Using Localization Microscopy to Quantify Calcium Channels at Presynaptic Boutons

Materials and reagents

Biological materials

1. C. elegans strain EG9617 (Jorgensen Lab, elks-1::Skylan-S, egl-19::SNAP IV; unc-2::HALO X)

2. C. elegans strain EG9667 (Jorgensen Lab, elks-1::Skylan-S, egl-19::SNAP IV; unc-68::HALO V)

3. Nematode growth media (NGM) plates seeded with OP50 (see Wormbook.org)

1. DMSO (Millipore Sigma, catalog number: D2650-5X5ML)

2. JF549cp-SNAPtag Ligand (Janelia Labs), store freeze-dried at -80 °C. Available by request from: https://janeliamaterials.azurewebsites.net/

3. JF646-HALOtag Ligand (Janelia Labs), store freeze-dried at -80 °C. Available by request from: https://janeliamaterials.azurewebsites.net/

4. Sodium azide (NaN₃) (Millipore Sigma, catalog number: S2002)

5. Agarose (Gold Biotechnology, catalog number: A-201-1000)

6. NaCl (Millipore Sigma, catalog number: S5886-5KG)

7. K₂HPO₄ (Millipore Sigma, catalog number: P8281-500G)

8. Na₂HPO₄ (Fisher Scientific, catalog number: S373-3)

9. MgSO₄ (Millipore Sigma, catalog number: M2643-500G)

1. M9 buffer (see Recipes)

2. 25 mM sodium azide in M9 (see Recipes)

3. 4% agarose in M9 (see Recipes)

1. M9 buffer

   Reagent	Final concentration	Amount
   K2HPO4	22 mM	3 g
   Na2HPO4	34 mM	6 g
   NaCl	86 mM	5 g
   MgSO4 (1 M)	1 mM	1 mL
   H2O	n/a	1,000 mL
   Total	n/a	1,000 mL

   M9 has a shelf-life of one year. Dispose if fungal or bacterial contamination occurs.




2. 25 mM sodium azide in M9

   Reagent	Final concentration	Amount
   Sodium azide (powder)	25 mM	16.26 mg
   M9 buffer	n/a	10 mL
   Total	n/a	10 mL

   Store at -20 °C in single-use aliquots to avoid repeated freeze-thaw cycles. The paralytic properties of sodium azide solution begin to degrade after one year in storage.




3. 4% agarose in M9

   Reagent	Final concentration	Amount
   Agarose	4%	2 g
   M9 buffer	n/a	50 mL
   Total	n/a	50 mL

   Heat in the microwave until dissolved. Store at 4 °C in aliquots to avoid evaporation from heating cycles. Agarose aliquots have a shelf-life of one year. Dispose if evaporation or fungal or bacterial contamination occurs.

Laboratory supplies

1. 1.5 mL microcentrifuge tubes (Life Science Products, model: Omniseal, catalog number: M-1700C-1M)

2. Pasteur pipettes (Fisher Scientific, catalog number: 13-678-20B)

3. Microscope slides (Fisher Scientific, model: Gold Seal, catalog number: 3048)

4. Zeiss 1.5 high-performance coverslips (Zeiss, catalog number: 474030-9000-000)

5. Worm pick (see Wormbook.org)

Equipment

1. Vutara SRX 352 (Bruker)

2. Benchtop centrifuge (Thermo Fisher, model: mySPIN6, catalog number: 75004061)

3. Heat block set to 95 °C (e.g., Eppendorf, model: ThermoMixer F1.5, catalog number: 5384000020)

Software and datasets

1. Vutara SRX (7.00.00rc39, 2020)

2. MATLAB (R2022a)

3. Microsoft Excel

Procedure

1. Synchronize worms by staging embryos

   1. Day 1

      1. At 9:00 am, use a worm pick to move 20 young adult animals onto a new NGM plate seeded with OP50; do this three times to create three plates total. This procedure takes approximately 10 min.

      2. At 5:00 pm, move the 20 adult animals from each plate onto new NGM plates seeded with OP50 to create three additional plates.

   2. Days 2–3

      1. At 9:00 am, move the 20 adult animals from each plate onto new NGM plates seeded with OP50 to create three additional plates.

      2. At 5:00 pm, move the 20 adult animals from each plate onto new NGM plates seeded with OP50 to create three new plates.

