2.1. 1H NMR Investigation of the Reactions of 2-Thiohydantoin Derivatives 2a and 2b with Cisplatin
The reaction of thiohydantoins
2a and
2b and cisplatin was monitored in an overnight kinetic
1H NMR experiment and then further tracked by recording spectra periodically until the completion of the reaction. In the reaction of
2a and cisplatin, two pairs of singlets that change in intensity throughout the experiment can be observed (
Figure 2). The first pair of singlets at 2.09 ppm (
a) and 2.59 ppm (
b) originates from –S-CH
3 protons of the free and coordinated
2a, respectively. The second pair of singlets at 3.95 ppm (
c) and 7.30 ppm (
d) is from the NH
3 protons of cisplatin and free ammonia. It is known that cisplatin has a high affinity towards DMSO and readily reacts with it, forming monosubstituted [PtCl(NH
3)
2(dmso)]
+ and disubstituted [Pt(NH
3)
2(dmso)
2]
2+ complexes [
25]. Coordination with the first DMSO molecule occurs through the sulfur atom and happens so fast that the reaction is completely finished within an hour, with the concentration of cisplatin at 1 mg/mL, which is five times lower than that used in this investigation. Monosubstituted [PtCl(NH
3)
2(dmso)]
+ then further reacts with a second DMSO molecule, forming [Pt(NH
3)
2(dmso)
2]
2+. Taking this into consideration, it can be concluded that the broad singlet at 3.95 ppm (
c) originates from the NH
3 ligand of the monosubstituted [PtCl(NH
3)
2(dmso-
d6)]
+, and not from cisplatin, and that the reaction is fast enough that all of cisplatin is converted to the monosubstituted [PtCl(NH
3)
2(dmso-
d6)]
+ before the reaction with
2a and
2b even began.
The presence of the first pair of singlets (a) and (b) in the spectra points to the coordination of Pt to the sulfur from the side chain. The intensity of the singlet of free 2a at 2.09 ppm decreases throughout the experiment, while the intensity of the singlet of the coordinated 2a at 2.59 ppm increases. The intensity of the broad singlet of the NH3 ligand of the monosubstituted [PtCl(NH3)2(dmso-d6)]+ at 3.95 ppm decreases throughout the experiment, while, after a few hours, a new broad singlet of free ammonia at 7.30 ppm appears. The delayed ammonia release indicates further coordination of DMSO in the later stages of the reaction. In the last spectrum of the overnight experiment, out of the signals of newly formed species, the singlet of the coordinated 2a at 2.59 ppm and the broad singlet of free ammonia at 7.30 ppm were dominant. There were no further changes observed after three days.
In the spectra of the reaction of thiohydantoin
2b and cisplatin (
Figure 3), similarly as with
2a, two pairs of singlets can be observed: (
a) singlet of the –S-CH
3 protons of free
2b at 2.03 ppm; (
b) singlet of the –S-CH
3 protons of coordinated
2b at 2.30 ppm; (
c) broad singlet of the NH
3 ligand of the monosubstituted [PtCl(NH
3)
2(dmso-
d6)]
+ at 3.95 ppm; (
d) broad singlet of free ammonia at 7.40 ppm. The intensity of the singlet of free
2b at 2.03 ppm decreases, while the intensity of the singlet of coordinated
2b at 2.30 ppm increases. Also, the intensity of the broad singlet of the monosubstituted [PtCl(NH
3)
2(dmso-
d6)]
+ at 3.95 ppm decreases, while a new broad singlet of free ammonia appears at 7.40 ppm. Out of the newly formed species in the last spectrum of the overnight kinetic experiment, the dominant signals were the singlet of the coordinated
2b at 2.30 ppm and the broad singlet of free ammonia at 7.40 ppm, and after a week, there were no further changes. Unlike with
2a, in the reaction of cisplatin with
2b, another broad singlet at 10.49 ppm (
e) can be observed, which belongs to the thiohydantoin ring NH group of the uncoordinated
2b. Its intensity seemingly decreases throughout the experiment, and it vanishes completely after a little less than two hours. The vanishing of this signal indicates the deprotonation of the thiohydantoin NH group and coordination to platinum through the nitrogen. The signal not only decreases in intensity throughout the experiment, but also stretches and broadens, which is a commonly observed phenomenon, and it occurs during metal coordination, as platinum can alter the chemical environment of the NH proton and influence its exchange rate with the residual water and HOD in the solvent, as well as its hydrogen binding interactions. This signal is missing from the spectra from the start of the experiment in the reaction of cisplatin with
2a, which can be explained by the apparent difference in the reaction speeds, because the reaction with
2a is significantly faster. Additionally, the absence of this signal (
e) from the spectra of the reaction of
2a can point to the multi-step nature of the reaction of cisplatin with
2b, which is supported by the fact that the singlet of the –S-CH
3 group of the coordinated
2b appears only after the broad singlet of the NH group of the free
2b disappears.
