pharmaceutics Article
Importance of Binding Affinity for the Activity of a Metallodendritic Chemical Nuclease Yi-Hsuan Tang 1 , Sodio C. N. Hsu 1 , Po-Yu Chen 1,2,3 , Si-ting Liou 1 , Hui-Ting Chen 4 , Carol Hsin-Yi Wu 5 and Chai-Lin Kao 1,2,3, * 1
2 3 4 5
*
Department of Medicinal and Applied Chemistry, Kaohsiung Medical University, Shih-Chuan 1st Road 100, Kaohsiung 80708, Taiwan;
[email protected] (Y.-H.T.);
[email protected] (S.C.N.H.);
[email protected] (P.-Y.C.);
[email protected] (S.-t.L.) Department of Medical Research, Kaohsiung Medical University Hospital, Kaohsiung 80708, Taiwan Department of Chemistry, National Sun Yat-sen University, 70 Lienhai Rd., Kaohsiung 80424, Taiwan Department of Fragrance and Cosmetic Science, Kaohsiung Medical University, Shih-Chuan 1st Road 100, Kaohsiung 80708, Taiwan;
[email protected] Division of Cellular and Immune Therapy, Department of Medical Research, Kaohsiung Medical University Hospital, Kaohsiung 80708, Taiwan;
[email protected] Correspondence:
[email protected]; Tel.: +886-7-3121101 (ext. 2620)
Received: 20 October 2018; Accepted: 22 November 2018; Published: 3 December 2018
Abstract: A family of bis(2-pyridyl)amino-modified poly(amidoamine) dendrimer Cu complexes was prepared, and their chemical nuclease activities and binding affinity (Kb ) levels for DNA plasmid were investigated. The Kb values of the G2 to G6 apodendrimers for DNA plasmid were found to be 7.4, 23, 48, 70, and 280 µM−1 , respectively, using ethidium bromide (EtBr) displacement experiments. The chemical nuclease activities of the corresponding complexes were determined by gel electrophoresis, and a clear positive dendritic effect was observed. Further analysis indicated a linear correlation between the Kb values of the G2 to G5 apodendrimers and the nuclease activity of the corresponding complexes. This observation indicated the importance of substrate binding affinity for macromolecular nuclease activity. In addition, an experiment using 30 -(p-hydroxyphenyl) fluorescein suggested that hydroxyl radicals formed under the tested conditions. Subsequently performed inhibition studies indicated that the hydroxyl radical was the active species responsible for the plasmid cleavage. Keywords: macromolecule nuclease; dendrimers; substrate binding affinity; Cu complexes
1. Introduction Metallodendrimers have attracted considerable interest in the past few decades because they have the advantages of organic materials while also possessing the properties of metallic ions [1–3]. In contrast to small-molecule complexes, metallodendrimers provide unique complexation environments and contain many coordination sites as well as a protective environment for the metallic ions. As a result, metallodendrimers have characteristics not displayed by small-molecule complexes and have been used in various applications [4–9]. One representative application of metallodendrimers has been their use as mimics of natural metallomacromolecules, such as metalloenzymes. Metallonucleases are metalloenzymes that contribute to the regulation of the cell cycle by participating in the editing of nucleic acids [10]. In such nucleases, catalytic metallic ions have been shown to function by either of two working mechanisms: hydrolytic or oxidative cleavage. For oxidative nucleases, the coordination environment [11], molecular charge [12], and size of the ligands [13] have been explored, and determined to be vital factors in this catalytic
Pharmaceutics 2018, 10, 258; doi:10.3390/pharmaceutics10040258
www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2018, 10, 258
2 of 9
process [14]. Various cofactors, such as H2 O2 , are usually necessary to break down nucleic acid chains: these cofactors generate reactive oxygen species during oxidative cleavage by the nuclease [15]. By contrast, hydrolytic nucleases operate using a catalytic mechanism involving assistance by a Lewis acid [16,17]. Although several types of metal ions have been shown to form metallonucleases [18,19], copper ions can act as either Lewis acids or oxidizing agents in nucleases. This unique character makes copper very attractive for the study of both natural and artificial nucleases [20]. In addition to proteins, numerous synthetic molecules, known as chemical nucleases, are capable of excising polynucleic acid backbones, and have been used for biomedical investigations and clinical applications [21,22]. These synthetic analogues of natural metalloenzymes demonstrate programmable nuclease activities and tolerate a wide range of conditions, features necessary for various investigations and applications. Various ligands, therefore, have been developed and characterized [23]. In contrast to aliphatic ligands, aromatic heterocycles can engage in π–π interactions with nucleic acids, and such interactions have been shown to improve the efficiency of the nuclease [24]. Of the various complexes of aromatic heterocycles with metal ions, aminopyridine metal complexes are particularly efficient catalysts due to their structural flexibility [25]. Macromolecular nucleases have been shown to provide advantages over their small-molecule analogues [26]. However, relatively few investigations of macromolecular nucleases have been carried out, due to the tedious nature of preparing monodisperse macromolecules and the difficulty in characterizing their active centers. Both issues have in particular hindered detailed investigations of macromolecular nucleases and their potential biomedical applications. Moreover, while the substrate binding affinities of small-molecule nucleases have been investigated, and reported to be a critical factor for the activity levels of these nucleases [27], the relationship between the activity levels of macromolecular nucleases and their binding affinities for nucleic acids has not been characterized [17] In contrast to other macromolecules, dendrimers are suitable candidates as nucleases because they are monodisperse and well characterized. These characteristic properties endow dendrimers with a well-controlled topology and make them very good models for investigation. Herein, we report our research results on the use of metallodendrimers as chemical nucleases. G2–G6 poly(amidoamine) (PAMAM) dendrimers with peripheral bis(2-pyridyl)amino groups were synthesized and then coordinated with copper (Cu) ions to obtain nuclease complexes. These Cu complexes were tested for their nuclease activities, and the results were analyzed to reveal the relationship between the metallodendrimer nuclease activity and binding affinity to DNA. 2. Materials and Methods 2.1. DNA Binding Affinity Ethidium bromide (EtBr) displacement assay studies were used to measure the binding affinities between dendrimer 2a to 2e and CT-DNA. The CT-DNA solution (0.46 mM) and ethidium bromide (EtBr, TCI, Tokyo, Japan) (50 µM) were dissolved in buffer (tris buffer 24 mM, pH = 7.24) and mixed in a florescence cuvette. A series of experiments were conducted by adding 0–14 µM of the compounds 2a–2e individually to the EtBr-bound CT-DNA in the florescence cuvette, and the florescence intensities were measured over time. The normalized florescence intensity was then plotted against the dendrimer concentration. 2.2. Nuclease Activity The cleavage activity was investigated by agarose gel electrophoresis. Complexes 3a (1.67 µM, 3.33 µM, 6.67 µM, 8.33 µM, 10.00 µM, 16.67 µM, 20.00 µM) were mixed with plasmid DNA in the presence of 1,4-dithiothreitol (DTT) (0.66 mM) in Tris buffer (pH = 7.24, 24 mM), and the fluorescence was monitored over 45 min. Thereafter, the plasmid DNA was stained with EtBr and separated on a 0.8% agarose gel in standard Tris/borate /EDTA (TBE) buffer at 110 V over 35 min. The fluorescence intensity of form II under various conditions was obtained by using the ImageJ software.
