Biophysical Insights into the Binding Interactions of Inhibitors (ICA-1S/1T) Targeting Protein Kinase C-ι

Abstract

The overexpression of atypical protein kinase C-iota (PKC-ι) is a biomarker for carcinogenesis in various cell types, such as glioma, ovarian, renal, etc., manifesting as a potential drug target. In previous in vitro studies, ICA-1S and ICA-1T, experimental candidates for inhibiting PKC-ι, have demonstrated their specificity and promising efficacy against various cancers. Moreover, the in vivo studies have demonstrated low toxicity levels in acute and chronic murine models. Despite these prior developments, the binding affinities of the inhibitors were never thoroughly explored from a biophysical perspective. Here, we present the biophysical characterizations of PKC-ι in combination with ICA-1S/1T. Various methods based on molecular docking, light scattering, intrinsic fluorescence, thermal denaturation, and heat exchange were applied. The biophysical characteristics including particle sizing, thermal unfolding, aggregation profiles, enthalpy, entropy, free energy changes, and binding affinity (K_d) of the PKC-ι in the presence of ICA-1S were observed. The studies indicate the presence of domain-specific stabilities in the protein–ligand complex. Moreover, the results indicate a spontaneous reaction with an entropic gain, resulting in a possible entropy-driven hydrophobic interaction and hydrogen bonds in the binding pocket. Altogether, these biophysical studies reveal important insights into the binding interactions of PKC-ι and its inhibitors ICA-1S/1T.

1. Introduction

The protein kinase C (PKC) contributes to the malignant progression of several cancers and pushes past some of the cellular defense mechanisms. It is an enzymatic family of proteins that contributes to carcinogenesis through the action of phosphorylating serine/threonine residues of proteins in cancer cascades [1]. These serine/threonine kinases are generally activated by various stimulating factors such as hormones, growth factors, and neurotransmitters [2,3]. There are three classifications within the PKC family: conventional PKC-α, PKC-βI and PKC-βII (splice variant), and PKC-γ; as well as novel PKC-δ, PKC-ε, PKC-η, and PKC-θ; these two isozyme groups need the neutral lipid diacylglycerol (DAG) as their common activator, where the conventional type needs calcium (Ca²⁺) additionally. The atypical PKC-ζ and PKC-ι/λ require acidic phospholipids, phosphatidylinositol-1, 3, 5-(PO₄)₃ (PIP₃), and phosphatidic acid (PA) instead of DAG for their activation [4,5].

Both atypical PKCs participate in modulating cancer-promoting cell signaling path-ways [6]. Several in vitro studies, focusing on glioma, melanoma, ovarian, and renal cancer cells, indicate the presence of overexpression and the relevance of the PKC-ι oncogene in various cellular mechanisms [7,8,9,10]. Although PKC-ι and PKC-ζ have a close homology (72% overall similar sequence and 84% similar in the catalytic domain), PKC-ι is heavily implicated in regulating tumorigenesis. These characteristics make it a known prognostic biomarker and a therapeutic target [6,11,12,13]. In a previous study, Pillai et al. demonstrated that ICA-1T (Figure 1A) an experimental drug, explicitly inhibits the kinase activity of PKC-ι over PKC-ζ in the presence of myelin basic protein (MBP), a common substrate of PKCs. In addition, ICA-1T exhibited an inhibitory effect inversely proportional to the varying concentrations of MBP, indicating a competitive mode of inhibition [12]. Moreover, in previous in vivo toxicological studies with acute and chronic murine models, ICA-1S (Figure 1B), an analog of ICA-1T, has shown reduced toxicity levels [14,15]. These prior developments necessitate further study to obtain insights into the binding interactions between PKC-ι and these specific inhibitors. Therefore, this study was conducted from a biophysical perspective to address this knowledge gap.

Publications reveal that the full-length and catalytic domain of atypical PKCs is best expressed in insect cell lines due to post-translational modifications (PTMs) and the abundance of eukaryotic protein synthesis machinery [16,17]. Messerschmidt et al. found that the expressed catalytic domain of PKC-ι had a higher molecular mass than predicted using mass spectrometry. However, treatment with the lambda-phosphatase followed by mass spectrometry resulted in a lower molecular mass determination of the purified protein domain. This observation suggested that the post-translational modifications resulted in phosphorylation at one, two, or three different sites within the protein structure [18]. Therefore, in this project, to maintain the optimal environment for PTMs of full-length PKC-ι, High Five Cells (hi5), derived from the ovaries of cabbage loopers (Trichoplusia ni, hereafter referred to as Tn cells), and SF9 cells, derived from the ovaries of Spodoptera frugiperda, were utilized [19]. The Tn cells showed better expression levels during the test expression phase (details in the Materials and Methods section); therefore, the PKC-ι obtained for this study was expressed accordingly from Tn cells.

