DNA oligonucleotide was synthesized and purified by Shanghai Sangon Biotechnological Co., Ltd. (Shanghai, China). T4 DNA ligase, Klenow Fragment (KF exo-), T4 Polynucleotide Kinase (PNK) and SG I (10,000× concentrated stock in H₂O) were purchased from Shanghai Sangon Biotechnological Co., Ltd. (Shanghai, China). Exonuclease I (Exo III) and Exonuclease III (Exo I) were purchased from New England Biolabs (Beijing, China). Other reagents and chemicals were of analytical grade and used as received. The reaction buffer solutions employed in this work were DNA ligase buffer (40 mM Tris-HCl, 10 mM MgCl₂, 10 mM DTT, 0.5 mM ATP, pH 8.0) and MOPS buffer (10 mM MOPS, 150 mM NaNO₃, pH 7.0). DNA sequence for dumbbell probe (DP) is as follows:

(5’-P-GCGTGGGAAATCCACGCCCCAACCCTAGGGTAGGGCGGGTTGGG-3’).

The italicized part and the underlined sequences are complementary, respectively.

DNA oligonucleotide was firstly dissolved in MOPS buffer, then pre-annealed by heating to 95˚C for 5 min and allowed to cool slowly to room temperature to form DP structure. For a typical detection experiment for DNA ligase, 5 μL of 10× DNA ligase buffer, 3 μL of 3 μM DP, 5 μL of T4 DNA ligase diluent, and 37 μL of pure water were mixed, kept at room temperature for 1 h. Then, 2 μL of 5 U/μL Exo III and 2 μL of 2.5 U/μL Exo I were added into the above mixture and incubated at 37˚C for 30 min. Subsequently, 240 µL of MOPS buffer, 6 µL of SG I (100×) were added into the above mixture and incubated at room temperature for 10 min.

Fluorescence measurements were performed on a Cary Eclipse fluorescence spectrophotometer (Agilent, USA) with an excitation wavelength (λ_ex) of 495 nm, and the fluorescence emission spectra were collected from 510 to 600 nm at room temperature. The fluorescence spectrum at λ_em = 527 nm was recorded, the PMT detector voltage was fixed at 800 V, and each measurement was carried out in a final volume of 300 μL.

A novel label-free fluorescence strategy for sensitive detection of DNA ligase was developed based on DP and DNA ligation reaction, which is presented schematically in Figure 1. As shown in Figure 1, the designed DNA strand can self-hybridize to form a dumbbell probe (DP) with a nick on its stem and a phosphate (P) at the 5' end. In the absence of DNA ligase, DP was hydrolyzed by exonuclease into single nucleotides from the nick, resulting in a weak background signal. However, the nick on DP was ligated to form a fully closed dumbbell DNA structure in the presence of DNA ligase, which prevented the hydrolysis of exonuclease. Finally, a strong fluorescence signal was observed after combining with fluorescent dye SG I. The proposed fluorescence strategy is sensitive, inexpensive and easy to operate.

Figure 2 depicts the typical fluorescence spectral responses of the DP-based strategy for detection of DNA ligase. Verification the feasibility of DP-based

strategy based on the difference in fluorescence spectra under varying conditions. As shown in Figure 2, a significant strong fluorescence signal (curve a) was observed in the both presence of DP and DNA ligase (0.01 U/μL). As expected, a lower fluorescence signal was obtained when less DNA ligase (0.001 U/μL) was present in the reaction mixture (curve b). Moreover, it was obvious that there was a low background signal in the absence of DNA ligase (curve d). Furthermore, one could get a weak signal similar to curve d when the denatured DNA ligase and DP were added in the reaction mixture. The results indicated that the proposed DP-based strategy was correct and feasible.

In this experiment, the concentrations of SG I, Exo I and Exo III have great impact on the performance of the fluorescence strategy. Those experimental conditions were investigated and optimized one by one, and the results are as follows.

Figure 3 depicts the relationship between the fluorescence signal ratio and the concentration of SG I. As shown in Figure 3, fluorescence signal ratio increased with SG I concentration up to its maximum peak then decreased gradually, the signal peak was obtained at the concentration of SG I (2×). Therefore, SG I (2×) was used as the optimal concentration of SG I and used for further experiments.

Figure 4 depicts the relationship between the fluorescence signal ratio and the concentrations of exonuclease. It can be seen from Figure 4, fluorescence signal ratio showed a trend of first increasing and then decreasing with the increase of exonuclease concentration, which indicated that excessive or too small concentration of exonuclease is detrimental to assay performance. Therefore, the optimal concentration of Exo I was chosen to be 0.1 U/μL, and the optimal concentration of Exo III was 0.2 U/μL.

The fluorescence spectra responses to DNA ligase of varying concentration were performed under the optimal experimental conditions. As shown in Figure 5, it

is observed that the fluorescence signals gradually increased as the concentrations of T4 DNA ligase varied from 0.00004 to 0.02 U/μL, which indicated the generation of more fully closed DP after ligation reaction.

Figure 6 illustrates the relationship between fluorescence intensity and ligase concentration. Notably, the fluorescence intensity of spectra response at 527 nm significantly increased with the increasing of DNA ligase concentration, but reached a plateau above the concentration of 0.01 U/μL. Fluorescence intensity showed a good linear correlation with the concentration of T4 DNA ligase in the range from 0.00004 - 0.004 U/μL. The correlation equation is F = 34901C + 13.24 (R² = 0.9969), where F represents the fluorescence intensity and C represents the concentration of T4 DNA ligase. The detection limit is estimated to be 0.00003

U/μL based on the principle of 3 times the standard deviation over the signal of the negative control. The proposed method is easy to operate and inexpensive, it may be helpful for novel drug screening.

Several irrelevant DNA modification enzymes were used as the negative control samples to evaluate the detection specificity. As shown in Figure 7, a remarkable fluorescence signal was observed in the presence of target DNA ligase (Target column), while no significant fluorescence signal was detected in response to the reaction buffer (Blank column). Under the identical conditions, similar weak fluorescence signals were obtained in response to other three enzymes. These results confirm the excellent specificity of the proposed strategy toward DNA ligase.