As previously mentioned, it is known that cisplatin reacts with DMSO, which raises the question of the competition of DMSO as a ligand with the investigated thiohydantoins, as well as its influence on the course and kinetics of their reaction with cisplatin. DMSO has a high affinity towards Pt(II) as a soft base. It was demonstrated that cisplatin undergoes solvolysis and reacts with DMSO in its pure form as well as in its aqueous solution [
26]. To investigate and describe the influence of DMSO on the reactions of
2a and
2b with cisplatin, a control kinetic experiment was conducted, where a reaction of cisplatin and deuterated DMSO, in which it was dissolved, was observed (
Figure 4). During the experiment, changes in two signals can be observed. The intensity of the broad singlet of the NH
3 ligand of the monosubstituted [PtCl(NH
3)
2(dmso-
d6)]
+ at 3.95 ppm (
a) decreases, while the intensity of the broad singlet of the NH
3 ligand of the disubstituted [Pt(NH
3)
2(dmso-
d6)
2]
2+ at 4.75 ppm (
b) increases throughout the experiment.
In the spectra of the reactions of
2a and
2b with cisplatin, the signal of the Pt-DMSO-
d6 adduct at 4.75 ppm (
Figure 4,
b) cannot be observed as a dominant species. The absence of this signal, together with the fact that ammonia is released in both of the reactions with
2a and
2b, but not in the control experiment, points to the conclusion that the disubstituted [Pt(NH
3)
2(dmso-
d6)
2]
2+ adduct is not the final product of the reaction. This secondary reaction of cisplatin and DMSO-
d6 does, however, make the system more complex and creates a competition of the DMSO-
d6 adducts in the reactions with
2a and
2b. In the spectra of the reaction of cisplatin with
2a, the broad singlet of the disubstituted [Pt(NH
3)
2(dmso-
d6)
2]
2+ adduct at 4.75 ppm is not present, while its appearance with a very small intensity can be noticed in the spectra of the reaction of cisplatin with
2b. The apparent difference in the speed of the reactions of cisplatin with
2a and
2b can explain the difference in the appearance of this signal in the spectra. Moreover, the difference in the appearance of the signal of the disubstituted [Pt(NH
3)
2(dmso-
d6)
2]
2+ adduct additionally points to the possible differences in the mechanisms of the reaction of
2a and
2b, in which the reaction with
2b would include an additional step, which was postulated earlier.
2.2. Mechanistic Insights Based on the Analysis of Reactant Concentrations and Potential Products
The results of the analysis of the
1H NMR spectra obviously demonstrate the reactions of [PtCl(NH
3)
2(dmso-
d6)]
+ with thiohydantoins
2a and
2b but do not give clear insight into the chemistry of these reactions. With this in mind, time-dependent changes in the concentrations of [PtCl(NH
3)
2(dmso-
d6)]
+,
2a,
2b, platinum complexes (marked as [Pt-
2a] and [Pt-
2b]), and free ammonia were analyzed. Graphs that illustrate time-dependent changes in the concentrations of the stated components are given in
Figure 5 and
Figure 6.
