Fast and multiplexed super resolution imaging of fixed and immunostained cells with DNA-PAINT-ERS

Abstract

Recent advances in super resolution microscopy have enabled imaging at the 10–20 nm scale on a light microscope, providing unprecedented details of native biological structures and processes in intact and hydrated samples. Of the existing strategies, DNA points accumulation in imaging nanoscale topography (DNA-PAINT) affords convenient multiplexing, an important feature in interrogating complex biological systems. A practical limitation of DNA-PAINT, however, was the slow imaging speed. In its original form, DNA-PAINT imaging of each target takes tens of minutes to hours to complete. To address this challenge, several improved implementations have been introduced. These include DNA-PAINT-ERS (where E = ethylene carbonate; R = repeat sequence; S = spacer), a set of strategies that lead to both accelerated DNA-PAINT imaging speed and improved image quality. With DNA-PAINT-ERS, imaging of typical cellular targets such as microtubules takes only 5–10 minutes. Importantly, DNA-PAINT-ERS also facilitates multiplexing and can be easily integrated into current workflows for fluorescence staining of biological samples. Here we provide a detailed, step-by-step guide for fast and multiplexed DNA-PAINT-ERS imaging of fixed and immunostained cells grown on glass substrates as adherent monolayers. The protocol should be readily extended to biological samples of a different format (for example tissue sections) or staining mechanisms (for example using nanobodies).

Basic Protocol 1: Preparation of probes for DNA-PAINT-ERS

Basic Protocol 2: Sample preparation for imaging membrane targets with DNA-PAINT-ERS in fixed cells

Alternate Protocol 1: Immunostaining of extracted U2OS cells

Basic Protocol 3: Super resolution image acquisition and analysis

INTRODUCTION:

Super resolution microscopy (SRM) has evolved into an increasingly powerful and flexible toolbox for biological imaging at the nanoscale (Liu et al., 2022). Among the many existing SRM modalities (Betzig et al., 2006; Gustafsson, 2000; Hell & Wichmann, 1994; Hess et al., 2006; Rust et al., 2006), DNA-PAINT is particularly useful when imaging multiple targets, which is frequently the case when studying complex biological processes (Jungmann et al., 2014). Like other single-molecule localization-based SRM techniques, DNA-PAINT achieves ‘super’ resolution through sub-diffractive localization of fluorescence signals from sparse single emitters. In DNA-PAINT, the localization signals arise from transient hybridizations between a pair of short DNA oligos (Jungmann et al., 2010). The target of interest is tagged with a single-stranded DNA oligonucleotide designated as the docking strand (DS), which is then probed using a fluorophore-conjugated, complementary DNA oligonucleotide designated as the imaging strand (IS). Typically, the DS-IS duplex is only ~10 base pairs, yielding transient hybridization events. Multiple targets can each be labeled with a distinct DS and probed using the corresponding IS in sequential imaging cycles (Jungmann et al., 2014).

A practical issue with DNA-PAINT in its original form, however, is the slow imaging speed, where each target can take tens of minutes to hours to complete (Jungmann et al., 2014). Several strategies have been introduced to address this problem (Auer et al., 2017; Lee et al., 2018, 2017; Schueder et al., 2019). In our lab, we have developed DNA-PAINT-ERS (Civitci et al., 2020) for fast and multiplexed SRM imaging. DNA-PAINT-ERS comprises three simple modifications to the DNA construct and the imaging buffer: E (ethylene carbonate or EC), R (repeat sequence), and S (spacer) (Figure 1). Of these, EC is an additive to the standard DNA-PAINT imaging buffer, and R and S are incorporated into the design of previously validated imaging constructs. Together, these modifications lead to 5–20x accelerated DNA-PAINT imaging, significantly shortening the acquisition time of single- and multi-target imaging (Civitci et al., 2020). With DNA-PAINT-ERS, we were able to complete imaging of common cellular targets such as microtubules in a matter of 5–10 minutes. Owing to optimized imaging kinetics, DNA-PAINT-ERS also improves the quality of the resulting SRM images.

Figure 1. Schematic of DNA-PAINT-ERS.

The top panel illustrates the standard DNA-PAINT process utilizing DS-conjugated antibodies bound to a target of interest and complementary IS diffusing above the sample. Binding of IS to DS followed by subsequent unbinding generates transient (~0.1 s) single-molecule signals, which accumulate over time and are then localized with sub-diffractive precision, typically on the order of 8–10 nm. The bottom panels depict the individual effects of ethylene carbonate (E), repeat sequence (R), and spacer (S) on DNA hybridization kinetics, respectively. DNA oligos and antibodies are not drawn to scale.

Here, we provide a detailed, step-by-step guide for DNA-PAINT-ERS. The guide comprises three basic protocols (Figure 2) and an alternate protocol. In Basic Protocol 1, we describe how to prepare DS-conjugated antibodies and fluorophore-labeled IS suited for DNA-PAINT-ERS. In Basic Protocol 2, we introduce the steps for immunostaining of cultured cells using DS-conjugated antibodies. Alternate Protocol 1 describes steps for immunostaining of extracted cells, which is advantageous for specific targets. Basic Protocol 3 shows the procedures for DNA-PAINT-ERS imaging and an example of image analysis pipeline. These steps are described in the context of imaging microtubules and clathrin in U2OS cells, but the procedures should be widely applicable to other targets of interest, provided that good affinity agents are available. As will be shown in detail, the additions of DS with repetitive sequences (DS 2x or DS 3x) as well as the PEG spacer involve only minor modifications to existing DNA-PAINT workflows, and the addition of EC to standard DNA-PAINT imaging buffers is also straightforward. Additionally, we will discuss several critical steps in immunostaining to ensure that the sample is prepared with high quality for best DNA-PAINT-ERS imaging results. These steps should be readily adapted to other types of affinity agents such as nanobodies (Koester et al., 2022) and sample formats such as clinical tissue sections (Rames et al., 2022).

Figure 2.

Overview of the protocols described in this article for imaging of fixed and immunostained cells with DNA-PAINT-ERS.

BASIC PROTOCOL 1

Preparation of probes for DNA-PAINT-ERS

DNA-PAINT-ERS is based on DNA-PAINT with important modifications, starting with the design of the imaging constructs. Additions of DS sequence repeats and a spacer between the DS and the affinity agent lead to optimized DNA hybridization kinetics and in turn accelerated imaging speed and improved image quality (Civitci et al., 2020). We describe the alternative DS design and the methodology to produce DS-conjugated secondary antibodies suitable for improved imaging using DNA-PAINT-ERS (Figure 3A). Conjugating DS to antibodies is achieved through copper-free click chemistry (Agard et al., 2004) (Figure 3B). DS oligos containing a primary amine are reacted with dibenzocyclooctyne (DBCO) N-hydroxysuccinimide (NHS)-esters, while antibodies are tagged with azide groups via the same mechanism. Following the conjugation reaction, ethanol precipitation is performed to purify the DS-DBCO conjugate, while the antibody-azide product is purified with the use of ultrafiltration. In the following step, DS-DBCO is then reacted with the azide-tagged antibodies to obtain DS-labeled antibodies. In parallel, IS complementary to DS is labeled with a fluorophore (e.g. CF660R or Atto643) that has a high single-molecule brightness for improved localization precision. Conjugated IS-dye product is purified by two rounds of ethanol precipitation. A number of different antibodies can be labeled with a variety of DS allowing for high multiplexing capabilities without cross-talk of the orthogonal DNA sequences.

Figure 3. Schematic of DS design and production of DS-labeled antibodies for DNA-PAINT-ERS.

(A) Schematic of DS modifications including repeat sequences (“docking site”) in blue with optional extra bases between the two binding sites, and a fluorophore such as Cy3 (orange) at the end for DNA-PAINT-ERS. Here Cy3 is not used for DNA-PAINT-ERS but instead to allow sample viewing and quality check after immunostaining and before super resolution imaging; (B) Schematic representation of site-specific copper-free click chemistry ligation reaction for the conjugation of azide-tagged secondary antibodies with single-stranded DBCO-modified DNA oligonucleotides. Prior to this reaction DS and antibodies are reacted with DBCO-PEG4-NHS-ester and azido-PEG4-NHS-ester, respectively. The linker moieties between the ssDNA and the affinity agent serve as a flexible ‘spacer’ that positions the ssDNA away from the bulky complex comprising the affinity agent and target.