      3. Move the original adult animals to fresh plates in the morning and evening for one more day.




2. Label SNAP and HALO

   1. Day 4, evening

      1. Inspect plates to select for broods with an abundance of L4 hermaphrodite animals. This is a synchronized plate.

      2. Wash animals off the three synchronized plates using ~1 mL of M9 buffer and a glass Pasteur pipette. Place animals into a microcentrifuge tube. Worms will stick to the plastic of a micropipette, so use glass to move animals.

      3. On a benchtop centrifuge, gently spin (10 s each) to remove bacteria, removing the supernatant and washing with 1 mL of M9 buffer.

      4. After three washes, remove the supernatant and resuspend the animals in approximately 195 μL of M9 buffer.

      5. Prepare JF dyes by resuspending at 1 mM in DMSO (e.g., 5 nmol of dye in 5 μL of DMSO). Move dye + DMSO suspension into the next dried dye aliquot to create a combination of dyes in suspension. This will create 5 μL of 1 mM JF646-HALOtag Ligand + JF549cp-SNAPtag Ligand dye.

      6. Add dye + DMSO solution to worms in M9 buffer to create a 25 μM final concentration of dye. Mix gently by vortexing.

         Note: DMSO above 5% is toxic to worms.

      7. Put the tube with the lid in a freezer box on an orbital shaker and stain for 120 min at RT. Gently flick and spin down worms every 15 min to prevent worms from forming clumps on the bottom of the tube or becoming stuck on the sides of the tube. It is essential that the dye and worms incubate in a dark place like a lidded freezer box.

      8. After 2 h, spin down the worm solution, aspirate off the supernatant, and wash four times with 1 mL of M9 buffer.

      9. With a Pasteur pipette, transfer worms to at least three OP50-seeded NGM plates. Let the M9 droplet dry completely before inverting the plate. This should be performed in the dark (we use a drawer). After the droplet has dried, cap the plates and invert. Let the worms recover in a lidded freezer box for 10–12 h. It is important to not expose the stained worms to light. Ideally, the L4 to adult molt will occur during recovery to minimize background staining of the cuticle and gut.

         Note: L4 hermaphrodite animals can be identified by the absence of a vulva, which appears as a white crescent at the center of the ventral side of the animal under a typical dissection microscope. Starting with synchronized plates is perhaps the most crucial step of the protocol because the dye sticks to the cuticle of the animal. If adults are exposed to dye, they will have non-specifically labeled cuticles forever, which interferes with imaging of the nervous system. L4s are the final larval stage of the animal and must molt their cuticle before adulthood. Thus, if L4s are stained and then imaged as young adults, a molt must have occurred during recovery.




3. Imaging preparation

   1. Day 5, morning

      1. Prepare a 4% agarose pad by heating an agarose aliquot in a 95 °C heat block and placing a small drop on a microscope slide. Then, flatten the pad with another microscope slide (Figure 2A).

      2. Trim the agarose pad to about 10 mm × 5 mm using a microscope slide. The agarose pad should fit entirely under the coverslip (Figure 2B).

      3. Add 2–4 μL of 25 mM NaN₃ to the pad.

      4. Use the worm pick to distribute the NaN₃ drop evenly across the agarose pad. If NaN₃ flows over the edges of the agarose pad, reduce the amount used.

      5. Transfer 20 young adult worms to the agarose pad. Older animals may have been young adults during staining and will be highly fluorescent because they have not shed their cuticle.

      6. Wait for animals to paralyze and straighten. This should take a few minutes. If the droplet of NaN₃ is drying out, place the slide into a humidity chamber. Seal the pad with a coverslip. Then, seal the coverslip with nail polish and image (Figure 2B).

         Notes:

         i. Do not image animals that have been in NaN₃for more than 60 min. The paralytic sodium azide causes neuron blebbing and disruption of cell morphology after approximately 1 h.

         ii. Avoid air bubbles in the agarose pad or trim them.

   
   

   
   Figure 2. Creating a worm imaging pad and slide. (A) The agarose pad acts as a support for the worm sample so it is not crushed by the microscope slide, also immobilizing the animal. (B) Once sealed with nail polish, the worms should last for hours. However, the sodium azide paralytic causes cell blebbing and disruption of cell morphology after about an hour. (C) Left: Under the brightfield illumination of the Vutara SRX 352, the worms will appear as pictured. Right: The roll can be in part determined by the position of the ventral nerve cord exiting the neuropil if there is a fluorescent marker in the nervous system or position of the vulva. Scale bar = 5 μm.