3.33 µM, 6.67 µM, 8.33 µM, 10.00 µM, 16.67 µM, 20.00 µM) were mixed with plasmid DNA in the presence of 1,4-dithiothreitol (DTT) (0.66 mM) in Tris buffer (pH = 7.24, 24 mM), and the fluorescence was monitored over 45 min. Thereafter, the plasmid DNA was stained with EtBr and separated on a 0.8% agarose gel in standard Tris/borate /EDTA (TBE) buffer at 110 V over 35 min. The fluorescence intensity of form II under various conditions was obtained by using the ImageJ software.
Pharmaceutics 2018, 10, 258
3 of 9
2.3. Reactive Oxygen Species (ROS) Responsible for DNA Cleavage
2.3. Reactive Species (ROS)chelators, Responsible for DNA Cleavage RadicalOxygen scavengers or metal such as sodium formate (53 mM), neocuproine (0.53 mM), EDTA (13 mM), and sodium azide (53 mM), treated with 5e nM) and DTT (0.66 mM). Radical scavengers or metal chelators, suchwere as sodium formate (53(33 mM), neocuproine (0.53 mM), Complex 5e was mixed with DTT and various radical scavengers and incubated for 2 h. The solution EDTA (13 mM), and sodium azide (53 mM), were treated with 5e (33 nM) and DTT (0.66 mM). Complex 5e was mixed then separated a 0.8% agarose gelscavengers in the standard Tris/borate/EDTA (TBE) bufferwas at 110 V was with DTTonand various radical and incubated for 2 h. The solution then over 35 min. separated on a 0.8% agarose gel in the standard Tris/borate/EDTA (TBE) buffer at 110 V over 35 min. 2.4. Determination Determination of of Hydroxyl Hydroxyl Radicals Radicals 2.4. Tris buffer buffer (24 (24 mM, mM, pH pH == 7.24), 7.24), was was degassed degassed by by bubbling bubbling nitrogen nitrogen through through the the solution solution over over Tris two h h to to form form an an anaerobic two anaerobic environment. environment. Compound Compound5e 5echelated chelated57 57Equations Equationscopper copperions ionstotoform forma in the the degassed degassed acopper coppercomplex. complex.The Thecopper coppercomplex complex(2.60 (2.60µM) µM)and andDTT DTT (161.00 (161.00 µM) µM) were were mixed mixed in 0 -(p-hydroxyphenyl) Trisbuffer bufferand andtransferred transferredto toaafluorescence fluorescencecuvette. cuvette.H H22O O22 (0.27 (0.27 mM, mM, TCI) TCI) and and 33′-(p-hydroxyphenyl) Tris fluorescein(HPF) (HPF)dye dye(Thermo (ThermoFisher, Fisher,Waltham, Waltham,MA, MA, USA) were added fluorescence cuvette fluorescein USA) were added to to thethe fluorescence cuvette to to obtain the reduced copper complex. The fluorescence intensity was used to determine the presence obtain the reduced copper complex. The fluorescence intensity was used to determine the presence of of hydrolytic radicals. hydrolytic radicals.
3. 3. Results Results and and Discussion Discussion The (PAMAM) dendrimers The bis(2-pyridyl)amino-modified bis(2-pyridyl)amino-modified (PAMAM) dendrimers (2) (2) were were prepared prepared according according to to the the literature literature [28] [28] (Scheme (Scheme 1). 1). To To study study their their binding binding affinity, affinity, an an ethidium ethidium bromide bromide (EtBr) (EtBr) displacement displacement assay thethe results indicated a clear correlation betweenbetween the fluorescence intensity assay was wasused usedand and results indicated a negative clear negative correlation the fluorescence of EtBr and the concentration of each of synthetic dendrimers 2. For any given of the intensity of EtBr and the concentration ofthe each of the synthetic dendrimers 2. For anyquantity given quantity dendrimer, the high-generation dendrimers displayed lowerlower fluorescence intensity levels thanthan the of the dendrimer, the high-generation dendrimers displayed fluorescence intensity levels low-generation ones. This observation suggested that the high-generation dendrimers bound more the low-generation ones. This observation suggested that the high-generation dendrimers bound strongly to the to nucleic acids (Figure 1). From computed binding constants of 7.4, 23, 48, more strongly the nucleic acids (Figure 1). these Fromdata thesewe data we computed binding constants of 7.4, 1 for 2a–2e, 70, 280 and µM−280 [19], i.e., [19], the larger thelarger dendrimer, the stronger binding its of 23, and 48, 70, µM−1 forrespectively 2a–2e, respectively i.e., the the dendrimer, theitsstronger the nucleic acid. binding of the nucleic acid.
