A Spectroscopic Study on PtCl4(2−) Binding to Rabbit Skeletal Muscle G-Actin

It was found that the binding of PtCl42− to G-actin and the consequent conformational changes are different with those for hard acids. It is a two-step process depending on molar ratio PtCl42−/actin (R). In the first step, R less than 25, the PtCl42− ions are bound to sulfur-containing groups preferentially. These high-affinity sites determined by Scatchard approach are characterized by n1 = 30 with average binding constant K1=1.0×107M-1. The conformational changes are significant as characterized by N-(1-pyrenyl) maleimide(NPM) labeled fluorescence, intrinsic fluorescence and CD spectra. EPR spectroscopy of maleimide spin labeled(MSL) actin demonstrated that even PtCl42−binding is limited to a very small fraction of high-affinity sites(R<1), it can bring about a pronounced change of conformation. In the range of R=25-40, high-affinity sites accessible are saturated. In the second step(R>40) , deep-buried binding sites turn out to be accessible as a result of the accumulated conformational changes. These new binding sites are estimated to be n2=26 with average binding constant K2=2.1×106M-1. Although in this step the quenching of intrinsic fluorescence goes on and the NPM-labled thiols moves to more hydrophilic environment, no change in α-helix content was found. These results suggested that with increasing in PtCl4(2−) binding, the G-actin turns to an open and loose structure in a discontinuous mode.


Introduction
DNA has been generally accepted to be the main target of anticancer complex, cisplatint'2. However, evidences accumulated in recent years support that there are targets other than DNA in the cell. Among them, cytoskeletal system has been found susceptible to the attacking platinum complexes. Kopf-Maier et al t3 and Su t reported separately the disorganization of microtubule, medium microfilament as well as microfilament network caused by cisplatin. Aggarwal also reported the effect of cisplatin on cytoplasmic actin filaments from qualitative histological and ultrastmctural observationst' 6. In our previous workt6, we found the platinum binding to F-actin causes changes in conformation and association, but the mechanism of these effects is still unclear. In the present work, the nature of binding of PtCI" to monomeric actin(G-actin) and the subsequent conformational changes were studied by spectroscopic methods.

Material and Methods
Chemicals ATP(disodium salt), sodium azide, NPM and N-(1-oxyl-2,2,6,6-tetra-methyl-4piperidinyl) maleimide were purchased from Sigma Chemical Co. and used without further purification. All other reagents were of analytical grade and all solutions were prepared with deionized water. A 1.02mM KPtCI solution was prepared and calibrated by means of AAS method. Buffer A was composed of 2mM Tris, 0.2mM CaCI, 0.005% NaN3 and pH8.0.
Purification of G-actin G-actin was extracted from rabbit muscle acetone powder and purified according to the method of Pardee and Spudich [7]. Free nucleotide and 2-mercaptoethanol were removed using a Sephadex G-25 column (preequilibrated with buffer A). G-actin concentration was determined according to the absorbance at 290nm with the absorption coefficient A:90g/m=0.63ts. The level of impurities in the actin is less than 0.5% as determined by SDS-polyacrylamide gel electrophoresis tgl.

Fluorescence Studies
All the steady fluorescence measurements were conducted with a Shimadzu RF-540 spectrophotofluorimeter at room temperature. Slits for both excitation and emission were set at 10nm.

Intrinsic Fluorescence
Samples were prepared by incubating aliquots of G-actin solution (0.4tM) with various amounts of K2PtCI4 for approximately 10h at 4C. The K2PtC1a concentration is varied in the range of 0-40gM. The intrinsic fluorescence was measured with ..=286nm and .=340nm.
The Reaction of PtCl2"-actin Complexes with NPM PtCl,--actin solution was prepared by incubating G-actin (0.41aM) with various volume of KPtCla solution. After standing at 4C for 10h, 71al 3mM NPM solution was added to each of the solutions, and set overnight at 4C, then fluorescence measurements were conducted. After the samples were dialyzed against the buffer A to remove the free NPM, the bound NPM was estimated t from the absorbance at 345nm (for pyrene chromophore) with the molar absorptivity as 4xl 0"cmM .
The Reaction of NPM Labeled G-actin with PtCI 2-G-actin was labeled with NPM by Kawasaki's methodtl, with molar ratio of NPM to actin 1.02. After standing 2-4h at 4C and dialyzing against the buffer A to remove the free NPM, the NPM-actin was diluted to 0.4tM and treated by addition of increasing volume of K2PtCI solution. After standing overnight at 4C, fluorescence intensity was measured with excitation wavelength at 342nm and emission wavelength at 375nm.