CAUTION: Sodium azide is highly toxic. Wear personal protective equipment/face protection, including safety glasses. Weighing/handling of solid sodium azide powder should be done on an enclosed balance inside a certified chemical fume hood, using a stainless-steel spatula (sodium azide reacts with metals to produce explosive metal azides). Avoid dust formation and ensure adequate ventilation, as handling of the powder can generate static electricity, spark and catch fire. Avoid ingestion and inhalation.

Materials:

Docking strand oligo with a 5’ amino modifier with C6 linker, and a 3’ Cy3 fluorophore; HPLC purified (Integrated DNA Technologies)

Invitrogen UltraPure DNase/RNase-Free Distilled water (Fisher Scientific, 10977023)

Sodium bicarbonate (Powder/Certified ACS) (Fisher Scientific, M-14636)

DBCO-PEG4-NHS ester (Click Chemistry Tools, A134)

DMSO (Sigma-Aldrich, 900645–4×2ML)

Ethanol 200 proof (100 %) (Fisher Scientific, 04355223)

Sodium Acetate (3 M), pH 5.2 (Quality Biological, 351-035-721)

AffiniPure Donkey anti-Rabbit IgG (H+L) (Jackson Immuno Research, 711-005-152)

AffiniPure Donkey anti-Mouse IgG (H+L) (Jackson Immuno Research, 715-005-150)

Gibco Dulbecco’s Phosphate-Buffered Saline (DPBS) no calcium, no magnesium (Fisher Scientific, 14190–144)

Azido-PEG4-NHS ester (Click Chemistry Tools, AZ103)

TWEEN 20 (Sigma-Aldrich, P7949)

Sodium azide (Sigma-Aldrich, S2002)

CF660R-succinimidyl ester (Biotium, 92137)

Complementary imaging strand oligo with a 3’ amino modifier; HPLC purified (Integrated DNA Technologies)

−80 °C freezer (or −20 °C) for DNA precipitation and long-term storage of reagents such as antibody-DS conjugates

Tabletop centrifuge operating at 4 °C that can reach 20,000 × g, such as the Eppendorf 5424R

UV-Vis spectrophotometer such as Nanodrop 2000 (ThermoFisher) for measuring concentrations of DNA, protein, and fluorophores

LoBind Microcentrifuge Tubes: Protein (Eppendorff, 022431081, or equivalent)

50 kDa Millipore Sigma Amicon Ultra Centrifugal Filter Units (referred to as 50 kDa spin filter, Fisher Scientific, UFC505096)

100 kDa Millipore Sigma Amicon Ultra Centrifugal Filter Units (referred to as 100 kDa spin filter, Fisher Scientific, UFC510024)

Eppendorf Research plus pipettes, variable volumes (or equivalent)

Preparation of docking strands tagged with DBCO (DS-PEG4-DBCO)

• 1Suspend commercially obtained DS oligos in ultrapure water to obtain a 1 mM concentration.
• 2Combine 20 μL of DS oligo (1 mM), 38 μL of ultrapure water, and 6.5 μL of freshly made 1 M sodium bicarbonate (NaHCO3) in a 1.5 mL tube. Vortex, then add 2.5 μL of DBCO-PEG4-NHS ester linker (200 mM stock in DMSO). Vortex and run reaction for 3 hours at room temperature in the dark with gentle shaking.
• 3Add 500 μL of 100 % ethanol (EtOH) and 50 μL of 3 M sodium acetate. Vortex, then leave at −80 °C overnight.
• 4Next day, cool down the centrifuge to 4 °C.
• 5Retrieve the reaction mixture from the-80 °C freezer and immediately spin it for 30 min at 15000 rpm at 4 °C.
• 6Remove supernatant, add 500 μL 100 % EtOH, vortex, and spin for 5 min at 15000 rpm at 4 °C. Repeat this washing step 3 times.
• 7After the last wash, remove supernatant and add 50 μL of ultrapure water, 500 μL 100 % EtOH and 50 μL of 3 M sodium acetate. Vortex, then leave at −80 °C overnight.
• 8Repeat DNA purification once more as described in steps 4 through 6.
• 9After the last wash, resuspend the pellet in 40 uL of ultrapure water.
• 10Measure concentration of the final product, DS-PEG4-DBCO, with a UV-Vis spectrophotometer and record ratio of absorbance values at 260 nm (for DNA) and 550 nm (for Cy3^™).

Preparation of secondary antibody tagged with azide

• 11Mix 110 μL of 1.3 mg/mL secondary antibody (either anti-rabbit or anti-mouse) with 18 μL of freshly made 1 M NaHCO3 and 10 μL DPBS. Use a 1.5 mL low protein binding tube.
• 12Add azido-PEG4-NHS ester in approximately 120 × molar excess (1 μL if using 100 mM stock in DMSO).
• 13Vortex and run reaction for 3 hours at room temperature with gentle shaking.
• 14Passivate the 50 kDa spin filter with 2 % TWEEN 20 in ultrapure water for 2 h.
• 15Cool down the centrifuge to 4 °C.
• 16Aspirate TWEEN 20 solution and rinse the spin filter with 0.5 mL ultrapure water twice.
• 17Add 0.5 mL ultrapure water and centrifuge the spin filter at 4°C, 7000 × g, 2.5 minutes.
• 18Repeat two times the wash from step 17 using DPBS.
• 19Add antibody-PEG4-azide conjugates to the spin filter and wash with DPBS 12 times via centrifugation. Centrifuge at 4°C, 7000 × g, 2.5 minutes each cycle.
• 20After the final wash, there should be at least 90 μL of antibody-PEG4-azide remaining.
• 21Measure antibody-PEG4-azide concentration with a UV-Vis spectrophotometer. Use either in-built mode (e.g. “Protein A280” in NanoDrop) or record absorbance value at 280 nm and use it together with the antibody extinction coefficient and molecular weight to calculate antibody-PEG4-azide concentration using the Beer-Lambert law.

Preparation of secondary antibodies conjugated to docking strands via azide-DBCO click chemistry

• 22Add DS-PEG4-DBCO in 5× molar excess to the antibody-PEG4-azide conjugate. Vortex and run reaction overnight at room temperature in dark with gentle shaking.

  Run reaction in the dark to prevent Cy3 fluorophores of DS from photobleaching.

• 23Prepare a 100 kDa spin filter as previously described in steps 14 through 18.
• 24Add sodium azide to a final concentration of 0.15 % to quench unreacted DBCO (e.g. add 5 μL of 3 % sodium azide per 100 μL reaction mixture). Keep the tube at room temperature with gentle shaking in the dark for one hour.
• 25Add antibody-PEG8-DS conjugates to a passivated 100 kDa spin filter and wash with DPBS five to six times via centrifugation, until the flow through is colorless. Centrifuge at 4 °C, 7000 × g, 2.5 minutes each spin. The final product (antibody-PEG8-DS) should be suspended in ~90 μL DPBS.
• 26Measure antibody concentration with a UV-Vis spectrophotometer. Calculate degree of labeling using peak absorbance values at 280 nm (for protein) and 550 nm (for Cy3^™), correcting for DNA absorbance contribution at 260 nm.

  DS-conjugated secondary antibodies can be stored long term at −20 °C in 40 % glycerol (v/v). The probes will be stable for multiple months. Working stocks can be kept at 4 °C for several weeks. Fluorescent probes should be kept in the dark at all times.

Imaging strand conjugation

• 27Make a 10 mg/mL stock of CF660R succinimidyl ester in DMSO; aliquot and store at −80 °C.

  In our published work we conjugated imaging strands to CF660R dye. In recent experiments we also used ATTO643 dye for conjugation (i.e. ATTO643 NHS-Ester from Atto-Tec, AD 643–31) for DNA-PAINT-ERS imaging experiments. Both dyes show high single-molecule brightness.

• 28Add 5 μL of IS oligo (1 mM) to 40.5 μL of freshly made 0.1 M NaHCO3. Vortex, then add 4.5 μL of dye (10 mg/mL). Vortex and run reaction for 3 hours at room temperature with gentle shaking.

  Keep the tube in the dark to prevent photobleaching.

