Novel Fluorescent Dyes for Single DNA Molecule Techniques

Abstract

To answer the demands of scientific and medical imaging issues, the family of nucleic acid fluorescent dyes is constantly enlarging. Most of the developed dyes reveal high qualities in bulk solution assays but are inefficient to produce a strong and sufficiently stable signal to enable the application of single-molecule techniques. Therefore, we tested 12 novel monomeric and homodimeric cyanine dyes for potential single DNA molecule imaging. Although their qualities in bulk solutions have already been described, nothing was known about their behavior on a single-molecule level. All 12 dyes demonstrated strong emission when intercalated into single DNA molecules and stretched on a silanized surface, which makes them the perfect choice for fluorescent microscopy imaging. A comparison of their fluorescence intensity and photostability with the most applicable dyes in single-molecule techniques, fluorescent dyes YOYO-1 and POPO-3, was carried out. They all exhibited a strong signal, comparable to that of YOYO-1. However, in contrast to YOYO-1, which is visualized under a green filter only, their emission permits red filter visualization. As their photostability highly exceeds that of similar spectrum POPO-3 dye, the studied dyes stand out as the best choice for a broad range of solid surface single-molecule applications when yellow to red DNA backbone fluorescence is needed.

IN RECENT STUDIES, stretching of fluorescent deoxyribonucleic acid (DNA) onto a pretreated solid surface by means of different techniques has been more commonly applied in various DNA molecule analyses.^1–4 The method is based on movement of the molecules in an aqueous solution at an interface between air and a substrate surface.^1,5 Usually, binding between DNA extremities and a variety of surfaces, such as hydrophobic surfaces (graphite, Teflon), glass coated with vinyl silanes, polystyrene, polydimethylsiloxane (PDMS), or hydrophilic surfaces (cleaned glass or glass coated with aminosilanes, polylysine, polyhistidine, or polymethylmethacrylate) ⁵ is used. One of the popular methods for the stretching of DNA molecules is the so-called “receding meniscus” method. ⁶ The movement of the meniscus applies a constant force on the DNA anchoring point, leaving behind a part of it combed dry and ahead the rest of the molecule in an aqueous solution. This force elastically stretches DNA molecules in the close vicinity of the meniscus. A characteristic feature of the receding meniscus approach is that higher nonspecific adsorption (high affinity between the surface and DNA) correlates with less stretching effectiveness. ⁶

Another important feature for successful binding of one or both DNA molecule extremities to the surface is the pH range of the surface. ⁵ On hydrophobic surfaces, the optimal pH for binding is assumed to be about 5.5. The typical reaction in that case is carried out in 2-morpholinoethanesulfonic acid (MES) buffer (pH 5.5). On surfaces coated with ionizable groups (such as clean glass, polylysine or polyhistidine, and aminosilanes), the optimal pH range depends on the pK_a (acid dissociation constant) of the surface (the charge of the free groups) and the density of the groups.

The most important part of such single-molecule experiments is the fluorescent dye to be applied for successful and stable staining of DNA backbone. The great variety of modern molecular biologic approaches require a broad range of fluorescent dyes, emitting in different spectral parts. That is why the development and examination of novel dyes for sensitive detection and measuring of double-stranded and single-stranded DNA, oligonucleotides and energy transfer primers for DNA sequencing and polymerase chain reactions, as well as dyes for microscopic single molecule assays, is a very important and constantly developing research area. Currently, there are several directions of development: microparticles (eg, phosphors, quantum dots, nanocrystals) and dyes that bind nucleic acids directly. The latter are basically fluorophores based on fluorescein (eg, cyanine dyes, rhodamine, carboxyfluorescein). There are commercially available (Invitrogen, Carlsbad, CA) families of monomeric (TO-PRO, YO-PRO, PO-PRO) and dimeric (TOTO, YOYO, BOBO, POPO) cyanine dyes that are found to be extremely useful in various applications^7–9 as they possess low intrinsic fluorescence in an unbound state, but perform with up to 1,000 times stronger emission when intercalated into nucleic acids to produce high-fluorescence quantum yields. ¹⁰ This makes cyanine dyes attractive as nucleic acid stains and labels as their strong and comparatively photostable signal protrudes over the weak and short-lived background fluorescence of the unbound dye.