Changes in the quantities of free and coordinated
2a were determined through the changes in the intensities of the signals at 2.09 ppm and 2.59 ppm, and quantities of [PtCl(NH
3)
2(dmso-
d6)]
+ and free ammonia were derived from the changes in the signals at 3.95 ppm and 7.40 ppm. The starting concentrations of [PtCl(NH
3)
2(dmso-
d6)]
+ and
2a were the same (around 0.015 M) because they were mixed in a 1:1 molar ratio. After 15 h (
Figure 5), their concentrations were also the same (around 0.004 M), indicating that the same quantities of [PtCl(NH
3)
2(dmso-
d6)]
+ and
2a have reacted, giving a mononuclear platinum complex with
2a coordinated in a bidentate manner. Taking into account the changes in the concentrations of reactants and the newly formed complex (marked on the graph as [Pt-
2a]), we can come to the conclusion that [Pt-
2a] forms in a 1:1 reaction of [PtCl(NH
3)
2(dmso-
d6)]
+ and
2a. From the
1H NMR spectra, it is known that [Pt-
2a] contains
2a bound to platinum through the sulfur atom from the cysteine residue. The concentration of released ammonia is two times higher than the concentration of the formed complex. Based on that, we can assume that the binding of
2a completely expels ammonia from [PtCl(NH
3)
2(dmso-
d6)]
+. Furthermore, it has to be taken into account that DMSO-
d6 can react with [PtCl(NH
3)
2(dmso-
d6)]
+, which additionally complicates the reaction mechanism as a whole.
The mechanism of the reaction of
2b with [PtCl(NH
3)
2(dmso-
d6)]
+ is more complex when looking at the changes in concentrations of reactants and products (
Figure 6). The concentrations of reactants decrease throughout the experiment, but not at the same rate. Namely, the concentration of
2b decreases by about 0.005 M, while the concentration of [PtCl(NH
3)
2(dmso-
d6)]
+ decreases by about 0.01 M. This two-fold difference in the concentration change indicates that a dinuclear complex forms in the reaction of
2b. This is also supported by the fact that the concentration of the formed [Pt-
2b] complex increases by about 0.005 M throughout the experiment. The concentration of [Pt(NH
3)
2(dmso-
d6)
2]
2+, detectable in the reaction with
2b, remains stable between 0.003 and 0.004 M throughout the experiment as a consequence of the balance between its formation in the reaction of [PtCl(NH
3)
2(dmso-
d6)]
+ with DMSO-
d6 and its depletion in the reaction with
2b. The concentration of the released ammonia is about 0.02 M, and, after accounting for the changes in the reactant concentrations, it can be concluded that during the reaction, four molecules of ammonia are released, which come from the substitution of four amine ligands for every single molecule of the formed [Pt-
2b] complex.
2.4. Determining the Mechanisms of the Reactions of 2a and 2b with Cisplatin Based on the Reaction Order
The graphs depicting the changes in the concentrations of reactants (
Figure 5 and
Figure 6) certainly give insights into the possible products of the reactions of
2a and
2b with cisplatin, but not the mechanisms themselves. In order to gain these insights, graphs of inverse concentration values vs. time were analyzed, which are used as models for second-order reactions, as the reactions of
2a and
2b with cisplatin are supposed to follow second-order kinetics.
