Effect of Flow Rate and Ionic Strength on the Stabilities of YOYO-1 and YO-PRO-1 Intercalated in DNA Molecules

Abstract

Single-molecule DNA studies have improved our understanding of DNAs’ structure and their interactions with other molecules. A variety of DNA labeling dyes are available for single-molecule studies, among which the bis-intercalating dye YOYO-1 and mono-intercalating dye YO-PRO-1 are widely used. They have an extraordinarily strong affinity towards DNA and are bright with a high quantum yield (> 0.5) when bound to DNAs. However, it is still not clear how these dyes behave in DNA molecules under higher ionic strength and strong buffer flow. Here, we have studied the effect of ionic strength and flow rate of buffer on their binding in single DNA molecules. The larger the flow rate and the higher the ionic strength, the faster the intercalated dyes are washed away from the DNAs. In the buffer with 1 M ionic strength, YOYO-1, and YO-PRO-1 are mostly washed away from DNA within 2 minutes of moderate buffer flow.

Graphical Abstract

Introduction:

Single-molecule techniques have allowed us to study many molecular interactions, which have both advantages and challenges to traditional methods. There are many techniques to study single-molecule DNA interactions.^1–4 For the fluorescence microscopies and spectroscopies, it is necessary to use fluorescent dye labeling to visualize the DNA molecules.

The class of small molecules which can intercalate reversibly in between the base pairs of DNA are called DNA intercalators.⁵ A group of cyanine dyes including YOYO, POPO, and TOTO dyes are useful to stain the DNAs because they have an extraordinarily strong binding affinity towards the double strands.⁶ When binding to the double-stranded DNA, these cyanine dyes show great enhancement in the fluorescent signal (~1000×),^6,7 whose mechanism is under debate.⁸ Thus, these cyanine dyes are especially useful for single-molecule as well as bulk studies of double-stranded DNA.⁹

Because mono intercalator YO-PRO-1 and its dimer bis intercalator YOYO-1 are widely used in double-stranded DNA staining, it is significant to know their stabilities of intercalation in the DNAs. In the literature, there have been many studies on the kinetics and mechanism of the interaction between DNA and DNA intercalators.^10–16 YOYO-1 binds DNA through bis-intercalation where two of its chromophore units stack in the base pair of DNA.¹⁷ In the literature, the binding affinity of YOYO-1 has been shown to depend on the ionic strength.^13,18,19 YOYO-1 is more ionic strength dependent than YO-PRO-1. It is suggested that since YOYO-1 has more (+4) positive charges as compared to the mono intercalator YO-PRO-1 (+2), their electrostatic interaction with the negatively charged backbone of the DNA is more sensitive to the ionic strength.¹⁹ At the salt concentration above 0.25 M, the fluorescence intensity of YOYO-1-DNA complex drops with increased salt concentration^18,14 suggesting dissociation of YOYO-1 from DNA molecules over high ionic strength.²⁰

In many DNA measurements, the desorbed dyes will reabsorb to the DNA making it difficult to analyze the desorption data from these measurements. Washing away the desorbed dyes in a flow cell is a good option to reduce the effect of this problem and flow cells have been widely used for DNA measurements in the literature.^21–25 However, the influence of the flow-induced sheering force on the DNA staining stability is largely unknown. In this report, we used a platform with immobilized DNAs in flow cells to study the effect of ionic strength and flow rate of buffer solutions on the staining stability of YOYO-1/YO-PRO-1 labeled DNAs. Briefly, DNA molecules are immobilized on a surface functionalized with polyethylene glycol (PEG) using biotin and neutravidin interaction. They are stained with YOYO-1 and YO-PRO-1 to visualize using a fluorescent microscope. These DNA molecules are washed at different ionic strength buffers and flow rates to systematically check their stability of fluorescence signal under these conditions. This platform will wash away the desorbed dyes and minimize the re-adsorption of them to the DNA molecules thus providing useful information about the desorption step of the dyes from the DNAs.

Materials and methods:

DNA, intercalators, and buffers:

Bacteriophage lambda DNA from New England Biolabs was functionalized with biotin at both 5’ recessive ends following the standard protocol.^26,27 YOYO-1 and YO-PRO-1 were obtained from the Invitrogen and stored at −20°C before use. All experiments were done in a buffer of 0.01M HEPES, pH 7.4 (Acros Organics) including NaCl (Sigma-Aldrich) at the concentration specified when used. All water used was from a Barnstead E-Pure ultrapure water purification system with a resistivity of 18 MΩ cm⁻¹.