• 29Add 500 μL of 100 % ethanol (EtOH) and 50 μL of 3 M sodium acetate. Vortex, then leave at −80 °C overnight.
• 30Next day, cool down the centrifuge to 4 °C.
• 31Spin the reaction mixture for 30 min at 15000 rpm at 4 °C.
• 32Remove supernatant, add 500 μL 100 % EtOH, vortex, and spin for 5 min at 15000 rpm at 4 °C. Repeat this washing step 3 times.
• 33After the last wash, remove supernatant and add 50 μL of ultrapure water, 500 μL 100 % EtOH and 50 μL of 3 M sodium acetate. Vortex, then leave at −80 °C overnight.
• 34Repeat DNA centrifugation and purification once more as described in steps 30 through 32.
• 35After the last wash, remove the supernatant and dry the pellet. Add 20 μL of ultrapure water to the final product.
• 36Measure IS concentration with a UV-Vis spectrophotometer. Calculate degree of labeling using 260 nm (for DNA) and 660 nm (for dye) absorbance values. Ideally, the molar ratio of dye to DNA is close to 1. Values substantially higher than 1 indicate insufficient removal of free dye.

  Dye-conjugated IS can be stored at −20 °C long-term where it will be stable for multiple months. Highly concentrated (>5 μM) working stocks can be kept at 4 °C for several weeks. Diluted working solutions (100 nM) for imaging experiments can be kept at 4 °C for multiple days. However, IS will stick to the tubes and actual concentration will decrease over time. Preparing fresh solutions weekly is important to ensure reproducible results.

  Probes should be kept in the dark at all times.

BASIC PROTOCOL 2

Sample preparation for imaging membrane targets with DNA-PAINT-ERS in fixed cells

Sample preparation for DNA-PAINT-ERS on cultivated and fixed cells is similar to a standard immunofluorescence workflow (Piña et al., 2022), but several important modifications are necessary. We add sheared Salmon Sperm DNA to block non-specific DS hybridization and to prevent attachment of probe to membrane surfaces. Image-iT FX Signal Enhancer is applied to the sample prior to staining to help reduce non-specific fluorescence background signals associated with the application of affinity agents with negatively charged modifications. Samples are stained with commercially available primary antibodies followed by staining with the previously generated DS-labeled secondary antibodies. Post-fixation using a combination of paraformaldehyde with glutaraldehyde is performed to best preserve the sample in high-salt imaging buffer and during storage. This protocol describes procedures used for adherent U2OS cells in culture, and it can be extended to other cell lines or sample formats (e.g. tissue sections) with proper optimizations.

CAUTION: Paraformaldehyde is a toxic chemical, suspected of causing cancer. Wear personal protective equipment/face protection, including safety glasses. Avoid dust formation and ensure adequate ventilation. Weighing/handling of solid paraformaldehyde powder should be done on an enclosed balance inside a certified chemical fume hood. Avoid ingestion and inhalation.

CAUTION: Glutaraldehyde is a toxic chemical. Wear personal protective equipment/face protection, including safety glasses. Avoid generation of vapors/aerosols and ensure adequate ventilation. Handling of glutaraldehyde should be done inside a certified chemical fume hood. Avoid ingestion and inhalation.

CAUTION: Sodium azide is highly toxic. Wear personal protective equipment/face protection, including safety glasses. Weighing/handling of solid sodium azide powder should be done on an enclosed balance inside a certified chemical fume hood, using a stainless-steel spatula (sodium azide reacts with metals to produce explosive metal azides). Avoid dust formation and ensure adequate ventilation, as handling of the powder can generate static electricity, spark and catch fire. Avoid ingestion and inhalation.

Materials:

U2OS cells (ATCC, HTB-96)

Gibco DMEM, high glucose, pyruvate (ThermoFisher, 11995073)

Gibco Fetal Bovine Serum, qualified, United States (FBS, Fisher Scientific, 26140079)

Sodium hydroxide reagent grade, ≥98 %, pellets (anhydrous) (NaOH, Sigma-Aldrich, S5881)

Gibco Dulbecco’s Phosphate-Buffered Saline, no calcium, no magnesium (DPBS, Fisher Scientific, 14190–144)

Gibco Trypsin-EDTA (0.25 %), phenol red (Thermo Fisher, 25200056)

Gibco Trypsin-EDTA (0.05 %), phenol red (Thermo Fisher, 25300054)

3.7 % Paraformaldehyde (PFA, see Reagents and Solutions)

Glycine ReagentPlus, ≥99 % (HPLC) (Sigma-Aldrich, G7126)

Saponin for molecular biology (Sigma-Aldrich, 47036)

Image-iT FX Signal Enhancer (Fisher Scientific, I36933)

Bovine serum albumin (BSA, Fisher Scientific, BP1600)

Salmon Sperm DNA, sheared (10 mg/mL) (Fisher Scientific, AM9680)

Anti-caveolin-1 antibody (abcam, ab2910)

Anti-clathrin heavy chain antibody (abcam, ab21679)

Donkey anti-mouse or anti-rabbit secondary antibody conjugated to DS (see Basic Protocol 1)

Sodium azide (Sigma-Aldrich, S2002)

Glutaraldehyde solution, Grade II, 25 % in H2O (Millipore Sigma, G6257)

50 nm gold colloid particles (BBI Solutions, EM.GC50/4)

Gibco Dulbecco’s Phosphate Buffered Saline, calcium, magnesium (DPBS+, Fisher Scientific, 14040182)

Potassium hydroxide, ACS reagent, ≥85 %, pellets (KOH, Sigma, 221473)

Titan3 Nylon Syringe Filters, 17mm, 0.22 μm pore size (Thermo Scientific, 42213-NN)

10 mL disposable Syringes Luer Lock (Air-Tite Products Co., ML10)

Corning TC-Treated Culture Dishes 20 × 100 mm (Fisher Scientific, 08-772-22)

Nunc Lab-Tek II 8-well Chambered Coverglass (ThermoFisher Scientific, 155409)

Eppendorf Safe-Lock microcentrifuge tubes, volume 1.5 mL, natural (Sigma-Aldrich, T9661)

Tissue culture microscope (Zeiss Axio Observer A1 Inverted Microscope or equivalent)

Class II, Type A2 Biosafety Cabinet (Thermo Fisher Scientific 1300 Series or equivalent)

Pipette Controller (PIPET-PAL Pipette Controller from WHEATON DWK Life Sciences or equivalent)

Corning Falcon Disposable Polystyrene Serological pipettes 5 mL and 10 mL (Fisher Scientific, 13-675-22 and 13-675-20)

Eppendorf Research plus pipettes, variable volumes (or equivalent)

Precision balance (Mettler Toledo NewClassic MS or equivalent)

Tissue culture and cell plating

• 1Prepare all solutions used in this protocol freshly (at least once per week) and filter them with a 0.22 μm syringe filter.
• 2Maintain U2OS cells in DMEM supplemented with 10 % FBS at 37 °C and 5 % CO2 in round culture dishes (⌀ 100 mm) and passage every three to four days when cells reach confluency.

  For super resolution imaging, it is recommended to use cells below passage number 20.

• 3Clean chambered coverglass the day before seeding cells by pipetting 100 μL 1 M NaOH into each well and incubate for 1 h.

  This step is critical for all super resolution imaging experiments to remove contaminants that could affect imaging from the coverglass.

• 4Wash coverglass twice by pipetting 100 μL DMEM into each well.
• 5Store cleaned chambered coverglass with 250 μL DMEM in the incubator at 37 °C overnight.

  It is important to clean the coverglass the day before seeding the cells and soak it in a buffered solution so that all remaining NaOH is neutralized completely as it is very harmful for the cells.

• 6Next day, passage the cells. Rinse cells with 5 mL warm DPBS followed by 1 mL 0.25 % trypsin for 30 s and then incubate the cells with 1 mL 0.05 % trypsin at 37 °C until all cells detach from the culture dish.

  Cells are only briefly rinsed with a solution containing high trypsin concentration to detach them from the culture dish. The cells are then incubated in a lower trypsin concentration solution so they do not round up too much before performing imaging experiments.