To be applicable for microscopy analysis of single DNA molecules stretched onto a pretreated solid surface, a dye should fulfill several requirements. Once bound to nucleic acids, dyes must emit a bright fluorescent signal, dispersed uniformly throughout the molecule backbone to allow its proper distinction from the background. To give time to researchers to carry out their tasks, the dye must also exhibit high photostability. Another advantage is lower dye consumption per base pairs. The green fluorescent nucleic acid stain YOYO-1 (Invitrogen, N-7565) finds a broad application in the contemporary single molecule techniques as it is strong and relatively stable over time. The red spectrum–shifted dyes, which are less photostable, are more applicable to the fluid flow stretching.

In the present study, a particular procedure was used to stain single DNA molecules from Saccharomyces cerevisiae total genome DNA with 12 monomeric and homodimeric cyanine dyes, based on monomethine cyanine chromophore with oxazolo[4,5-b]pyridinium and quinoline end groups. ¹¹ The procedure was followed by a simple stretching method on glass slides silanized with aminoalkylsilane, to test them for potential application in fluorescent microscopy visualization of single DNA molecules. Their intensity and photostability were compared to those of the well-known green spectrum emitting YOYO-1 iodide and yellow spectrum emitting POPO-3 iodide dimeric cyanine nucleic acid dyes. Our study demonstrates that bound to DNA, the novel 12 cyanine dyes give strong and stable fluorescence that can be visualized by a red filter set. Their qualities make them a perfect choice for solid surface single DNA molecule visualization when yellow to red DNA backbone fluorescence is needed.

Materials and Methods

Materials

Twelve monomeric and homodimeric cyanine dyes, based on monomethine cyanine chromophore with oxazolo[4,5-b]pyridinium and quinoline end groups, ¹¹ are tested for their ability to be used for visualization of single DNA molecules. We refer to those dyes as dye 1 to dye 12. The dimeric cyanine dyes YOYO-1 iodide (N-7565, Invitrogen) and POPO-3 iodide (P3584, Invitrogen) were used to compare the fluorescent properties of the examined dyes. All 12 cyanine dyes were stored as 1 mM solutions in dimethylsulfoxide (DMSO) and kept at −20°C. Some physical and chemical characteristics of dyes 1 to 12, YOYO-1 iodide, and POPO-3 iodide are provided in Table 1. The emission spectra of dyes 1 to 12 are provided in the online version only (Figure S1).

Table 1.

Physical and Chemical Characteristics of Dyes 1 to 12, YOYO-1 Iodide, and POPO3-Iodide

			In DNA Presence (nm)
Dye	MW	Chemical Structure	λ abc	λ ex	λ em
1	650.29		531	533	547
2	825.31		531	533	548
3	839.33		532	533	547
4	853.36		531	533	549
5	843.36		531	533	546
6	857.39		531	533	547
7	871.42		532	534	548
8	951.54		531	533	546
9	965.57		532	534	547
10	979.60		532	534	546
11	1300.67		530	532	547
12	1408.20		532	534	547
POPO-3 iodide	1222.61		533	534	570
YOYO-1 iodide	1270.64		491	488	509

MW = molecular weight; λ_abs, λ_ex, λ_em ⁼ maximum wavelengths of absorption, excitation, and emission spectra.

Methods

DNA Staining

Total genome DNA from yeast S. cerevisiae was used. DNA probes were prepared according to standard procedure. ¹² Genome DNA was lightly digested via Bgl I (New England BioLabs, Ipswich, MA) restriction endonuclease. To dye DNA, a ratio of 1:20 dye molecules to DNA base pairs was applied. When POPO-3 iodide was used, 1:5 dye molecules to DNA base pairs was used. Dyeing reactions were carried out in 150 mM MES buffer (pH 5.5) or TE (10 mM Tris, pH 8.0 and 1 mM ethylenediamine tetraacetic acid [EDTA]) buffer (pH 7.5) for 30 minutes at room temperature in a dark place. Twenty percent β-mercaptoethanol (β-ME) was added before microscopy to reduce the photobleaching by scavenging oxygen from solution. ¹³