Surface functionalization:

Cleaning steps:

Glass coverslips (25×25 mm, thickness #1, Fisher Scientific) were cleaned by sonicating in 1% detergent (Liquinox) for 10–15 minutes followed by washing with ultrapure (18 MΩ) water several times. Afterward, glass coverslips were base piranha cleaned in a mixture of 1:1:1 (v/v/v) of hydrogen peroxide, ammonium hydroxide, and water for 45 minutes at 70–75°C. These glass coverslips were washed with water and dried with N₂ gas.

Coverslip surface functionalization:

Cleaned glass coverslips were further amine functionalized by immersing in a solution of 100:5:3 (v/v/v) methanol, glacial acetic acid, and (3-Aminopropyl) triethoxysilane (APTES, Sigma Aldrich) for 30 minutes at room temperature (Figure 1). Amine functionalized glass coverslips were cleaned with HPLC grade methanol (fisher) and were further functionalized/passivated by using the reaction mixture of 10000:1 molar ratio of polyethylene glycol (PEG)-Succinimidyl Valerate (SVA) (MW 5,000, Laysan Bio Inc.) and Biotin-PEG-SVA (MW 5,000, Laysan Bio Inc.) prepared in 0.1 M NaHCO₃ buffer (pH, 7–8).

Figure 1.

Coverslips surface functionalization process.

Preparation of flow/microfluidic channel:

Microfluidic flow cells were prepared from polydimethylsiloxane (PDMS) blocks, double-sided tapes (Adhesives Research, Inc.), and the functionalized glass coverslip. PDMS was prepared by curing the mixture of 10:1 (w/w) Sylgard 184 silicon elastomer base (Dow Corning) and curing agent. Cured PDMS were cut into smaller sizes as glass coverslips. Dispensing tips (CML) which were connected to the plastic tubing (Tygon) were inserted through the PDMS blocks. Double-sided tape cut with multiple rectangle grooves was used to attach the glass coverslips on the PDMS block (supporting information, SI Figure S1). The dimension of the flow channel was about (11.7 ± 0.1) × (1.09 ± 0.01) × (0.076 ± 0.005) mm (average length, width, and height measured). See SI Table S1a for the estimated errors. The speed of the flow in the channel is thus estimated at ~0.020 m/s for every flow rate of 100 μL/min (SI Table S1b). The sheer flow near the surface is not estimated but should be smaller than this rate.

DNA immobilization and dye labeling:

In the flow cells, two-end biotin-functionalized DNA molecules are stretched using buffer flow and attached to the biotin-functionalized surface with neutravidin and biotin interaction (Figure 2a). This method immobilizes the DNAs with their middle parts not attached to the surface. The solutions are pumped using an automated syringe pump (New Era, NE-1000). DNA molecules are either stained with YOYO-1 at 1: 5 (dye: base pair) or YO-PRO-1 at 1: 5 (dye: base pair) ratio. Two approaches are used to label the DNA. One is mixing the DNA and dye > 0.5 hour before each experiment. The other is mixing the DNA and dye and incubating at 50°C for > 3 hours before each experiment to make the DNA labeling more homogeneous.²⁸ The incubation of dye with DNA at this condition improves the DNA labeling and does not affect the stability of the DNA and dye.¹⁸ Before the DNA is immobilized, the flow channel was washed with buffer solution (0.01M HEPES, 0.1M NaCl, pH 7.4) at a flow rate of 0.04 m/s for 2 minutes. Then 200 μg/mL neutravidin was flowed at 0.01 m/s for 4 minutes. Afterward, unbound neutravidin was washed with buffer at 0.04 m/s for 2 minutes. Dye-labeled DNA was flowed at 0.01 m/s for 4 minutes followed by buffer at 0.04 m/s for 2 minutes. The DNA molecules are visualized at room temperature (~22 ± 1°C) with a home-built fluorescence microscope reported before (SI Figure S2)²⁹ under total internal reflection fluorescence (TIRF) mode (Fig. 2b). Blue laser (wavelength 473 nm) with a power density of 0.32 ± 0.05 W/cm² was used as the excitation light. Videos of length 1000 to 4000 frames with the camera (Andor iXon 897) EM gain 200× and exposure time 125 milliseconds were taken for the dye washing experiments.