After an in silico screening of 300,000 compounds (MW < 500), 30 potential candidate molecules (Hit), including ICA-1T, a small molecule inhibitor, were selected for specific structural pockets of PKC-ι. In previous investigations, PKC-ι was found to be inhibited by ICA-1T, as determined via in vitro assays and computational analysis. According to the molecular docking results, it was found that ICA-1T interacts with the 469-475 amino acid residues of the catalytic domain of PKC-ι (PDB: 1ZRZ) [12]. In this investigation, besides ICA-1T, we studied the intricate binding interactions and conformational stability profile of PKC-ι while interacting with ICA-1S. The compound is a non-phosphorylated nucleosidic homolog of ICA-1T, which contains a hydroxymethyl group attached in the ‘cyclopentyl’ moiety. We have applied micro scale thermophoresis (MST), nano differential scanning fluorimetry (nanoDSF), dynamic light scattering (DLS), and isothermal titration calorimetry (ITC) to understand the biophysical properties of the interactions and further corroborate our previously published in vivo, in vitro, and in silico findings [10,15]. The pre-conditional validation and protein–ligand complex stability characterization were studied with the state-of-the-art instrument PR.Panta by NanoTemper Technologies. The instrument combines multiple characterization factors such as thermal stability, protein folding status, aggregation, turbidity information, and protein self-association during the protein–ligand conjugate formation [20,21]. This multi-parameter information on protein conformational and colloidal stability is crucial in determining the optimum conditions for the protein ligand affinity study. In PR.Panta control outputs software (NanoTemper Technologies), Onset (T_ON) is a parameter that marks the temperature at which a physical change in the protein molecule, such as unfolding, size increase (cumulant radius), or aggregation, initiates. It must not be exceeded for any subsequent experiments. The inflection point (IP) is the temperature at which these biophysical properties change to their maximum, and the corresponding first derivative plots are used to visualize.

The conformational stability profile of the protein domains due to the introduction of the ligand in the bound state is a crucial factor to understand [22]. Moreover, the buffer system used during experiments must suit the protein-drug affinity study [23]. PR.Panta provides this information on the target protein by utilizing DLS. In brief, the DLS technique (also known as photon correlation spectroscopy) quantifies the Brownian motion of a protein in a solution as the macromolecules are constantly being collided with solvent molecules. The omnidirectional nature and light intensity of the incident laser light on the macromolecule in motion are detected. The continuous motion of the large-sized molecules, such as proteins in solution, will cause a Doppler broadening of both mutually constructive and destructive phases to produce detectable signals. The mathematical correlation of light intensity fluctuations of the scattered light with respect to time in ns-µs level determines the level of rapid intensity fluctuations. This intensity change in scattered light gives insight into the diffusion behavior directly correlated to the homogeneity of the macromolecule [24]. In PR.Panta DLS analysis, the cumulant fit model, first developed by Koppel in 1972, is a frequently used method for DLS data fitting. The autocorrelation function (ACF) was used to obtain a single averaged diffusion coefficient, which results in a single averaged hydrodynamic radius (r_H) and a polydispersity index (PDI). Mathematically, the ACF is fit by a polynomial series expanding around the mean decay rate:

g₂ (τ) – (B + β exp(−2Γτ) (1 + µ₂/2 τ²)²)

where B is the baseline, β is the amplitude, µ₂ is a factor for polydispersity, and τ is the time increment. The mean decay rate Γ is proportional to the mean translational diffusion coefficient D, which calculates r_H [25].

The nano-differential scanning fluorimetry (nanoDSF) mechanism utilizes the intrinsic fluorescence of the aromatic side chains of hydrophobic residues like tryptophan and tyrosine in the protein structure. Unlike conventional DSF, the nanoDSF is a dye-free method that monitors the thermal unfolding as the microenvironment polarity of these hydrophobic residues changes [26]. Moreover, the gradual unfolding of the protein structure exposes the tryptophan amino acid residues from the hydrophobic core, and the fluorescence emission spectra result in a redshift profile. The thermally denatured protein provides the intrinsic fluorescence signals at 350 nm (unfolded state emission maximum) and 330 nm (folded state emission maximum), plotted as a ratio against a temperature range. As the gradual exposure to the fluorescence source intensifies, a fluorescence intensity ratio of 350/330 nm is registered. Therefore, the unfolding profile of a protein based on fluorescence signals against temperature provides insight into the structural stability of a protein in the presence and absence of a ligand in a specific buffer [27]. Consequently, the plot provides the biophysical perspective for understanding the transitions from folded to unfolded state with an inflection point (IP), a parameter often interchangeably used with the melting points (T_m). A multi-domain protein may exhibit more than one unfolding event or transition, as indicated by the IPs. If a protein’s stability changes due to a reason, such as introducing a foreign molecule, then its IP will likely change [28]. In general, the higher the IP, the more stable the protein under investigation. Here, we have shown the protein unfolding and aggregation profiles of the PKC-ι with or without ICA-1S in phosphate-buffered saline, PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, and 1.8 mM KH₂PO₄, pH = 7.4).