Microscopy

In this study, two different protocols for DNA stretching were used. The first stretching method is based on a published protocol. ¹⁴ Four microliters of the stained DNA solution was pipetted onto a glass slide (Sigma, St. Louis, MO, S4651) silanized with aminoalkylsilane. The coverslip was positioned so that it touched the edge (free end) of the slide. The angle between the coverslip and slide was gradually decreased by tilting the coverslip progressively until it came in contact with the drop of DNA solution. Holding the two slides simultaneously, they were slightly pressed together for 10 to 15 seconds. Small wafer bands were positioned on the edges of the coverslips.

The “moving dropletn” stretching method was also used. ¹⁵ Again, 4 μL of the stained DNA solution was spotted onto a glass slide silanized with aminoalkylsilane that was stably tilted at approximately 45°. The droplet was led to slowly roll down the surface, causing DNA adsorption and stretching. For maximum effectiveness, the same droplet can be reloaded at the top of the glass slide. The microscopy was carried out on an Axiovert 200M Zeiss inverted microscope by EC-Plan Neofluar 100×/1,3 oil-immersion objective. Pictures were taken by an AxioCam MRm charge-coupled device camera (Carl Zeiss, Germany). Three types of fluorescent filters were used: green: filter set 38 high fidelity (HE) (Zeiss) (excitation band pass [BP] 470/40; beamsplitter farb Teiler (color splitter [FT]) 495; emission BP 525/50); red: filter set 15 (Zeiss) (excitation BP 546/12; beamsplitter FT 580; emission long pass [LP] 590); and blue: filter set 01 (Zeiss) (excitation BP 365/12; beamsplitter FT 395; emission LP 397). Images were acquired and processed by Carl Zeiss AxioVision Rel.4.7 software. ImageJ software (National Institutes of Health, Bethesda, MD) was used for measurements of the fluorescent signal intensities.

Results

Imaging of Single DNA Molecules Pretreated with Fluorescent Dyes

Our study was focused on testing 12 (dyes 1–12) recently synthesized and characterized monomeric and homodimeric cyanine dyes, based on monomethine cyanine chromophore with oxazolo[4,5-b]pyridinium and quinoline end groups. ¹¹ They were examined for potential application in fluorescent microscopy visualization of single DNA molecules. All results were compared to those obtained after YOYO-1 iodide (N-7565) and POPO-3 iodide (P3584) DNA staining. Their similar chemical structure and physical properties (see Table 1) made both dyes proper controls for the purposes of our study.

YOYO-1 iodide is also a dimeric nucleic acid dye with high molar excitation coefficient (117,000 cm⁻¹M⁻¹) and quantum yield for DNA complexes (0.52) and possesses high sensitivity and affinity for nucleic acids. Its bright emission signal and low background fluorescence make it the primary choice for many single-molecule applications, such as staining of nucleic acids on solid support (eg, microarrays),^1,16 DNA combing,^17,18 gel or capillary electrophoresis,^7,19 flow cytometry, ²⁰ and probing biologic structure, function, and interactions.^3,17,21 As it is relatively stable in time, YOYO-1 iodide is most routinely applied when stretching of DNA on a solid surface is needed.

POPO-3 iodide was used as a control as its spectral characteristics (see Table 1) are close to those of the 12 examined dyes. When bound to nucleic acids, it exhibits excitation/emission maxima at ≈ 534/570 nm, but when unbound, it is essentially nonfluorescent. Its quantum yield for DNA complexes is measured to be 0.46. POPO-3 stain has been used to detect single DNA fragments by flow cytometry, ²² microarrays, ²³ DNA stretching between beads with optical tweezers, ²⁴ and stretching of single DNA molecules in an elongational flow. ²⁵