4. Superresolution imaging of C. elegans

   1. Image worms by 3-color imaging

      1. Organize laser lines in descending order of wavelength, e.g., 1) 646 nm, 2) 549 nm, 3) 488 nm. This reduces the photobleaching of the sample.

      2. In widefield mode: Locate worms on the pad and coverslip.

      3. In super-resolution mode:

         i. Use only extremely low laser powers (< 1%).

         ii. Set the exposure time to 300 ms.

         iii. Focus both imaging planes as much as possible on the dorsal nerve cord (Figure 2A).

         iv. Turn lasers off as soon as possible. We try to have the laser on for under 10 s while focusing on the synapses we plan to image.

      4. After the nerve cord is in focus, reduce the exposure time to 30 ms and proceed to record. Tune laser powers for optimal blinking vs. the lifespan of the fluorophores. An example of a properly tuned laser power is shown in Figure 3B.

         Notes:

         i. The best samples are worms that are oriented with their dorsal nerve cord tilted toward the objective.

         ii. Be very cautious with the laser power while finding your sample. The worm is not in oxygen scavenging buffer; thus, fluorophores will bleach. To focus on the cord, use very low laser power and a 300 ms exposure time.

         iii. Optimizing laser power while imaging: A laser power that is too low will not induce blinking of the fluorophores. A laser power that is too high will bleach the sample in a few frames. The user should find these limits and note them for their microscope. The localization counts over time displayed in Figure 3B is an example of an ideally tuned laser power.

   2. Image worms—counting channels

      1. Organize laser lines by starting with the protein you wish to quantify, followed by fiducial markers. Fiducial markers label a known structure, like the synaptic dense projection, and are used to spatially orient the experimental channel.

      2. In widefield mode: Locate worms on the pad and coverslip.

      3. In super-resolution mode:

         i. Use only extremely low laser powers (< 1%).

         ii. Set the exposure time to 300 ms.

         iii. Focus both imaging planes as much as possible on the dorsal nerve cord.

         iv. Turn lasers off as soon as possible.

      4. After the cord is in focus, reduce the exposure time to 30 ms and proceed to record. Increase the laser powers for optimal blinking vs. lifespan of the fluorophores. This is typically between 4% and 12% with 200 mW lasers and the 40 μm biplane module. For counting emitters, the dyes should be completely bleached after imaging; this typically occurs within 4,000 frames.

         Notes:

         i. The best samples are worms that are oriented with their dorsal nerve cord tilted toward the objective.

         ii. Optimizing laser power while imaging: A power that is too low will not induce blinking of the fluorophores. A laser power that is too high will bleach the sample in a few frames. The user should find these limits and note them for their microscope. The localization counts over time displayed in Figure 3B is an example of an ideally tuned laser power.




5. Localizing

   1. Proceed to the Localization tab.

   2. Set the background threshold. This should be optimized to minimize background localizations (false positive) but not exclude real emitters (false negative). Adjust the background threshold to produce cutouts (colored boxes) around real signals but minimal cutouts in areas with no signal. An example of the effect of background thresholds is displayed in Figure 3C. For C. elegans, we find the ideal threshold to be between 3 and 5 and rely on denoising (detailed below) to remove false positives.

   3. Select images that demonstrate blinking by adjusting the region of interest frame (Probe x) slider. This is primarily to exclude frames where fluorophores are not blinking. Blinking can be determined by moving 10 frames forward and 10 backward (600 ms range). If emitters are turning on and off between those frames, then they are suitable to analyze.

      Notes:

      1. At the beginning of imaging, all fluorophores will be active, and separate light flashes will not be visible. Blinking may not occur for a few hundred frames into the acquisition. Non-blinking frames should not be analyzed.

