The exDNA degradation induced by DNase I in the presence of PAEs was first performed via agarose gel electrophoresis. As shown in Fig. 1a-e, the inhibitory effect of DNA enzymatic degradation exposed to a series of concentrations of PAEs was presented. In CK (without DNase I), all degraded DNA fragments appeared at the same position (> 2000 bp), suggesting that the undegraded DNA fragments were more than 2000 bp, which was consistent with our previous research (Qin et al. 2019). Furthermore, the brightness of DNA fragments (more than 2000 bp) was gradually strengthened (that was, the amount of undegraded DNA gradually increased), and the brightness of DNA fragments in the range of 100–2000 bp gradually became weakened (meaning that the degraded DNA amount gradually decreased) with the PAEs concentration increasing from 0 to 5.0 (DMP), 5.0 (DEP), 4.0 (DBP), 3.0 (DEHP), 3.0 (BBP) µmol·L− 1, respectively. This result indicated that PAEs at the concentration of this study could inhibit the enzymatic degradation process of DNA and enhanced the persistence of DNA. This result may arise because PAEs weakened the secondary bonds between DNA polymer molecules, that was van der Waals forces, and increased the distance and activity space between polymer molecular chains, making them less susceptible to degradation by the degrading enzyme (DNase I). The related morphologies of DNA before and after degradation influenced by DMP are shown in Fig. 2.
It was reported that DNA degradation led to the exposure of bases, causing the absorbance to increase at 260 nm (Yang et al. 2019). Therefore, the DNA absorbance increase (ΔA) at 260 nm was closely related to the DNA degradation degree. Thus, the ΔA values of DNA after degradation were also exactly recorded to confirm the dose-dependent inhibition effect of PAEs on DNA degradation (Fig. 1f-j). Interestingly, the ΔA values were gradually decreased in a nonlinear mode from 0.34 to 0.0084 with the DMP concentrations increasing from 0 to 5 µmol·L-1. For DEP, DBP, DEHP, and BBP, the ΔA values showed the same nonlinear downward trends. Thus, we again confirmed that PAEs inhibited the enzymatic degradation of DNA in a dose-dependent manner and these spectroscopic results were consistent with the above agarose gel electrophoresis results.
To confirm whether PAEs binding with DNA molecules can occupy the enzymatic reaction sites of DNase I, we added more DNase I (0.002 ~ 0.014 U·µL-1) into the condition that PAEs could inhibit DNA degradation. This was the best method to confirm whether PAEs had completely occupied the reaction sites of DNase I at the test concentrations (Fig. S2). The results illustrated that DNA degradation was restarted with the increasing addition of DNase I, indicating that PAEs did not occupy the enzymatic reaction sites of DNase I in the DNA molecular structure. This is probably because PAEs and DNase I bind more strongly to different sites on DNA–PAEs interacting primarily with the DNA bases, while DNase I interacts with the sugar-phosphate backbone (Engavale et al. 2021). Since PAE binding to DNA did not induce any structural change of DNA (Cheng et al. 2022), we assumed that the inhibition by PAEs of DNA degradation was probably due to the interaction between PAEs and DNase I. This speculation was based on one previous study reporting that synthetic melittin binding with secretory phospholipase A2 the structure of which resembles DNase I inhibited its enzymatic activity (Saini et al. 1997).
To examine the influence of PAEs on DNase I activities, we detected the enzymatic activity change of DNase I after the addition of PAEs. The results are shown in Fig. 3a. With the DMP concentrations increased from 0 to 17.5 µmol·L-1, the enzymatic activity of DNase I decreased accordingly from 1.00 U to 0.69, 0.55, 0.28, 0.20, 0.17, 0.17, 0.17 U, respectively. Similarly, the DNase I activity decreased from 1.00 U to 0.94, 0.82, 0.69, 0.62, 0.57, 0.47, 0.36 U after adding 0 ~ 15.4 µmol·L-1 DEP. The same phenomenon also appeared when DBP, DEHP, and BBP were added to the DNA degradation system because the enzymatic activity of DNase I decreased from 1.00 U to 0.40 U (DBP), 0.23 U (DEHP), 0.12 U (BBP) with the concentrations of these PAEs increased from 0 to 12.6 µmol·L-1 (DBP), 9.1 µmol·L-1 (DEHP), 11.2 µmol·L-1 (BBP), respectively. All these results suggested that PAEs inhibited the enzymatic activity of DNase I. The maximum enzymatic activity inhibition rates of PAEs on DNase I in our study were up to 83% (DMP), 64% (DEP), 60% (DBP), and 77% (DEHP), 88% (BBP), respectively (Fig. 3a). Nonetheless, the agreement between the PAEs’ inhibition of DNA degradation and our enzyme activity change induced by PAEs suggested that the above speculation was quite plausible. Therefore, this was the first work that indicated PAEs could inhibit DNA enzymatic degradation by reducing the activity of nuclease.