Scheme 1. 1. Synthesis Synthesis of of compounds compounds 2. 2. Scheme
To investigate the nuclease activity, we first saturated dendrimer 2a with Cu and then subjected the resulting complex 3a to a DNA cleavage experiment. 3a displayed nuclease activity toward the supercoiled form (form I) of the plasmid to produce the circular form (form II) in the presence of 1,4-dithiothreitol (DTT) as a reducing agent (Figure 2). In this investigation, bleomycin was used as a reference compound (lane 3). Different amounts of the complex (3a) were used to study its nuclease activity, and the results are indicated in lanes 7–14. The quantity of the circular form (form II) of the DNA plasmid increased in the presence of larger quantities of the complex. The concentration-dependent nuclease activity was used to characterize the catalytic efficiency of this complex. Complex 3a alone showed no nuclease effect (lane 6). This observation implied that
Normalized fluoresence intensity
Pharmaceutics 2018, 10, 258
0.8
0.6
4 of 9 0.4 G2-2py
Normalized fluoresence intensity
G3-2py no role was played by the 0.2 hydrolytic mechanism in this reaction. To understand the possible effect G4-2py G5-2py of ligand and copper alone, a comparable amount of bipyridine analogue (4, Figure S1) and CuSO4 G6-2py were subjected to similar experiments and the results are shown in lanes 4 and 5; here the plasmid 0.0 2 4 6 8 10 12 remained in the loading well. 0Presumably, the negative charges of the plasmid were neutralized by the Concentration compound () excess quantity of copper ions, making the complexofstationary on the gel. The monomeric dendrimer and the free copper ions have reported to show nodisplacement ROS generation Figurebeen 1. Ethidium bromide (EtBr) assay.activity under the given conditions [28]. Therefore, the monomer was assumed to have no nuclease activity in the present studies. observed was therefore concluded to have arisen 2a from theCu macromolecular Cu To The investigate theactivity nuclease activity, we first saturated dendrimer with and then subjected Pharmaceutics 2018, 10, x FOR PEER complex rather than from Cucleavage ions in the medium. 3a displayed nuclease activity toward the4 of 8 the resulting complex 3a the toREVIEW afree DNA experiment. supercoiled form (form I) of the plasmid to produce the circular form (form II) in the presence of 1,4dithiothreitol (DTT) as a reducing agent (Figure 2). In this investigation, bleomycin was used as a 1.0 reference compound (lane 3). Different amounts of the complex (3a) were used to study its nuclease activity, and the results are indicated in lanes 7–14. The quantity of the circular form (form II) of the 0.8 DNA plasmid increased in the presence of larger quantities of the complex. The concentrationdependent nuclease activity was used to characterize the catalytic efficiency of this complex. Complex 0.6 3a alone showed no nuclease effect (lane 6). This observation implied that no role was played by the hydrolytic mechanism in this reaction. To understand the possible effect of ligand and copper alone, 0.4 bipyridine analogue (4, Figure S1) and CuSO4 were subjected to similar a comparable amount of experiments and the results are G2-2py shown in lanes 4 and 5; here the plasmid remained in the loading G3-2py well. Presumably, the negative charges of the plasmid were neutralized by the excess quantity of 0.2 G4-2py G5-2py copper ions, making the complex stationary on the gel. The monomeric dendrimer and the free G6-2py copper ions have been reported to show no ROS generation activity under the given conditions [28]. 0.0 0 2 6 8 10 Therefore, the monomer was assumed to4 have no nuclease activity in 12 the present studies. The Concentration of compound () the macromolecular Cu complex observed activity was therefore concluded to have arisen from rather than from the free Cu ions in the medium. Figure1.1.Ethidium Ethidium bromide assay. Figure bromide(EtBr) (EtBr)displacement displacement assay.
To investigate the nuclease activity, we first saturated dendrimer 2a with Cu and then subjected the resulting complex 3a to a DNA cleavage experiment. 3a displayed nuclease activity toward the supercoiled form (form I) of the plasmid to produce the circular form (form II) in the presence of 1,4dithiothreitol (DTT) as a reducing agent (Figure 2). In this investigation, bleomycin was used as a reference compound (lane 3). Different amounts of the complex (3a) were used to study its nuclease activity, and the results are indicated in lanes 7–14. The quantity of the circular form (form II) of the DNA plasmid increased in the presence of larger quantities of the complex. The concentrationdependent nuclease activity was used to characterize the catalytic efficiency of this complex. Complex 3a aloneFigure showed no nuclease effect (lanewere 6). This observation implied no role waswhich played by the 2.2.Electrophoresis experiments performed to the conditions under Figure Electrophoresis experiments were performed tocharacterize characterize thethat conditions under which reactive oxygen (ROS) radicals generated observed. Lane 1: reactive oxygenspecies species (ROS) radicals were generatedand andnuclease nucleaseactivity activity was observed. Lane 1: hydrolytic mechanism in this reaction. Towere understand the possible effectwas of ligand and copper