PCNA as Protein‐Based Nanoruler for Sub‐10 nm Fluorescence Imaging

Super‐resolution microscopy has revolutionized biological imaging enabling direct insight into cellular structures and protein arrangements with so far unmatched spatial resolution. Today, refined single‐molecule localization microscopy methods achieve spatial resolutions in the one‐digit nanometer range. As the race for molecular resolution fluorescence imaging with visible light continues, reliable biologically compatible reference structures will become essential to validate the resolution power. Here, PicoRulers (protein‐based imaging calibration optical rulers), multilabeled oligomeric proteins designed as advanced molecular nanorulers for super‐resolution fluorescence imaging are introduced. Genetic code expansion (GCE) is used to site‐specifically incorporate three noncanonical amino acids (ncAAs) into the homotrimeric proliferating cell nuclear antigen (PCNA) at 6 nm distances. Bioorthogonal click labeling with tetrazine‐dyes and tetrazine‐functionalized oligonucleotides allows efficient labeling of the PicoRuler with minimal linkage error. Time‐resolved photoswitching fingerprint analysis is used to demonstrate the successful synthesis and DNA‐based points accumulation for imaging in nanoscale topography (DNA‐PAINT) is used to resolve 6 nm PCNA PicoRulers. Since PicoRulers maintain their structural integrity under cellular conditions they represent ideal molecular nanorulers for benchmarking the performance of super‐resolution imaging techniques, particularly in complex biological environments.


Introduction
Super-resolution microscopy techniques have become essential tools for studying the molecular organization and composition of cellular organelles and complex molecular interactions.They allow us to shed light on the molecular organization and composition of cellular organelles and complex molecular interplays that support life with approximately tenfold higher spatial resolution as provided by standard fluorescence microscopy. [1,2]However, the realistic and reliable resolution estimation of super-resolution images has sparked debates in recent decades. [3,4]Typically, the spatial resolution has been estimated by measuring either the minimum resolvable distance between two fluorophores or from the cross-sectional profiles of super-resolved filamentous structures.In this context, immunolabeled microtubules have often served as a cellular reference structure.If the 2D projection of the fluorescence intensity distribution measured from microtubule filaments shows a bimodal distribution, the peak-to-peak distance can be fitted with a sum of two Gaussians and used as an estimate of spatial resolution. [5]Alternatively, Fourier ring correlation (FRC) can be employed to estimate image resolution by calculating a cross-correlation histogram in the frequency domain between two images of the same region. [6,7]However, since microtubules are not always uniformly immunolabeled, and suitable microtubule filaments or image areas are often subjectively identified for resolution analysis, these methods can be prone to errors.Several of these drawbacks can be bypassed using objective threshold-free resolution estimation by decorrelation analysis. [8]onsequently, reference structures or so-called rulers have been introduced to estimate the spatial resolution of superresolution microscopy methods.Ideal reference structures should have a well-defined geometric structure composed of identical or different molecules arranged at subdiffractionresolution distances that can be efficiently and specifically labeled with fluorescent probes.The first cellular reference structure introduced for super-resolution microscopy was the nuclear pore complex (NPC). [9]The NPC is a wheel-shaped cylindrical assembly with a core structure ≈125 nm in diameter and a height of around 70 nm.It contains eight spokes arranged in radial symmetry and a central channel 35-50 nm in diameter.Using direct stochastic optical reconstruction microscopy (dSTORM), the structure of the eightfold radial symmetry of the NPC and the diameter of the central channel were resolved with a spatial resolution of about 15 nm as early as 2012. [10]13] For cell-free super-resolution microscopy demonstrations, DNA origami structures have been widely used as flexible and modular rulers.They allow for the specific attachment of fluorophores at well-defined positions and distances.DNA origamis are synthesized by folding a long circular singlestranded DNA (a scaffold strand typically several thousand nucleotides in length) programmatically through hybridization with hundreds of short (around 10-15 nucleotides) single-stranded DNA oligonucleotides, known as staple strands. [14]The DNA origami technique also facilitates the construction of intricate 2D and 3D geometries. [15]As a result, DNA origami have been utilized as standard reference to demonstrate the resolving power of new super-resolution microscopy methods.18][19][20] Nevertheless, the appropriateness of using negatively charged DNA origami as rulers to demonstrate the performance of refined super-resolution microscopy methods, particularly for cellular imaging applications, is debatable.Ideally, emerging superresolution microscopy methods should be evaluated using biologically relevant reference structures or rulers, such as multiprotein complexes.However, using multiprotein complexes as natural rulers for sub-10 nm fluorescence imaging remains challenging, primarily due to the size, steric hindrance, and significant linkage error of common fluorescent probes.For instance, antibodies are 10-15 nm in size, and even fluorescent proteins, protein-based self-labeling tags, and camelid antibodies (nanobodies) are still too large for efficient fluorescence labeling of multiprotein complexes with the minimal linkage error required for sub-10 nm fluorescence imaging. [2,21]This necessitates the creation of protein-based reference systems that can be labeled site-specifically and quantitatively with small fluorophores with minimal linkage error.