• 7Add 3 mL warm DMEM supplemented with 10 % FBS to the trypsinized cells.
• 8Plate the cells (approximately 25–30 uL of cell suspension mixed with 275 uL DMEM with 10 % FBS per well) into 8-well chambered coverglass and let them grow overnight until they reach 50–60 % confluency on the day of the experiment.

  Check the cells under a light microscope to make sure they are evenly distributed across each well. Gentle pipetting while plating the cells might help to distribute cells more evenly.

Immunostaining of unextracted U2OS cells with clathrin or caveolin

• 9For immunostaining of clathrin or caveolin wash cells with warm DPBS and apply warm 3.7 % paraformaldehyde (PFA) for 30 min for fixation at room temperature (without rocking).
• 10Wash the sample 3 times for 5 min with DPBS whilst gently rocking.

  Following washing steps rigorously is important to achieve the best staining results.

• 11Simultaneously quench and permeabilize the cells with 300 mM glycine and 0.5 % saponin in DPBS for 40 minutes whilst gently rocking.

  Using a mild detergent for permeabilization is vital to preserve morphology of membrane targets.

• 12Wash the sample 3 times for 5 min with DPBS whilst gently rocking.
• 13Incubate the sample with Image-iT FX Signal Enhancer for 40 min in the dark without rocking.

  The signal enhancer reduces binding of negatively charged modifications commonly found in dyes that can lead to nonspecific association with cells. Use of signal enhancer reduces nonspecific fluorescence signal (background).

• 14Wash the sample 3 times for 5 min with DPBS whilst gently rocking.
• 15Block the sample with 3 % BSA in DPBS supplemented with 5 % salmon sperm DNA for 40 min on a rocker.

  Salmon sperm DNA will prevent background fluorescence of the sample more effectively.

• 16Incubate the sample with 1 % BSA, 5 % salmon sperm DNA and 0.005 mg/mL primary antibody (e.g. anti-caveolin or anti-clathrin) in DPBS on a rocker at room temperature for 2 h.
• 17Wash the sample 3 times for 5 min with DPBS whilst gently rocking.
• 18Whilst performing the staining (step 16) mix 1 % BSA, 5 % salmon sperm DNA, ~0.008 mg/mL DS-conjugated secondary antibody and 0.15 % sodium azide. Preincubate the mixture on a rocker at room temperature for 30 min in dark.

  The DBCO-azide reaction has been quenched with supplemental (sodium) azide during preparation of the secondary antibodies. However, we have observed that repeating this step during the staining procedure can lead to improved results and therefore highly recommend this additional step.

• 19Incubate the sample with DS-conjugated secondary antibody in DPBS buffer containing 1 % BSA and 5 % salmon sperm DNA prepared in step 18 on a rocker at room temperature for 2 h.

  Keep sample in the dark from now on.

• 20Wash the sample 3 times for 5 min with DPBS whilst gently rocking.

Post-fixation and final preparations for imaging

• 21For all cell staining procedures, post-fix the cells for 15 minutes with 3.7 % PFA and 0.2 % glutaraldehyde (GA) (without rocking).

  This is essential to preserve the sample from degradation by imaging buffers due to high salt content. Depending on the target some samples are stable at 4 °C for multiple days following post-fixation. If samples are stored for extended periods of time addition of sodium azide (0.01 %) is highly recommended to reduce microbial growth.

• 22Wash the sample 3 times for 5 min with DPBS whilst gently rocking.
• 23Before imaging, add 150 μL of 2.5 % 50 nm gold particles in DPBS+ to cells for 10 minutes. Wash with DPBS.

  Gold fiducials can also be added to the sample whilst it is mounted on the microscope stage before imaging. This helps to lay down the right amount of gold nanoparticles into spaces between cells.

ALTERNATE PROTOCOL 1

Immunostaining of extracted U2OS cells

Basic protocol 2 describes the steps needed for the preparation of unextracted U2OS cell samples for DNA-PAINT-ERS imaging. However, some targets benefit from extraction of cells prior to immunostaining (Piña et al., 2022). In this protocol the cells are pre-permeabilized with Triton X-100 before fixation. Furthermore, fixation is carried out with the addition of glutaraldehyde which is a stronger cross-linking agent. This results in better preservation of the cytoskeleton and is highly beneficial when staining microtubules with anti-beta tubulin antibodies. Additional quenching steps will be performed after applying the stronger cross-linking fixation agent to ensure appropriate antibody binding. This protocol can also be used for staining of clathrin but may not be suitable to sensitive targets such as caveolae.

Additional Materials to Basic Protocol #2:

Triton X-100 (Sigma, X100)

Sodium borohydride (Sigma, 452882)

Beta tubulin monoclonal antibody (ThermoFisher Scientific, 32–2600)

Immunostaining of extracted U2OS cells with beta tubulin

1. Maintain and plate U2OS cells as described in Basic Protocol #2, Section 1.

2. Wash cells with warm DPBS once.

3. For immunostaining of extracted samples (microtubules alone, or microtubules co-labeled with clathrin), pre-permeabilize cells with ice-cold 0.1 % Triton X-100 in 1x PHEM buffer and DPBS for 45 s.

   Incubation time might be varied between 30 s and 60 s, depending on the density of cells (higher confluence requires slightly longer pre-permeabilization time, while more dispersed cells are easier accessible and thus should not be incubated for longer than 30 s).

   It is essential to keep the cells cold (at 4 °C) at this stage and treat them very gently. Permeabilization of cell walls is performed whilst the cells are alive and mechanical disturbances result in increased cell lysis and consequently death.

4. Subsequently fix cells with ice-cold 3.7 % PFA and 0.1 % glutaraldehyde (GA) for 20 minutes without rocking.

   Perform this step quickly but without disturbing the fragile cells. Pipette and aspirate buffers slowly from one side of the well only.

5. Gently wash the sample 3 times for 5 min with DPBS whilst rocking slowly.

6. Quench samples with fresh 0.1 % sodium borohydride in DPBS for 7 min whilst gently rocking.

7. Gently wash the sample 3 times for 5 min with DPBS whilst rocking slowly.

8. Perform a second quenching and permeabilization step by incubating the sample with 300 mM glycine with 0.2 % Triton X-100 in DPBS for 40 min whilst gently rocking.

9. Gently wash the sample 3 times for 5 min with DPBS whilst rocking slowly.

10. Incubate the sample with Image-iT FX Signal Enhancer for 40 min in the dark without rocking.

11. Gently wash the sample 3 times for 5 min with DPBS whilst rocking slowly.

12. Block the sample with 3 % BSA in DPBS supplemented with 5 % salmon sperm DNA for 40 min on a rocker.

13. Incubate the sample with anti-beta tubulin antibody and/or anti-clathrin heavy chain antibody (0.005 mg/mL or 1:200 dilution from stock) in DPBS buffer containing 3 % BSA and 5 % salmon sperm DNA on a rocker at room temperature for 2 h.

14. Gently wash the sample 3 times for 5 min with DPBS whilst rocking slowly.

15. Incubate the sample with DS-conjugated secondary antibody at a final concentration of ~0.008 mg/mL in DPBS buffer containing 1 % BSA and 5 % salmon sperm DNA. Let incubation take place on a rocker at room temperature for 2 h in dark.

    Keep sample in the dark from now on.

16. Wash the sample 3 times for 5 min with DPBS whilst gently rocking.

17. Post-fix for 15 minutes with 3.7 % PFA and 0.2 % GA (without rocking).

18. Wash the sample 3 times for 5 min with DPBS whilst gently rocking.

BASIC PROTOCOL 3.

Super resolution image acquisition and analysis

This protocol describes the steps for performing DNA-PAINT-ERS imaging on the samples prepared above. The protocol is developed for using a custom SRM setup as detailed below, and the user should adapt this protocol according to the configuration of the exact microscope system.