alone, marker; Lane 2: DNA; Lane 3: bleomycin (10.00 µM); Lane 4: CuSO (1.6 + 4 (16.7 µM) + DTT (0.66 marker; Lane 2: DNA; Lane 3: bleomycin (10.00 µM); Lane 4: CuSO 4 mM) (1.6 mM) + 4 (16.7 µM) + DTT 4 a comparable amount of bipyridine analogue (4, Figure S1) and CuSO4 were subjected to similar mM); CuSO (1.6 mM) DTT (0.66 mM); 3a (1.67 µM); Lane 7: Lane DTT (0.66 mM); Lane 8: (0.66 Lane mM);5:Lane 5:4CuSO 4 (1.6+mM) (0.66Lane mM);6:Lane 6: 3a (1.67 µM); 7: DTT (0.66 mM); experiments and the results are shown+ DTT in lanes 4 and 5; here the plasmid remained in the loading 3a (1.67 µM) + DTT (0.66 mM); Lane 9: 3a (3.33 µM) + DTT (0.66 mM); Lane 10: 3a (6.67 µM) + DTT Lane 8: 3a (1.67 µM) + DTT (0.66 mM); Lane 9: 3a (3.33 µM) + DTT (0.66 mM); Lane 10: 3a (6.67 µM) + well. Presumably, the negative charges of the plasmid were neutralized by the excess quantity of (0.66 3a (8.33 + DTT mM); 12:Lane 3a (10.00 µM) + DTT (0.66 mM); Lane 13: DTTmM); (0.66 Lane mM);11: Lane 11: 3aµM) (8.33 µM) (0.66 + DTT (0.66Lane mM); 12: 3a (10.00 µM) + DTT (0.66 mM); copper ions, making theµM) complex stationary on µM) the3a The monomeric dendrimer 3a (16.67 µM) + DTT (0.66 14: 3a (20.00 +gel. DTT (0.66 mM). DTT(0.66 = 1,4-dithiothreitol. Lane 13: 3a (16.67 +mM); DTTLane (0.66 mM); Lane 14: (20.00 µM) + DTT mM). DTT = and 1,4- the free copper ions have been reported to show no ROS generation activity under the given conditions [28]. dithiothreitol. To explore the relationship between dendrimer binding affinity and nuclease activity, the number Therefore, the monomer was assumed to have no nuclease activity in the present studies. The of Cu ions in each dendrimer must be the same to prevent any possible bias associated with a different observed activity was therefore concluded to have arisen from the macromolecular Cu complex number of Cu ions in each dendrimer. In previous work [29], each G2-dendrimer (2a) molecule was rather than from the free Cu ions in the medium. found to bind to no more than six copper ions. Accordingly, six equivalents of copper ions were coordinated to each dendrimer generation (2a–2e) to form hexaCu complexes 5a–5e, respectively, and these complexes were subjected to nuclease activity assays. The fluorescence intensities obtained from form II indicated that the nuclease activity was positively correlated with the dendrimer size (Figure S2). High-generation dendrimers exhibited stronger plasmid DNA cutting activities than did
ions were coordinated to each dendrimer generation (2a–2e) to form hexaCu complexes 5a–5e, respectively, and these complexes were subjected to nuclease activity assays. The fluorescence intensities obtained from form II indicated that the nuclease activity was positively correlated with the dendrimer size (Figure S2). High-generation dendrimers exhibited stronger plasmid DNA cutting Pharmaceutics 2018, 10, 258 5 of 9 activities than did the low-generation analogues. Dendrimers 5d and 5e exhibited, respectively, 2.15fold and 2.02-fold higher nuclease activities than did 5a. Meanwhile, 5b and 5c exhibited, the low-generation analogues. Dendrimers 5d and 5e exhibited, respectively, 2.15-fold and 2.02-fold respectively, only 1.39-fold and 1.64-fold higher nuclease activities than did 5a. Surprisingly, the higher nuclease activities than did 5a. Meanwhile, 5b and 5c exhibited, respectively, only 1.39-fold normalized nuclease activities of metallodendrimers G2 through G5 were found to be highly and 1.64-fold higher nuclease activities than did 5a. Surprisingly, the normalized nuclease activities of positively correlated (R2 = 0.97) with the Kb values of the corresponding apodendrimers for the DNA metallodendrimers G2 through G5 were found to be highly positively correlated (R2 = 0.97) with the plasmid (Figure 3).corresponding These results indicated that binding affinity of the dendrimer for the DNA Kb values of the apodendrimers for the the DNA plasmid (Figure 3). These results indicated plasmid directly contributed nuclease for activity. While 5e was expected, based on thenuclease trend of the that the binding affinity of to theits dendrimer the DNA plasmid directly contributed to its results for 5a–5d, higher based activity 5d, itofwas to display only 94% ofactivity the activity activity. Whileto 5eexhibit was expected, on than the trend the obser resultsved for 5a–5d, to exhibit higher than form 5d, it was ved toS1) display 94% of the activity form of greater 5d. (Table S1) This of towards II ofobser 5d. (Table This only observation may havetowards been due to II the degradation may have due to5e. the greater degradation of form II in the presence of complex 5e. formobservation II in the presence ofbeen complex 2.4
Normalized nuclease activity
2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0
20
40
60 5
80
-1
1/K b ( x10 M )
Figure 3. Plot of the normalizednuclease nuclease activities activities ofofhexaCu complexes 5a–5d as a as function of 1/Kof b 1/Kb Figure 3. Plot of the normalized hexaCu complexes 5a–5d a function for apodendrimers 2a–2d toward the plasmid. The fluorescence intensity of form II in the presence of for apodendrimers 2a–2d toward the plasmid. The fluorescence intensity of form II in the presence of each compound 5 was obtained by using ImageJ software. each compound 5 was obtained by using ImageJ software.