In this study, we utilized genetic code expansion (GCE) [22,23] to design and synthesize a protein-based imaging calibration optical ruler (PicoRuler) for sub-10 nm fluorescence imaging.We achieved site-specific attachment of organic dyes with minimal linkage error by incorporating noncanonical amino acids (ncAAs) into proteins, which can then be efficiently labeled through bioorthogonal click chemistry with small tetrazinedyes. [21,24]We chose the proliferating cell nuclear antigen (PCNA) [25,26] as our protein-based reference system and inserted ncAAs into the homotrimeric protein at well-defined 6 nm distances using GCE. [27,28]After site-specific click labeling with tetrazine-dyes and tetrazine-functionalized oligonucleotides, we demonstrate the performance of our PCNA PicoRuler using photoswitching fingerprint analysis and DNA-PAINT. [18,20,29]

Synthesis and Labeling of a PCNA PicoRuler
PCNA plays a pivotal role in DNA replication and repair.It forms a homotrimeric ring with inner and outer diameters of 3.4 and 8.0 nm, respectively.This ring encircles the DNA, anchoring DNA polymerases and other DNA editing enzymes as it slides along the DNA strand. [25,26,30]Given its structure, PCNA is a suitable candidate for creating a PicoRuler for sub-10 nm superresolution fluorescence imaging.When a single ncAA is inserted into a PCNA monomer, it automatically introduces three equidistant click sites in the assembled trimer.Importantly, PCNA maintains its structure without undergoing conformational changes, acting as a rigid scaffold for the anchoring of three fluorophores at precise positions (Figure 1a).
To construct the PCNA PicoRuler, we produced recombinant PCNA in bacterial culture.We added several tags at both termini for enhanced protein purification and detection in western blots.An N-terminal Strep-and HA-tag were appended via a TEV protease cleavage site and flexible linker.Conversely, a His-tag was added to the C-terminus via a thrombin protease cleavage site and another flexible linker (Figure S1a, Supporting Information).At first, we focused on establishing the expression and purification protocol for the wild type PCNA (PCNA WT).This protein was abundantly expressed, achieving optimal purity after four stages of chromatographic purification (Figure 1b; Figure S1b-d, Supporting Information).
Notably, it is known that PCNA can form a transient double trimer complex under certain conditions. [31]Two amino acids, R5 and K110, have been identified as responsible for PCNA dimer formation.Previous studies have shown that mutating one of these amino acids to alanine can prevent dimer formation. [32]owever, in our hands, mutations R5A and K110A were destabilizing the PCNA trimer and hardly expressed (Figure S1c, Supporting Information).Based on this observation, we utilized protein stability prediction tools, which proposed the K110I mutation to stabilize PCNA homotrimers. [33]This single mutation not only preserved protein expression but also successfully inhibited the formation of double trimer complexes under glutaraldehyde crosslinking conditions (Figure S1c,e-g, Supporting Information).Considering the potential dissociation of PCNA trimers into monomers at concentrations below 50 × 10 −9 м and the typical concentration of <1 × 10 −9 м used in single-molecule experiments, we crosslinked PCNA with BIS(NHS)PEG-9 at a higher concentration before dilution (Figure S1h, Supporting Information). [34]To validate the synthesis of the homotrimeric PCNA molecules and verify their expected structure, we employed transmission electron microscopy (TEM) using negative staining.The TEM images distinctly depict ring-shaped trimeric structures, with an average diameter of 5.5±1.1 nm (s.d.), confirming the successful synthesis of PCNA homotrimers (Figure 1c; Figure S2, Supporting Information).
To enable fluorescence labeling of PCNA via bioorthogonal click labeling with tetrazine-dyes, we site-specifically integrated ncAAs into PCNA through GCE.Seven potential click-sites were assessed for efficiency in inserting the lysine derivatives Norb-K, BCN-K, and TCO*A-K, using the NORS3T suppressor plasmid (Figure S3, Supporting Information). [35]Although optimized for Norb-K, the aminoacyl-tRNA-synthetase enabled the insertion of BCN-K and TCO*A-K in response to the amber stop codon during translation.Of the sites tested, position 6 (S186TAG) showed the most efficient ncAA insertion in PCNA, resulting in interfluorophore distances of 6 nm (Figure 1a).Subsequent absorption spectroscopy confirmed successful labeling of PCNA.Using a high tetrazine-dye to protein ratio, we achieved an average degree of labeling (DOL) of ≈3.0 for PCNA (S186Norb) labeled with H-Tet-Cy5.As the reference sample (singly labeled Pi-coRulers), we used a PCNA sample with an average DOL of ≈0.4 applying a low tetrazine-dye to protein ratio (Figure 1d; Figure S4a, Supporting Information).Here it has to be considered that a PCNA sample with a DOL of ≈0.4 may contain also duallabeled PCNAs besides an excess of unlabeled and singly labeled PCNAs.