      2. Background threshold should be optimized across the acquired frames.

      3. Localization cutouts can be turned on in Expert Options under the Localizations tab.

      
      

      
      Figure 3. Biplane, blinking, and background thresholding. (A) The Vutara SRX simultaneously images in two focal planes (“biplane”), which splits the light path into two focal planes separated by about 700 nm. To focus on a structure like the dorsal nerve cord, it should be equally in focus on both planes—which ensures that the emitters are in between the two focal planes. Scale bar = 5 μm. (B) Graph of localization cutouts over time. At the start of imaging, there will be many fluorophores in an excitable state. However, as imaging progresses, fluorophores will begin to enter the dark state. This change can be confirmed by the drop in localization cutouts over time. Stochastic blinking is required for super-resolution and can be confirmed by the spikes and troughs in the number of cutouts over time. (C) Examples of background thresholding on image reconstruction of CaV2 in the dorsal nerve cord (background threshold = 2, 10, or 30). Upper: Magenta boxes represent successful PSF fitting. Red arrow shows an example of a false positive. Black arrow shows an example of a successful localization. White arrow shows an example of a false negative. Lower: Magenta points represent protein localizations. Scale bar = 1 μm.




6. Visualizing

   Proceed to the Visualization tab.

   1. Discard localizations that are not accurate in the Advanced Particle Filters dialogue. For live worms, we set an arbitrary threshold of 50 nm radial precision and display localizations on the visualization plot as a 50 nm particle size (Figure 4A).

   2. Denoise the image using Method: Mean Distance and denoise by Percent. Lower the slider until most of the signal outside the cord is removed but the signal inside the cord appears unperturbed (Figure 4B).

   3. Save this image to the View Manager.

   4. For analysis of proteins at individual synaptic boutons, select a region of interest 700 nm radius from the center of an in-focus dense projection. Save this view in View Manager. Export the data and save it as a particle .csv file. Out-of-focus synapses and synapses with chromatic aberrations in the XY plane are discarded. Some chromatic aberrations will be present in the Z plane. Aim to analyze five synapses per animal from five different animals, with a total of 25 synapses analyzed. We recommend saving particle files ordered by “animal#_synapse#”, e.g., “3_1.csv.”

      
      

      
      Figure 4. Data filtering. (A) First, reconstructions are filtered to exclude imprecise localizations. (B) Next, localizations are trimmed based on nearest neighbors using the denoising tool. Finally, a region of interest is drawn around a synapse to be analyzed to exclude localizations that cannot be within the synapse based on the size of the bouton. Numbering and green arrows show the click order for the user. Magenta = CaV2, Yellow = CaV1.




7. Cluster analysis

   1. Click on Advanced Statistics Dialogue and go to the Cluster Analysis tab.

   2. Set the parameters for calling clusters based on the size of the protein and resolution of the image. See note.

   3. Click Compute.

   4. Export the resulting output as a .csv file.

      Note: We assume that multiple localizations within a close distance are the same channel blinking multiple times. That distance is defined by the size of the calcium channel and the resolution. Voltage-gated calcium channels are approximately 10 nm in size, and localizations worse than 50 nm in precision were discarded. Thus, any localizations within 60 nm could be from the same channel. Therefore, the Maximum particle distance is set to 60 nm. The Minimum particle count to form cluster is set to 2. An example of clustering is available in Figure 5A.

Data analysis

Mapping distributions of proteins from particle files (produced in section F):

1. Open MATLAB and load the following script into a command window.

   https://github.com/bdmscience/probeDistances/blob/main/probeDist.m

2. Change the address field to the folder where the particle files are saved.

3. Modify the underlined section of the first line of code to load the particle file you wish to analyze:

   particles = readtable('____.csv')

4. Change “mkdir” to the worm and synapse you are analyzing, e.g., mkdir 3_1. Then, cd to that new directory.

5. Lines 4–9 set the x and y coordinates of each imaging channel. “particles.vis_probe == 0” is channel 1 in SRX, “==1” is channel 2, and “== 2” is channel 3. Keep this in mind when considering which channels to analyze. The script default is to compare channel 3 to channel 1.

6. Run each section of the code to measure the distances between the localizations of the first channel to the center of the second channel, and the localizations of the second channel to the center of the second channel.

7. Each section will output two variables called “dist”; these are your measurements.

   Note: After each section of code, save the “dist” distance outputs before analyzing a new set of channels. We recommend you plot the data in MATLAB. Other graphing software has a maximum number of points that can be displayed on a plot.

Counting calcium channels

Foreword: Presented here are two methods to quantify the number of calcium channels at synaptic boutons. Both methods rely on clustering localizations into possible single protein domains. A caveat for both of these methodologies is that tight clusters of proteins can skew the estimate for blinks or for photon yield for a single protein. These methods of counting are more accurate for dispersed proteins than densely localized proteins but likely produce an undercount of the total number of proteins.