The turbidity profile describes the changes in the light scattering pattern due to the cloudiness of a solution caused by the suspended particles [29]. The part of incident light from the light source is transmitted (non-scattered light) and is used for turbidity measurements. In the PR.Panta system, the turbidity is measured using the back-reflected part of the incident light (residual light) attenuation to understand the aggregation behavior of a protein sample. The back reflection technology mainly depends on a non-constant parameter called polarizability, other than constant parameters such as the angle between the detector and light source, the distance between the detector and the particle, the wave number and incident light intensity. Moreover, the intensity of the residual light back-reflected after a light scattering caused by a single particle depends on the refractive index and the diameter of the particle. In a protein sample with aggregates, the incident light will be scattered differently depending on the parameters above for each aggregate. As a result, the rest of the light will proportionally decrease based on the aggregate concentration. Therefore, the back-reflected light was measured in the PR.Panta system is influenced by the refractive index of the protein, particle diameter and aggregate concentrations. An increase in detected mAu indicates an increase in particle size and a decrease in the aggregate concentration. A higher IP in the turbidity profile indicates higher stability of the protein [21]. The turbidity profile of a protein reflects the turbidity average, and the inflection increase parameter for both the free-state and bound-state. The turbidity average is the average of all data points after the inflection point.

MST uses infrared laser-induced localized temperature fields exerted on a molecule of interest during protein–ligand interaction analysis. The (bio-)molecules in motion due to the temperature fields become highly sensitive to the molecule–solvent interface (thermophoretic effect). During a binding event, minuscule changes in conformation, charge, and molecular size induced by the interaction are quantified [30]. When a molecule displays this directed movement due to the temperature gradient, then molecular size, effective surface charge, and hydration entropy are affected. During a binding event, a local change in concentration results in a quantifiable fluorescence variation in the given sample, which is detected in MST as signals. Hence, when a protein–ligand reaches a binding equilibrium, an affinity constant can be measured where a fluorescence excitation/emission module is coupled with an infrared (IR) laser in the experimental setup. A K_d is calculated for the interaction of two hetero molecules from measured normalized fluorescence (F_norm) [31]. Usually, for MST analysis, the ratio of fluorescence values after 1 s and 30 s IR laser heating is recorded with respect to the concentration of the ligand. In this way, the normalized fluorescence values are free from the initial temperature jump (T-jump) and only fluorescence changes due to the change in the concentrations are considered [32,33,34].

In brief, ITC is a label-free method that measures the intricate heat transfer during a protein–ligand or protein–protein interaction in an adiabatic system. In the instrument, changes in heat energy are quantified based on a reference cell filled with water when the other cell contains the sample. The binding partner is injected from an injection syringe. A binding event results in a temperature difference between cells that is compensated by the instrument’s heaters. This heat compensation is represented by the differential power and utilized to obtain the thermodynamic parameters of the binding reaction. The binding event can be exothermic (heat-releasing) or endothermic (heat-absorbing) in nature. Depending on the heat exchange, the thermodynamic parameters such as enthalpy, entropy, free energy of reaction, and affinity can be calculated [35,36].