Photoswitching Fingerprint Analysis of PCNA PicoRulers
To validate the performance of PCNA PicoRulers for sub-10 nm fluorescence imaging, we immobilized dilute concentrations on coverslips and attempted to resolve the 6 nm interfluorophore distances by direct stochastic optical reconstruction microscopy (dSTORM). [36]In addition, we synthesized DNA origami containing three Cy5 dyes with ≈6 nm interfluorophore distance for comparison.Imaging was conducted in standard photoswitching buffer with 640 nm irradiation exclusively.As expected, dSTORM failed to resolve PCNA PicoRulers and DNA origami with 6 nm interfluorophore distances (Figure 2a,b).However, a strong dimer contribution in the absorption spectrum indicates the presence of multiple contact-forming labels on individual PCNA molecules.In addition, due to multiple resonance energy transfer reactions between Cy5 fluorophores separated by less than 10 nm, which result in an accumulation of fluorophores in their on-state (in photoswitching buffer), time-resolved fluorescence detection paired with photoswitching fingerprint analysis can provide insights about spatially unresolvable fluorophores in the sub-10 nm range. [18]hotoswitching fingerprint analysis of PCNA PicoRuler signals under dSTORM conditions revealed shorter off-state lifetimes and a greater number of blinking events for PCNA Pi-coRulers labeled with three Cy5 fluorophores.The observed reduction in median off-state lifetime from 23.5 s for a DOL ≈0.4 to 13.9 s for a DOL ≈3.0 can be attributed to energy transfer between fluorophores in the on-and off-state, leading to fluorophore accumulation in the on-state (Figure 2c). [18]In addition, the median number of on-events detected for a DOL ≈3.0 was 11 and 7 for a DOL ≈0.4 (Figure 2c).Finally, the number of on-events detected per frame as a function of time demonstrate faster blinking of DOL ≈3.0 PicoRulers during the first minutes of the experiment (Figure 2c).Taking together our data measured under different excitation intensities show the successful synthesis of Cy5-labeled PicoRulers with 6 nm interfluorophore distance (Figure 2c; Figure S4d,e, Supporting Information) .
Fluorescence lifetime (FLIM) images further confirm the interactions of Cy5 fluorophores within sub-10 nm distances via multiple energy transfer routes. [18]While reference PCNA PicoRulers with a DOL ≈0.4 exhibit an average fluorescence lifetime of ≈1.5 ns, PCNA PicoRulers with a DOL ≈3.0 exhibit a shorter lifetime of ≈1.1 ns (Figure 2d).These observations align with the average fluorescence decays measured ) and 3.2 ± 1.9 nm (s.d.), respectively. [38]) Temporal degradation profile of PCNA PicoRulers in FCS medium.
across the sample (Figure 2e).DNA origami with 6 nm interfluorophore distance showed a slightly more pronounced photoswitching fingerprint signature (Figure S5a-c, Supporting Information).
An alternative way of demonstrating the successful synthesis of a multiple-labeled protein that emphasizes the underlying energy transfer pathways between identical Cy5 fluorophores separated by less than 10 nm is photon antibunching. [18,37]Photon antibunching experiments enable the investigation of the number of emitting fluorophores contributing to the detected fluorescence signal per PCNA PicoRuler (Figure 2f).This so-called photon coincidence analysis takes advantage of the fact that the probability of emitting two consecutive photons drops to zero for a single emitter for time intervals shorter than the excitedstate lifetime.For sufficiently short laser pulses, the number of photon pairs detected per laser pulse can thus be used to determine whether the emission is from one or more independently emitting quantum systems. [18,37]Since the intensity of the central peak contains information about the number of independently emitting molecules, the number of photon pairs detected in the central peak, N c , at delay time zero, relative to the average number of the lateral peaks, N l,av , can be used to determine the number of independently emitting fluorophores.For example, neglecting the background, N c /N l,av ratios of ≈0.0, ≈0.5, ≈0.67, and ≈0.75 are expected for 1-4 independently emitting fluorophores.In our experiments, we detected N c /N l,av ratios of 0.3±0.2 and 0.6±0.3 for PCNA PicoRulers with a DOL of ≈0.4 and ≈3.0, respectively, measured in photostabilizing buffer containing trolox and an oxygen scavenging system (Figure 2f; Figure S4f,g, Supporting Information).The strong scattering of N c /N l,av ratios shows that the PicoRuler samples with an average DOL of ≈3.0 and ≈0.4 (reference) contain also PCNA molecules that are labeled with less than three and more than one fluorophores, respectively.
Unfortunately, the number of on-events (Figure 2c; Figure S4d,e, Supporting Information) does not directly reflect the number of fluorophores attached per PicoRuler because in the initial period, i.e., during the first seconds of irradiation, too many fluorophores are simultaneously active impeding the detection of each individual on-event.Nevertheless, all information including the off-state lifetime, the number of on-events, the temporal evolution of localizations, the photon antibunching signature, and the fluorescence lifetimes unequivocally demonstrate the successful synthesis of PCNA Pi-coRulers carrying three fluorophores separated by less than 10 nm.

