131I-C19 Iodide Radioisotope and Synthetic I-C19 Compounds as K-Ras4B–PDE6δ Inhibitors: A Novel Approach against Colorectal Cancer—Biological Characterization, Biokinetics and Dosimetry

In 40–50% of colorectal cancer (CRC) cases, K-Ras gene mutations occur, which induce the expression of the K-Ras4B oncogenic isoform. K-Ras4B is transported by phosphodiesterase-6δ (PDE6δ) to the plasma membrane, where the K-Ras4B–PDE6δ complex dissociates and K-Ras4B, coupled to the plasma membrane, activates signaling pathways that favor cancer aggressiveness. Thus, the inhibition of the K-Ras4B–PDE6δ dissociation using specific small molecules could be a new strategy for the treatment of patients with CRC. This research aimed to perform a preclinical proof-of-concept and a therapeutic potential evaluation of the synthetic I-C19 and 131I-C19 compounds as inhibitors of the K-Ras4B–PDE6δ dissociation. Molecular docking and molecular dynamics simulations were performed to estimate the binding affinity and the anchorage sites of I-C19 in K-Ras4B–PDE6δ. K-Ras4B signaling pathways were assessed in HCT116, LoVo and SW620 colorectal cancer cells after I-C19 treatment. Two murine colorectal cancer models were used to evaluate the I-C19 therapeutic effect. The in vivo biokinetic profiles of I-C19 and 131I-C19 and the tumor radiation dose were also estimated. The K-Ras4B–PDE6δ stabilizer, 131I-C19, was highly selective and demonstrated a cytotoxic effect ten times greater than unlabeled I-C19. I-C19 prevented K-Ras4B activation and decreased its dependent signaling pathways. The in vivo administration of I-C19 (30 mg/kg) greatly reduced tumor growth in colorectal cancer. The biokinetic profile showed renal and hepatobiliary elimination, and the highest radiation absorbed dose was delivered to the tumor (52 Gy/74 MBq). The data support the idea that 131I-C19 is a novel K-Ras4B/PDE6δ stabilizer with two functionalities: as a K-Ras4B signaling inhibitor and as a compound with radiotherapeutic activity against colorectal tumors.


Introduction
Colorectal cancer (CRC) ranks third in terms of cancer incidence and second in terms of mortality, with over 1.8 million new cases and 881,000 deaths estimated in 2018 worldwide [1]. Approximately 40-50% of colon cancers present activating mutations in the In order to evaluate the binding affinity as well as the anchorage sites of the C19 and I-C19 compounds on the wild-type (K-Ras4B WT -PDE6δ) and mutant (K-Ras4B G13D -PDE6δ) system, docking and molecular dynamics (MD) were performed as previously reported [7]. The complexes formed in molecular docking studies were analyzed by molecular dynamics (MD) simulations to determine the stability. An analysis of the root means squared deviation (RMSD) showed that K-Ras4B WT -PDE6δ-C19, K-Ras4B G13D -PDE6δ-C19, K-Ras4B WT -PDE6δ-I-C19, K-Ras4B G13D -PDE6δ-I-C19 complexes reached stable values between 40 and 60 ns, with values oscillating between 1.9 and 3.2 Å ( Figure 1A). The evaluation of the radius of gyration (Rg) also showed equilibration at similar simulation times, with values of about 22.5 Å ( Figure 1B). Therefore, a further analysis was performed excluding the first 40 ns. To obtain representative protein-ligand conformers, a clustering analysis was performed over the equilibrated simulation time (see methods) considering a 2.5 Å cut-off.
The free binding energies and individual energy terms of complexes started from docked conformations (kcal/mol). The polar (∆Epolar = ∆Eele + ∆Gele,sol) and nonpolar (∆Enon-polar = ∆Evwd + ∆Gnpol,sol) contributions are shown. All energies are averaged over 400 snapshots at time intervals of 100 ps from the last 40 ns MD simulations and are given in kcal/mol (±standard deviation).

I-C19 Radiolabeling with 131 I and Evaluation of the Radiochemical Purity
Radioactive labeling of I-C19 with 131 I (named 131 I-C19) was performed through an isotopic exchange reaction. The radiochemical purity of 131 I-C19 was 95% as evaluated by reversed-phase radio-HPLC, and the retention time was 15.3 ± 2 min ( Figure 3).

I-C19 Radiolabeling with 131 I and Evaluation of the Radiochemical Purity
Radioactive labeling of I-C19 with 131 I (named 131 I-C19) was performed through an isotopic exchange reaction. The radiochemical purity of 131 I-C19 was 95% as evaluated by reversed-phase radio-HPLC, and the retention time was 15.3 ± 2 min ( Figure 3).