Identifying the active species responsible for the nuclease activity is also an important issue for Identifyingbiomedical the activeapplications species responsible for the nuclease activity is alsoactive an important developing of the nuclease. To determine the major species ofissue the for developing biomedical applications of their the nuclease. determine theconsidered major active species of the synthesized dendrimers responsible for cleavage ofTo plasmid, we only oxygen-derived synthesized dendrimers for possible their cleavage wepossible only considered oxygenspecies, since hydrolyticresponsible cleavage is not here [26].ofOfplasmid, the various reactive oxygen derived species, hydrolytic cleavage is not possible [26]. Offor the possible reactive species (ROS),since the hydroxyl radical has been found to be here responsible thevarious activities of various nucleases [27]. Therefore, the hydroxyl radicalhas wasbeen hypothesized be responsible involved in the of the of oxygen species (ROS), the hydroxyl radical found totobe foractivity the activities dendrimer nuclease. This hypothesis was tested by determining the presence of hydroxyl radicals various nucleases [27]. Therefore, the hydroxyl radical was hypothesized to be involved in the usingof30 the -(p-hydroxyphenyl) fluorescein which reacts with hydroxyl activity dendrimer nuclease. This(HPF), hypothesis was predominantly tested by determining the radicals presence of or superoxide anion radicals via dearylation to form a fluorescent product. Superoxide, however, hydroxyl radicals using 3′-(p-hydroxyphenyl) fluorescein (HPF), which reacts predominantly with can induce re-oxidation reactions that disturb the identification of the hydroxyl radical. Tris buffer hydroxyl radicals or superoxide anion radicals via dearylation to form a fluorescent product. was, therefore, degassed to avoid the formation of metallodendrimer-generated superoxide anion Superoxide, however, can induce re-oxidation reactions that disturb the identification of the hydroxyl radicals in the aerobic environment used. Instead, H2 O2 was used as the oxygen source and mixed radical. was,environment, therefore, degassed to avoidofthe formation of metallodendrimer-generated withTris 5e inbuffer an anoxic and the formation hydroxyl radicals was monitored using HPF. superoxide anion radicals in the aerobic environment used. Instead, H 2O2 was usedthan as the oxygen The fluorescence intensity was much higher in the presence of the metallodendrimer in the source and of mixed with 5e in an anoxic and theusing formation absence the metallodendrimer (Figureenvironment, S3). This observation complexof5ehydroxyl indicated radicals that this was monitored using HPF. likely The 5a–5d fluorescence intensity much higher in the radical. presence of the dendrimer, and hence as well, can catalyze was the formation of the hydroxyl To further confirm presence of the hydroxyl radical under the given conditions, we carried outusing metallodendrimer than inthe the absence of the metallodendrimer (Figure S3). This observation inhibition studies using various ROS scavengers or metal chelators, including EDTA, sodium complex 5e indicated that this dendrimer, and hence likely 5a–5d as well, can catalyze the azide, formation neocuproine, and sodium formate, which served as a copper chelator, singlet oxygen inhibitor/copper of the hydroxyl radical. chelator, specific copper (I) chelator, and hydroxyl radical scavenger, respectively. The inhibition activities of these compounds were examined by introducing each inhibitor into the DNA cleavage experiments and measuring the resulting nuclease activities. As shown in Figure 4, in lanes 3, 5, and 6, the plasmid remained in form I when sodium formate, EDTA, and sodium azide, respectively,
Tofurther furtherconfirm confirmthe thepresence presenceof ofthe thehydroxyl hydroxylradical radicalunder underthe thegiven givenconditions, conditions,we wecarried carried To out inhibition inhibition studies studies using using various various ROS ROS scavengers scavengers or or metal metal chelators, chelators, including including EDTA, EDTA, sodium sodium out azide, neocuproine, neocuproine, and and sodium sodium formate, formate, which which served served as as aa copper copper chelator, chelator, singlet singlet oxygen oxygen azide, inhibitor/copperchelator, chelator,specific specificcopper copper(I) (I)chelator, chelator,and andhydroxyl hydroxylradical radicalscavenger, scavenger,respectively. respectively. inhibitor/copper Theinhibition inhibitionactivities activitiesof ofthese thesecompounds compoundswere wereexamined examinedby byintroducing introducingeach eachinhibitor inhibitorinto intothe the The Pharmaceutics 2018, 10, 258 6 of 9 DNA cleavage experiments and measuring the resulting nuclease activities. As shown in Figure 4, in DNA cleavage experiments and measuring the resulting nuclease activities. As shown in Figure 4, in lanes3,3,5,5,and and6,6,the theplasmid plasmidremained remainedin inform formIIwhen whensodium sodiumformate, formate,EDTA, EDTA,and andsodium sodiumazide, azide, lanes respectively, were included, clearly indicating that each of them inhibited DNA cutting (Figure 4). respectively, were included, clearly indicating that each of them inhibited DNA cutting (Figure 4). were included, clearly indicating that each of them inhibited DNA cutting (Figure 4). EDTA removed EDTAions removed metal ions from fromThe the dendrimers. dendrimers. The results nuclease activitythat results indicated that the EDTA removed metal ions the The nuclease activity results that the metal from the dendrimers. nuclease activity indicated theindicated copper ions were copper ions were crucial to the nuclease activity. Also note that azide has been shown to chelate copper ions were crucial to the nuclease activity. Also note that azide has been shown to chelate crucial to the nuclease activity. Also note that azide has been shown to chelate copper ions and copperions ions andabrogate abrogatecopper copper re-oxidation incatecholases catecholases[30]. andtyrosinases tyrosinasessodium [30].Moreover, Moreover, sodium copper and re-oxidation in and [30]. sodium abrogate copper re-oxidation in catecholases and tyrosinases Moreover, azide could also azide also act as asscavenger. singlet oxygen oxygen scavenger. However, the presence presence of singlet singlet oxygen was azide also act aa singlet scavenger. However, the of was act as could acould singlet oxygen However, the presence of singlet oxygen was highlyoxygen unlikely in highly unlikely in this reaction, as it did not involve photo-irradiation. The presence of reducing highly unlikely in this reaction, as it did not involve photo-irradiation. The presence of reducing this reaction, as it did not involve photo-irradiation. The presence of reducing agents was necessary agents wasnecessary necessary forthe the nuclease activity, andsodium sodium azide appeared toact actas asabut acopper copper chelator agents for and azide to for the was nuclease activity, andnuclease sodiumactivity, azide appeared to act as aappeared copper chelator not achelator singlet but not a singlet oxygen scavenger in this reaction. Nuclease inhibition by sodium formate implied but not a singlet oxygen scavenger in this reaction. Nuclease inhibition by sodium formate implied oxygen scavenger in this reaction. Nuclease inhibition by sodium formate implied that the hydroxyl that the the hydroxyl radical was the ROS ROS involved in DNA DNA cleavage. cleavage. Nevertheless, theto failure of that hydroxyl was involved in Nevertheless, the failure of radical was the ROSradical involved inthe DNA cleavage. Nevertheless, the failure of neocuproine remove neocuproine to remove remove the copper(I) copper(I) ions and and hence stop the reaction may have have been due toions the neocuproine the ions hence reaction may been to the the copper(I) to ions and hence stop the reaction may havestop beenthe due to the strong binding ofdue Cu(I) strong binding of Cu(I) ions to the bis(2-pyridyl)amino moieties, indicating that neocuproine could strong binding of Cu(I) ions to the bis(2-pyridyl)amino moieties, indicating that neocuproine could to the bis(2-pyridyl)amino moieties, indicating that neocuproine could not remove the copper(I) ions not remove thecopper(I) copper(I)ions ionsor orstop stopthe thereaction. reaction. not remove the or stop the reaction.