Figure 2.

(a) Scheme of DNA immobilization on the functionalized surface and washing intercalator dyes from DNA molecules (b) an example fluorescent image of surface-immobilized DNA molecules.

Data analysis:

Data analysis involves multiple steps. In the first step raw data, videos of 1000 to 4000 frames were collected using a fluorescence microscope and were converted into MATLAB files using software Andor SOLIS (64-bit, Oxford Instruments). Each MATLAB file then was background corrected and used to generate the images for further analysis (Figure 3a). From our observations and the example image, most DNA has a similar length with a relatively narrow distribution of a few μm except for those pre-broken ones that are much shorter. In the second step, individual two-end immobilized DNA molecules with lengths 7μm or longer were manually picked for further analysis (Figure 3b). The third step generates the photon counts decay profile over time for each picked DNA normalized to the length of the DNA in the unit photocounts per nm length of DNA (Figure 3c). Because we normalize the signal with the length of each DNA, the difference in physical (bp) length is less of a concern. But stretching degree contributes to the error. In this report, we chose those DNAs that are longer than 7 μm for decay analysis and average >20 DNA molecules for each data point to reduce the effect of these errors. The final step fits the decay profile of each DNA using the double exponential decay function to estimate the decay rate over time (Figure 3d). Home-written MATLAB codes were used to analyze each video.

Figure 3.

Process of data analysis. (a) Fluorescence image (b) DNA pick for further analysis, red and green markers are used to represent the different DNA molecules and the area to integrate the total fluorescence signal. (c) Decay profile for each DNA molecule. (d) The decay profile of each DNA is fitted with a double-exponential decay function.

Results and Discussion:

As shown in Figure 4 DNA molecules have been successfully immobilized on the PEG-modified coverslip surface and most of them survive in flowing buffer solutions. We observed that over time the fluorescence intensities of the DNA molecules decrease under the flow and the DNAs are invisible after 2 minutes of flow with buffer containing 0.01 M HEPES and 1 M NaCl (pH, 7.4) (Figure 4b). The DNAs are confirmed to stay in the same location without moving during the washing by restrained with flowing 100 μL buffer containing 100 nM dye molecules through the channel (Figure 4c). The re-stained DNAs have different brightness from the original stained ones because the in-situ staining is dynamic, and some dyes are photobleached. This decrease is also confirmed to be irrelevant with photobleaching or auto-desorption because when imaged with no flow in the same buffer solution, DNAs are observed for 5 minutes under the same continuous illumination (Figure 4d, 4e). The washing effect is confirmed to be an additional fluorescence decay pathway than photobleaching in the control measurements with the laser being turned ON and OFF during the measurements (SI Figure S3). These observations confirm that the YOYO-1 is washed away by the flowing buffer and the binding of YOYO-1 to DNA is reversible. We hypothesize that the speed of dyes being washed away is related to buffer conditions such as the ionic strength, the flow rate, the temperature, and the pH of the buffer. It is important to obtain quantitative values of the speed and calculate how much YOYO-1 should be added to maintain the fluorescence intensities of the DNAs. Special attention will be given to comparing the bidental (YOYO-1) and mono-dental (YO-PRO-1) intercalation dyes that have different binding strengths to DNAs. We will maintain the pH to 7.4 and room temperature ~22°C for this study.

Figure 4.

DNA and YOYO-1 washing versus control. (a) DNA YOYO-1 before and (b) after washing with a buffer of ionic strength of 1.1 M for 2 minutes. (c) Same location of DNA after flowing buffer solution containing 100 nM YOYO-1 for 2 minutes. (d) DNA before and (e) DNA after 5 minutes under no flow.

The dissociation reaction is measured in this report by minimizing the re-adsorption reaction with a continuously flowing buffer.

$$DNA\_dye+\mathit{\text{flow}}\stackrel{k}{\to }DNA+dye$$ (1)

The auto-dissociation rate constant (K_d = 1/K_a) without flowing buffer is expected to be low, reported in the literature to be 0.33×10⁻⁵ M in 0.2M NaCl in Tris buffer at pH 7.5 for YO-PRO-1.³⁰ In case YOYO-1 the auto-dissociation rate constant (K_d = 1/K_a) value is 3.9×10⁻⁶ M at 0.15M NaCl in Tris buffer at pH 7.5.¹² The strength of the binding will be adjusted by ionic strength and tuning the flow rate will change the shearing force that pulling them apart.

