine complexes can be added to the culture medium in the presence or absence of serum/antibiotics. The manufacturer recommends using Opti-MEM Reduced Serum media for transient transfections. However, we found no difference in the transfection efficiency using Neurobasal media.

Transient transfection of cultured neurons
Note: Use an endotoxin-free DNA Maxi Kit to prepare concentrated and high-quality DNA plasmids. If dissolving plasmids in TAE, be cautious when working with low DNA concentrations, as too much EDTA can decrease transfection efficiency.
30. Add Tube 2 to Tube 1, gently mix by pipetting, and incubate at 20 C-22 C for 20 min.
Note: The manufacturer recommends adding Tube 1 to Tube 2 and incubating for 5 min. We found no difference in the transfection efficiency using our method. In addition, the increased incubation time allows for more DNA-Lipofectamine complexes to form. 31. Add transfection complexes dropwise to neuron cultures and incubate at 37 C in 5% CO 2 for 45 min. 32. During the incubation, add equal parts of saved media from step 28 with fresh MM and equilibrate at 37 C in 5% CO 2 . 33. After 45 min, replace the neurons' media with the conditioned media from step 32.

Mitochondrial damage and SNAP pulse-chase labeling assays
Timing: 2 d Within this protocol, mitochondrial damage is induced via antioxidant deprivation using the AOF MM, which initiates mild oxidative stress and results in low levels of mitochondrial damage. Mitochondria are damaged for 6 h while SNAP and Halo labeling typically require 1 h (Figures 3C-3E; troubleshooting 4). Thus, labeling occurs during the final hour of the damaging incubation period. The ligand is added for 30 min, followed by two quick washes and a 30-min washout. The first SNAP OPTN is recruited to damaged mitochondria and facilitates engulfment into an autophagosome, which fuses with a lysosome for downstream acidification to degrade the damaged organelle fully. (C and D) Timeline (C) and schematic (D) of the mitochondrial pulse-chase experiment. The day after transfection, mitochondria are damaged for 6 h and labeled with the first SNAP pulse (''Old'' Mito, magenta). Next, a nonfluorescent SNAP Block was added for 2 h to saturate the remaining SNAP binding sites, and neurons were left for $16 h. Before imaging, a second SNAP pulse (''Young'' Mito, green) was performed. (E) Representative images of ''Old'' and ''Young'' mitochondrial populations. The dynamic mitochondrial network is dual labeled (white). Scale bar, 5 mm.

OPEN ACCESS
pulse identifies ''Old'' mitochondria, while the second SNAP pulse labels ''Young'' mitochondria ( Figures 3C and 3D). For SNAP-Cell Block, the ligand is added for 120 min, followed by two quick washes and a 30-min washout. To complete all required washes, it is necessary to warm ample media to 37 C in the 5% CO 2 incubator. Mitochondrial damaging and pulse-chase labeling assays must be performed in the Biosafety cabinet. Before imaging, turn on the temperature and CO 2 controller and allow conditions to stabilize ($1 h). If using a microscope with a 5% CO 2 environmental chamber, image neurons in MM; if not, use a media that can maintain cells in ambient CO 2 , such as Hibernate E Low Fluorescence by BrainBits supplemented with B27. When performing quantitative live-cell imaging, design appropriate configurations where all channels have good signal-to-noise and avoid pixel saturation, fluorophore crosstalk, and accidental FRET.
42. Laser-scanning confocal setup using a Leica STELLARIS 8 equipped with a 633/1.40 oil objective and LAS X Software ( Figure 4A