DNA-PAINT Resolves PCNA PicoRulers
Our data show that PCNA PicoRulers with an interfluorophore distance of 6 nm cannot be entirely resolved by dSTORM experiments with a localization precision of 6.5 ± 0.98 nm (s.d.). [38]herefore, single-molecule localization microscopy methods that achieve higher localization precisions have to be used to resolve the PicoRulers.MINFLUX has shown spatial resolutions of a few nanometers using DNA origami as reference structures. [12,16]owever, when using Cy5 fluorophores in photoswitching buffer, MINFLUX also suffers from interfluorophore energy transfer at distances <10 nm.As previously described, this leads to the accumulation of fluorophores in the on-state, causing increased photobleaching subsequently resulting in diminished localization probabilities. [18]In response to these findings, latest MIN-FLUX demonstrations substituted photoswitchable by photoactivatable fluorophores ensuring that on average only one fluorophore remains in its fluorescent state within every 10 nm area during the experiment. [39]n alternative method to bypass fluorophore interactions in the sub-10 nm regime and to attain nanometer spatial resolutions involves the use of DNA-based points accumulation for in nanoscale topography (DNA-PAINT). [20,29,40]DNA-PAINT uses the transient binding of a fluorophore-labeled single-stranded short oligonucleotide probe (imager strand) to a complementary oligonucleotide (docking strand) that is covalently bound to the target of interest.To harness the superior localization precision of DNA-PAINT and discern the 6 nm distances, we labeled the PCNA PicoRuler with a tetrazinefunctionalized standard DNA-PAINT docking strand (H-Tet-P1), [41] using the same concentration ratios optimized for PCNA labeling with H-Tet-Cy5, targeting a DOL of ≈3.0 and ≈0.4 for completely-and singly labeled reference PicoRulers, respectively.
Using Cy3B labeled P1 imager strands DNA-PAINT allowed us to resolve the regular triangle structure of the 6 nm PCNA PicoRuler (Figure 3a; Figure S6, Supporting Information).Unfortunately, but in line with other sub-10 nm fluorescence imaging studies, [12,[16][17][18][19][20] triangular structures can be resolved in only a fraction of nanorulers.For our PCNA PicoRuler this is mainly due to (i) incomplete click-labeling with highly negatively charged docking strands, (ii) a localization precision of ≈3 nm, which is insufficient to unequivocally resolve all 6 nm distances (Figure 3b), and (iii) the nonplanar orientation of PicoRulers on the glass surface in combination with 2D DNA-PAINT imaging (Figure 1c; Figure S2, Supporting Information).While the average DOL of fluorophore-labeled PCNA molecules can be easily determined at the ensemble level by absorption spectroscopy it is extremely error-prone for DNA-labeled PCNA molecules because the absorption spectra overlap strongly.Hence, the optimization of the labeling efficiency for DNA-PAINT imaging remains difficult.
A critical but mostly neglected characteristic of synthetic rulers is their stability, in particular in a cellular environment.Ultimately, super-resolution microscopy methods have to demonstrate their resolving power under realistic conditions in order to be used advantageously for imaging in cells.Therefore, we evaluated the stability of both, PCNA PicoRuler and DNA origami in degradation experiments.Here, the rulers were incubated in PBS containing 10% fetal calf serum (FCS) to mimic cellular environment and their stability was examined by gel electrophoresis at different times.Most mammalian cell growth media contain FCS as it provides several growth factors and hormones involved in growth promotion and cell function.The stability test revealed, that PicoRulers show only low degradation in 48 h (Figure 3c), whereas DNA origami degraded within the first hours of the experiment as has already been observed in previous studies (Figure S5d,e, Supporting Information). [42,43]To prevent degradation and explore real-life applications of DNA origami, e.g., in biological environments, coating of the ruler with calcium phosphate or silica can be used. [44]

Conclusion/Discussion
With the advent of cutting-edge super-resolution microscopy methods that approach resolutions in the one-digit nanometer range such as MINFLUX, [12,16,39] MINSTED, [19] DNA-PAINT, [20,40] and Expansion Microscopy (ExM), [45,46] robust molecular rulers are urgently needed to benchmark their performance.So far, DNA origami have been successfully used to demonstrate spatial resolutions in the sub-10 nm range.44] Here, we successfully demonstrated the development, optimization, and characterization of a protein-based imaging calibration optical ruler (PicoRuler) to benchmark the resolution power of methods in the one-digit nanometer range.These protein-based nanorulers are not merely alternatives to traditional tools like DNA origami but offer distinct advantages for specific applications, especially where precision in functionality and stoichiometry is paramount.Built on the homotrimeric protein PCNA and equipped with identical fluorophores through GCE, our Pi-coRuler combines the precise control of synthetic design with site-specific labeling.Apparently, other stable protein complexes can be used as nanorulers that contain known numbers of subunits and exhibit uniform stoichiometry. [47] hallmark of PicoRulers is their modular design enabling the replacement of natural amino acids by ncAAs.Given that any natural amino acid can be replaced by a ncAA and with over 200 ncAAs available, the possibilities are vast.[48] Due to its modular conception ncAAs can be labeled or functionalized with organic fluorophores, oligonucleotides, and biotin among many other small molecules with minimal linkage error.Thus, PicoRulers can also be exploited for the synthesis of other nanometer-sized systems including plasmonic devices.[49] In addition, the possibility of site-specific anchoring of ncAAs into a hydrogel using multifunctional linkers containing, e.g., a tetrazine group, a fluorophore and an acryloyl unit, cannot only be used advantageously for ExM experiments but can likewise pave the way for enhancing and broadening the horizons of super-resolution microscopy techniques.