2. Materials and Methods

2.1. Expression and Purification of PKC-ι

The expression and purification protocol of PKC-ι is proprietary to Bio Bench Technologies (Shijiazhuang Quan Jing Biotechnology Co. Ltd., Shijiazhuang, China). In brief, PKC-ι (Uniprot: P41743) with a 6x-His tag, GS linker, and Strep-tag II was expressed and purified according to the following protocol. First, the gene codon optimization and synthesis were performed. The sub-cloning of the gene into an expression plasmid was performed, followed by a standard maxi prep for further transfection and cell culture. The pFastBac1 was used as the expression plasmid, and the Spodoptera frugiperda (SF9) cell line was used for P1 generation (Figure 2A,B). In this test expression and rapid purification phase, 3 mL of the transfected (Bac to Bac) SF9 cells were cultured at 27 °C and collected after 2–3 days post-induction. After 2–3 days of transfection, 1.5 mL of the transfected culture of SF9 cells was used for P2 generation. For the P2 generation, the SF9 and Tn cell (obtain from Bio Bench, Shijiazhuang Quan Jing Biotechnology Co. Ltd., Shijiazhuang, China) lines were transfected (Figure 2C,D, in the expression and rapid purification test phase). Here, 15 mL of SF9 or Tn cell cultures were induced and kept under the same conditions as above. After 2–3 days of post-transfection, 1.5 mL of the cell culture medium and cells were collected. The cell culture medium was run through rapid purification against the His-tag column. The sample was collected and named as ’Medium’. The cells were collected and lysed using a supersonic system and then centrifuged. The collected supernatant was named ’Nature Protein Extract (NPE). The insoluble fraction was denatured by 8M urea in PBS. The denatured protein was centrifuged and collected as a denatured Protein Extract (DPE). All these samples were detected by SDS-PAGE (AccuPAGE^™ Precast Gel, Bio Bench Technologies) with Coomassie blue staining. P2 generation was harvested and used for the large-scale production of the expression of PKC-ι in the transfected Tn cell line. The best multiplicity of infection (MOI) of approximately 1 was observed at 30 µL after 2 days of cell culture transfection upon testing of the medium, NPE and DPE samples at 30, 150, and 300 µL of volume for 2–3 days of incubation time (Figure 2C,D). For large-scale expression, 1 L of cell culture media of the transfected Tn cell line (P2 generation virus) was incubated at 27 °C and collected after 2 days as a post-transfection sample. The NPE was collected as described above and purified first through a Strep-tag affinity purification system. Sample fractions were collected and loaded onto a His-tag affinity column with Ni²⁺ resin pre-equilibrated with PBS, pH = 7.5. The column was washed with 0 mM, 10 mM, and 30 mM imidazole and wash fractions were collected accordingly for analysis. The column was then eluted with 300 mM imidazole in a PBS buffer, and the protein was collected. His-tag was removed at the TEV site by TEV protease, and the protein purity was analyzed by SDS-PAGE (Figure 2E), and Bradford assays were performed to measure the protein quantity (1.38 mg/L of PKC-ι). The pure protein was then lyophilized and kept at −80 °C until further experiments.

2.2. Molecular Docking

The protein structure of PKC-ι (PDB ID: 3ZH8, resolution 2.74 Å) was prepared using Protein Preparation Module of the Schrödinger Suite [37,38]. All small molecules, crystallization additives and ions present in the original crystal structures were removed. Hydrogen atoms were added, and bond orders and disulfide bonds were assigned. Protein protonation states were assigned by PROPKA (v3.1, University of Copenhagen, Copenhagen, Denmark) at pH 7.4 [39]. Energy minimization was performed in OPLS4 force field (v4, Schrödinger, LLC, New York, NY, USA), with restraint applied to heavy atoms [40]. Following minimization, crystallographic water molecules were removed. Missing loops (residues 448-455 and 538-555 in 3ZH8) were modeled in Prime using the homologous template of PKC-ι (PDB ID: 9EJK8) [41,42,43]. Small molecules ICA-1T and ICA-1S were prepared with LigPrep (Schrödinger Release 2023-4: LigPrep, Schrödinger, LLC, New York, NY, USA, 2025) tool to generate 3D coordinates, assign protonation states at pH 7.4 using Epik (v7, Schrödinger, LLC, New York, NY, USA) and perform compound minimization in OPLS4 force field [44].

Molecular docking was performed using AutoDock Vina and Glide [45,46]. For AutoDock Vina, the prepared protein and ligand structures were converted to PDBQT format using AutoDockTools 1.5.7 (Schrödinger, LLC, New York, NY, USA) [47]. The docking grid was centered on protein residues 469-475, corresponding to an allosteric site previously identified [12]. AutoDock Vina was run 100 times per ligand with an exhaustiveness parameter of 50 to ensure extensive sampling of binding modes. The top-ranked poses from all runs were clustered based on heavy-atom RMSD, using a 2 Å threshold to identify distinct binding conformations. The representative pose was selected as the lowest-energy structure from the most populated cluster. To account for protein flexibility, Induced Fit Docking (IFD) [48,49] with extended sampling was performed using Glide. Initial docking was conducted in Glide to generate poses for the ligand, and residues within 6 Å of the ligand were treated as flexible during Prime side-chain and backbone refinement. The refined binding site conformation underwent Glide redocking using the standard precision (SP) mode. The final binding pose was selected based on the lowest Glide score. The visualizations were prepared using PyMOL (v3.0, Schrödinger, LLC, New York, NY, USA) [49].