Figure Identification of the active species. Lane1, plasmid alone; Lane 2,2,plasmid plasmid with 5e (33 nM) Figure4. Identificationof ofthe theactive activespecies. species.Lane1, Lane1,plasmid plasmidalone; alone;Lane Lane2, plasmidwith with5e 5e(33 (33nM) nM)+++ Figure 4.4.Identification DTT (33 nM) ++DTT (0.66 mM) + +sodium formate (53 mM); Lane 4, DTT(0.66 (0.66mM); mM);Lane Lane3,3,3,plasmid plasmidwith with5e 5e (33 nM) +DTT DTT (0.66 mM) +sodium sodium formate (53 mM); Lane (0.66 mM); Lane plasmid with 5e (33 nM) (0.66 mM) formate (53 mM); Lane plasmid with 5e (33 nM) + DTT (0.66 mM) + neocuproine (0.53 mM); Lane 5, plasmid with 5e (33 nM) 4, plasmid with 5e (33 nM) + DTT (0.66 mM) + neocuproine (0.53 mM); Lane 5, plasmid with 5e (33 4, plasmid with 5e (33 nM) + DTT (0.66 mM) + neocuproine (0.53 mM); Lane 5, plasmid with 5e (33+ DTT mM) +mM) EDTA (13 mM); 6,Lane plasmid with 5ewith (33 nM) + nM) DTT mM) sodium azide nM)+(0.66 +DTT DTT (0.66 mM) EDTA (13Lane mM); Lane plasmid with 5e(33 (33 nM)+(0.66 +DTT DTT (0.66+mM) mM) sodium nM) (0.66 ++EDTA (13 mM); 6,6,plasmid 5e (0.66 ++sodium (53 mM); plasmid with 5e (33 5e nM). azide (53Lane mM);7,Lane Lane plasmid with 5e(33 (33nM). nM). azide (53 mM); 7,7,plasmid with
In previous [31], Cu dendrimer complexes apparently generated Inview viewof ofprevious previousstudies studies[31], [31],the theCu Cudendrimer dendrimercomplexes complexesapparently apparentlygenerated generatedsuperoxide superoxide In view of studies the superoxide anion radicals in the presence of reducing agents. The superoxide anion radicals could then anionradicals radicalsin inthe thepresence presenceof ofreducing reducingagents. agents.The Thesuperoxide superoxideanion anionradicals radicalscould couldbe bethen thenbe further anion further further reduced and protonated to form hydrogen peroxide. The hydroxyl radical could have reducedand andprotonated protonatedto toform formhydrogen hydrogenperoxide. peroxide.The Thehydroxyl hydroxylradical radicalcould couldhave havebeen beenobtained obtained reduced + ion conjugated to been obtained from either one of two routes. With excess reducing agent, Cu + ion conjugated + from either one of two routes. With excess reducing agent, Cu to hydrogen peroxide from either one of two routes. With excess reducing agent, Cu ion conjugated to hydrogen peroxide hydrogen mayaahave produced a highly reactive hydroxyl radical copper Fenton may have haveperoxide produced highly reactive hydroxyl radical through thethrough copperthe Fenton reaction. may produced highly reactive hydroxyl radical through the copper Fenton reaction. reaction. Otherwise, the Haber–Weiss reaction might have been involved while reductant was Otherwise, the the Haber–Weiss Haber–Weiss reaction reaction might might have have been been involved involved while while reductant reductant was was being being Otherwise, being 2+ + consumed. Under this circumstance, Cu reduced by superoxide anion radicals to give the 2+ was 2+ consumed. Under this circumstance, Cu was reduced by superoxide anion radicals to give the Cu++ consumed. Under this circumstance, Cu was reduced by superoxide anion radicals to give the Cu Cu ion, ion,which whichcould couldlead leadto tothe theformation formationof ofthe thehydroxyl hydroxylradical radical(Figure (Figure5). 5). ion, which could lead to the formation of the hydroxyl radical (Figure 5). + + CuCu + + CuCu
Cu++ Cu
2+ 2+ CuCu
2+ 2+ CuCu
+ + CuCu
Cu + +
Cu
+ + CuCu
Cu2+2+ Cu
2+ 2+ CuCu 2+ 2+ CuCu
+ + CuCu
2+ 2+ CuCu
2+ 2+ CuCu
33
OO22
HH OO . OO22-.-. 22 22 .OH OH++OH OH- -++OO22
OO22 L-Cu2+2+ L-Cu
DTT ++ DTT
++ -. L-Cu++ ++ OO22-. ++ 22HH L-Cu + HH2O L-Cu++ + 2 O22 L-Cu
L-Cu++ ++ OO22-.-. L-Cu 2+ L-Cu2+ L-Cu ++ HH22OO22 OH- - ++ OH L-Cu2+2+ ++ OH OH. . L-Cu
Figure 5.5.Hypothetical Hypothetical mechanism for the dendrimer-induced Figure5. Hypotheticalmechanism mechanismfor forthe thedendrimer-induced dendrimer-inducedproduction productionof ofhydroxyl hydroxylradicals. radicals. Figure production of hydroxyl radicals.