The SRM setup is built on a Nikon Ti-U frame. Two laser beams emitting at 561 nm and 639 nm are combined and introduced into the back of the microscope equipped with a 60× TIRF objective. A focusing lens (f=400 mm) is placed at the back port of the microscope to focus the collimated laser light to the back aperture of the objective to achieve through-objective total internal reflection (TIR) illumination (Fish, 2009). The excitation light can be continuously tuned between epi-fluorescence and strict TIR modes by shifting the incident laser horizontally with a translational stage before entering the back port of the microscope. A custom focus stabilizing system based on detection of the reflected excitation laser stabilizes the focus during data acquisition. A multi-edge polychroic mirror is used to reflect the lasers into the objective and clean up fluorescence signals from the sample. Emission filters are placed in front of the camera to select specific (Cy3, CF660R, etc.) fluorescence signals. Fluorescence signals are collected through the objective by an electron-multiplied charge-coupled device using a typical EM gain setting of 200–300 in frame transfer mode. In this configuration, the 561 nm channel records signals from the Cy3-labeled DS and is routinely used to view the sample and find regions of interest, and the 639 nm channel is used for actual DNA-PAINT-ERS imaging (Figure 4).

Figure 4. Raw images of microtubules in the green and red detection channels.

(A) Green detection channel: COS7 cell microtubules labeled with a secondary antibody conjugated to a DS labeled with Cy3. (B) Red detection channel: DNA-PAINT-ERS single-molecule ‘blinking’ events of IS labeled with Atto643 binding to the microtubules shown in (A). Scale bars: 10 μm

DNA-PAINT-ERS achieves fast SRM imaging by optimizing DS-IS hybridization and dehybridization kinetics. The ‘Repeat sequence’ and ‘Spacer’ elements speed up hybridization and are incorporated in Basic Protocol 1. Accelerating dehybridization relies on the addition of EC to the DNA-PAINT imaging buffer. Typical EC concentrations between 5–15 % (v/v) increase the dehybridization rate by up to 10x (Civitci et al., 2020).

Here, we also describe the procedure for using our in-house software packages to process the acquired raw image frames, post-process the resulting coordinates, and render a super-resolved image (Figure 5A). First, single-molecule localizations need to be detected by fitting a Gaussian point-spread function, extracted and saved to a coordinates file using the Matlab package (Nan, 2020) wfiread (Figure 5B). Then, the coordinate file is loaded into another Matlab script (Nan, 2020) palm (Figure 5C), where drift correction and sorting of the localizations is applied. Finally, the coordinates and calculated precision of the selected localizations are used to render a super-resolved image (see examples in Figures 6 and 7B). The use of these software packages was previously documented (Nickerson et al., 2015), and both packages have now been updated to handle DNA-PAINT images. Of note, many other software packages exist for processing single-molecule SRM data, for example SMAP (Ries, 2020), picasso (Schnitzbauer et al., 2017), the ImageJ plug-in ThunderSTORM (Ovesný et al., 2014), and DAOSTORM (Holden et al., 2011).

Figure 5. Super resolution data analysis.

(A) Flow chart depicting data processing steps. (B) Detection and extraction of single-molecule localizations is done using the in-house Matlab script wfiread. (C) Drift correction, filtering, sorting of localizations and finally rendering is done using the in-house Matlab script palm. Data processed corresponds to the images in Figure 4.

Figure 6. Example of super-resolved microtubules in COS7 cells.

(A) Microtubules were imaged with 30,000 frames at 40 ms exposure time. The images correspond to the raw data in Figure 4B. (B) Only 15,000 frames from the data in (A) were processed and rendered, showing that a 10 min acquisition is already sufficient to achieve high quality super resolution images. Scale bars: 10 μm (A, left), 2 μm (A, right top panel, and B, top panel), 500 nm (A, right bottom panel and B, bottom panel)

Figure 7. Two-color imaging of clathrin and microtubules with DNA-PAINT-ERS.

(A) Primary clathrin antibodies (1° ab, light green) are applied to the sample and bound to secondary antibodies (2° ab, dark green) conjugated to DS1 (dark green) and sampled with IS1 bound to a fluorophore (yellow star) during the first imaging cycle. Washing with 15 – 20 % EC removes the imaging buffer containing IS1. Subsequently, imaging buffer containing IS2 is added to the sample to bind to DS2 conjugated to 2° ab (dark blue) that recognizes 1° ab (light blue) targeted to microtubules. (B) Two-color image showing co-labeled microtubules (magenta) and clathrin (green) in U2OS cells obtained using two different DS-IS pairs. Microtubules were imaged with 30,000 frames and 20 ms exposure time, while clathrin was imaged with 30,000 frames and 25 ms exposure time. Image reproduced with permission from (Civitci et al., 2020). Scale bars: 2 μm (left) and 500 nm (right)

Materials:

Sample with a target of interest immunolabeled following Basic Protocol 2 or Alternate Protocol 1

2.5 % (v/v) 50 nm gold nanoparticles (see Reagents and Solutions)

Gibco Dulbecco’s Phosphate Buffered Saline 1X (DPBS, Fisher Scientific, cat. no. 14190–144)

Gibco Dulbecco’s Phosphate Buffered Saline, calcium, magnesium (DPBS+, Fisher Scientific, 14040182)

Buffer C (see Reagents and Solutions)

25 % (v/v) ethylene carbonate (EC, see Reagents and Solutions)

Dye-labeled imaging strand (IS-CF660R, IS-Atto643, etc., prepared following Basic Protocol 1)

Laser emitting at 561 nm (Coherent, Sapphire 561 LP, 200 mW)

Laser emitting at 639 nm (Opto Engine LLC, MRL-FN-639–500mW)

60× TIRF objective (Nikon, TIRF 60XC Oil, oil immersion, NA 1.49)

Focusing lens (Thorlabs, LA1725-A, N-BK7 Plano-Convex Lens, Ø2”, f = 400 mm, AR Coating: 350 – 700 nm)

Translational stage used to tune the excitation light between epi-fluorescence and strict TIR modes (e.g. Thorlabs, M30X or similar)

Custom focus-stabilizing system based on detection of the reflected excitation laser (ASI, CRISP Autofocus system)

Multi-edge polychroic mirror (Semrock, BrightLine^® quad-edge laser dichroic beamsplitter Di01-R405/488/561/635)

Emission filter for the 561 nm excitation (Semrock, BrightLine^® single-band bandpass filter FF01–605/64)

Emission filter for the 647 nm excitation (Semrock, BrightLine^® single-band bandpass filter FF01–708/75)

Electron-multiplied charge-coupled device (EM-CCD, Andor, iXon Ultra 897)

Immersion oil Type HF (Cargille Laboratories, cat. no. 16245)

Lens cleaning paper (Tiffen, Type 1, Class 1 product Specification AA-50177)

(optional) Gel loading tips (Thermo Scientific, ART Gel Loading Pipette Tips, cat. num. 2155P)

Data acquisition

• 1Bring the immunolabeled sample prepared according to Basic Protocol 2 or Alternate Protocol 1 to room temperature if it has been stored at 4 °C.

  Temperature variations due to thermal equilibration introduce strong sample drift during imaging.

• 2If gold fiducials have not been added as part of the immunostaining protocol, add 150 μL of 2.5 % (v/v) 50 nm gold nanoparticles in DPBS+ to each well of the sample and incubate for 10 minutes. Wash with 250 μL DPBS.

  Gold nanoparticles serve as fiducials that enable drift correction during data processing. This step can be fine-tuned by applying the gold fiducials after mounting the sample on the microscope. Once the 639 nm laser is turned on the gold nanoparticles can be seen as they settle onto the glass. Incubation time can be adjusted to achieve an ideal density of gold fiducials.

• 3Add imaging buffer with desired ratios of Buffer C, 25 % (v/v) ethylene carbonate (EC) (usually between 10 % and 15 %), and IS (at 1–2 nM final concentration).

  The exact concentrations of EC and IS may need to be adjusted depending on the target and based on the imaging kinetics, which depend on IS-DS complementary sequence as well as the ambient temperature. We recommend to equilibrate imaging buffer to room temperature as cold imaging buffer increases sample drift.

• 4Turn on the microscope, lasers and EMCCD camera. Select the 60× Apochromat TIRF objective with 1.49 NA and mount the sample by applying index matching oil between the objective and sample.

  Make sure the sample is mounted perfectly flat and that there are no bubbles in the immersion oil. If sample illumination appears uneven remove the sample and clean the bottom of the coverglass with isopropanol. Clean the objective with lens cleaning paper and remount the sample with fresh immersion oil.

• 5Initialize software that controls the camera and enables data acquisition. In our custom setup μManager (Edelstein et al., 2014) software is used.
• 6Open the laser shutter and bring the sample to focus.
• 7Adjust TIRF angle by shifting the incident laser beam horizontally with a translational stage before entering the back port of the microscope such that desired penetration depth is reached.