The average YOYO-1 and YO-PRO-1 washing lifetime of the individual DNA was calculated. Bis-intercalating dye YOYO-1 is known to bind very strongly to the DNA with a binding constant value (K_a = 10⁸–10⁹ M⁻¹), and mono-intercalating dye YO-PRO-1 has a lower binding affinity (K_a = 10⁶ M⁻¹) reported in the literature,³¹ both remain in the DNA for an extended period which helps to visualize DNA molecules under the fluorescence microscope. Low laser intensity was kept constant at 0.32 ± 0.05 W/cm², so there is a low photobleaching effect during the washing experiments and no photocleavage has been observed under this laser power.

The rate constant of photobleaching k_pb was measured from the fluorescent decay of the YOYO-1 and YO-PRO-1 labeled DNA molecules over time under a buffer flow rate of 0 m/s. The average photobleaching lifetime is fitted to be 1249 s for YOYO-1 and 1210 s for YO-PRO-1 (SI Table S2) using a single exponential decay function:³²

$$I\left(t\right)=A{e}^{-t/\tau }$$ (2a)

and the photobleaching rate constant is k_pb = 1/τ.

Fluorescence decay curves for DNA and YOYO-1 under different flow rates were fitted with a single exponential function (Eq. 2b) (SI Table S3a and S3b).

$$I\left(t\right)=A{e}^{-t/{\tau }_1}$$ (2b)

where A is the normalized pre-exponential factor, t is time, and τ₁ is the decay lifetime. The washing rate constant is k₁ = 1/ τ_1. The effective rate constant is calculated by using the equation because washing and bleaching are two parallel reactions:

$$k_{eff}=k_1-k_{\text{pb}}$$ (3)

All other DNA-dye-washing fluorescence signal decay data can be fitted with a double exponential decay function (Eq. 4)

$$I\left(t\right)=A_1{e}^{-t/{\tau }_1}+A_2{e}^{-t/{\tau }_2}$$ (4)

Where, As are normalized pre-exponential factors, t is time, and τs are decay lifetimes. During the data fitting, τ₂ is confined to near the photobleaching lifetime with ~10% allowed variation during the data fitting, and the photobleaching rate k₂ = 1/τ₂. The assumption is that there are two different types of dyes in the DNA-dye complex, washable and “nonwashable” (relatively speaking respecting the photobleaching lifetime), represented by fluorescence decay lifetime τ₁ and τ₂, respectively.

The rate constant of dye washing is calculated by using the equation.

$$k_{eff}=k_1-k_2$$ (5)

Where the fluorescence decay rate constant of washable dyes k₁ = 1/τ₁, and the fraction of washable dyes is A₁/(A₁+A₂).

The effect of flow rate on DNA-YOYO-1 and DNA-YO-PRO-1 washing:

The average intensity decay profile of each DNA over time is measured and plotted against the time to see the effect of buffer washing rate on the decay of the fluorescence signals of YOYO-1 and YO-PRO-1 on the DNA molecules (Figure 5). Here, we keep the ionic strength fixed by maintaining the buffer 0.01 M HEPES and 0.1 M NaCl and pH = 7.4.

Figure 5.

Effect of the flow rate of buffer on YOYO-1 and YO-PRO-1 washing from DNA molecules (a) DNA and YOYO-1 washing (b) washing of DNA and YOYO-1 annealed for 3 hours at 50°C. (c) DNA and YOPRO-1 washing (d) washing of DNA and YOPRO-1 annealed for 3 hours at 50°C. The buffer is maintained the same for all experiments 0.01 M HEPES and 0.1 M NaCl and pH = 7.4., F1, F2, F3, F4, F5, and F6 represent the decay profile for flow rates 0, 0.01, 0.02, 0.04, 0.06, 0.08 m/s respectively.

The effect of flow rate on washing was observed by changing the flow rate of the buffer in the flow channel at fixed ionic strength ~0.1 M. The buffer was flowed at 0, 0.01, 0.02, 0.04, 0.06, and 0.08 m/s. Then live videos were taken while washing at the flow rates to monitor the washing events from each DNA molecule. The fluorescence intensity of YOYO-1 labeled DNA molecules was fitted with a single exponential function (Equation 2) while data for YO-PRO-1 labeled DNA molecules were fitted with a double exponential function (Equation 3) to get the averages decay parameters (SI, Table S3a, S3b, S4a, and S4b). These rate constants and the effect of annealing will be further analyzed later.