CD Measurements
CD spectra were recorded at room temperature on a Jasco-500c Polarizing Spectrophotometer. G-actin concentration was 0.41aM and the KPtCI concentration varied in the range of 0-40tM. The spectra were scanned from 240nm to 190nm and repeated four times at sensitivity 2mcm".

MSL-EPR Measurements
Labeling of G-actin with N-(1-oxyl-2,2,6,6-tetra-methyl-4-piperidinyl) maleimide was carried out as described by Galazkiewicz tl with some modifications. G-actin was treated with maleimide (0.5mol/per mole of G-actin) at 4C for 8h. KCI and MgC12 solutions were added to induce G-actin association and F-actin was then sedimented by ultracentrifugation. The pellet was homogenized and dialyzed against buffer A to remove the unbound dye. After dialysis, the G-actin solution was clarified by ultracentrifugation and concentrated to 5mg/ml. In 100tl concentrated G-actin solution, 0-9tl K2PtCI solution was added and various volume of buffer A was used to keep the final volume at 110tl of all samples. The samples were incubated at 22C for 4h. EPR measurements were performed at 22C using a Bruker ESP-300 instrument. It was operated at scan range 10mT, radiation power 10mW, modulation amplitude 0.23mT, and field modulation 100KHz.

Intrinsic Fluorescence
The titration curve based on intrinsic fluorescence measurement (Fig.l) shows that the fluorescence was quenched by PtCl " continuously with the increasing of PtCl " concentration.

NPM Labeled Fluorescence
The Reaction of PtCl4'-Actin Complexes with NPM It is known that NPM binds to cysteine-374 residue of G-actin specifically. NPM itself is nonfluorescent in aqueous solution but forms strongly fluorescent condensation product with sulfhydryl groups of organic compounds or proteins. If G-aetin was treated previously with K2PtCI and then labeled with NPM, NPM reacted with the sulfhydryl of cys374 unoccupied by PtCI '. The dependence of fluorescence intensity on the concentration of PtCI ', as shown in Fig.2A., has a two-step profile. In the first step, a monotonous decrease in fluorescence intensity was found in the range of molar ratio PtCl-/aetin(R)=0-25, and after then, a plateau appeared up to R=40. The data before the plateau were processed by Scatchard approacha'u as following.
From the linearity of this part, it is reasonable to assume that at low PtCl-concentration, the PtCIbinding sites fall into one category and are mutually independent, the number (v) of PtCI " bound to one G-actin molecule might be expressed as: Here, n=number of PtCl, " binding sites in the first step, k= binding constant of thes binding sites and Cozoneentrations of the unbounded PtCl,2". C is the sum of concentration of the bound and unbound ions: where Cb refers to the concentration of the bound ions and depends on v and the analytical concentration of G-actin, C,: Combining the above equation with Eqns. and 2, we obtain: C=v/k/(n-v)+vC.
(4) We define NPM fluorescence intensity(F) as the conformational change parameter, and assume that all bound ions contribute equally to the change in F. Consequently, there must exist a linear relation between v and F: AF=Fo-F=av (5) The parameter, a, is a measure of the structure change in G-actin by the binding of a single PtCI4 " to G-actin. F0 is the fluorescence intensity at zero PtCI, " concentration. At very high PtCI 2" concentrations, v becomes v.,, and is given by" Vmax=-l'l "-AFmax/a (6) Combining this expression with Eqn.4 and 5, we have C=AF/k,/(AFmax-AF)+C,AF/a (7) AF and C, are variables. A least-square fit of the experimental results gives k,=l.0xl07M " with a=2.1 and n=30. From the data in the second step, the binding constant k and the number of binding sites n2 were obtained as follows.
Cb was supposed to be equal to the analytical PtCI " concentration at the beginning of PtCI binding to those new sites. Fluorescence intensity (F) and Q can be expressed as" Cb=( 106-F)/4.55 With the C values obtained from the equation, the Ct and v were calculated. The best fit analysis showed that PtCI, " binding sites fall in one group in this step (linear regression coefficient r=-0.90) with k=2.1xl06M " and n=26. With increasing PtCI, " concentration from 0 to 10tM (R=0-25), the molar ratio of the conjugated dye to actin was reduced significantly in NPM labeling PtCl'-actin complexes, which is reflected in the reduction of fluorescence intensity. When Ccu" >10pM (R>25), the labeling stoichiometry is essentially unchanged (Fig.2B). The results reveal that Cys374 is involved in the first step(but not in the second step). The number of free thiols accessible to NPM decreased as the result of the occupation of a fraction of Cys374 thiol groups by previous platinum binding.
The Reaction of NPM Labeled G-actin with PtCI, " When labeled G-actin was titrated by KPtCI, the profile of the change in fluorescence intensity (Fig.3) resembles that given in Fig.2A, but the plateau appeared at lower R range (R= 0-20). The [wo-step binding data were analyzed separately by [he me[hods mentioned above. The parameters obtained are as fllowing" in first step, k'=2.3xl 0M", a=2.5 and n'=l 0 in second step, k'=3.2x 0M", n'=20 (r=-0.92)