  Structures such as the microtubules are probed adequately with moderately relaxed TIRF angle, whereas imaging membrane targets such as clathrin and caveolae benefits from stricter TIRF angles.

• 8Initially set the exposure time (20–50 ms), laser power (~500 W/cm²), EMCCD gain (200–300), and adjust these parameters as necessary depending on the single-molecule brightness and localization kinetics.

  Strive to achieve signal-to-noise ratio of five or more.

  Adjust IS concentration such that the overall number of localizations is high while no overlapping binding events are observed. If necessary, adjust EC concentration to match expected kinetics.

• 9If available, use Cy3 signal to find regions of interest to be imaged.

  Make sure there are at least five gold nanoparticles in the field of view if a drift-correction procedure based on stable fiducials (as opposed to cross-correlation analysis, which does not require fiducials to be present in an image) is going to be applied as a post-processing step.

• 10Engage the auto-focus system to correct in real time for sample drift in the z-direction and start the acquisition.

  The number of frames that needs to be collected depends on each target specifically. It helps to start with acquisitions between 20,000 – 50,000 frames and reconstruct the images to see if the target structure is well resolved. We typically acquire 25,000 – 30,000 frames for microtubules (see Figure 6 for an example), clathrin, and caveolae, and up to 800,000 frames for dense targets such as the actin cytoskeleton.

• 11To perform data acquisition for another target, wash the sample with 15 – 20 % EC in DPBS after completing acquisition of data for the first target (Figure 7A).

  Perform washes and changes of solutions gently to not move the coverglass on the microscope stage or disturb the sample or gold fiducials. To minimize the chances of accidentally hitting the coverglass, gel-loading tips can be used. Perform 2–3 washes until all IS has been removed and no blinking events are observed anymore.

• 12Add imaging buffer containing a different IS for an alternative target (Figure 7A) and optimize EC and IS concentration before restarting image acquisition detailed in steps 6–10.

  To store the sample in the fridge following image acquisition remove imaging buffer and add DPBS. The lower salt concentration in DPBS will help to preserve the sample better over time.

Data analysis

• 13Process the data using in-house Matlab scripts wfiread and palm, available at Github (Nan, 2020), following the steps described below.
• 14Open raw data (see Figure 5B) in wfiread.

  If your data has more than 50,000 number of frames, use “Open Large Video” mode. In this mode, only every 100th frame is loaded into wfiread to reduce the memory usage during image viewing, however all frames are processed.

• 15Vary “Contrast factor”, “Threshold”, and “PSF size” such that single (non-overlapping) localizations are detected. Use “Find particles” button to preview the selection results and test the parameters.

  “Sigma range” defines the fitted widths of the point spread functions. For most datasets, leave at default values 0.3 – 3.

  “Contrast factor” is defined as the contrast between signals at the center and all the peripheral pixels. Higher values help remove spurious localizations (e.g. those resulting from diffusing probes). The default value is 1.5, and the typical working range is 1.2 – 1.7.

  “Threshold” is defined as minimum intensity of a localization over the surrounding background to be considered a real event. The default value is 5 times RMS, and the typical working range is 4–6 times RMS.

  “PSF size” is defined as the (set-up specific) number of pixels on each side of the center pixel that comprise the fitting box for each candidate localization event. The default value is 2 pixels, corresponding to 5 × 5 pixels fitting box. Typical working range is 2–4 pixels. If the data was acquired in a 3D astigmatic mode using cylindrical lens in the detection path, the value of PSF size might need to be increased to correctly represent the ellipsoid shape of astigmatic PSF.

• 16Extract and save coordinates of the selected localizations (and their intensities) by pressing “Make Coord File”. By default, the data is saved as a .cor file.
• 17Open .cor file in palm (see Figure 5C).
• 18Apply “Drift correction”.

  Confirm that fiducials pre-selected by the software show comparable and relatively small drift.

• 19Set sorting parameters that best represent your data. In our typical DNA-PAINT-ERS experiments this would be:

  “Combine Frames” combines localizations present in N consecutive frames and assigns average coordinates (typically used value 2–3).

  “Combining Distance (nm)” combines localizations that are less than x nm apart and also assigns average coordinates (typically used values 80–100 nm).

  “Minimum RMS” defines the minimal ratio between fitting amplitude and the standard deviation of residue noise after the fitting (typically used value ≥ 5).

  “Minimum Fit Goodness” selects localizations that can be well fitted to a theoretical point-spread function (typically used value 0.15, the lower the value, the more stringent data filtering).

  “Max. Eccentricity” selects localizations that have similar width of point-spread function in both, X and Y, directions (typically set to 1.4).

• 20Filter and sort the localizations according to the parameters selected in step 7.
• 21Set rendering parameters (“Raw Pixel Size” is specific to experimental setup used to acquire the data).

  Small punctate structures such as caveolae and clathrin benefit from being rendered at a lower feature size (i.e., 10–20 nm). Continuous, filamentous structures such as microtubules benefit from being rendered at higher feature size (20–40 nm). It is important to adjust parameters of image rendering for each target independently to obtain the best results.

REAGENTS AND SOLUTIONS:

PHEM buffer (2X, pH 7)

Mix:

• 0.060 M PIPES (Sigma-Aldrich, P6757)

• 0.025 M HEPES (Fine White Crystals/Molecular Biology) (Fisher Scientific, BP310–500)

• 0.010 M EGTA (Fisher Scientific, O2783–100)

• 0.008 M Magnesium Sulfate Anhydrous (MgSO4 Powder/Certified (Fisher Scientific, M65–500)

• distilled water

Mix all reagents thoroughly until dissolved before adjusting the pH with 10 M potassium hydroxide. Store buffer at 4 °C for up to 12 months.

PFA fixative in PHEM (3.7 %, pH 7.4)

Mix:

• 0.37 g of PFA (Paraformaldehyde, crystalline (Sigma, P6148)

• 30 μL 1 M NaOH (freshly prepared)

• 4.8 mL ultrapure dH20

• 5 mL PHEM buffer (2X)

Prepare fresh NaOH before making the 3.7 % PFA solution to ensure sufficient breakdown of formaldehyde chains into monomers.

First, add 1 mL distilled water to 0.37 g PFA and 30 μL 1 M NaOH in a 1.5 mL Eppendorf tube and heat up to 70 °C whilst vigorously shaking to dissolve the PFA. After the solution becomes clear combine it with 3.8 mL water and 5 mL PHEM buffer (2X) to a final volume of 10 mL. Filter the 3.7 % PFA solution with a 0.22 μm syringe filter. Store the solution at 4 °C for up to 1 week.

2.5 % (v/v) gold nanoparticles

Mix (per one well with sample):

• 5 μL Gold colloid 50 nm (BBI Solutions, cat. no. EM.GC50)

• 195 μL Dulbecco’s Phosphate Buffered Saline + calcium + magnesium 1X (DPBS+, Gibco, cat. no. 14040141)

Before pipetting, vortex gold colloid solution to assure uniform nanoparticle size. Vortex once more after mixing with DPBS+.

Use immediately, discard any leftover solution.

25 % (v/v) EC

Mix:

• 250 μL ethylene carbonate (Sigma-Aldrich, cat. no. 676802)

• 750 μL Buffer C (see Reagents and Solutions)

For convenience, prepare 250 μL aliquots of 100 % ethylene carbonate (EC) ahead of time. EC is solid at room temperature, with a melting point of 34 – 37 °C. Thus, first warm up EC at 60 °C overnight (if using a big volume bottle) and work fast when aliquoting into smaller volume tubes.

When preparing the imaging buffer, handle the aliquots with 100 % EC very gently, as it readily solidifies at room temperature upon the slightest mechanical disturbance. It can be however re-solubilized by warming up to 60 °C.

Store at room temperature for up to 12 months.

Buffer C

Mix:

• 50 mL DPBS

• 1.461 g sodium chloride (NaCl, Fisher Scientific, cat. no. BP358)

Store at room temperature for up to 12 months.