In Vitro Cytotoxicity Studies
In this study, we investigated the cytotoxic activity and specificity of the analogous compound I-C19 on the colorectal cancer cell lines HCT116, LoVo and SW620, because all of them contain the K-Ras4B mutation, even though the cell line HCT116 is considered K-Ras4B-independent, while the cell lines LoVo and SW620 are K-Ras4B-dependent [9]. The CCD-18Co cell line was also used as a control of healthy human colon cells, which presented an I-C19 IC 50 of 171.1 µM (data not shown). The HCT116 K-Ras4B-independent cell line showed an IC 50 of 88.24 µM ( Figure 4A), while the K-Ras4B-dependent cell lines had an IC 50 of 18.39 µM in LoVo cells and 68.28 µM in the SW620 cell line ( Figure 4B,C, respectively). The HCT116 and SW620 cancer cell lines did not produce cytotoxic effects in high concentrations (i.e., more than 100 µM; data not shown). This result allowed us to corroborate that the I-C19 compound has a higher affinity than the C19 compound by inhibiting the viability of K-Ras4B-dependent cells at a lower concentration [7]. As shown in the calculation of the binding free energy, I-C19 presented a higher binding energy than C19 (−95.82 vs. −149.23 kcal/mol for C19 and I-C19 in the molecular complex K-Ras4B-PDE6δ, respectively) ( Table 1).
Cancer cells were exposed to 1.44 Bq/cell of 131 I-C19 at a concentration of 8.8, 1.8 and 6.8 µM for HCT116, LoVo and SW620, respectively, and their viability significantly decreased compared to exposure to the I-C19 compound without labeling at the same concentrations after 72 h of treatment ( Figure 4D). Therefore, a synergistic cytotoxic effect of chemotherapy (I-C19) and radiotherapy ( 131 I-C19) was demonstrated when 131 I-C19, at absorbed radiation doses of 0.19, 0.29 and 0.43 Gy (total dose from the cytoplasm and cell membrane contributions to the nuclei) significantly affected the viability of HCT116, LoVo and SW620 cells, respectively ( Table 2). The inhibition of K-ras4B and its signaling pathways is directly proportional to the radiosensitizing effect on cancer cells. Absorbed radiation doses at 0.2 Gy are enough to induce deterministic effects in human cells, for example, the phenotypic changes observed in blood cells. Therefore, our results demonstrated that a dose of approximately 0.2 Gy is sufficient to effectively potentiate I-C19 cytotoxicity (synergistic effect between I-C19 and I-131 radiation).
compound I-C19 on the colorectal cancer cell lines HCT116, LoVo and SW620, because all of them contain the K-Ras4B mutation, even though the cell line HCT116 is considered K-Ras4B-independent, while the cell lines LoVo and SW620 are K-Ras4B-dependent [9]. The CCD-18Co cell line was also used as a control of healthy human colon cells, which presented an I-C19 IC50 of 171.1 µM (data not shown). The HCT116 K-Ras4B-independent cell line showed an IC50 of 88.24 µM ( Figure 4A), while the K-Ras4B-dependent cell lines had an IC50 of 18.39 µM in LoVo cells and 68.28 µM in the SW620 cell line ( Figure 4B,C, respectively). The HCT116 and SW620 cancer cell lines did not produce cytotoxic effects in high concentrations (i.e., more than 100 µM; data not shown). This result allowed us to corroborate that the I-C19 compound has a higher affinity than the C19 compound by inhibiting the viability of K-Ras4B-dependent cells at a lower concentration [7]. As shown in the calculation of the binding free energy, I-C19 presented a higher binding energy than C19 (−95.82 vs. −149.23 kcal/mol for C19 and I-C19 in the molecular complex K-Ras4B-PDE6δ, respectively) ( Table 1).
Cancer cells were exposed to 1.44 Bq/cell of 131 I-C19 at a concentration of 8.8, 1.8 and 6.8 µM for HCT116, LoVo and SW620, respectively, and their viability significantly