Comparatively, the examined dyes also intercalate into DNA/ribonucleic acid backbone. In the presence of DNA, they reveal a fluorescence intensity increase up to 370 times for monomers (dyes 1–10) and up to 1,670 times for homodimers (dyes 11 and 12). ²⁶ All 12 dyes are practically nonfluorescent alone (have a very low fluorescence quantum yield), but when bound to nucleic acids, their quantum yield rises to 0.75. ¹² Their usefulness for nucleic acid agarose gel electrophoresis is already demonstrated. ²⁶ Although their high qualities in bulk solutions have already been described, nothing was known about their behavior on a single-molecule level. Many described and commercially available DNA stains can be successfully applied in DNA electrophoresis, flow cytometry, and other bulk nucleic acid experiments (eg, acridine orange, propidiun iodide, 4,6-diamidino-2-phenylindole [DAPI]), but they are inefficient in producing a strong and sufficiently stable signal to enable single-molecule imaging. Therefore, our work focused on testing of the 12 studied cyanine dyes for applicability in single-molecule techniques.

We applied a simple and basic method for DNA stretching onto a glass slide silanized with aminoalkylsilane. The achievement of proper visualization by means of that method would be sufficient to guarantee the potential usefulness of the studied cyanine dyes for a broad range of single DNA molecule techniques.

For visualization, we used total genome DNA from the yeast S. cerevisiae. Genome DNA was lightly digested via rarely cutting restriction endonuclease Bgl I (New England BioLabs) to randomly shorten the length of the DNA fragments. As some of the cationic dyes (such as POPO-3 iodide) appear to be readily adsorbed out of aqueous solutions onto surfaces (such as glass) but are very stable once intercalated into nucleic acids, DNA was subjected to staining before stretching in a plastic tube. To dye DNA, a ratio of 1:20 dye molecules to DNA base pairs was applied. Data from YOYO-1 experiments indicate that the critical staining ratio is 0.31 ± 0.06 dye/bp. ²⁷ The significantly lower ratio we used ensured that all dye molecules are bound to DNA, allowing adequate comparison among dyes. An exception was made for dyeing with POPO-3 iodide, where a ratio of 1:5 was applied as POPO-3 in smaller ratios seemed to be much more photounstable.

Two types of buffers were used to check the influence of the pH range for the staining and stretching of prestained DNA molecules. Dyeing reactions were carried out in MES buffer (pH 5.5) and in TE buffer (pH 7.5) for 30 minutes at room temperature in a dark place. The slightly acidic 150 mM MES buffer was chosen as it is one of the most commonly used in standardized reactions for nucleic acid stretching techniques (predominantly when stretching molecules onto hydrophobic surfaces). ⁵ The ionic conditions of the reaction were close to the optimal, described by Gunter and colleagues, ²⁷ to support stable cyanine dyes binding to DNA. As we were stretching genome DNA fragments onto a hydrophilic surface, exposing ionizable amino groups (aminosilane), we decided to test another pH variation option by carrying out the reaction in TE buffer (pH 7.5).

A typical single DNA molecule technique uses the compounds called photoreductors, which reduce photobleaching by scavenging oxygen from solution. The process of photobleaching is related to the sudden decomposition of the emitter. The low signal intensity limits the correctness of exact spatial positioning of the examined stretched DNA molecule. That is why the more photostable the fluorescent dye is, the broader the range of experimental procedures it can be applied to. A characteristic reductor is β-ME. We added 20% (standard) ¹³ β-ME before microscopy.

Two simplified procedures for stretching DNA were used in our study, based on published methods.^14,15 Both methods performed with equal effectiveness. Chan and colleagues described their procedure as one of the fastest for DNA stretching that is applicable for minimal amounts of nucleic acid (10–20 ng). ¹³ For detection, only a fluorescent microscope is needed. Stained DNA solution was pipetted onto a ready-to-use glass slide (Sigma, S4651) silanized with aminoalkylsilane and slowly covered by the coverslip. Holding the two slides simultaneously, they were slightly pressed against each other for 10 to 15 seconds. The stretching force is that of the liquid flow that acts between the slide and the coverslip.