Based on the changes in the concentration of
2a, it is apparent that this is not purely a second-order reaction, as expected, because two linear slopes can be observed (
Figure 8a) and the reaction of
2a consists of two phases. This points to the complexity of the reaction system. In the beginning, the reaction of
2a and [PtCl(NH
3)
2(dmso-
d6)]
+ has a linear slope with the following equation: y = 7.704x + 61.303. Then, after 2.5 h, a tipping point occurs, after which the reaction of
2a speeds up, which is shown by the coefficient in the equation of the second phase (y = 16.31x + 34.267). The coefficient for the second phase of the reaction (
k2 = 16.31) is significantly higher than for the first phase (
k1 = 7.704), which reveals that there is a change in the reaction system and that a new chemical species has formed, the concentration of which became significant after 2.5 h. A strong possibility is that the disubstituted [Pt(NH
3)
2(dmso-
d6)
2]
2+ formed in the reaction of [PtCl(NH
3)
2(dmso-
d6)]
+ with DMSO-
d6, which replaces the chlorido ligand. Based on the
1H NMR spectra, two conclusions can be made. First, the signal of the thiohydantoin ring NH group is already missing from the first spectra of the experiment, indicating relatively fast coordination to the nitrogen atom. Second, based on the downfield shift in the signal of S-CH
3 protons, subsequent coordination to the side-branch sulfur atom is observed. Obviously, a bidentate coordination of
2a occurs in the resulting [Pt-
2a] complex. While analyzing the second-order graph for [Pt-
2a] (
Figure 8b), an intensive downward slope can be observed up until 2.5 h, after which the plot asymptotically approaches the
x-axis. After 11 h, the plot is linear and parallel to the
x-axis, indicating that there is no further change in the quantity of [Pt-
2a], but also the possibility of a change in the structure of [Pt-
2a] due to a reaction with the solvent, DMSO-
d6.
When looking at the second-order graph for [PtCl(NH
3)
2(dmso-
d6)]
+ (
Figure 8c), two phases can also be observed, the same as with
2a. The first, faster phase lasts about 5 h and is marked with the equation: y = 19.465x + 60.889. If only [PtCl(NH
3)
2(dmso-
d6)]
+ reacts with
2a, then the slope coefficient (
k1) would have to be 7.704, because [PtCl(NH
3)
2(dmso-
d6)]
+ and
2a are in a 1:1 molar ratio. However, the slope coefficient,
k1, is much higher, equaling 19.465. This indicates a parallel reaction of [PtCl(NH
3)
2(dmso-
d6)]
+, which is faster than the reaction with
2a. It is assumed that in this phase [PtCl(NH
3)
2(dmso-
d6)]
+ reacts with the solvent, giving [Pt(NH
3)
2(dmso-
d6)
2]
2+. In order to confirm this, a graph of the logarithm of concentration vs. time for [PtCl(NH
3)
2(dmso-
d6)]
+ was analyzed, which is used to depict first-order reactions (
Figure 8d). On the graph, it is clear that there is a linear slope up until 5 h, after which there is a deviation from linearity. This, of course, implies that a pseudo-first-order reaction with the solvent occurs in the first phase. Now the issues regarding the changes in the second-order graph for
2a become clearer. Namely,
2a reacts more readily with [Pt(NH
3)
2(dmso-
d6)
2]
2+ than with [PtCl(NH
3)
2(dmso-
d6)]
+, which is the reason why the second phase in the graph in
Figure 8c is slower. In other words, two parallel competitive processes occur, which deplete
2a, and the consumption of [PtCl(NH
3)
2(dmso-
d6)]
+ is lower in the first phase because
2a has a higher affinity towards [Pt(NH
3)
2(dmso-
d6)
2]
2+. The preferred reaction gives [Pt(
2a)(dmso-
d6)
2]
+ as the product. However, if we take into consideration the graph for [Pt-
2a] in
Figure 8b, where, after 11 h, a reaction of the complex with the solvent, DMSO-
d6, can be assumed. DMSO-
d6 substitutes an amine ligand, as, from the reaction of
2a with [PtCl(NH
3)
2(dmso-
d6)]
+, two molecules of ammonia are released in total, as seen in
Figure 5. As the stoichiometry of the total reaction of
2a with [PtCl(NH
3)
2(dmso-
d6)]
+ indicates the release of two ammonia molecules, it can be concluded that [Pt(
2a)(dmso-
d6)
2]
+ is the final product of the reaction.