2.3. Dynamic Light Scattering (DLS)

DLS measurements were performed using the PR.Panta (Nanotemper Technologies, Watertown, MA, USA). For DLS measurements with a good signal-to-noise, 1 mg/mL of PKC-ι in PBS was used. Samples in the presence and absence of 100 µM ICA-1S were loaded in high-sensitivity premium capillaries and heated with a 1 °C /min temperature ramp from 25 °C to 110 °C with a 2 s DLS integration time. The cumulative radius and polydispersity index (PDI) values were determined using the Prometheus Panta analysis software V1.7.4 (Nanotemper Technologies, Watertown, MA, USA).

2.4. Nano-Differential Scanning Fluorimetry (nanoDSF)

The nanoDSF coupled to scattering measurements were performed with the Prometheus Panta (Nanotemper Technologies, Watertown, MA, USA). PKC-ι was diluted to the indicated concentrations in PBS with a final concentration. Samples were loaded in high-sensitivity premium capillaries and heated with a 1 °C/min temperature ramp from 25 °C to 110 °C.

2.5. Microscale Thermophoresis

The target protein PKC-ι was labeled (lysine-specific) with the fluorophore dye RED-NHS 2nd generation (Cat #MO-L011, NanoTemper Technologies) at a ratio of 1:1.25 (dye: protein). For the initial pre-conditional screening, the concentration of the non-labeled binding partner (ICA-1T) was kept constant at 5 µM, while the concentration of the labeled PKC-ι was tested at 30 nM and 100 nM. The conjugate was incubated for 30 min at room temperature in the dark to complete the protein labeling with the dye, the samples were loaded into premium Monolith NT.115 Capillaries (NanoTemper Technologies, MOK025), and the MST measurements for K_d determination were performed using the Monolith X (NanoTemper Technologies). Using 100 nM of PKC-ι labeled with 80 nM of the dye resulted in complete saturation, and no separation of the unbound fluorophore was observed. For the binding affinity studies, PKC-ι was labeled similarly using the protein labeling kit RED-NHS 2nd Generation. The labeling reaction was performed according to the manufacturer’s instructions in NHS labeling buffer, using a concentration of 10 µM protein (molar dye: protein ratio = 3:1) at room temperature for 30 min. Unreacted dye was removed using the supplied dye removal column, which was equilibrated with PBS containing 0.005% Tween-20, pH 7.4. After 30 min of incubation, followed by centrifugation at 15000 g for 10 min at 4 °C, the degree of labeling was determined using UV/VIS spectrophotometry at 650 and 280 nm. A degree of labeling of 0.8–1.0 was typically achieved. The labeled PKC-ι was adjusted to 30 nM with PBS with 0.005% Tween-20. The ligand was dissolved in PBS with 0.005% Tween-20, and a series of 16 dilutions were prepared using the same buffer, producing ligand concentrations ranging from 2000 µM to 61 nM. For the measurement, each ligand dilution was mixed with 60 nM of one volume of labeled protein PKC-ι, resulting in a final concentration of 30 nM and final ligand concentrations ranging from 1000 µM to 30.5 nM. After 1 h incubation followed by centrifugation at 2500 rpm or 850 g for 3 min, the samples were loaded into Monolith Premium Capillaries. MST was measured using a Monolith X instrument at an ambient temperature of 25 °C. Instrument parameters were adjusted to 100% LED power and medium MST power. Data from three independently pipetted measurements were analyzed (MO. Affinity Analysis software version 3.0, Nanotemper Technologies, Watertown, MA, USA) using the signal from an MST on-time of 5 s.