As an alternative, the moving droplet stretching method was also applied. ¹⁵ Stained DNA solution was pipetted onto a tilted (approximately 45°) glass slide silanized with aminoalkylsilane and led to slowly roll down. The stretching force is caused by the moving interface among the air, the droplet, and the substrate. The microscopy was carried out on an Axiovert 200M Zeiss inverted microscope on a 100×/1.3 oil-immersion objective and consequently documented. Three types of fluorescent filters were used (see Materials and Methods) in blue, green, and red spectra to allow emitted signal separation (Figure 1). All 12 examined cyanine dyes revealed a strong fluorescent signal of stretched DNA backbone. An exposure time of 1,000 ms was used in all cases with red and green filter sets (filter set 15 and filter set 38 HE) and 500 ms with the blue set (filter set 01). As expected, YOYO-1 iodide–stained DNA backbone is easy to visualize via the green filter (filter set 38 HE [excitation BP 470/40; beamsplitter FT 495; emission BP 525/50]), but the emitted fluorescence does not reach the red or the opposite blue zone. After analysis by means of our different spectra color filter sets (see Figure 1), a signal shift toward the red zone was visualized for the novel dyes. Dyes 1 to 12 in complex with DNA gave the strongest signal when visualized under the red filter (filter set 15 [excitation BP 546/12; beamsplitter FT 580; emission LP 590]). The signal strongly decreased under the green filter set and was missing in the blue one. As expected, POPO-3 iodide–stained and stretched DNA molecules were also better visualized under the red filter set used (filter set 15).

Figure 1.

Single DNA molecules pretreated with fluorescent dyes. Fluorescent microscopy analysis of genome DNA fragments, pretreated with the examined cyanine dyes and stretched on glass slides, silanized with aminoalkylsilane (Sigma, S4651), was carried out under three sets of filters: red: filter set 15 (Zeiss) (excitation BP 546/12; beamsplitter FT 580; emission LP 590); green: filter set 38 HE (Zeiss) (excitation BP 470/40; beamsplitter FT 495; emission BP 525/50); and blue: filter set 01 (Zeiss) (excitation BP 365/12; beamsplitter FT 395; emission LP 397). Stretched DNA molecules, prestained with YOYO-1 iodide (N-7565, Invitrogen) and POPO-3 iodide (P3584, Invitrogen), are used as controls. All reactions are carried out in MES buffer (pH 5.5).

For all studied dyes as well as YOYO-1 iodide and POPO-3 iodide, better results were obtained in MES buffer (pH 5.5). When TE buffer (pH 7.5) was used, DNA staining seemed to be performed to the same extent, but stretching of the molecules was altered (data not shown).

To summarize and quantify the results, we performed an analysis of the intensity of stretched DNA molecules emitting fluorescence using ImageJ software. The intensity of the fluorescent signal was measured for 10 stretched DNA molecules, and the corresponding background fluorescence was subtracted. The results were summarized for every dye, and average values were calculated. The results are presented as a diagram on Figure 2. Comparatively, dyes 1 and 7 to 11, when observed under a fluorescent microscope, revealed fluorescence intensity calculated above 30, which is higher than that measured for POPO-3 iodide (under the red filter set). The results of the dye 7 red filter were even measured above 60, which is higher than the green filter value of YOYO-1. The same experiment was also performed on bulk solutions (Figure S2, available in the online version only). Dyes were excited at 546 nm, and fluorescence emission was measured after 590 nm (corresponding to filter set 15). Similar to single-molecule data (see Figure 2), the highest fluorescence intensity maxima in bulk solution measurements revealed dyes 1 and 7 to 10 (Figure S3, available in the online version only). In contrast, the homodimeric cyanine dyes 11 and 12 revealed stronger fluorescence intensity when applied in DNA stretching assays than in solution (probably as a result of the stretching procedure and/or the interaction of the dye-DNA complex with the aminoalkylsilane solid surface).

Figure 2.

Intensity of the emitted fluorescent signal of stretched DNA molecules. The intensity of fluorescent signals was measured by means of ImageJ software for several stretched DNA molecules, and the corresponding background fluorescence was subtracted. The calculated summarized data for the 12 examined cyanine dyes and the control YOYO-1 iodide and POPO-3 iodide are shown as individual bars for the analysis carried under red (filter set 15; Zeiss) and green (filter set 38 HE; Zeiss) filter sets. The standard error of the mean is indicated as error bars.