DNA Double-Stranded Breaks: γ-H2AX Foci Assay
It is well known that radioactive compounds promote DNA double-stranded breaks (DSBs). DSBs are usually triggered by exogenous chemical agents or ionizing radiation and can lead to genetic instability and apoptosis [10]. DNA damage response (DDR) to ionizing radiation or DNA-damaging chemotherapeutic agents rapidly results in the phosphorylation at serine 139 of histone H2A variant H2AX, named γH2AX. Because γ-H2AX is abundant and correlates well with DSBs, it has been widely used as the most sensitive marker for DSBs for examining the DNA damage produced and the subsequent repair of the DNA lesion [11,12]. Thus, to demonstrate whether the iodinated analogs I-C19 and 131 I-C19 promote DNA DSBs, immunofluorescence assays were performed to detect the foci of the phosphorylated H2AX histone variant (γ-H2AX). The results showed that treatment with I-C19 increased the number of γ-H2AX foci in the cancer cell lines compared with the vehicle ( Figure 5A). However, the DSBs foci were significantly higher when HCT116, SW620 and LoVo cells were treated with 131 I-C19 than with the unlabeled I-C19 ( Figure 5A). The radiotoxic effect of 131 I-C19 was observed at the same level in the K-Ras4B-independent cell line HCT116 as in the LoVo and SW620 K-Ras4B-dependent cells. In K-Ras4B-dependent cell lines (i.e., the SW620 and LoVo cell lines), the γ-H2AX foci increased by 2.3 and 2.1 times with 131 I-C19 than with the I-C19 compound at the same concentration. However, the HCT116 K-Ras4B-independent cell line showed almost three times more γ-H2AX foci under similar conditions ( 131 I-C19 vs. I-C19) ( Figure 5B). Consequently, the effect of I-C19 was combined with the effect of the radioactively to promote DNA damage in the K-Ras4B-dependent and independent colorectal cancer cells as well.

GTP-RAS Pull-Down and Human Phospho-Kinase Array
To determine the impact of the I-C19 analog compound on preventing the binding of RAS to GTP as well as on the activation of important kinases in K-Ras4B signaling, a GST-RAS pull-down and human phospho-kinase array were used. The LoVo and SW620 cell

GTP-RAS Pull-Down and Human Phospho-Kinase Array
To determine the impact of the I-C19 analog compound on preventing the binding of RAS to GTP as well as on the activation of important kinases in K-Ras4B signaling, a GST-RAS pull-down and human phospho-kinase array were used. The LoVo and SW620 cell lines treated with the IC 50 of the I-C19 drug concentration showed an 80% decrease in the concentration of RAS-GTP in the LoVo and the SW620 cell lines ( Figure 6A; right panel). These results correlate with the results of the binding free energy calculations in which the compound I-C19 presented a ∆G binding of −95.82 kcal/mol when it interacted with the molecular complex K-Ras4B-PDE6δ; however, when the compound I-C19 interacted with the complex K-Ras4Bb G13D -PDE6δ (KRAS mutated), the ∆G binding increased (−149 kcal/mol) ( Table 1). As expected, the K-Ras4B-independent HCT116 cell line did not show a decrease in active RAS ( Figure 6A; left panel). Therefore, a human phosphokinase array test was carried out to detect whether I-C19 decreases the phosphorylation of kinases involved in K-Ras4B signaling. In the HCT116 cell line, I-C19 treatment reduced the phosphorylation of AKT1/2/3 by 50% ( Figure 6B). Nevertheless, the I-C19 compound induced an increase in JNK1/2/3 phosphorylation by 155% which, in turn, promoted the phosphorylation of the CREB and c-Jun transcription factors in HCT116 cells. Additionally, I-C19 treatment reduced the phosphorylation of AKT1/2/3 (−57.7%), ERK 1/2 (−53.4%) and AMPka1 (−58.1%) in the LoVo cell line and reduced the inhibitory phosphorylation of GSK3α/β (S21/S9) involved in WNT signaling ( Figure 6C). Moreover, the I-C19 compound reduced the phosphorylation of AKT1/2/3 S473 (−92.6%), AKT1/2/3 T308 (−25%) and AMPka1 (−70%) in the SW620 cell line ( Figure 6D). Most notably, the I-C19 treatment reduced the phosphorylation of the WNK1 oncogene (which is overexpressed in colorectal cancer and is mainly involved in the regulation of metabolism) at the levels of 38.6 and 32.6% in the LoVo and SW620 cell lines, respectively [13]. Indeed, I-C19 treatment reduced the concentration of GTP-RAS and, therefore, decreased the phosphorylation of kinases involved in K-Ras4B signaling in K-Ras4B-dependent cell lines.