COMMENTARY:

Background Information:

DNA-PAINT is a single-molecule localization microscopy (SMLM) technique (Hess et al., 2006; Rust et al., 2006), in which the imaging speed and image quality both depend on the localization kinetics (van de Linde et al., 2010). In DNA-PAINT, localization kinetics depend on the association (‘on’) and dissociation (‘off’) rates between the DS and the IS. In its original form, the bright localization events typically last seconds or longer (‘on’ >2 s) (Stehr et al., 2019), much longer than in other SMLM (‘on’ ~0.1 s) (Dempsey et al., 2011). This in turn requires a slower ‘on’ rate of events in DNA-PAINT to reduce overlapping localizations. For SMLM, the slower probe turn-over also degrades image quality, especially in sample regions of high target density due to higher chances of overlapping localizations. Hence, the rationale for developing DNA-PAINT-ERS for faster imaging speed was to shorten the binding events and accelerate the ‘on’ and ‘off’ rates between the DS and the IS (Civitci et al., 2020).

Each of the three modifications that DNA-PAINT-ERS comprises, namely ethylene carbonate (EC), repeat sequence (R), and spacer (S), serves an important role in accelerating DS-IS kinetics (Figure 1). Originally used as a replacement for formamide in fluorescence in-situ hybridization (FISH) (Matthiesen & Hansen, 2012), EC was found to accelerate the dissociation between DS and the IS in our tests, reducing the duration of ‘on’ events by 5–15 times in the typical 5–15 % (v/v) concentration range without significantly affecting the association rate (Civitci et al., 2020). In practice, the concentrations of both the EC and the IS can be continuously tuned to achieve the desired imaging kinetics for each target or to compensate for changes in experimental conditions (such as room temperature).

Next, the use of multiple, repetitive docking sequences on the DS significantly increased the speed of IS binding to DS, to an extent greater than the number of repeats. For example, a 2x repeat sequence was able to increase IS binding ~3.5 times and a 3x repeat ~5 times (Civitci et al., 2020). This prompted us to hypothesize that the docking sequence further away from the affinity agent (such as an antibody) may be more accessible to the IS for faster binding, leading to the inclusion of a spacer between the affinity agent and the DS in our strategies. Indeed, a short spacer such as 8-unit polyethylene glycol (or PEG₈) was sufficient for a ~50 % gain in DS-IS hybridization frequency (Civitci et al., 2020).

Together, these modifications lead to 5–20x accelerated DNA-PAINT imaging depending on the exact imaging constructs and buffer composition, significantly shortening the acquisition time of single- and multi-target imaging (Civitci et al., 2020). With optimized imaging kinetics independent of the laser intensity, DNA-PAINT-ERS also improves the quality of the resulting SRM images despite inhomogeneous sample illumination (e.g. using a Gaussian beam on most single-molecule setups).

Since the modifications (E, R, and S) involve steps that are fully compatible and readily incorporated into existing DNA-PAINT workflows, DNA-PAINT-ERS can be directly used with DS-IS pairs previously validated for DNA-PAINT. To date, more than 15 orthogonal DS-IS pairs have been experimentally validated for multiplexed DNA-PAINT imaging, all 9 bps long with melting temperatures around ~5 °C in the standard buffer C and at oligo concentrations of 1–2 nM (Agasti et al., 2017). In this protocol, we chose to use DS-IS pairs that are ~10 bps by extending the 9 bp DS-IS pairs used in the original DNA-PAINT work to achieve faster ‘on’ rates (Civitci et al., 2020). For examples, DS1 and DS2 are based on the original P1 and P2 constructs (complementary sequences underlined and in upper case):

• DS1(2x):	tATACATCTAA ATACATCTAAt		paired with
  IS1:	TTAGATGTAT	(Tm ~ 12 °C)	which is derived from
  P1 (imager):	TAGATGTAT	(Tm ~ 5 °C)

• DS2(2x):	ttATCTACATAT ATCTACATAT		paired with
  IS2:	ATATGTAGAT	(Tm ~ 12 °C)	which is derived from
  P2 (imager):	TATGTAGAT	(Tm ~ 5 °C)

As can be seen from the two examples, we empirically added either an ‘A’ or a ‘T’ to the complementary sequence to obtain the desired DS-IS pairs with a target Tm around 12 °C (±1 °C) in the standard imaging buffer. We also commonly use ‘t’ or ‘a’ in non-complementary positions as ‘linker’ or ‘spacer’ sequences (e.g. between the oligo and the dye). Inevitably, there will be variations in the Tm and the on/off rates among the DS-IS pairs. Even for the same DS-IS pair, the localization kinetics will depend on the room temperature. These differences could be easily compensated by tuning the EC concentration as discussed earlier. Before experimentally validating the oligo designs, we first analyze the modified sequences to eliminate oligo dimers (both self and hetero-dimers) and hairpins for example by using the IDT oligo analysis tool at https://www.idtdna.com/calc/analyzer.

In general, we design IS and the complimentary DS sequences by following a few simple guidelines: a) long complimentary sequences correspond to longer binding times, and thus also longer acquisition times and potentially deterioration of the imaging quality; b) similarly, high GC content will increase stability of the IS-DS duplex, causing longer binding events; c) very short complimentary sequences correspond to very short binding events that might not be compatible with time resolution of a camera used; d) longer oligos (especially DS) might form secondary structures in the absence of the complementary sequence, which can negatively affect ‘on’ rate once the complementary oligo is present.

Both DNA-PAINT and DNA-PAINT-ERS use a sample preparation procedure similar to that for standard immunofluorescence (Piña et al., 2022), with some important, additional steps including the use of salmon DNA and signal enhancer to eliminate nonspecific background staining due to DNA-conjugated affinity agents, and the need for sodium azide quenching steps due to the use of reagents with the reactive DBCO group. With these precautions in mind, one can essentially image any targets that have high-quality, validated affinity agents with DNA-PAINT-ERS and benefit from both the fast imaging speed and improved image quality.

Critical Parameters:

High quality super resolution images depend on three key elements which are sensitive to user manipulation, namely sample preparation, image acquisition and image reconstruction.

First, it is critical to ensure that the immunostaining protocol is optimized for a particular target of interest (Piña et al., 2022). Some targets can benefit from cell extraction, others do not; labeling of some targets is possible only if a harsh permeabilization agent like Triton TX-100 is used, others need to be treated in a gentler manner (i.e., with saponin). We present here two protocols (Basic Protocol 2 and Alternate Protocol 1) that address this issue. In Alternate Protocol 1 it is important to treat the cells gently and avoid mechanical disturbances to preserve intracellular structures. Three important steps are recommended during reagent preparation and cell labeling. First is proper quenching of a DBCO reactive species used in the production of the DS-conjugated secondary antibody (Agard et al., 2004). Such quenching can be achieved by adding sodium azide that reacts with non-reacted DBCO during the DS-PEG8-antibody preparation, as well as during immunostaining. Second is adding salmon DNA to the solution during blocking and incubation with secondary antibody to minimize non-specific interactions with DS oligo. Finally, using signal enhancer limits non-specific interactions of cells with negatively-charged dyes. Imaging the cells at lower resolution gives a good indication if the quality of immunostaining is sufficient to perform high resolution imaging and should be done for each individual experiment.

Second, to acquire high quality DNA-PAINT-ERS data it is important to assure bright molecule localizations with minimal background fluorescence signal by optimization of binding kinetics (Jungmann et al., 2014). It is important to remember that each IS-DS pair will show different, sequence-dependent kinetics that might be fine-tuned by varying EC concentration in the imaging buffer. Next, it is crucial to know that ambient conditions can influence the IS-DS hybridization/dehybridization reactions as well. Therefore, we recommend to monitor temperature in the experimental room and use adjusted EC concentrations to compensate for day-to-day thermal fluctuations. It is important to mention here that differences in EC concentration as small as 0.5 % can evidently affect the IS-DS binding kinetics. Therefore, fine-tuning of EC concentration may be necessary following the initial, quick tests. Finally, using highly diluted IS stock solution for extended period of time might result in decreased number of binding events due to the loss of IS molecules to the tube wall. Thus, we advise to refresh IS stock solutions regularly.

Lastly, data analysis is critical for successful generation of SRM images. Choosing the right parameters is crucial for systematic estimation of the positions of molecules with sub-diffractive precision (Nan, 2020). Post-processing eliminates molecules with poor localization precision, merges reappearing molecules in subsequent frames, and corrects molecular positions for lateral drift. Image rendering provides visualization of the high-resolution image and also offers an opportunity for parameter adjustment.