OPEN ACCESS
c. Acquire Z-stack images in ''Sequential Scanning'' mode.
CRITICAL: Use ''Sequential Scanning'' mode and built-in crosstalk function in LAS X to determine fluorophore crosstalk (troubleshooting 6). The spectra of the fluorophores used in this protocol are close to each other. Thus, the detection window of SNAP-Cell 430 was narrowed to avoid crosstalk with EGFP.
43. Acquire images at 1024 3 1024 format (0.103 mm/pixel; image size 105.10 mm 3 105.10 mm), 0.3 mm step size (averaging 35-50 steps per image), using bidirectional scanning, 600 Hz scan speed, zoom factor of 1.75, and line average of 2. The pinhole size was 1 AU ( Figures  4B-4E and Methods video S1). a. Use the Over/Under look-up table to avoid pixel saturation throughout all Z-planes and for all channels (troubleshooting 6). 44. Deconvolve images using LIGHTENING Process with adaptive strategy, a refractive index of 1.33, and water as the mounting medium. All channels used 20 iterations, an optimization of 1, a regularization parameter of 0.05, and ''Very High'' smoothing. Contrast enhancement and cut-off were set to ''Auto.'' 45. Save images files for subsequent analysis.
Note: Pixel misalignment may occur when imaging and can be corrected during post-processing. Perform the correction before the analysis, as it can affect marker colocalization (troubleshooting 7).
Alternatives: Imaging times are reduced using media that maintains cells in ambient CO 2 . To determine how long a plate can be kept in ambient CO 2 , consult the manufacturer for the buffering capacity.

Timing: 2 h
We use spectrally distinct fluorophores to identify ''Old'' and ''Young'' mitochondria. To distinguish between these two populations, it is essential to note which channel corresponds to each fluorophore when analyzing the data. The ''Rectangle'' and ''Line'' functions in Fiji are used to identify OPTN-positive ''Old'' mitochondria and quantify the fluorescence intensity. a. Use the ''Line'' function to draw a line (width = 1) through the Halo-OPTN ring ( Figure 5D). The line should be drawn through a representative area, avoiding high-fluorescence artifacts and saturated pixels (troubleshooting 6). b. For each channel, use ''Plot Profile'' and ''List'' to obtain the fluorescence intensity values for the drawn line ( Figures 5E and 5F). c. Use the ''Rectangle'' function to draw a box and randomly measure the fluorescence intensity of the image background using ''Measure'' in the ''Analyze'' tab. Repeat in multiple representative areas and calculate the average background intensity for each channel. d. Correct the fluorescence signal for each channel by subtracting the average background intensity from the measured fluorescence intensity values from the line scan. e. For each channel, normalize the intensity values by dividing each value by the overall highest intensity value for that channel. Normalization sets the highest value to 1 and additional values to less than one. (Figures 5G,6A, and 6B). 51. Data can be presented as bar graphs, scatter plots, or pie charts with representative images.