The mechanism of the reaction of
2b with cisplatin in DMSO-
d6 is significantly more complex compared to
2a. Graphs for the assessment of reaction orders are depicted in
Figure 9. The change in the inverse concentration value for
2b has two phases, in which it has a linear character (
Figure 9a). In the first 2 h of the reaction, the change can be delineated with the equation: y = 0.3037x + 60.031. The slope coefficient for this phase is substantially lower (
k1 = 0.3037) than for the first phase for
2a (
k1 = 7.704), which means that
2a reacts significantly faster with [PtCl(NH
3)
2(dmso-
d6)]
+, about 25 times. In the first 2 h of the experiment, there is no signal of the coordinated S-CH
3 group, and only the coordination of the ring NH group nitrogen can be observed. It can be assumed that
2b coordinates through nitrogen and sulfur atoms of the thiohydantoin ring to platinum from [PtCl(NH
3)
2(dmso-
d6)]
+, while replacing the chlorido ligand and one amine ligand. This is the reason why
2b reacts more slowly than
2a. Compared to the reaction with
2a,
2b does not coordinate to platinum through the sulfur atom from the methionine side branch initially. The methionine side branch is one methylene longer than the cysteine side branch, which increases the number of rotatable bonds and gives more conformational freedom. Bidentate coordination through the ring nitrogen and side-branch sulfur is less favorable entropically for
2b because there is a bigger loss of conformational freedom by coordination to the side-branch sulfur than with the ring thiocarbonyl sulfur. Besides that, the formation of a five-membered ring by bidentate coordination of
2a is kinetically more favorable than the formation of the six-membered ring, in the case of
2b.
The second phase starts after 2.4 h, and the changes in the inverse concentration values can be delineated with the equation: y = 1.9228x + 55.83. The second phase is six times faster than the first phase (
k1 = 1.9228). Obviously,
2b reacts with [Pt(NH
3)
2(dmso-
d6)
2]
2+ significantly faster than with [PtCl(NH
3)
2(dmso-
d6)]
+. The increased reactivity can be explained by the increased affinity of platinum towards nucleophiles. Namely, by coordination of the second DMSO-
d6 molecule, a chlorido ligand is replaced, which, as a consequence, increases the charge of the complex from +1 to +2 and increases the partial positive charge of platinum. In this case, coordination of
2b occurs through the methionine side-branch sulfur atom, which can be seen in the spectra (
Figure 3), as, after 2 h, the signal of the coordinated S-CH
3 group starts to appear. As after 2 h, the concentration of [Pt(NH
3)
2(dmso-
d6)
2]
2+ is significant, it can also react with
2b and also with the already formed [Pt(NH
3)(dmso-
d6)
2(
2b)]
+ species, yielding a dinuclear complex in which
2b has the role of a bridging ligand.
Change in the inverse concentration values for the newly formed [Pt-
2b] complex has an asymptotic shape up until 8 h, after which it turns linear (
Figure 9b). In the linear part, the curve is parallel to the
x-axis, indicating a pseudo-first-order reaction, that is, a reaction of the formed dinuclear complex with DMSO-
d6. Coordination of the DMSO-
d6 molecule would result in the final dinuclear complex, with the formula [Pt(dmso-
d6)
2(
2b)Pt(dmso-
d6)
3]
3+. Upon analyzing the second-order plot for [PtCl(NH
3)
2(dmso-
d6)]
+, three phases of the reaction course can be noticed (
Figure 9c). The first phase can be marked with the slope equation: y = 23.76x + 73.012. It is apparent that [PtCl(NH
3)
2(dmso-
d6)]
+ has a higher slope coefficient for the first phase (
k1 = 23.76) than
2b (
k1 = 0.3037). As they are in a starting molar ratio of 1:1, it should be expected that their coefficients would be similar. However, a 78-times higher coefficient for [PtCl(NH
3)
2(dmso-
d6)]
+ indicates a reaction with another species, DMSO-
d6 in this case. This is also confirmed by the first-order plot for [PtCl(NH
3)
2(dmso-
d6)]
+ (
Figure 9d), as there is a linear change in the time interval that matches the first phase, which indicates a pseudo-first-order reaction, or a reaction of [PtCl(NH
3)
2(dmso-
d6)]
+ with DMSO-
d6.