MSL-EPR Studies
Like many other proteins labeled with maleimide spin label, G-actin exhibits an EPR spectrum composed of two components, indicating an extremely immobile and a highly mobile fraction of the spin label. As shown in Fig.5, P and P are mainly due to very rapidly tumbling spins, whereas P and Ps are due to very slowly umbling labels. In the middle of the spectrum (P) both components are superimposed. A change in the protein structure can be detected either by a change in the correlation times of the two components or by a change in the amplitude ratio of these components. We chose the ratio(Q) of the height of peaks P and P, and the correlation time of rapidly tumbling spin labels to observe the effect of PtCl'on G-actin.
: (in second)is obtained from the following equationS: x=6.5x 0"AHo[(h(0)/h(-1 ))'a-l here AHo is the peak-to-peak distance of P, and h(0), h(-l) are the peak-to-peak heights of the P. and P,, respectively.
Since MSL-EPR is a very sensitive way to get the information of conformational changes induced by the complexes such as PtClfl, it can be used to study interactions between PtCI-'and G-actin in R<I. As shown in Fig.6, Q decreased with increasing R. x varied from 10.5xl0s to 7.9xl0ts.

Discussion
The clear-cut two steps with a plateau in between, as shown in the fluorescence intensity-PtCI v concentration curves( Fig.2A and 3) implicate that the interaction between PtClfl and G-actin has a two-step feature. In the first step, PtClfl" binds to high-affinity sites, which can be characterized by k=l.0xl07M and n=30. The platinum binding induces reduction in m-helix content from the beginning and the exposure of tryptophane residue and the labeled Cys374 to hydrophilic environment. All these changes are related to molar ratio R. At this step, Cys374 residue is involved as deduced by comparing the curves in Fig.2A and   3. The previous occupation of Cys374 by NPM(in Fig.3) blocks PtClflbinding and thus n < n,. Another evidence supporting Cys374 to be one of high-affinity groups is that the labeling molar ratio (NPM/actin) was reduced significantly with increasing PtCIconcentration in this stage.The MSL-EPR enables us to get insight into the effect of PtClfl binding in very low R value. The decreased Q value and x which show that the freedom of thiols increased even in R below 1. It suggests the conformation of G-actin is becoming open from the initial stage of bound PtCI .
The plateau in Fig.2A and 3 suggest that the non-specific binding in this stage did not make Cys374-related eonformational change. The tryptophane residues is facing the more hydrophilie environment; nevertheless, the reduction in t-helix content becomes less pronounced. That means after the high-affinity sites accessible in the first stage are practically saturated, the C-terminal is little affected by further binding, but the whole molecular conformation becomes more opened continuously. The effect is accumulated and the deep-buried binding sites turn out to be accessible. Then the second step PtCIbinding and eonformational changes begin. Generally speaking, in this step, PtCI, binds to sites of low-affinity (k=2. xl06M , n=26). As the results show the NPM-labeled thiols move to hydrophilie environment again. While the quenching in intrinsic fluorescence goes on, no change in helix content was found.
In summary, with increasing in PtCI " binding, the G-acin turns from a somewhat compact protein to a loose one in a discontinuous mode. With NPM-labeled Cys374 and tryptophane residue as the conformation markers, we can observe this two-stepness.
It is quite interested that the two-step process also has been lbund in Cd*(soft acid too)spectrin systemi, but not in Mg , Ca , Ce/(hard acids)-actin systemsa'9. The relation between the two-stepness and the hardness of metal ions is open to be testified.
For the Cd+-spectrin system, lhe behavior is different, although there is two-step feature too. Cd -'* ions mainly bind to the non-thiol groups at the first sep Io lead to the exposure of sulfhydryl groups and further stronger binding occured. Thus we postulate the rigidity of proteins determines the behavior of conformational change. The relatively loose structure of actin facilitates the PtCI binding to sulfur-containing groups at the beginning of binding. However spectrin is a rod shape and compact protein, and it is difficult for metal ions to directly bind to deep-buried thiols, though they are of high affinity to soft acids.