Due to their high stability, PicoRulers might be useful also for intracellular super-resolution reference experiments once their performance has been optimized in extracellular environment.To enable intracellular resolution performance demonstrations with PicoRulers, they have to be delivered into cells possibly by microinjection or functionalization with cell penetrating peptides (CPPs). [50]In addition, intracellular experiments will require the use of 3D single-molecule localization microscopy methods and, in particular, the localization precision might suffer from hindered anisotropic emission due to nonspecific fluorophore interactions with intracellular molecules.In addition, we acknowledge other challenges that PicoRulers face, such as controlling their assembly and ensuring their uniformity.But ongoing advancements in protein engineering promise to overcome these hurdles.

Experimental Section
Expression Plasmids and Suppressor Plasmids: Human PCNA (UniProt: P12004) was commercially synthesized (Eurofins Genomics) and cloned into pRSET A vector via NdeI (5′) and EcoRI (3′) restriction sites.This plasmid served as the expression plasmid.Plasmids for expressing all the other PCNA variants (K110I and the amber mutants) were generated through site directed mutagenesis.The suppressor plasmid NORS3T was a gift from Thomas Carell.
PCNA Expression: PCNA WT and K110I was expressed according to a previously published literature with several modifications. [51]In brief, plasmid was transformed into C41(DE3) cells (Sigma Aldrich) and plated on 2XYT agar with ampicillin (100 μg mL −1 ).A single colony was used to prepare an overnight preculture in 2XYT-ampicillin media at 37 °C and 200 rpm.The next day, the preculture was diluted 100 × in fresh 2XYTampicillin media (100 mL) and shaken at 37 °C until the cells reached OD600 of 0.8.The cells were chilled at 4 °C for 10 min.Protein expression was induced by IPTG at a final concentration of 0.3 × 10 −3 м.After 24 h, the cells were pelleted by centrifugation and stored at −20°C until purification.For evaluating the insertion of ncAAs at different click-sites, the amber mutants of PCNA were expressed in a similar way, but with a few differences.First, both the expression and the suppressor plasmids were cotransformed into C41(DE3) cells by electroporation and maintained in cells by ampicillin and chloramphenicol (34 μg mL −1 ).Next, Norb-K (Iris Biotech), BCN-K, or TCO*A-K (SiChem) was added at a final concentration of 1 × 10 −3 м when the OD600 reached 0.5.And the expression was performed only in 10 mL media to keep the use of ncAAs minimum.The cell pellets were lysed and purified (see below) using the Ni-NTA spin column (ThermoFisher).The first eluted fraction then was analyzed via SDS PAGE.PCNA-6 (S186Norb) was expressed in a similar way as the amber mutants of PCNA, but right before the induction the media was replaced with 100 mL bactomedia (Thermofisher) supplemented with 0.75% glycerol.Furthermore, Norb-K was added at 3 × 10 −3 м after the media replacement.
PCNA Purification: All the PCNA variants were purified in the same way.First, the cell pellets were gently resuspended in lysis buffer ( 10 PCNA Characterization: Western blot analysis was performed against PCNA WT using an anti-HA tag antibody conjugated with AF488 (Invitrogen) as the primary antibody at 1:500 dilution.
PCNA Crosslinking: A mild crosslinking by hanging drop method was applied to prevent the intermolecular crosslinking of PCNA (Figure S1e Transmission Electron Microscopy (TEM) and Analysis: The synthesis of PicoRulers were validated by TEM (JEM 1011, JEOL) under negative staining.Carbon coated 300 Mesh TEM-grids were used and glowed freshly.The prepared grids were incubated with 20 μL of 500 × 10 −9 м PCNA solution for 2 min.Afterwards, the solution was blotted using a filter paper.The grid was dipped 3× into a 0.75% uranyl acetate solution (EMS) and blotted immediately.Finally, the sample was incubated with 0.75% uranyl acetate solution (EMS) for 45 s.The solution was blotted and air-dried for 24 h prior imaging.All TEM images were analyses with ImageJ 2.3.0/1.53t.