Amino acid residues of DNase I can generate fluorescence at certain excitation and emission wavelengths, which are highly sensitive to organic matters and lead to their fluorescence quenching (Meng et al. 2021). Since PAEs-DNase I binding is the first step for the inhibition of DNase I enzymatic activity caused by PAEs, we next studied the possibility of PAEs-DNase I binding interaction by traditional fluorescence quenching tests. With the addition of various PAEs, the maximum fluorescence intensity of DNase I at the best excitation (295 nm) and emission (355 nm) wavelength was decreased appreciably (Fig. 3c), indicating that PAEs-DNase I binding interaction led to the alterations of the environment around amino acid residues (tryptophan) and finally induced its fluorescence quenching. To further identify the mechanism of how PAEs induced the DNase I fluorescence quenching, we plotted the Stern-Volmer curves (Fig. 3c). All fitted data (Table S1) showed an obvious linear relationship to the Stern-Volmer equation (R2 > 0.9), and the quenching constant (Ksv) values of PAEs-DNase I binding were 2.3×103 (DMP), 1.5×103 (DEP), 1.8×103 (DBP), 1.6×103 (DEHP), 2.0×103 (BBP) L·mol-1, respectively. The calculated quenching rate constant (Kq) values were 1.55×1011 (DMP), 1.03×1011 (DEP), 1.20×1011 (DBP), 1.07×1011 (DEHP), 1.34×1011 (BBP) L·mol-1·s-1. All these Kq values were larger than the highest diffusion collision Kq (2.0×1010 L·mol-1·s-1) (Lange et al. 1998), indicating the coexistence of static quenching (Mahmoudpour et al. 2020). Figure 3c shows the double-log plots of PAEs-DNase I and the corresponding fitted parameters are listed in Table S2. Log (KA) values of DMP, DEP, DBP, DEHP, BBP were 7.59, 7.01, 8.00, 5.93, 7.71, respectively. All these KA values were relatively close to the noncovalent binding interaction between cyanidin-3-O-glucoside and whey protein isolate/β-lactoglobulin (Meng et al. 2021). The binding sites numbers were close to 1 (Table S2), suggesting that there was one binding site between PAEs and DNase I, which was similar to the binding interaction between small aromatic molecules and DNA (Qin et al., 2021). These results further confirmed the PAEs-DNase I binding interaction. Besides, dynamic quenching and static quenching both occurred in the binding interaction.
PAEs-DNase I binding sites were further investigated by the Fourier transform infrared spectroscopy (FTIR). The IR spectra of PAEs-DNase I complexes are illustrated in Fig. S3. The band located at 1632 cm-1 of the free DNase I spectrum was assigned to the amide I region that was often reported in the range of 1600 ~ 1700 cm-1, which corresponded primarily to C(N) = O stretching (Souillac et al. 2002). The appearance of this band unambiguously indicated the β-sheet structure. Additionally, the peaks at 1554 and 1462 cm-1 were assigned to the stretching vibration of C-N and N-H in amide II. Besides, the band at 1400 cm-1 was attributed to C-N bending, and the band at 1297 cm-1 was attributed to N-H bending in amide III (Haris and Severcan 1999). Besides, the IR bands at 1162, 1081, and 930 cm-1 were presented as unsaturated bands in DNase I structure (Haris and Severcan 1999; Kang et al. 2010). After binding with PAEs, the band at 1297 cm-1 was moved to 1271 cm-1 (DMP), 1273 cm-1 (DEP), 1282 cm-1 (DBP), 1271 cm-1 (DEHP), 1281 cm-1 (BBP), respectively. This obvious IR band shift indicated the N-H bending change in amide III. By analyzing the molecular structure of DNase I, we believed that the main PAEs-DNase I binding sites were its amino groups.