Troubleshooting:

During each of the procedures explained in our Basic protocols 1–3 a variety of challenges may be encountered. Table 1 provides a summary of the most common anticipated problems and an explanation to what may cause them. Furthermore, we suggest solutions to address these challenges.

Table 1.

Troubleshooting guide for imaging of fixed and immunostained cells with DNA-PAINT-ERS

Problem	Possible Cause	Solution
Low conjugation efficiency for DS-DBCO or IS-dye	Loss of NHS-ester activity or improper pH for the reaction buffer	Use fresh reagents Confirm pH of the reaction buffer
Low recovery of DS-DBCO or IS-dye conjugates	Potential problem with DNA precipitation	Repeat DNA precipitation with DNA LoBind tubes Keep sample at −80°C degrees for longer Spin sample down at higher rpm or for prolonged time
Concentration of azide-tagged antibody is low	Problem with ultrafiltration	Passivate the spin filter with TWEEN 20 Check spin filter’s expiration date – the membranes used in spin filters occasionally fail
Low recovery of DS-conjugated antibody	Problem with ultrafiltration Failed conjugation	Check spin filter’s expiration date – the membranes used in spin filters occasionally fail Use fresh reagents Check pH of the reaction buffer
Unspecific staining	DBCO non-specifically bound to cellular targets Dirty coverglass	Quench DBCO with azide repeatedly Validate specificity of primary antibody Optimize blocking steps (% BSA, salmon DNA, signal enhancer) during low resolution imaging experiments Etch and clean coverglass
Low staining signal	Insufficient permeabilization Insufficient concentration of antibodies Too much blocking Solutions not fresh	Use triton instead of saponin (disadvantageous for membrane targets) Increase antibody concentration Block with lower % BSA Prepare fresh reagents
Nuclear staining	DBCO bound to nucleus	Add signal enhancer incubation step (see Basic protocol 2 step 13) Add salmon sperm DNA for additional blocking (see Basic protocol 2 step 15)
Uneven illumination after sample mounting	Bubbles in immersion oil Old oil smear on bottom of the coverglass from previous imaging session	Clean coverglass with isopropanol and objective lens with lens cleaning paper Remount sample with new immersion oil
Low signal to noise ratio during imaging	High IS concentration Wrong TIRF angle	Lower IS concentration to reduce background signal Adjust TIRF angle to select signal relevant to imaging plane of the target
Too slow/Fast DS/IS kinetics	Low/high room temperature EC concentration Orthogonal DS-IS pairs	Adjust EC concentration (5–15 %) Mix EC properly Monitor room temperature Check specific kinetics of DS-IS sequences Confirm the correct DS-IS pair is used
Unstable temperature in microscope room	Central climate control	Adjust EC concentration (5–15 %) Monitor room temperature and image during certain time windows
Focus drift	No focus stabilization Temperature changes during image acquisition Missing lid on the 8-well chambered coverglass	Introduce focus stabilization Equilibrate samples and imaging buffers at room temperature if previously stored in the fridge Place correct lid on top of the chambered coverglass to minimize evaporation of imaging solution
Lack of gold fiducials	Gold solution needs to be more concentrated or incubated for longer	Apply gold nanoparticles when sample is mounted on microscope, with laser illumination to assess proper amount of gold fiducials for a successful drift correction

Understanding Results:

In Basic Protocol 1, reagents for DNA-PAINT-ERS are prepared. When conjugating a dye to IS (which is short), we typically reach a 50 % recovery rate, with an expected 1:1 labeling ratio. Conjugation of DBCO-PEG4-NHS to DS (which is >20 bases long) results in a better, 80 % - 90 % recovery rate. Azido-PEG4-NHS to secondary antibody conjugation has a 60 % - 70 % recovery rate, and the final conjugation reaction between DBCO-PEG4-DS and Azido-PEG4-antibody usually yields a 60–70 % product recovery, with a typical labeling ratio of 2–3 DS oligos per antibody.

Basic Protocol 2 and Alternate Protocol 1 describe steps necessary to obtain sample with very high labeling yield and resulting in minimal background when imaging. Failure to apply crucial blocking reagents (salmon DNA, Image Enhancer) or quenching step using sodium azide will cause non-specific interactions of DS-conjugated antibody with the cells. This in turn can lead to increased fluorescence background during data acquisition and thus decreased precision of single-molecule localizations, resulting in a blurred final image. Before collecting super resolution data, we recommend to inspect the labeling quality with a standard fluorescence microscope.

When a target of interest is labeled following procedures described in Basic Protocol 2 or Alternate Protocol 1, DNA-PAINT-ERS imaging can typically be completed in a matter of minutes when using recommendations provided in Basic Protocol 3. As an example, Figure 6A shows microtubules imaged in COS7 cells in 20 minutes (30,000 frames with 40 ms exposure time, 1.5 mM IS, and 12.5 % EC), Figure 6B shows the same COS7 cell imaged in 10 minutes, and Figure 7B shows both microtubules and clathrin imaged in U2OS cells in 5 minutes or less for each target. If, however, the super-resolved image of a well-stained target is of sub-par quality, we advise to repeat the data acquisition and examine the binding kinetics at specific IS and EC concentrations and (if possible) to confirm specific interactions between DS and IS by overlaying the Cy3 (reference) and DNA-PAINT signal. During data post-processing, it is important to inspect calculated drift correction by looking at the RMS values. Finally, image resolution can be roughly estimated by measuring the width of microtubules in the super-resolved images. We typically achieve lateral resolutions on the order of 22–35 nm.

Time Considerations:

The full procedure described here can comfortably be completed in 8–10 days. Once DS-conjugated secondary antibodies and dye-conjugated IS have been prepared as described in Basic protocol 1, staining and imaging as described in Basic Protocol 2 and 3 can be performed in only 3–4 days overall. Computationally extensive downstream data processing is subjective and depends on the specific research question and has to be added on top of the time considerations described here.

DAY 1 (3.5 h total):	Preparation of DBCO tagged DS and dye conjugated IS 10 min of hands-on time for mixing of the reactions 3 h incubation time
DAY2 (1 h total):	45 min hands-on time for ethanol precipitation (DS-DBCO and IS-dye)
DAY3 (5.5 h total):	45 min hands-on time for ethanol precipitation (DS-DBCO and IS-dye) 1 h hands-on time for preparation of azide-tagged antibody 3 h incubation time 30 min hands-on time for determining the concentration of each reagent
DAY 4 (3 h total):	40 min hands-on time to harvest final antibody-PEG8-DS conjugates 2 h incubation time 20 min hands-on time to etch coverglass for imaging experiments
DAY 5 (1 h total):	30 min hands-on time to seed cells
DAY 6 (8 h total):	7–8 h hands-on time to perform immunostaining
Day 7 (8 h total):	Variable hands-on time for imaging and start of data processing
Day 8 (4 h total):	Variable hands-on time for data processing

Flexible stopping points:

Ethanol precipitation is a flexible stopping point in Basic Protocol 1 as incubation at −80 °C can be easily extended for multiple days. Primary antibody incubation can be performed overnight at 4 °C and provides some flexibility in Basic Protocol 2. Data analysis described in Basic Protocol 3 can be performed with great flexibility.

Most time sensitive steps:

For this protocol quality of the samples is more important than quantity as single cells and not populations are imaged and analyzed in the following. Labeling of cells is the most time sensitive part of the described procedures and washes and incubation times should be followed rigorously to obtain the best results. Especially cell extraction and fixation have to be performed with great attention to detail to ensure recovery of targets of interest.

Additional comments:

Usually, 2–4 chambers of an 8-well coverglass are used for each experiment. Acquisition times depend on the target and need to be optimized. Experienced users can perform data acquisition with as little as 10–15 min hands-on time needed to set up correct imaging conditions (IS concentration, EC concentration, TIRF angle, etc.) Less experienced users require more time (on the order of 1–2 h) to optimize DNA hybridization kinetics, TIRF angle, focus as well as to find an interesting target cell to image.

DATA AVAILABILITY STATEMENT:

Most of the data including images and videos relevant to this protocol were published with our previous work (Civitici et al.). New image data were already shown as Figures in this work.