DNA Origami Design and Hybridization: DNA origami rectangle structures were designed, hybridized, and purified like previously described. [18]he DNA origami ruler used in this study includes 3 dye modified staple strand, which are arranged triangularly, with an average interfluorophore distance of 6 nm.For DNA origami reference sample, only one staple strand sequence was modified with dye.All dye modified staple strands were ordered at biomers.netGmbH, whereas all biotinylated strands were ordered at Sigma-Aldrich.All unmodified staple strands were ordered at Merck KGaA.The phage M13mp18 derivate DNA type p7560 were used as scaffold DNA (tilibit nanosystems, M1-32).Briefly, 10 × 10 −3 м scaffold DNA was mixed with 15 eq.unmodified staple strands and 30 eq. modified staple strands in hybridization buffer (5 × 10 −3 м Tris, 5 × 10 −3 м NaCl, 1 × 10 −3 м EDTA, and 12 × 10 −3 м MgCl 2 (AppliChem)).Hybridization was performed using a ThermoCycler (C1000 Thermal Cycler, BioRad) with a linear thermal gradient of −1 °C min −1 from 90 to 4 °C.Afterwards, the hybridized samples were purified by electrophoresis in a 1.5% agarose gel (Sigma) in 1× TBE buffer (4.5 × 10 −3 м Tris, 4.5 × 10 −3 м boric acid (Merck), and 10 × 10 −3 м EDTA) with 0.5× TBE containing 12 × 10 −3 м MgCl 2 as running buffer.The agarose powder was dissolved using a microwave.The solution was cooled down to ≈60 °C before adding 12 × 10 −3 м MgCl 2 .Afterward, the gel was poured immediately.A small amount of sample (≈10 μL) was picked as reference, which was mixed with intercalating dye (2 μL, Safe-Green, Applied Biological Materials).A small amount of pure scaffold as well as pure staple strands were used as references and mixed in hybridization buffer and also mixed with intercalating dye respectively.The rest of the hybridized origami samples were not treated with intercalating dye.All samples were mixed with loading dye (10 × 10 −3 м Tris, 60% glycerol v/v (Merck), 0.03% w/v bromophenol blue (Carl Roth)).Electrophoresis was performed using a programmable d.c.voltage source (PowerPac Basic, BioRad) at 70 V, for roughly 2 h in water/ice bath.The part of the gel including the references were separated by cutting across the length of the gel and the bands were marked at an ultraviolet transilluminator (UST20M-8E, INTAS).Afterward, the marked gel was combined with the nonilluminated part of the gel containing the DNAorigami structures unmixed with intercalating dye.These DNA origamis were cut out according to the height of the references.These gels were purified separately by Freeze N' Squeeze columns (BioRad) according to the manufacturer's instructions using a benchtop centrifuge (Biofuge fresco, Heraeus) at 13 000 g.For all microscopy measurements, the DNA origami was produced freshly on the same day.
Single-Molecule Surface Preparation: The surfaces of 8-well chambered cover glass with high performance cover glass (Cellvis, C8-1.5H-N) were washed once with PBS (Sigma-Aldrich) and treated with 2% Hellmanex (Hellma) for 1 h.Afterwards, the treated chambers were washed 3× with PBS, incubated with 1 м KOH (Fluka) for 20 min, and rewashed with PBS.For PicoRuler measurements, the crosslinked PCNA samples were diluted to ≈1 × 10 −9 m (dSTORM and DNA-PAINT) or ≈100 × 10 −12 m (confocal lifetime measurements, FLIM) in crosslinking buffer.Briefly, the cleaned surfaces were incubated with the diluted protein samples for ≈5 s and washed 3× with PBS.The samples then were washed and covered in appropriate imaging buffers.For DNA origami measurements the surfaces were incubated with 10% polyethylene glycol 400 (Fluka) overnight at 4 °C.Thereafter, the surfaces were rinsed 3× with PBS and incubated with 0.5 g L −1 BSA-Biotin (ThermoFisher) in PBS overnight at 4 °C.On the next day, the chambers were washed 3× with PBS before incubation with 0.5 g L −1 Neutravidin (ThermoFisher) in PBS for 20 min.Afterwards, the treated chambers were washed 2× with PBS and 2× with PBS containing 50 × 10 −3 м MgCl 2 .The prepared surfaces were incubated with purified DNA origami solution (diluted 5× in PBS containing 50 × 10 −3 м MgCl 2 ) for 10 min.Next, the surfaces were washed 3× in PBS containing 50 × 10 −3 м MgCl 2 before transferred to imaging buffers.