2.6. Coordination of DMSO to Platinum, Quantum Mechanical and Kinetic Aspects
In
Section 2.5, it was shown that in the mechanism of the reactions of
2a and
2b with cisplatin, there is a pseudo-first-order phase in which a platinum complex reacts with DMSO-
d6. In order to determine which mode of coordination of DMSO to cisplatin is kinetically favored, a model system for this reaction was constructed (
Figure 11), with the purpose of assessing its Gibbs free energy (Δ
Gr).
The calculations show that the reaction in which DMSO monodentately coordinates to platinum through the oxygen atom is more spontaneous, with a lower reaction Gibbs free energy value of −7.98 kcal/mol. Kinetically somewhat less favorable is the coordination of DMSO to platinum through the sulfur atom (−6.81 kcal/mol).
The quantum mechanical calculations, however, do not fully align with realistic outcomes of these reactions, as it is known that DMSO does indeed coordinate to platinum through sulfur in many cases. The answers to this dilemma might be found in a published
1H NMR kinetic study, which investigates the reaction of the [PtCl
4]
2− complex with DMSO [
27]. The [PtCl
3(dmso-
S)]
− complex, formed in this reaction, further reacts with a second DMSO molecule (
Figure 12). This reaction yields a
trans-[PtCl
2(dmso)
2] complex, in which the two DMSO ligands are in a
trans position one to another because of the strong
trans-effect of the sulfur atom from the first DMSO ligand. The complex isomerizes and transforms over time to a thermodynamically more stable
cis-[PtCl
2(dmso)
2] complex. Based on the chemical shifts in the
1H NMR spectra, it was concluded that, in the
trans isomer, the second DMSO ligand is coordinated through oxygen, which is a kinetically favored product,
trans-[PtCl
2(dmso-
O)(dmso-
S)]. This complex isomerizes to the thermodynamically more stable
cis-[PtCl
2(dmso-
S)
2], in which both DMSO ligands are coordinated through sulfur atoms.
2.7. Final Structures of the Complexes
Based on previous considerations, a conclusion can be made about the structures of the platinum complexes with
2a and
2b. In the first phase of the reaction,
2a does not react with cisplatin but with its DMSO derivative,
cis-[PtCl(NH
3)
2(dmso-
d6-
S)]
+. This phase lasts 2.5 h, and the reaction can be described with Equation (3).
In the obtained complex,
2a is a bidentate ligand, coordinated through nitrogen from the thiohydantoin ring and from the side-branch sulfur atom, forming a five-membered chelate. As the reaction has a pseudo-first-order character for the first 5 h, it is obvious that the main reaction in this interval is the one occurring in parallel to the reaction from Equation (3), which is a reaction of the
cis-[PtCl(NH
3)
2(dmso-
d6-
S)]
+ complex with DMSO-
d6, according to Equation (4).
When looking at the reactivity of
2a, a breaking point is noticed at 2.5 h, after which the speed of its reaction increases. It is obvious that, after 2.5 h, enough
cis-[Pt(NH
3)
2(dmso-
d6-
S)
2]
2+ was formed and that
2a has a higher affinity towards this complex. This reaction can be described with Equation (5).
On the second-order plot for [Pt-
2a] in
Figure 8b, after 11 h, there is a straight line, parallel to the
x-axis, indicating that there is no further change in the quantity of the [Pt-
2a] complex. However, as the concentration of ammonia still continues to increase, it is clear that there is a reaction in which ammonia is released that does not include
2a nor
cis-[Pt(NH
3)
2(dmso-
d6-
S)
2]
2+. As ammonia exists as a ligand in the [Pt(
2a-
N,
Sa)(NH
3)(dmso-
d6-
S)]
+ complex, formed in the first phase of the reaction, it is obvious that this complex reacts with the solvent. The reaction can be described with Equation (6), in which a complex with two coordinated DMSO ligands forms. The coordination of the second DMSO ligand occurs through oxygen, as there is no possibility of interconversion in this case, in which an isomer with two
S-coordinated DMSO ligands would form.