Molecular dynamics simulations results revealed the dynamic process of PAEs-DNase I binding interaction (Fig. 4 and Table S3) (Kawamura and Ohta 2022). For DMP, its positions, and binding amino acids were changed from 0 ps (Arginine, Asparagine, Glutamate, Histidine, Proline), to 29798 ps (Alanine, Asparagine, Glutamine, Isoleucine), and 98122 ps (Asparagine, Threonine) in a relatively stable state (Fig. 4a). However, for DEP, DBP, DEHP, and BBP, the binding amino acids were various (Fig. S4). The appearance of the blue-green dots indicated by the yellow arrows suggested that the driving forces for PAEs-DNase I binding were van der Waals forces and hydrogen bonds. This result was consistent with the binding force analysis results of PAEs binding with human serum albumin (HSA) (Lv et al. 2021). Additionally, 1 ~ 3 hydrogen bonds were formed during the binding process. It was similar to PAEs-DNA binding interaction which formed 1 ~ 4 hydrogen bonds (Cheng et al. 2022). In addition, the Root-Mean-Square-Deviation (RMSD) values of PAEs-DNase I binding varied greatly from 5.00×10− 7 to 6.65 nm (Table S4), and the average RMSD values were in the range of 2.34 ~ 3.05 nm, which were slightly smaller than the RMSD values of PAEs-DNA binding. Since the smaller RMSD values mean more stable complex formation, this result indicated that PAEs were more likely binding with DNase I than DNA, which could also be confirmed by the comparison of KA values (Cheng et al. 2022). In addition, the relative stability could only be achieved when hydrogen bonds were formed (Fig. 4b). Therefore, PAEs-DNase I binding interaction was a dynamic and unstable process.
Because the secondary structure change of endonuclease can have a close relationship with its biological activity (Okabe et al. 1982), the residue ellipticity change at specific peaks can reflect the helical structure deformation of DNase I (Qin et al. 2021; Okabe et al. 1982; 1993). The circular dichroism (CD) spectra of DNase I in the specific range from 185 to 250 nm in the presence of PAEs at various concentrations are shown in Fig. 3b. PAEs-DNase I binding caused the characteristic negative CD bands of DNase I near 215 nm to decrease, revealing that PAEs deformed the ordered DNase I structure by binding interaction. More importantly, the positive peaks at 199 nm were transformed to negative peaks at 197 nm (DMP), 199 nm (DEP), 201 nm (DBP), 200 nm (DEHP), and 197 nm (BBP), respectively. This characteristic peak change was probably due to the structural deformation of DNase I. However, the observed CD changes do not appear to be greatly influenced by the increasing PAEs concentrations. Then, we analyzed the CD change in the 250–300 nm range which could be more useful as it would indicate how phenylalanine (257 nm), tyrosine (278 nm), and tryptophan (295 nm) residues are affected by PAEs (Fig. 3b) (Sreerama et al. 1999). The characteristic peaks of DNase I at 257 nm, 278 nm, and 295 nm were changed by PAEs in intensity and position, indicating the binding of PAEs with phenylalanine, tyrosine, tryptophan residues changed the molecular structure of DNase I. The observed perturbations appeared to depend on the PAEs concentrations. Thus, we concluded that PAEs binding with DNase I induced structural deformation of DNase I which was the main mechanism for the inhibition of DNA degradation.In summary, we found the inhibition effect of coexistent PAEs on the degradation of extracellular DNA in a dose-dependent manner. This inhibition effect was mainly due to the binding interaction between PAEs and DNase I at amino groups via van der Waals forces and hydrogen bonds. Since PAEs-DNase I binding interaction could induce structural deformation of DNase I, PAEs decreased the activity of DNase I. To the best of our knowledge, this is the first report indicating the molecular ecological effects of PAEs pollution. We truly believe that coexistent PAEs can influence the persistence and the dissemination of specific genes such as ARGs.