Qualitatively, the results show that with an increased flow rate of buffer, the washing of dye from DNA molecules also increased but overall, there is a much weaker dependence on the flow rate of buffer solution for YOYO-1 compared to YO-PRO-1 in the chosen flow rate region. At 0.1 M ionic strength and various flow rates, the fluorescence decay rates of YOYO-1 in DNA are very close to just photobleaching (Figure 5a and 5b). These results suggest that YOYO-1 is strongly bound to DNA and is difficult to wash away at moderate ionic strength. The fluorescence decay rates of YO-PRO-1 in DNA significantly increase over the increase of flow rate suggesting its binding is much weaker than YOYO-1 and washable under mild flow rates (Figure 5c and 5d).

The effect of ionic strength on DNA-YOYO-1 and DNA-YO-PRO-1 washing:

More extreme conditions can be achieved by increasing the ionic strength of the buffer solutions. When increasing the ionic strength, the interaction between the DNA and YOYO-1 or DNA and YO-PRO-1 is weakened, and dissociation becomes more significant for both (Figure 6). Since dissociation only depends on the stability of the complex, it is likely that high ionic strength either weakens the electrostatic attraction between the dye and the DNA or weakens the base pair stacking strength that holds the dyes in between.

Figure 6.

Effect of ionic strength of buffer on the washing of YOYO-1 and YO-PRO-1. (a) DNA and YOYO-1 washing (b) washing of DNA and YOYO-1 annealed at 50°C for 3 hours. (c) DNA and YOPRO-1 washing (d) washing of DNA and YO-PRO-1 annealed at 50°C for 3 hours. The flow rate is fixed at 0.02 m/s. C1, C2, C3, C4, and C5 represent the decay profile for each sample in presence of 0.1, 0.2, 0.4, 0.5, and 1 M NaCl in 0.01M HEPES buffer (pH = 7.4).

To study the effect of ionic strength on the washing of YOYO-1 and YO-PRO-1 from DNA molecules. The washing buffer solutions of different ionic strength was prepared. The ionic strength of the washing buffer solution was varied by changing the concentration of the NaCl in 0.01 M HEPES. The washing was carried out with 0.1, 0.2, 0.4, 0.5, and 1 M NaCl in 0.01M HEPES buffer (pH 7.4) respectively at a fixed flow rate of 0.02 m/s. Then, live videos were taken during each washing to monitor the washing effect of each DNA. The average intensity decay profile of each DNA over time is measured and plotted against the time to see the washing rate (Figure 4a, 4b). The average decay profile shows that with increased ionic strength the washing becomes faster. Then the normalized fluorescence intensity is fitted with a double exponential function to get the average decay parameters (SI, Table S5a, S5b, S6a, and S6b). From the average decay lifetime values the rate constant for each process is calculated using Equation 4.

Summary of rate constants:

Using a buffer with higher ionic strength than 0.1 M, YOYO-1 is easier to be washed away (Figure 6). YO-PRO-1 is significantly easier to be washed than YOYO-1 (Figure 6). We assumed that at higher ionic strength the DNA strand becomes more flexible as the strength of the hydrogen bond becomes weaker resulting in weak binding between DNA and intercalating dye. This makes it easy to release dye molecules from their positions in the DNA strands during washing. Also, we confirmed that this DNA-dye washing process is reversible, which means DNA is intact and can be visualized again after flowing extra dye (SI, Video S1).

The photobleaching rate for samples under buffer flow at various conditions is shown in SI Figure S4. We found that fixing τ₂ around the photobleaching lifetime τ_pb can fit the curves similarly well compared to letting τ₂ free. Thus, we set an initial guess of τ₂ = τ_pb with a small boundary (~10%) and allow free variations on the other parameters to fit the fluorescence decay curves of the dyes under different conditions (details in SI section S7). We hypothesized that part of the intercalated dyes can be washable with a lifetime shorter than the photobleaching lifetime, and part of the dyes are “nonwashable” with a lifetime much longer than the photobleaching lifetime. The fractions of the washable dyes are shown in SI Figure S5 and the dependence of the decay rates of the washable dyes on the buffer conditions is shown in Figure 7. We can see a trend that the fraction of the washable dyes do not change significantly over annealing and slightly increase under more extreme washing conditions at high flow rates and high ionic strength (Figure S5). This observation suggests that the stability of the dyes in DNA may be location specific.