EXPECTED OUTCOMES
Cellular antioxidants, such as superoxide dismutase, catalase, and glutathione, protect against oxidative damage by maintaining the oxidative balance. In this protocol, we described a mitochondrial damaging paradigm where antioxidant deprivation (AOF) induced mild oxidative stress (Joselin et al., 2012;Evans and Holzbaur, 2020a). Removing antioxidants from the cell culture media resulted in low levels of reactive oxygen species sufficient to damage mitochondria and initiated the mitophagy pathway. Within this pathway, autophagy receptors bind to phospho-ubiquitinated mitochondrial fragments and recruit autophagosomes to sequester and degrade damaged organelles via lysosome fusion (Evans and Holzbaur, 2020b;Youle and Narendra, 2011). Mitophagy events were identified by the translocation of OPTN, a known autophagy receptor, while lysosomal fusion was monitored using LAMP1 ( Figure 3B).
Here, we exploited a mitochondrially targeted SNAP-tag construct and SNAP ligands to monitor longterm mitochondrial degradation in primary hippocampal neurons (Grimm et al., 2015(Grimm et al., , 2017. Pulsechase labeling allowed for the visualization of distinct ''Old'' and ''Young'' mitochondrial populations (Figure 3). The ''Old'' mitochondria were labeled during the first SNAP pulse and comprised Mito-SNAP expressed from the time of transfection until SNAP Block labeling ($24 h). Meanwhile, the ''Young'' mitochondria were labeled during the second SNAP pulse and represented Mito-SNAP expressed from the time after SNAP Block until imaging ($22 h; Figure 3C). Since mitochondrial damage via AOF treatment coincided with the first SNAP pulse, we identified mitophagy events where OPTN translocated to ''Old'' fragmented mitochondria. Visualizing OPTN-positive ''Old'' mitochondria that were negative for the ''Young'' mitochondria marker demonstrates that these ''aged'' mitochondria were sequestered and engulfed before the addition of the second SNAP pulse. Next, we identified OPTN-positive ''Old'' mitochondria that were negative for ''Young'' mitochondria and quantified the ratio of events that were LAMP1-negative or LAMP1-positive. In control and AOF conditions, the majority of ''Old'' mitochondria were LAMP1-positive. Colocalization of ''Old'' mitochondria with LAMP1 illustrates that the autophagosome containing the damaged mitochondria fused with the lysosome for degradation. Surprisingly in both treatment groups, a population of OPTN-positive ''Old'' mitochondria was LAMP1-negative 24 h after initial damage, suggesting that this population of ''Old'' damaged mitochondria failed to fuse with lysosomes to undergo degradation ( Figure 6). Using this pulse-chase paradigm, we demonstrated that dysfunctional mitochondria persist for hours to days following mitochondrial insult. Thus, the degradation of engulfed mitochondria takes much longer than the minutes-to-hours time scale that was previously reported (Ashrafi et al., 2014;Hsieh et al., 2016). Inefficient degradation of damaged organelles may contribute to decreased neuron viability and neurodegeneration.
Our protocol could investigate mitochondrial turnover in disease backgrounds using primary cultures from OPTN transgenic knockout or mutant mice. Mutations in OPTN are linked to amyotrophic lateral sclerosis (ALS), a neurodegenerative disease caused by the gradual degeneration of motor neurons (Maruyama et al., 2010;Maruyama and Kawakami, 2013). In cell-based assays, expression of OPTN-linked mutations affected mitochondrial network health and either decreased the efficiency or disrupted the mitophagy pathway (Wong and Holzbaur, 2014;Evans and Holzbaur, 2020a). Therefore, our protocol could be easily adapted to study mitochondrial turnover in disease-related contexts. Additionally, modifying the timeframe between the two pulses would enable monitoring of the mitochondria over extended periods.
While this protocol was designed for primary hippocampal neurons, we anticipate that it will be easily adapted for other primary culture systems and non-neuronal cell lines. SNAP-tag

OPEN ACCESS
cloning vectors have been used to generate plasmids encoding fusion proteins localized to the nucleus, endoplasmic reticulum, and many other cellular structures. Thus, this SNAP pulse-chase labeling protocol could be used to monitor the long-term turnover of additional organelles.

LIMITATIONS
Primary hippocampal neuron cultures are isolated and do not have the same accompanying cells or architecture as in vivo, which could impact physiology. However, neuron cultures are routinely used to investigate numerous aspects of the cell biology of the neuron. In this protocol, live confocal microscopy was used to study mitochondrial damage. Care must be taken with the laser power and exposure time when imaging to prevent phototoxicity. Transient transfection can produce a range of protein concentrations that vary from cell to cell. Thus, it is vital to avoid high-expressing cells as it may alter the biology. Furthermore, the time scale of mitochondrial turnover observed is specific to primary hippocampal neurons. We expect the rates of mitochondrial turnover observed here to be consistent across neuronal cell types, but this has yet to be directly demonstrated.

TROUBLESHOOTING Problem 1
Sample contamination during embryonic dissection and hippocampal isolation.

Potential solution
There are multiple places where contamination can occur. It is best to wipe down all dissection surfaces and tools with 70% ethanol before starting the dissection. Animal-generated contaminants can be a significant source of contamination. Use a clean pair of scissors and forceps for each step of the embryo isolation during the dissection. Switching to a fresh pair of gloves between working with animals and isolating hippocampi can help prevent animal-generated contamination.

Problem 2
Poor survival or quality of primary hippocampal cultures. Steps 8 and 26.