The methods differ in the quantification of signal within those protein clusters. The first calculates the number of localizations per protein. The second calculates the number of photons per protein. The advantages of one over the other are unclear; we recommend users attempt both methods and compare results. For a full discussion of counting methods refer to the methods section titled “Counting calcium channels” in Mueller et al. [14].

1. Counting proteins by position and particle count (Figure 5B and 5C):

   1. Open the cluster file in Excel from section G in Microsoft Excel.

   2. The ID column refers to cluster ID. Each row is a different cluster of localizations. Cluster 0 are single blinks. “particleCount” contains the number of localizations from channel 1 in each cluster. Using this information, we can calculate how many blinks each protein emitted.

   3. Create a new Excel file called spreadsheet2. Name columns as Number of blinks, Count, Particles in Clusters, Total particles, and Single particles.

   4. Create rows numbered 1 through the max number of “particleCount” in a cluster ID.

   5. Manually count how many clusters contain each number of particles and input the total count in the “count” column. Do this for each synapse and replicate.

   6. Combine these data across synapses or replicates and fit a Poisson distribution to the cumulative data. Lambda equals the mean number of blinks emitted from a protein in the dataset. Record this value on spreadsheet2.

   7. Take the sum of the total number of localizations at each synapse and then divide the total number of blinks by the lambda value to yield how many proteins are present at each synapse. Plot the mean and SEM of these data. We recommend analyzing at least six synapses from three different animals.

2. Counting proteins by photon flux (Figure 5D and 5E):

   1. Open the particle file from section G in Microsoft Excel.

   2. Sort particles by clusterID. Note that clusterID = 0 are non-clustered localizations.

   3. Sum the photons (“photon-count”) for each non-clustered localization and each clusterID.

   4. Sum the total photons for all ClusterIDs.

   5. Repeat 2a–2d for each synapse.

   6. Take the average “photon-count” per cluster across the dataset.

   7. Divide the “photon-count” from each synapse by the average “photon-count” per cluster (total photons divided by the average number of photons per channel). The result is the number of channels present at that synapse by photon flux. Plot the mean and SEM of these data. We recommend analyzing at least six synapses from three different animals.

      Note: We found that counting by photon flux had higher variability than counting by particle count; however, there was agreement between the results. An example of blink number and photon yield for clusters is present in Figure 5F.

      
      

      
      Figure 5. Counting calcium channels. (A) CaV2 (magenta) and CaV1 (yellow) localizations with cluster shells. (B–C) Quantification output from counting by particle count. Poisson curve fitting revealed that each protein blinked 2.7 times. (D–E) Quantification output from counting by photon flux. Each protein emitted on average 1996 photons. (F) Example statistics from voltage-gated calcium channel clusters at C. elegans synapses. Scale bar = 250 nm. Figure adapted from Mueller et al. [14].

Data availability and example data

All data mentioned in this study is available as source data files with the original publication. For ease of access, we have included particle files, cluster files, and channel count files as supplementary files with this protocol.

Validation of protocol

This protocol or parts of it has been used and validated in the following research article(s):

1. Mueller et al. [14]. CaV1 and CaV2 calcium channels mediate the release of distinct pools of synaptic vesicles. eLife (Figure 4, panels B, C; Figure 5, panels A–C; Figure 5 supplement 1, panels A–E; Figure 6, panels A–D; Figure 7, panels A–D; Figure 8, panels A–D; Figure 10, panels A–G).

Acknowledgments

We thank Luke Lavis and the Lavis lab for providing all JF dyes used in this study. We thank Brittany Graham for feedback on the manuscript, and Josh Kaplan who suggested we attempt to count calcium channels from these datasets. An application of this protocol was originally described in Mueller et al. [14]. Funding sources for this work were the NSF NeuroNex 2014862 to E.M.J., and NIH R01 NS034307 to E.M.J., F31 NS084826 to S.A.M. E.M.J is an investigator at Howard Hughes Medical Institute.

Competing interests

E.M.J holds a patent on optical biplanes, which are integral to the design of the microscope discussed in this study. However, E.M.J does not have any financial stake in Bruker.