It can be concluded that in the reaction of
2a with cisplatin in DMSO-
d6, two mononuclear complexes are formed, with a bidentately coordinated
2a and two DMSO ligands. The major product is the [Pt(
2a-
N,
Sa)(dmso-
d6-
S)
2]
+ complex, with both DMSO ligands coordinated to platinum through sulfur atoms, while the minor product is the [Pt(
2a-
N,
Sa) (dmso-
d6-
O)(dmso-
d6-
S)]
+ complex, in which one of the DMSO ligands is coordinated through the oxygen atom (
Figure 13). The major product is kinetically favored, while the minor product is thermodynamically more stable.
In the reaction of
2b, the first phase lasts about 2.1 h, in which
2b reacts with the
cis-[PtCl(NH
3)
2(dmso-
d6-
S)]
+ complex, according to Equation (7). In the formed complex,
2b forms a four-membered chelate, coordinating to platinum through the nitrogen and sulfur from the thiohydantoin ring.
In the second phase, the reaction is faster and is followed by coordination through the methionine side-branch sulfur atom, which is confirmed from the
1H NMR spectra. The reason for the increased reaction speed is the formation of the
cis-[Pt(NH
3)
2(dmso-
d6-
S)
2]
2+ complex, which can react with
2b (Equation (8)), but also with the newly formed [Pt(
2a-
N,
S)(NH
3)(dmso-
d6-
S)]
+ complex (Equation (9)).
The mononuclear complex, formed in the reaction in Equation (8), can also react with
cis-[Pt(NH
3)
2(dmso-
d6-
S)
2]
2+, forming a dinuclear complex (Equation (10)) that is different from the one formed in the reaction in Equation (9).
In the second-order plot for [Pt-
2b] in
Figure 9b, there is a straight line, parallel to the
x-axis, starting from 8 h, indicating a reaction with the solvent. In addition to that, there is also a straight line in the second-order plot for
cis-[PtCl(NH
3)
2(dmso-
d6-
S)]
+ in
Figure 9c after 5 h, indicating that this species is not involved in the reaction anymore. Both of the formed dinuclear species undergo reactions with the solvent, which are underlined in Equations (11) and (12).
DMSO ligand substitutes ammonia and coordinates through the oxygen atom because isomerization is impossible. Structures of the final complexes are shown in
Figure 13. The major product is the dinuclear complex with five DMSO ligands in total, one of which is coordinated through the oxygen atom, with the formula [{Pt(dmso-
d6-
S)
2}(
2b){Pt(dmso-
d6-
O)(dmso-
d6-
S)
2}]
3+. The minor product is the dinuclear complex with the formula [{Pt(dmso-
d6-
O)(dmso-
d6-
S)}(
2b){Pt(dmso-
d6-
O)(dmso-
d6-
S)
2}]
3+, in which two of the five DMSO ligands are coordinated through oxygen atoms.