dSTORM and DNA-PAINT Imaging and Analysis: Super-resolution imaging was performed on an inverted wide-field fluorescence microscope (IX-71, Olympus).For excitation of Cy5 (dSTORM), a 641 nm diode laser (Cube 640-100 C, Coherent) in combination with a clean-up filter (Laser Clean-up filter 640/10, Chroma) was used.For excitation of Cy3B imager strands (DNA-PAINT), a 561 nm diode laser (Genesis MX561-500 STM, Coherent) with irradiation intensity of ≈0.25 -≈1 kW cm −2 in combination with a clean-up filter (Laser Clean-up filter 561/14, Chroma) was used.All measurements were performed using circular polarized light by mounting a quarter-wave plate (Thorlabs,) within the excitation path.The laser beam was focused onto the back focal plane of the oil-immersion objective (×60, NA 1.45, Olympus).For measurements, emission light was separated from the illumination light using a dichroic mirror (HC 560/659 (dSTORM) or FF580-FDi01 (DNA-PAINT), Semrock) and spectrally filtered by a bandpass filter (FF01-679/41-25 (dSTORM) or BrightLineHC-600/50 (DNA-PAINT), Semrock).Images were recorded with an electronmultiplying CCD camera chip (iXon DU-897, Andor).Pixel size for data analysis was measured to 128 nm.For dSTORM measurement 60 000 images with an exposure time of 5 ms (frame rate 200 Hz) and irradiation intensity of roughly ≈1.5 -≈5 kW cm −2 were recorded using epi illumination.For DNA-PAINT measurement, 18 000 images with an exposure time of 10 ms (frame rate 100 Hz) were recorded by total internal reflec-tion fluorescence microscopy illumination.All dSTORM experiments were performed in PBS-based photoswitching buffer containing 100 × 10 −3 м -mercaptoethylamine (Sigma-Aldrich) adjusted to pH 7.6.DNA-PAINT experiments were performed with 5 × 10 −9 м imager strand concentration (5′-3′: CTA GAT GTA T, Eurofins), 3′-modified with Cy3B, in PBSbased buffer containing 5 × 10 −3 м Tris, 1 × 10 −3 м EDTA, and 10 × 10 −3 м MgCl 2 adjusted to pH 7.6.All SMLM results were analyzed with rapidSTORM3.3and the highly resolved pictures were reconstructed with ThunderSTORM. [52,53]The localization precisions were calculated according to Mortensen et al. [38] Therefore, the frame rates were binned to 20 ms (dSTORM) and 100 ms (DNA-PAINT) matching the average on-times of blinking fluorophores and binding time imager strands, respectively.For analyzing the photoswitching fingerprints, fluorescent spots containing more than 115 (dSTORM)/1500 (DNA-PAINT) photons per frame were analyzed.The estimation of the number of localizations per fluorophore was calculated by using the tracking function (Kalman filter) of rapid-STORM3.3.Fluorescent spots were tracked over the whole image stack within a tracking radius of 200 nm and exported as tracked localization file.A custom written python script was used to calculate the number of frames between on-time events of the same fluorescent spot within the defined tracking radius (off-time) as well as the number of on-time events per tracked spot.
Fluorescence Lifetime Imaging Microscopy and Analysis: All fluorescence lifetime and photon antibunching measurements were performed on a time-resolved confocal fluorescence microscope setup consisting of a FLIMbee galvo scanner (PicoQuant), a MicroTime200 (PicoQuant), an Olympus IX83 microscope including an oil-immersion objective (60×, NA 1.45; Olympus), two single-photon avalanche photodiodes (SPADs) (Excelitas Technologies, 75 154 K3, 75 154 L6), as well as a TimeHarp300 dual-channel board.For pulsed excitation, a white-light laser (NKT Photonics, SuperK extreme) was coupled into the MicroTime200 system using a glass fiber (NKT Photonics, SuperK FD PM, A502-010-110).For all measurements, a 100 μm pinhole was used.A 50:50 beamsplitter (Pi-coQuant) was used to split the emission light onto the SPADs.To filter out after glow effects as well as scattered and reflected light, two identical bandpass filters ET700/75 M (Semrock) were installed in front of the SPADs.The measurements were performed and analyzed using the Sym-PhoTime64 software (PicoQuant).The protein and DNA origami measurements were performed in Trolox/PBS buffer including an oxygen scavenger system (5% w/v glucose, 10 U mL −1 glucose oxidase, and 200 U mL −1 catalase; Sigma-Aldrich) adjusted to pH 7.6, as well as an irradiation intensity of ≈2.0-2.5 kW cm −2 in T3 mode with 25 ps time resolution, 25 μs pixel dwell time and a 642 nm laser repetition rate of 40 MHz, whereas all single-molecule trajectories for photon antibunching measurements were performed in T2 mode for 60 s per fluorescence spot.For photon antibunching experiments, the Sync cable was disconnected and replaced by the SPAD 2 cable.For analyzing the fluorescence lifetime, the decay parameters were determined by least-squares deconvolution, and their quality was judged by the reduced 2 values and the randomness of the weighted residuals (2 of roughly 1).A multiexponential model was used to fit the decay ( av =  1 a 1 +  2 a 2 ) in the case a monoexponential model was not adequate to describe the measured decay.