References

1. Dani, A., Huang, B., Bergan, J., Dulac, C. and Zhuang, X. (2010). Superresolution Imaging of Chemical Synapses in the Brain. Neuron. 68(5): 843–856.
2. Südhof, T. C. (2012). The Presynaptic Active Zone. Neuron. 75(1): 11–25.
3. Tang, A. H., Chen, H., Li, T. P., Metzbower, S. R., MacGillavry, H. D. and Blanpied, T. A. (2016). A trans-synaptic nanocolumn aligns neurotransmitter release to receptors. Nature. 536(7615): 210–214.
4. Vyleta, N. P. and Jonas, P. (2014). Loose Coupling Between Ca²⁺ Channels and Release Sensors at a Plastic Hippocampal Synapse. Science. 343(6171): 665–670.
5. Huang, B., Wang, W., Bates, M. and Zhuang, X. (2008). Three-Dimensional Super-Resolution Imaging by Stochastic Optical Reconstruction Microscopy. Science. 319(5864): 810–813.
6. Juette, M. F., Gould, T. J., Lessard, M. D., Mlodzianoski, M. J., Nagpure, B. S., Bennett, B. T., Hess, S. T. and Bewersdorf, J. (2008). Three-dimensional sub–100 nm resolution fluorescence microscopy of thick samples. Nat Methods. 5(6): 527–529.
7. Mlodzianoski, M. J., Juette, M. F., Beane, G. L. and Bewersdorf, J. (2009). Experimental characterization of 3D localization techniques for particle-tracking and super-resolution microscopy. Opt Express. 17(10): 8264–8277.
8. Pavani, S. R. P., Thompson, M. A., Biteen, J. S., Lord, S. J., Liu, N., Twieg, R. J., Piestun, R. and Moerner, W. E. (2009). Three-dimensional, single-molecule fluorescence imaging beyond the diffraction limit by using a double-helix point spread function. Proc Natl Acad Sci USA. 106(9): 2995–2999.
9. von Diezmann, L., Shechtman, Y. and Moerner, W. E. (2017). Three-Dimensional Localization of Single Molecules for Super-Resolution Imaging and Single-Particle Tracking. Chem Rev. 117(11): 7244–7275.
10. Nieuwenhuizen, R. P. J., Lidke, K. A., Bates, M., Puig, D. L., Grünwald, D., Stallinga, S. and Rieger, B. (2013). Measuring image resolution in optical nanoscopy. Nat Methods. 10(6): 557–562.
11. Hellmeier, J., Strauss, S., Xu, S., Masullo, L. A., Unterauer, E. M., Kowalewski, R. and Jungmann, R. (2024). Quantification of absolute labeling efficiency at the single-protein level. Nat Methods. doi.org/10.1038/s41592-024-02242-5.
12. Kurshan, P. T., Merrill, S. A., Dong, Y., Ding, C., Hammarlund, M., Bai, J., Jorgensen, E. M. and Shen, K. (2018). γ-Neurexin and Frizzled Mediate Parallel Synapse Assembly Pathways Antagonized by Receptor Endocytosis. Neuron. 100(1): 150–166.e4.
13. Li, P., Merrill, S. A., Jorgensen, E. M. and Shen, K. (2016). Two Clathrin Adaptor Protein Complexes Instruct Axon-Dendrite Polarity. Neuron. 90(3): 564–580.
14. Mueller, B. D., Merrill, S. A., Watanabe, S., Liu, P., Niu, L., Singh, A., Maldonado-Catala, P., Cherry, A., Rich, M. S., Silva, M., et al. (2023). CaV1 and CaV2 calcium channels mediate the release of distinct pools of synaptic vesicles. eLife. 12: e81407.
15. Adler, J. and Parmryd, I. (2010). Quantifying colocalization by correlation: The Pearson correlation coefficient is superior to the Mander's overlap coefficient. Cytometry Part A.: 733–742.
16. Bayas, C. A., Wang, J., Lee, M. K., Schrader, J. M., Shapiro, L. and Moerner, W. E. (2018). Spatial organization and dynamics of RNase E and ribosomes in Caulobacter crescentus. Proc Natl Acad Sci USA. 115(16). doi.org/10.1073/pnas.1721648115.

Supplementary information

The following supporting information can be downloaded here:

1. Cluster files
2. Particle files
3. Counting channels by photon flux.xlsx
4. Counting channels by position and particle count.xlsx