2.9. Molecular Docking
In order to predict binding modes of the major and minor complex products, molecular docking analysis was performed on two different DNA chains. The first DNA target is a dodecamer structure (PDB ID 1BNA) that does not contain an intercalation gap and consists of three regions. The first and last regions consist of four G≡C pairs, while between them is a region with four A=T pairs. Interactions of the complexes and DNA are depicted in
Figure 15. The results indicate that both minor and major [Pt-
2a] products bind in the minor groove of DNA, with [Pt-
2a] minor having a slightly higher binding affinity (−3.01 kcal/mol) than [Pt-
2a] major (−2.89 kcal/mol). However, both dinuclear [Pt-
2b] complexes do not bind in the groove but bind to the phosphate backbone instead, while bridging the minor groove. This binding mode is referred to as “minor groove spanning” in the literature and is considered to be a consequence of strong electrostatic interactions between positively charged complexes and the negatively charged DNA phosphate backbone [
28]. The differences in the binding of mononuclear [Pt-
2a] and dinuclear [Pt-
2b] complexes can possibly originate from the big difference in their total charge. The dinuclear complexes have a significantly greater total positive charge (+3) than the mononuclear complexes (+1), which would be responsible for the difference in their binding affinity towards the negatively charged phosphate backbone. Although the mononuclear complexes are positively charged, they have a higher binding affinity towards minor groove binding. Due to their size, the mononuclear complexes could not bind to the phosphate backbone of two DNA chains simultaneously, but only one instead. This binding mode would considerably reduce its contact (interaction) surface compared to minor groove binding. As binding energy depends on the size of the contact surface of interacting species, it is clear that the inclusion of more fragments of the mononuclear complexes in the interactions with DNA during minor groove binding raises their binding affinities. The mononuclear complexes seem to bind in regions rich in G≡C pairs, with which they form hydrogen bonds, which are additionally strengthened by electrostatic interactions.
Besides the two mentioned DNA binding modes, there are intercalation and covalent binding. The anticancer action of cisplatin is based on the covalent binding of platinum to the nitrogen bases. Intercalation is characteristic of certain compounds that, due to their specific nature and size, can fit between nucleobase pairs and strengthen the stacking architecture of DNA. These compounds mostly have a planar, aromatic, or polycyclic structure that enables them to penetrate between nucleobases. In order to examine whether the investigated complexes have an affinity towards intercalation, a docking analysis with a DNA target with an intercalation gap (PDB ID 1XRW) was performed. The results show that neither the mononuclear nor the dinuclear complexes have an affinity towards intercalation (
Figure 15).
The dinuclear complexes [Pt-
2b] major and minor, which bind to the phosphate backbone of DNA, would not have access to nucleobases and would not have a possibility of platinum coordination. It would then be reasonable to expect that the dinuclear complexes exhibit reduced anticancer activity compared to cisplatin. This is, however, not always the case, as phosphate backbone binders, like TriplatinNC, have been reported to exhibit potent anticancer activity [
29]. Through a binding mode called “backbone tracking”, these complexes form so-called “phosphate clamps” via hydrogen binding to the oxygens of the phosphate groups along the backbone [
30]. This bidentate hydrogen bonding along the backbone causes significant DNA structural distortion and inhibition of normal function, which in turn causes anticancer activity. Nevertheless, this type of binding would not be possible for dinuclear [Pt-
2b] complexes, as they are unable to form these types of hydrogen bonds.
In the case of mononuclear [Pt-
2a] complexes, the picture is a lot less clear. In a related study, based on mass spectrometry, the results pertaining to the binding of cisplatin and its DMSO derivative to double-stranded and single-stranded DNA were reported [
25]. A covalent cisplatin-DNA adduct was identified, but not the DNA adduct with the DMSO derivative. However, a covalent adduct of the DMSO derivative with single-stranded DNA was identified with mass spectrometry. The binding site of the DMSO derivative to DNA did not have a guanine ring, with which it would form a coordination covalent bond, which was possible in the case of single-stranded DNA. Analysis of the binding sites of [Pt-
2a] major and minor revealed that both complexes have a guanine nucleoside in their binding sites (
Figure 16).
Most minor groove binding drugs have a higher binding preference towards A=T-rich sequences [
31,
32,
33]. Minor groove binders that target G≡C-rich sequences have been shown to exhibit reduced activity compared to those that target A=T regions because the presence of guanine’s 2-amino group creates steric hindrance and requires different binding mechanisms, such as side-by-side binding, that can be less stable and efficient [
34]. This might explain the reduced anticancer activity of mononuclear [Pt-
2a] complexes. Overall, there might be other factors that impede their anticancer action, as DNA may not be the primary biomolecular target that [Pt-
2a] complexes interact with within cancer cells.