Figure 7.

The rate constant measurement under different conditions. Correlations between the rate constants and flow rate or ionic strength were fit with the linear equation y= ax empirically. Rate constants of (a) YOYO-1-DNA, (b) annealed YOYO-1-DNA, (c) YO-PRO-1-DNA, and (d) annealed YO-PRO-1-DNA washed by buffer with fixed ionic strength (0.1 M) at different flow rates. Rate constants of (e) YOYO-1-DNA, (f) annealed YOYO-1-DNA, (g) YO-PRO-1-DNA, and (h) annealed YO-PRO-1-DNA washed by buffer with different ionic strengths at a fixed flow rate (0.02 m/s).

YOYO-1 is more stable than YO-PRO-1 in DNA (Figure 7), which has been expected because YOYO-1 has two arms to insert between DNA bases while YO-PRO-1 only has one, and YOYO-1 is +4 charged while YO-PRO-1 is +2 charged. In Figure 7, the smaller the rate constant, the more stable the complex. Comparing YOYO-1 and YO-PRO-1, YOYO-1 is washed away ~5 times slower than YO-PRO-1 in the same buffer (0.01 M HEPES, 0.1 M NaCl, pH = 7.4) at a different rate. YOYO-1 is ~2 times slower to be washed than YO-PRO-1 under the same flow rate (0.02 m/s) and different ionic strength. This slower rate indicates that YOYO-1 bind to DNA stronger than YO-PRO-1.

The effect of annealing on the dissociation rate of YOYO-1 and YO-PRO-1 is observed (Figure 7). Annealing slightly increases the stability of YOYO-1 in DNA. This effect is more clearly observed for YOYO-1 at higher ionic strength when the slope is reduced ~1/2 for annealed samples than the non-annealed samples. But no effect on the stability of the intercalated YO-PRO-1 in DNA is observed, although annealing may have increased the ratio between the intercalated YO-PRO-1 and non-specifically absorbed YO-PRO-1 on DNA. From these results, we conclude that annealing has increased the ratio between the two-arm intercalation of YOYO-1 and one-arm intercalation. No other significant effects on the stability of the binding are observed, for example, optimizing the distribution of dyes in the DNA sequences, the binding geometries, or the DNA structures. If these effects are significant, we should have observed significant changes in the stability of YO-PRO-1.

Empirically in the region of our measurements, the dissociation rate increases with the increase in flow rate and the increase of ionic strength. The dependence of rate constant k in unit s⁻¹ on flow rate and ionic strength is empirically formulated for non-annealed samples.

$$k\approx 0.02F+0.1I-0.01$$ (6)

for YOYO-1 and

$$k\approx 0.5F+0.1I-0.01$$ (7)

for YO-PRO-1, where F is the flow rate in unit m/s, and I is the ionic strength in unit M.

Conclusion:

In this report, we have studied the effect of ionic strength and flow rate of the buffer on the stability of YOYO-1 and YO-PRO-1 intercalated in DNA molecules. Briefly, we immobilized dye-stained DNA on the PEGylated inner surface of flow cells using a two-end immobilization method such that the middle parts of the DNAs are relatively free from the surface. Then buffers with different ionic strengths are flowed to wash the dyes at various flow rates and the dye washing rates are measured. As expected, bidental YOYO-1 is about five times more stable than monodentate YO-PRO-1 when intercalated in DNAs. Annealing at relatively mild conditions at 50°C significantly increases the ratio between two-arm intercalated and one-arm intercalated YOYO-1 and do not affect the stability of YO-PRO-1. We also confirmed that this process of washing is reversible, and DNAs remain intact during the experiments and can be visualized again by flowing extra intercalating dye.

Empirical equations for the dissociation rate constants under different buffer and flow conditions are derived from the measurements. The washing rate constant increases with the increase of flow rate and/or increase of ionic strength. The larger the flow rate, the larger the sheering force, and the larger the ionic strength, the weaker the binding strength. Thus, both can significantly affect the dissociation rate constants between the dyes and the DNAs. These measurements may help in designing stronger binders or calculating the amount of additional dyes added to the buffer to maintain stable staining for fluorescence imaging under relatively harsh conditions of high ionic strength and flow rate.