We found three of the stains (dyes 7, 8, and 11) of special interest as our single-molecule results indicated strong red filter signal predominantly and very week green filter signal penetration. This makes them extremely useful for various single-molecule applications when “red zone” DNA backbone fluorescence is needed.

Photostability of Dyed DNA Molecules

To test in practice the qualities of the 12 cyanine dyes for an interval of time that is sufficient for a researcher to carry out microscopy visualization and documentation of single molecules, we examined the photostability of the dyes in comparison with YOYO-1 iodide and POPO-3 iodide. DNA genome fragments stained with all dyes were constantly illuminated for 4 minutes under the red filter set. The studied period of time was divided into 17 frames. Pictures were taken applying 1,000 ms exposure time. The results were compared to those with POPO-3 iodide and YOYO-1 iodide (under the green filter set) in the same conditions. By means of ImageJ software, the intensity of the fluorescent DNA-dye complex signal was measured for every time frame (subtracting the background fluorescence). The summarized data are shown as diagrams in Figure 3. (Representative movies are provided with the online version only.) All of the bleaching profiles of the examined dyes were approximately in the range of the one obtained with YOYO-1 iodide. In comparison with POPO-3 iodide, all of the examined dyes revealed significantly higher photostability. Dyes showing higher fluorescent light intensity values (dyes 1, 7–11) of stretched DNA molecules (see Figure 2) exhibited an elevated profile of the photostability diagram even in comparison with YOYO-1 iodide. The profile of some dyes (eg, 7 and 8) indicated an initial steep decay (similar to YOYO-1), but their fluorescence intensity values still stayed higher than those measured for YOYO-1 iodide throughout the studied interval of time. This indicates that those dyes are not only strong but also highly photostable and can be successfully applied in solid surface single-molecule techniques.

Figure 3.

Photostability of the stained DNA molecules in comparison with YOYO-1 iodide and POPO-3 iodide. DNA fragments stained with all dyes were time-lapsed for 4 minutes under the red filter set (15), applying 1,000 ms exposure time. The studied period of time was divided into 17 frames. The results were compared with those for YOYO-1 iodide (under green filter set 38 HE) and POPO-3 iodide (under red filter set 15) in the same conditions. By means of ImageJ software, the intensity of fluorescent DNA-dye complex signal was measured for every frame (subtracting the background fluorescence). On the horizontal axis of all panels, the individual time frames of the 4-minute interval are presented. The vertical axes present the fluorescence intensity.

Discussion

Our study demonstrated that 12 novel monomeric and homodimeric cyanine dyes enable successful visualization of DNA molecules distinct on fluorescent microscopy after application of a simple and time-saving staining and stretching procedure. Although monomeric dyes are described as binding more weakly to DNA than the corresponding dimeric dyes, ²⁸ our results for the monomeric dyes 1 to 10 show significant fluorescence intensity, comparable (and even higher for dyes 7 and 8) to that of the dimeric dyes 11 and 12 used. One possible explanation is that those dyes are di- and tricationic. As previously described,^29,30 polycationic dyes are significantly stronger emitters than monocationic dyes on intercalation. Additionally, pyrido-oxazole dyes are described to possess a high extinction coefficient that significantly enlarges their sensitivity.^11,26 The best results in our study were obtained for dyes 1 and 7 to 11 as the calculated intensity of the DNA backbone fluorescence was in the range of that produced by the popular amino acid dye YOYO-1 iodide. The others (dyes 2–6 and 12) exhibited in some measure a weaker but still strong signal intensity when dyed DNA molecules were observed under a fluorescent microscope. We found three of the stains (dyes 7, 8, and 11) of special interest as our results indicated a predominantly strong red filter signal and a very weak green filter signal penetration. Their fluorescence intensity was near that calculated for the YOYO-1 iodide and higher than that measured for the POPO-3 iodide. This makes them extremely useful for various single-molecule applications when yellow to red DNA backbone fluorescence is needed.