4. Conclusions The most significant discovery of this investigation was the establishment of a correlation between the binding affinity of copper dendrimer-based nucleases and their nuclease activity. For the first time, a linear correlation between Kb and chemical nuclease activity was found among the complexes
Pharmaceutics 2018, 10, 258
7 of 9
tested. The higher-generation dendrimers displayed higher binding affinities for the DNA plasmid, and the nuclease activity of the dendrimer complex was found to depend on the size of the dendrimer. These results revealed the importance of binding affinity for macromolecular nuclease activity and could be used for the development of new macromolecular nucleases. We also identified the formation of hydroxyl radicals to be the major working mechanism for this cleavage reaction. Our observations here have provided a fundamental understanding of how macromolecules can serve as nucleases and pave the way for the design of efficient chemical nucleases. Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4923/10/4/258/s1, Procedures used to synthesize compounds 2 and 4, complexes 3a and 5; H-NMR spectra of compounds 2. Figure S1. Structure of compound 4; Figure S2. Nuclease activity of hexaCu complexes 5; Figure S3. Fluorescence spectra of HPF in the presence and absence of 5e; Table S1. Relative fluorescence intensity of II. Author Contributions: Conceptualization, Y.-H.T. and C.-L.K.; Data curation: Y.-H.T.; S.-t.L., and C.H.-Y.W.; Methodology, Y.-H.T.; C.H.-Y.W.; S.C.N.H.; H.-T.C. and P.-Y.C.; Validation, Y.-H.T. and C.-L.K.; Formal Analysis, Y.-H.T. and S.-t.L.; Investigation, Y.-H.T.; Data Curation, Y.-H.T. and S.-t.L.; Writing—Original Draft Preparation, Y.-H.T.; Writing—Review and Editing, C.-L.K. Supervision, C.-L.K.; Funding Acquisition, C.-L.K. Funding: This work was supported by the Ministry of Science and Technology (MoST), Taiwan (MOST 106-2632-M-037-001- and 106-2113-M-037-008-) Kaohsiung Medical University (KMU-DK108009) and Kaohsiung Medical University Hospital (KMUH106-M625). Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.
References 1. 2. 3. 4.
5.
6. 7. 8.
9.
10.
11.
Astruc, D.; Ornelas, C.; Ruiz, J. Metallocenyl Dendrimers and Their Applications in Molecular Electronics, Sensing, and Catalysis. Acc. Chem. Res. 2008, 41, 841–856. [CrossRef] [PubMed] Newkome, G.R.; He, E.; Moorefield, C.N. Suprasupermolecules with Novel Properties: Metallodendrimers. Chem. Rev. 1999, 99, 1689–1746. [CrossRef] [PubMed] Tang, Y.-H.; Huang, A.Y.; Chen, P.-Y.; Chen, H.-T.; Kao, C.-L. Metallodendrimers and Dendrimer Nanocomposites. Curr. Pharm. Des. 2011, 17, 2308–2330. [CrossRef] [PubMed] Corbin, D.A.; Shircliff, D.M.; Reeves, B.J.; Boardman, B.M. Metallopolymers from direct polymerization of functionalized cobalt chalcogenide clusters and thiophene comonomers. Polym. Chem. 2017, 8, 3801–3809. [CrossRef] Greenfield, J.L.; Rizzuto, F.J.; Goldberga, I.; Nitschke, J.R. Self-Assembly of Conjugated Metallopolymers with Tunable Length and Controlled Regiochemistry. Angew. Chem. Int. Ed. 2017, 56, 7541–7545. [CrossRef] [PubMed] Zhang, K.Y.; Liu, S.; Zhao, Q.; Huang, W. Stimuli–responsive metallopolymers. Coord. Chem. Rev. 2016, 319, 180–195. [CrossRef] Vishwanath, R.S.; Kandaiah, S. Metal ion-containing C3N3S3 coordination polymers chemisorbed to a copper surface as acid stable hydrogen evolution electrocatalysts. J. Mater. Chem. A 2017, 5, 2052–2065. Liang, Y.; Strohecker, D.; Lynch, V.; Holliday, B.J.; Jones, R.A. A Thiophene-Containing Conductive Metallopolymer Using an Fe(II) Bis(terpyridine) Core for Electrochromic Materials. ACS Appl. Mater. Interfaces 2016, 8, 34568–34580. [CrossRef] Caminade, A.-M.; Ouali, A.; Laurent, R.; Turrin, C.-O.; Majoral, J.-P. Coordination chemistry with phosphorus dendrimers. Applications as catalysts, for materials, and in biology. Coord. Chem. Rev. 2016, 308, 478–497. [CrossRef] Verma, S.; Srivatsan, S.G.; Claussen, C.A.; Long, E.C. DNA strand scission by a Cu(I)·adenylated polymeric template: Preliminary mechanistic and recycling studies. Bioorg. Med. Chem. Lett. 2003, 13, 2501–2504. [CrossRef] Melvin, M.S.; Tomlinson, J.T.; Saluta, G.R.; Kucera, G.L.; Lindquist, N.; Manderville, R.A. Double-Strand DNA Cleavage by Copper·Prodigiosin. J. Am. Chem. Soc. 2000, 122, 6333–6334. [CrossRef]
Pharmaceutics 2018, 10, 258
12.
13.
14. 15. 16. 17.
18.
19. 20.
21. 22.
23.
24. 25.
26.
27. 28.
29.