3. Results

3.1. Molecular Docking of ICA-1T and ICA-1S to PKC-ι

Molecular docking of ICA-1T and ICA-1S was performed to study their binding mode in the allosteric binding pocket of PKC-ι, comprising residues 469-475 [12]. The crystal structure PDB ID: 3ZH8 (resolution 2.74 Å), which is the kinase domain of PKC-ι bound to a thieno [2,3-d]pyrimidine-based chemical inhibitors of aPKCs, was used for AutoDock Vina and Glide [37,45,50]. From the AutoDock Vina docking analysis, ICA-1S and ICA-1T showed binding free energies of –5.31 kcal/mol and –5.50 kcal/mol, respectively, suggesting comparable binding affinities. ICA-1T forms seven interactions with PKC-ι (Figure 3A), among them are five hydrogen bonds (Pro443, Lys468, Arg471 and Ile472). Additionally, the phosphate group of ICA-1T forms a π-cation and a salt bridge interaction to Arg441. ICA-1S forms four hydrogen bonds to PKC-ι residues Pro443, Val447, Lys468, and Arg471 (Figure 3C). According to the Glide standard precision docking study, the Induced Fit Docking (IFD) allowed for flexibility of residues in the binding site, which enclosed around the ligands and improved the compound fit. ICA-1T and ICA-1S showed Glide docking scores of −8.58 kcal/mol and –8.50 kcal/mol, respectively. ICA-1T forms thirteen interactions with PKC-ι (Figure 3B). Among them are seven hydrogen bonds to Arg441, Pro443, Tyr460, Gln469, Arg471, and Arg474. Additionally, the phosphate group forms six salt bridge interactions to three arginine residues: 441, 471, and 474. ICA-1S forms seven hydrogen bonds to PKC-ι residues Ser442, Pro443, Lys468, Gln469, Arg471, and Arg474 (Figure 3D).

3.2. The DLS Analysis Indicates PKC-ι and ICA-1S Homogeneity in the Protein–Ligand Complex

The DLS data on PKC-ι shows a predominantly single polydispersity index (PDI) of 0.30 ± 0.02, indicating a vastly monodispersed and homogeneous status of the PKC-ι (Figure 4). As a macromolecular particle, in the presence of ICA-1S, the target PKC-ι mostly retains its macromolecular homogeneity in solution.

3.3. The NanoDSF Results Indicate a Stable PKC-ι and ICA-1S Complex Formation

Upon observing the intrinsic fluorescence of PKC-ι during thermal denaturation represented by the first derivative of the ratio 350 nm/330 nm, indicates a 0.55 °C shift in the domain-specific unfolding (on Figure 5B, left side, red peak, first domain). In contrast, the other peak remains unperturbed at 50.45 °C (red peak, right side, second domain) before and after the conjugate formation. The triplicate data resulted in a thermal shift that is 3 times larger than the standard deviation of the protein-only sample and can be assumed statistically significant, according to resources from Nanotemper Technologies. This thermal shift indicates a change in protein stability, as observed after a potential binding event. The dashed lines indicate the inflection points, and the dotted lines indicate the T_ON points.

3.4. Turbidity Profile of PKC-ɩ and ICA-1S Indicates a Ligand-Induced Thermal Stability Enhancement

According to the turbidity profile, in the case of the free-state of PKC-ι, the inflection point peaks at 71.90 °C (Figure 6, black dashed line), while the same for the ICA-1S bound state of PKC-ɩ is at 73.30 °C (Figure 6, red dashed line). The higher turbidity IP indicates that the aggregation started 1.4 °C earlier for the free state of PKC-ɩ than in the bound state. The increased steepness of the slope of the inflection points and higher amplitude of the turbidity signal in the free state compared to the complex indicate that PKC-ɩ in the free state aggregates slightly faster than the PKC-ι and ICA-1S conjugate. This characteristic also indicates that a ligand-induced thermal stability enhancement may have occurred due to the binding. In addition, the cumulant radius profiles for the free and bound states were nearly identical, indicating that in both cases, the change in the average particle size began at the same temperature despite the enhanced stability observed in the bound state of PKC-ι (Figure 7).

3.5. Determination of Binding Affinity by MST Using NHS-Ester Labeled PKC-ι

The efficient stoichiometric NHS-ester (N-hydroxysuccinimidyl) labeling of PKC-ι is the prerequisite for a binding interaction study with MST [51]. The NHS-ester group of the labeling dye covalently bonds with the primary amino acid lysine. In MST analysis, if present, the aggregated biomolecules may indicate a protein sample’s poor quality and functionality. The aggregate can form for various reasons, such as the hydrophobic interaction of the incoming molecule (drug) with the partially denatured protein sample, exposing its hydrophobic core. In addition, the pH level equal to the isoelectric point (pI) of the protein may result in nonspecific binding due to the reduction of electrostatic repulsion between the proteins, causing aggregation. An aggregated sample deteriorates the MST analysis and affects the reproducibility of the experiments by increasing the noise level. Here, an initial single-point protein–ligand titration of 30 nM of PKC-ι with 5 µM ICA-1T reveals a clear binding event with sufficient fluorescence intensity.