Potential solution
Timed-pregnant female rats typically have 8-12 embryos. It is essential to inspect embryos when dissecting to ensure health. Avoid using too small/large or discolored embryos, as they can result in neuron death when plated. Performing successful dissections and obtaining high-quality cultures is skill-dependent (Figure 1). The survival and quality of cultures also increase with practice and repetition.
Additional factors that can affect the quality or survival of primary hippocampal cultures include 1) dissecting in a timely manner to maintain tissue quality, 2) performing procedures (e.g., transfections, feedings, labeling) efficiently to decrease the amount of time primary cultures spend outside the incubator, 3) using fresh or recently made media, and 4) carefully monitoring the lot-to-lot variability of media components (e.g., GlutaMAX, B27, and Neurobasal media).

Problem 3
Low transfection efficiency of primary hippocampal neurons.

Potential solution
If experiencing low transfection efficiency, increase the DNA concentration or the amount of Lipofectamine 2000 used. In this protocol, 1.1 mg total DNA and 4 mL Lipofectamine 2000 were used for transient transfection, which is in the lower range for concentrations specified by the manufacturer. Additionally, alternative transfection reagents or techniques may be used.
Additional constructs could be used to study mitochondrial turnover. If generating new SNAPtagged constructs, it is important that the tagged protein 1) is routinely used, 2) is specifically targeted to mitochondria, and 3) does not affect non-target pathways. While highly unlikely, low expression of Mito-SNAP could affect the feasibility of the protocol. Alternatively, Mito-Halo or Mito-Timer constructs could be used in place of Mito-SNAP.

Problem 4
Poor neuron survival following SNAP pulse-chase.

Potential solution
It is best to use prewarmed media to 37 C and equilibrated to 5% CO 2 in an incubator. Gently add and remove media to the imaging dish's side to avoid disturbing the neurons. We have found that adding the reserved media (step 34) back to the imaging dish after the SNAP block (step 39) helps with the viability of the neurons.
Moreover, the SNAP-Cell 430 labeling occurs in 500 mL of media to save on costly reagents. Low volumes of liquid could affect the relative oxygen tension. If researchers observe that the low volumes affect neuronal health, we recommend performing the SNAP labeling in 2 mL of media instead of 500 mL.

Problem 5
Difficulty visualizing SNAP or Halo ligands.

Potential solution
Difficulty visualizing ligands is typically due to inadequate labeling. To ensure equal distribution, we have found it is best to premix ligands in a small volume of media and then add it to the imaging dishes. Higher concentrations of ligands can be used if necessary. In addition, store ligands in the dark and avoid using ligands past the recommended storage time.

Potential solution
Select fluorophores with non-overlapping excitation and emission spectra to avoid fluorescence crosstalk and accidental FRET. We recommend using the ''Sequential Scanning'' Mode to separately excite individual fluorophores or pairs of fluorophores with non-overlapping spectra. Minimize the detection window to reduce the signal from other fluorophores or increase the number of settings to mitigate crosstalk. To check for the presence of fluorescence or accidental FRET, create new image parameters excluding one fluorophore while leaving the remaining laser lines intact. Comparing this null image with the image containing all laser lines will identify potential crosstalk generated while imaging. Several microscope software programs have built-in functions that determine the amount of crosstalk, such as Leica's LAS X.
When acquiring images, it is critical to avoid pixel saturation. Saturated pixels can significantly alter the normalized fluorescence intensity for line scans. Use the Over/Under look-up table to identify pixel saturation, visualized as blue pixels. It is necessary to adjust the settings for each channel and scroll through the Z-planes to ensure there are no saturated pixels.

Potential solution
It is possible that pixel misalignment can occur and is most easily visualized when comparing SNAP-Cell 430 and SNAP JF549cp channels. The offset of the lasers and mirrors in the microscope's optical path leads to image misalignment. To correct this, we recommend using optical beads to determine the relative alignment of the channels in X, Y, and Z. If the alignment is significantly off, we suggest having the microscope realigned. Minor misalignments can be corrected during post-processing. LAS X software can be used to fix misalignment in XY, but not Z. To alter the registration in X, Y, and Z, we recommend using the ''MultiStackReg'' plugin in Fiji.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Chantell Evans (chantell.evans@duke.edu).

Materials availability
This study did not generate new unique reagents. The recombinant DNA used in this study is available from the lead contact upon request.
Data and code availability https://doi.org/10.7554/eLife.50260 Any additional information is available from the lead contact upon request.