Photon Antibunching Measurements and Analysis: The interphoton time data can be used to determine the number of independent emitters by calculating the ratio of the number of photons in the central peak (N c ) to the average number in the neighboring lateral peaks (N l,av ).For determination of N l,av , we used the average number of events in the nearest 8 peaks, 4 to each side of the zero-time peak.
Stability Test of PCNA and DNA-Origami in Serum Rich Environments: PCNA K110I was dissolved in PBS (Dulbecco) containing 10% fetal calf serum (FCS, Sigma Aldrich) at the final concentration of 5 × 10 −6 m and incubated at 37 °C for the following time points: 0, 1, 2, 4, 8, 12, 24, 48 h.The samples then were analyzed by western blot against anti-PCNA mouse monoclonal antibody conjugated with AF 647 (Abcam) as the primary antibody at 1:200 dilution.Similarly, freshly purified DNA origami in TBE buffer was incubated in PBS containing 10% FCS at 37 °C for: 1, 2, 3, 6, 12, 24, 48 h.Additionally, three samples were stored at −20, 4, and 37 °C without FCS as references.After 48 h, all samples were mixed with 0.16 eq.loading dye and intercalating dye and the stability of the DNA-origamis was examined by electrophoresis in a 1.5% agarose gel in 1× TBE buffer for 2 h in water/ice bath.Afterwards, the agarose gel was illuminated with an ultraviolet transilluminator (UST20M-8E, INTAS).
Statistical Analysis: Spectroscopic graphs were analyzed and generated using OriginPro2021b (OriginLab, Northampton, MA).All dSTORM and DNA-PAINT measurements were analyzed and images reconstructed (including intensity rescaling by histogram equalization) with rapid-STORM3.3,and the highly resolved pictures were reconstructed with ThunderSTORM (Version 1.3).Fluorescence lifetime and antibunching measurements were analyzed using SymPhoTime64 (Version 2.7).All experiments regarding imaging and analysis were carried out at least three times if not stated otherwise.

Figure 1 .
Figure 1.Engineering a PCNA-based PicoRuler for sub-10 nm super-resolution microscopy.a) Depiction of the PCNA ruler design, showing the positions of three identical fluorophores at intervals of 6 nm.This was achieved through bioorthogonal click labeling of ncAAs, which were site-specifically introduced into PCNA through GCE.As a reference sample we used PCNA labeled with a DOL ≈0.4.b) SDS-PAGE illustrating the purity of the PCNA WT after each step of purification.L: protein ladder, His: after nickel affinity chromatography, Strep-Trap: after Strep-tag affinity chromatography and SEC: after size exclusion chromatography.The PCNA monomer band is marked by a red triangle.c) PCNA-6 (S186Norb) imaged by TEM under negative staining with uranyl acetate.d) SEC chromatogram of PCNA-6 (S186Norb) after labeling with H-Tet-Cy5, confirming the successful conjugation of the fluorophores to the PCNA protein.Fraction containing the labeled PicoRuler is shown in blue shading.Scale bar 5 nm (c).

Figure 2 .
Figure 2. Photophysical evaluation of PCNA-based PicoRulers and DNA origami.a,b) Selected dSTORM images of triple-labeled (DOL ≈ 3.0) PicoRulers (a) and DNA origami (b) with interfluorophore distances of 6 nm, along with their singly-labeled reference samples (Ref).While each DNA origami contains a single fluorophore, the PCNA reference sample exhibits a DOL ≈0.4 and thus contains also unlabeled and dual-labeled PCNA molecules (pixel size 2 nm).Scale bars, 10 nm.c) Relative occurrence of lifetimes of the off-state (OFF-time), number of on-states (Events) detected from individual PicoRulers in dSTORM experiments and number of events (localizations) detected per frame as a function of time (Fingerprint Analysis).d) FLIM images of PicoRulers click-labeled with H-Tet-Cy5 (the singly labeled reference is shown in a grey and the triple-clicked PicoRuler in a magenta box) measured by confocal TCSPC imaging in photoswitching buffer at an irradiation intensity of ≈2.5 kW cm −2 .To minimize photobleaching of fluorophores, FLIM images were recorded at 25 μs of integration time per pixel.No intensity threshold was applied.Scale bars, 2 μm.e) Average fluorescence decays from PCNA PicoRulers labeled with one (grey) or three (magenta) Cy5 fluorophores.f) N c /N l,av ratios determined for singly-labeled reference PCNA molecules and triple-labeled PicoRulers in photon antibunching experiments (n = 13).

Figure 3 .
Figure 3. Performance and stability of PCNA-based PicoRulers.a) Selected DNA-PAINT images of PCNA PicoRulers labeled with a low docking strand to PCNA ratio (reference) and high docking strand to PCNA ratio (pixel size 2 nm).Scale bars, 10 nm.b) Localization precisions in x-and y-direction are 3.1 ± 1.8 nm (s.d.) and 3.2 ± 1.9 nm (s.d.), respectively.[38]c) Temporal degradation profile of PCNA PicoRulers in FCS medium.