Initial screening of MST fluorescent count and binding check for ICA-1S and PKC-ι were performed. These pretest MST experiments were performed to obtain a qualitative yes/no answer for developing an ideal MST binding assay. An MST on-time of 1.5 s was used for analysis and a K_d of 1.13 ± 0.23 µM. The MST experiments were conducted at medium MST power at 25 °C with the LED (light-emitting diode) power set to 100%. The signal obtained from this pre-conditional screening exhibits no irregularity, indicating the absence of a large cluster of biomolecules (aggregates) and particles that could traverse through the detection volume of the sample (Figure 8). The absence of any aggregation also indicates that no fluorescence variation, adsorption, or ligand-induced photo-bleaching rate changes were present. Moreover, the addition of detergent (0.005% Tween-20) and the pH value of the buffer were found to be suitable for performing additional MST experiments.

3.6. Assessing the Biophysical Properties of PKC-ι and ICA-1S by Isothermal Titration Calorimetry (ITC)

In this study, buffer suitability, pH and temperature for MST were followed while performing experiments on ITC. Therefore, the PKC-ι and ICA-1S remained in this identical buffer system, and the temperature for ITC experiments was kept at room temperature (25 °C). ITC provides ample information on the direct binding event, in addition to the thermodynamic insights into the nature of the noncovalent forces involved in the binding. In a protein–ligand complex, if a polar interaction is present, the thermodynamic data, such as enthalpy, tend to contribute favorably. In contrast, a hydrophobic interaction is indicated by the entropically driven outcome. Here, the thermodynamic data from PKC-ι and ICA-1S titrations indicate a spontaneous binding interaction with a possible entropy- driven hydrophobic interaction at the binding site (Figure 9).

4. Discussion

Since the development of the experimental drug ICA-1T as a PKC-ι specific inhibitor, several in vitro and in vivo studies have been performed. The early efficacy report of the ICA-1T highlighted its potential as a chemotherapeutic agent in inhibiting neuroblastoma cell proliferation and inducing apoptosis [12]. This potential further inspired more research on the ICA-1T, and ICA-1S was introduced in the scenario. In an in vitro study, the T98G and U87MG glioblastoma cells were treated with ICA-1S in combination with TMZ (Temozolomide), demonstrating a significant inhibitory effect on cellular invasion pathways [7]. Moreover, the acute and sub-acute toxicity studies in the murine models have shown a reduced toxicity level in prostate carcinoma tumors while treated with ICA-1S [14]. Therefore, in these studies, both ICA-1T and ICA-1S demonstrated a potent inhibitory effect on PKC-ι, which regulates biological markers in crucial signaling pathways in various cancer cells. Despite these in vitro and in vivo successes, as a potential therapeutic compound, ICA-1S involved studies were missing vital information on binding characteristics and biophysical properties. In the present study, we report the chemical evidence of binding interactions between PKC-ι and ICA-1S/1T as evidenced by the dissociation constants and biophysical characterizations of the PKC-ι protein in complex with ICA-1S.

Both the AutoDock Vina and Glide molecular docking studies of ICA-1T and ICA-1S to the PKC-ι protein structure (PDB: 3ZH8) identified potential hydrogen bonds as binding interactions. Rigid docking was more successful with this protein structure, where the ICA-1T and ICA-1S were fully accommodated within the binding site of PKC-ι. The Induced Fit Docking (IFD) enabled conformational adaptation of the binding site, facilitating better accommodation of the ligands and mimicking potential ligand-induced fit under physiological conditions. Binding affinities predicted by IFD (Glide) were approximately 3 kcal/mol more favorable than those from rigid docking, indicating that hydrophobic closure and protein flexibility significantly enhance ligand binding.

According to the DLS analysis, the PDI of 0.30 ± 0.02 (Table S1) is close to the monodisperse range (PDI 0.1–0.2). The analysis indicates that the protein size has a hydrodynamic radius of 42.19 ± 0.44 nm in the free state and 43.58 ± 0.46 nm in the ICA-1S-bound state. A slight aggregation or oligomerization of the protein may also be present due to the additional peaks observed in the bound state. This observation may result from the addition of ICA-1S, which may cause the aggregate formation as indicated by a minor peak observed in the bound state (Figure 4, red peak at 626.33 nm). According to NanoTemper Technologies, the PDI value is appropriate for the quantitative DLS analysis, as the higher limit of a sample being inappropriate for DLS analysis is PDI = 1.0. In the nanoDSF data analysis of PKC-ι in the presence or absence of ICA-1S interaction, we observe the presence of domain-specific stages in the unfolding profile. Moreover, the protein thermostability may be increased due to interaction with ICA-1S as indicated by the derived IP (/Tm) shifts of 0.55 °C in the conjugate state (Figure 5A,B). This ligand-induced protein stability in PKC-ι may result from modifications attained in the protein structure. In particular, changes in the binding pocket amino acid residues side chain orientations may contribute to the conformational flexibility of PKC-ι [26,28,53].