The conditions of DNA staining and stretching procedure were determined. For all 12 dyes, as well as the control staining and stretching with YOYO-1 iodide and POPO-3 iodide, MES buffer (pH 5.5) proved to be appropriate. When dyeing reactions were carried out in TE buffer (pH 7.5), DNA molecules seemed to be stained to a similar extent, but the subsequent stretching onto glass slides silanized with aminoalkylsilane was altered.

Our in situ study of photostability of the examined 12 homodimeric cyanine dyes (when equal amounts of photoreductor is used) demonstrated that all of the studied dyes are more photostable than the other yellow spectrum–emitting dye used, POPO-3 iodide. The bleaching profile diagram of all dyes was in the range of that of YOYO-1 iodide–dyed DNA molecules. Dyes 1 and 7 to 11, which produce higher values of fluorescent light intensity (see Figure 2) of stretched DNA molecules, revealed photostability diagrams (see Figure 3) that were more elevated than the others.^2–6,12 Their bleaching profile is even higher than that of YOYO-1. Those results demonstrate in practice the eligibility of dyes for prolonged microscopy work on single DNA molecules stretched on a solid surface. The diagram values of the fluorescence intensity of dyes 1 and 7 to 11 remained higher even than that of YOYO-1 until the last time frame.

Another substantial advantage of the studied dyes in comparison with POPO-3 iodide (the dye currently used in that spectral zone) is their efficiency; they were applied at a four times smaller ratio of dye molecules to DNA base pairs than POPO-3. The higher photostability and bright fluorescent signal of the studied dyes, dispersed uniformly throughout the molecule backbone, allow better distinction and longer visualization duration even when molecules are stretched on a solid surface and subjected to continuous illumination.

Currently, single-molecule analysis targets the development of applications for visualization and analysis of protein–DNA interactions.^31–34 In contrast to bulk assays, this approach can achieve more substantial information about the formation and dynamics of macromolecular complexes. As many of the Gen, Clone, and Cell Line Collections are based on green fluorescent protein (GFP) tagging of proteins, the single-molecule microscopy assays require a green filter to be used for the protein of interest. Therefore, another fluorescent color must be applied for the DNA backbone imaging. The fluorescent yellow to red dyes 1 to 12, which can be distinguished from green sources of fluorescence (such as GFP and enhanced GFP, possessing the same emission maxima as YOYO-1) by means of particular filter sets, such as filter set 15 used in our study, can be perfectly implemented in protein-DNA multichannel microscopy applications. Stretching of whole human genome ¹ and the development of sequence-targeted fluorescent labeling of combed DNA ³⁵ also indicate that in future science and medicine, more colors of fluorescent dyes will be needed to carry out various multichannel single-molecule tasks. The studied dyes undoubtedly have qualities for DNA molecule imaging, which requires the use of nucleic acid stains with a stable and strong signal to permit the execution of more sophisticated and time-consuming analysis.

Conclusion

All of the studied dyes, especially dyes 1 and 7 to 11, are not only strong but also highly photostable. Their fluorescent intensity in a DNA-bound state is similar to that of YOYO-1. In contrast to YOYO-1, their higher wavelength allows emission signal to be caught by a red filter set. Being more photostable than the currently applied red filter–penetrating dye POPO-3, they can broaden the horizon of DNA research via various solid surface single-molecule applications when yellow to red zone DNA backbone fluorescence is needed.

The following movies demonstrate the photostability of dyes 7, 8, and 11 when bound to single DNA molecules. DNA fragments stained with dyes were constantly illuminated for 4 minutes under filter set 15 (Zeiss) (excitation BP 546/12; beamsplitter FT 580; emission LP 590), applying 1,000 ms exposure time. The studied period of time was divided into 17 frames. For comparison, movies of POPO-3 iodide–stained DNA molecule (filter set 1) and YOYO-1 iodide–stained DNA molecule (filter set 38 HE; Zeiss) (excitation BP 470/40; beamsplitter FT 495; emission BP 525/50) are also shown, documented under the same conditions.