8 of 9
Zandarashvili, L.; Esadze, A.; Vuzman, D.; Kemme, C.A.; Levy, Y.; Iwahara, J. Balancing between affinity and speed in target DNA search by zinc-finger proteins via modulation of dynamic conformational ensemble. Proc. Natl. Acad. Sci. USA 2015, 112, E5142–E5149. [CrossRef] [PubMed] Wang, X.-L.; Chao, H.; Li, H.; Hong, X.-L.; Ji, L.-N.; Li, X.-Y. Synthesis, crystal structure and DNA cleavage activities of copper(II) complexes with asymmetric tridentate ligands. J. Inorg. Biochem. 2004, 98, 423–429. [CrossRef] [PubMed] Jiang, Q.; Xiao, N.; Shi, P.; Zhu, Y.; Guo, Z. Design of artificial metallonucleases with oxidative mechanism. Coord. Chem. Rev. 2007, 251, 1951–1972. [CrossRef] Li, Y.; Wu, Y.; Zhao, J.; Yang, P. DNA-binding and cleavage studies of novel binuclear copper(II) complex with 1,10 -dimethyl-2,20 -biimidazole ligand. J. Inorg. Biochem. 2007, 101, 283–290. [CrossRef] [PubMed] Liu, C.; Wang, M.; Zhang, T.; Sun, H. DNA hydrolysis promoted by di- and multi-nuclear metal complexes. Coord. Chem. Rev. 2004, 248, 147–168. [CrossRef] Hurley, A.L.; Mohler, D.L. Organometallic Photonucleases: Synthesis and DNA-Cleavage Studies of Cyclopentadienyl Metal-Substituted Dendrimers Designed To Increase Double-Strand Scission. Org. Lett. 2000, 2, 2745–2748. [CrossRef] Teyssot, M.-L.; Jarrousse, A.-S.; Chevry, A.; De Haze, A.; Beaudoin, C.; Manin, M.; Nolan, S.P.; Díez-González, S.; Morel, L.; Gautier, A. Toxicity of Copper(I)–NHC Complexes Against Human Tumor Cells: Induction of Cell Cycle Arrest, Apoptosis, and DNA Cleavage. Chemistry 2009, 15, 314–318. [CrossRef] Mack, D.P.; Iverson, B.L.; Dervan, P.B. Design and chemical synthesis of a sequence-specific DNA-cleaving protein. J. Am. Chem. Soc. 1988, 110, 7572–7574. [CrossRef] Suntharalingam, K.; Hunt, D.J.; Duarte, A.A.; White, A.J.P.; Mann, D.J.; Vilar, R. A Tri-copper(II) Complex Displaying DNA-Cleaving Properties and Antiproliferative Activity against Cancer Cells. Chemistry 2012, 18, 15133–15141. [CrossRef] Mancin, F.; Scrimin, P.; Tecilla, P.; Tonellato, U. Artificial metallonucleases. Chem. Commun. 2005, 20, 2540–2548. [CrossRef] [PubMed] Tan, J.D.; Hudson, S.E.; Brown, S.J.; Olmstead, M.M.; Mascharak, P.K. Syntheses, structures, and reactivities of synthetic analogs of the three forms of cobalt(III)-bleomycin: Proposed mode of light-induced DNA damage by the cobalt(III) chelate of the drug. J. Am. Chem. Soc. 1992, 114, 3841–3853. [CrossRef] El Amrani, F.B.A.; Perelló, L.; Real, J.A.; González-Alvarez, M.; Alzuet, G.; Borrás, J.; García-Granda, S.; Montejo-Bernardo, J. Oxidative DNA cleavage induced by an iron(III) flavonoid complex: Synthesis, crystal structure and characterization of chlorobis(flavonolato)(methanol) iron(III) complex. J. Inorg. Biochem. 2006, 100, 1208–1218. [CrossRef] [PubMed] Wende, C.; Lüdtke, C.; Kulak, N. Copper Complexes of N-Donor Ligands as Artificial Nucleases. Eur. J. Inorg. Chem. 2014, 16, 2597–2612. [CrossRef] Soler, M.; Figueras, E.; Serrano-Plana, J.; González-Bártulos, M.; Massaguer, A.; Company, A.; Martínez, M.Á.; Malina, J.; Brabec, V.; Feliu, L.; Planas, M.; et al. Design, Preparation, and Characterization of Zn and Cu Metallopeptides Based On Tetradentate Aminopyridine Ligands Showing Enhanced DNA Cleavage Activity. Inorg. Chem. 2015, 54, 10542–10558. [CrossRef] [PubMed] Zhiryakova, M.V.; Izumrudov, V.A. Interaction of Astramol Poly(propyleneimine) Dendrimers with DNA and Poly(methacrylate) Anion in Water and Water–Salt Solutions. J. Phys. Chem. B 2014, 118, 8819–8826. [CrossRef] [PubMed] Jin, Y.; Cowan, J.A. DNA Cleavage by Copper−ATCUN Complexes. Factors Influencing Cleavage Mechanism and Linearization of dsDNA. J. Am. Chem. Soc. 2005, 127, 8408–8415. [CrossRef] Kao, C.-L.; Tang, Y.-H.; Lin, Y.C.; Chiu, L.-T.; Chen, H.-T.; Hsu, S.C.N.; Hsieh, K.-C.; Lu, C.-Y.; Chen, Y.-L. Copper complex of a pyridine-modified poly(amidoamine) dendrimer as a chemical nuclease: Synthetic and catalytic study. Nanomedicine 2011, 7, 273–276. [CrossRef] Tang, Y.-H.; Lin, Y.C.; Hsu, S.C.N.; Liou, S.-T.; Chen, H.-Y.; Hsieh, K.-C.; Chuang, W.-J.; Chiu, L.-T.; Chen, Y.-L.; Kao, C.-L. Cooperative Effects in Copper Polyamidoamine Dendrimer Complexes Catalyzing the Reduction of Molecular Oxygen. Eur. J. Inorg. Chem. 2015, 29, 4839–4847. [CrossRef]
Pharmaceutics 2018, 10, 258
30. 31.
9 of 9
Healey, D.F.; Strothkamp, K.G. Inhibition of the catecholase and cresolase activity of mushroom tyrosinase by azide. Arch. Biochem. Biophys. 1981, 211, 86–91. [CrossRef] Tang, Y.-H.; Cangiotti, M.; Kao, C.-L.; Ottaviani, M.F. EPR Characterization of Copper(II) Complexes of PAMAM-Py Dendrimers for Biocatalysis in the Absence and Presence of Reducing Agents and a Spin Trap. J. Phys. Chem. B 2017, 121, 10498–10507. [CrossRef] [PubMed] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