In MST analysis (Figure 10 and Figure 11), the binding of the low molecular weight ICA-1S/1T to the full-length multi-domain PKC-ι exhibited a change in the thermophoretic property of the protein due to the fluorophore dye labeling (NHS labeling method) without any optimization. Moreover, the PKC-ι concentration was maintained at 30 nM which was sufficient to perform the assays. The initial concentrations ranged from 500 µM (ICA-1T) and 100 µM (ICA-1S) to the 10⁻⁸ M range. The K_d obtained here for both ligands was under 1 h of incubation. Further experiments optimized for more extended periods (8 and 24 h) were also performed, and the effect of the incubation time was observed. However, longer incubation times did not improve (not reported here) the K_d (s) for the binding interaction between ICA-1S/1T and PKC-ι. MST data reported a direct binding event for both ICA-1S and ICA-1T with the fluorescently labeled PKC-ι. The PKC-ι and ICA-1S complex exhibited a tighter binding (K_d = 9.14 ± 0.19 µM, Figure 10) compared to PKC-ι and ICA-1T (K_d = 173 ± 2.32 µM, Figure 11), as indicated by their dissociation constants.

The thermodynamics data obtained from ITC indicate a binding event in an endothermic manner during the calorimetric titrations. The reactions were found to be spontaneous, as indicated by negative free energy differences (∆G = −10.14 Kcal/mol), while the K_d was found to be in the nM range (36.20 nM). The data indicate that the binding event falls within the strong to moderate range, as per Malvern Panalytical (PEAQ-ITC, Worcestershire, UK). [35,54]. The initial water–water and ligand–buffer control experiments showed no binding event, confirming that the actual protein–ligand binding event was free from any influence of heat, dilution, or buffer mismatch. The binding equilibrium was achieved, as indicated by the inverse sigmoidal shape of the heat of injection vs. molar ratio plot (Figure 9B). In addition, the calculated thermodynamic data from the 20 µM PKC-ι and 2 µM ICA-1S ITC experiment exhibits a positive entropic gain of +299.80 cal/mol.K indicating a possible entropy-driven hydrophobic interaction in the binding pocket. Moreover, a significant positive enthalpy (+79.20 Kcal/mol) contribution was noted, which consists of heat changes due to conformational changes (∆H_conformation), water displacement from the protein surface (∆H_solvation), and a simultaneous noncovalent hydrogen bond and/or Van Der Waals interactions among the interacting molecules (∆H_interaction) [52,55].

The K_d value determined from ITC is much lower than that determined by MST. The differences might arise due to the inherent differences in the ways a reaction is observed. Although in both techniques the observable signals obtained are proportional to the progression of the binding events or the achievement of the point of saturation by the target molecule. However, the experimental setup, data fitting algorithm, and binding models utilized were different in both cases. Moreover, in ITC, the change in energy due to the binding event between PKC-ι and ICA-1S was detected based on the first encounter, and from that point, the concentration of the ICA-1S was considered less than the total ligand, as part of the ligand remained in the bound state. Based on this consideration, the Levenberg–Marquardt algorithm (LMA) resulted in the K_d. However, in the case of PKC-ι and its inhibitors, the MST measurements were taken after a specific incubation period. The Langmuir-like equation was used (K_d model) to determine the dissociation constants. Further experiments with multiple timeframes of incubation periods (up to 24 h) yielded no meaningful K_d. These differences may influence the determination of the actual K_d from these two techniques. Regardless, the MST-derived affinity constants indicate that the functional group differences from ‘methyl dihydrogenphosphate’ (ICA-1T) to ‘hydroxymethyl’ (ICA-1S) in the ‘cyclopentyl’ moiety resulted in a relatively stronger binding affinity (approximately 18-fold) to PKC-ι. This key information on the chemical structure of these PKC-ι specific inhibitors will be further explored for derivatization to create the next generation of ICA-1S/1T analogs. We aim to explore the analogs with long-chain alkyl dihydrogenphosphate, alkyl carboxylic acid and alkyl sulfonic acid functional groups. These analogs are being synthesized in collaboration and will be studied in future. Therefore, for rational drug design, these derivatives of ICA-1S/1T may open a new avenue of structure-activity relationship (SAR) study and enhance our understanding of these inhibitors.

5. Conclusions

ICA-1S and ICA-1T both bind with the target PKC-ι. The conjugate of PKC-ι with ICA-1S results in a thermo stability, which may occur from a local conformational change upon binding. The molecular docking and thermodynamic results support the binding events, indicating entropy-driven, spontaneous hydrophobic interactions and hydrogen bonds between PKC-ι and ICA-1S.