I-C19 Compound Suppressed Tumor Growth In Vivo
To determine whether the I-C19 compound induced a decrease in neoplastic cells in vivo, the in situ model of colorectal cancer with AOM/DSS was used. At week 50 after injection of 10 mg/kg of AOM and two cycles of DSS, the histological analysis showed the development of malignant neoplastic lesions of glandular epithelial lineage in mice colorectal tissue. Aberrant glands formed these lesions, and the cells that formed them presented atypical nuclei with evident nucleoli and loss of mucus production. The percentage of mice colorectal tissue affected with these lesions was considered as the percentage of colon adenocarcinoma in mice.
After 12 days of treatment with the vehicle (i.e., PBS 1X with 10%DMSO), the percentage of colon adenocarcinoma in mice was 7-8%, and mice treated with C19 (30 mg/kg) showed a percentage of colon adenocarcinoma of 5-6%. However, I-C19 (30 mg/kg) treatment significantly reduced the rate of colon adenocarcinoma (3-4%) ( Figure 7A). In the AOM/DSS mice model treated with I-C19 for 12 days, no changes were observed in body weight ( Figure 7B). To evaluate the effect of I-C19 on the degree of tumor malignancy and the proliferation of colon cancer cells, the immunohistochemistry assay using carcinoembryonic antigen (CEA) and intrinsic proliferation marker Ki67 was conducted. I-C19 treatment significantly reduced the CEA marker (7.5% of neoplastic tissue) compared to treatment with C19 (12.5% of neoplastic tissue) and the vehicle (65% of neoplastic tissue) ( Figure 7C  To further investigate this point, the LoVo CRC xenograft model was developed in nude mice. Mice were subsequently treated with 30 mg/kg I-C19 and C19 for 12 days, and the tumors were measured with a caliper. As can be seen in Figure 8, from days 10 to 12, I-C19 (~160 mm 3 ) and C19 (~200 mm 3 ) treatments significantly inhibited tumor growth ( Figure 8A), and I-C19 effectively reduced the tumor size by six-fold relative to the control that reached ~1000 mm 3 ( Figure 8B). Moreover, the I-C19 compound did not reduce body weight during the 12 days of treatment ( Figure 8C). In the immunohistochemical analysis, CEA-positive cells increased in tumors treated with the vehicle, and I-C19 treatment significantly reduced the expression of the CEA marker ( Figure 8D,E; upper panel). The treatment of tumor-bearing mice with I-C19 suppressed the proliferation of Ki67-positive cells ( Figure 8D,E; middle panel). Based on our previous study, which found that C19 triggers apoptosis in vivo, we hypothesized that the I-C19 compound might exert To further investigate this point, the LoVo CRC xenograft model was developed in nude mice. Mice were subsequently treated with 30 mg/kg I-C19 and C19 for 12 days, and the tumors were measured with a caliper. As can be seen in Figure 8, from days 10 to 12, I-C19 (~160 mm 3 ) and C19 (~200 mm 3 ) treatments significantly inhibited tumor growth ( Figure 8A), and I-C19 effectively reduced the tumor size by six-fold relative to the control that reached~1000 mm 3 ( Figure 8B). Moreover, the I-C19 compound did not reduce body weight during the 12 days of treatment ( Figure 8C

Biodistribution
Biodistribution studies showed both renal and hepatobiliary excretion of 131 I-C19. The radiopharmaceutical study showed good specific tumor uptake of LoVo cells in athymic mice. In all other organs, the average percentage of injected radioactivity (%IA) of 131 I-C19 decreased in parallel after the administration of 5 MBq (135 µCi/mice) of 131 I-C19 ( Figure 9A,B). The 131 I-C19 radiopharmaceutical was cleared from the blood according to the A h (t) = 89.8e −1.524t + 10.2e −0.033t fitted model, which showed that 89.8% of the activity was rapidly cleared from the blood within 0.45 h (ln2/1.524 = 0.45 h) and the remaining 10.2% was cleared within 21 h (ln2/0.033 = 21 h). I-C19 was also cleared rapidly from the blood as shown in the fitted biokinetic model: q h (t) = 89.8e −1.520t + 0.260e −0.029t in which 89.8% of the activity was rapidly cleared from the blood within 0.46 h (ln2/1.520 = 0.46 h) and the remaining 10.2% was cleared within 24 h (ln2/0.029 = 24 h). The lung and liver were the organs that received the highest absorbed doses of radioactivity: 0.40 and 0.34 Gy, respectively; the colon received 0.14 Gy and the kidney received 0.12 Gy (Table 3). Because the reported maximum tolerated dose for the lungs and liver is 25-40 Gy, severe radiotoxicity was not expected after 131 I-C19 intraperitoneal administration. The radiation dose delivered to the colon was lower than that of the spleen (0.19 Gy) and pancreas (0.18 Gy). The highest absorbed dose was observed in the tumor (3.55 Gy/5 MBq) (

Biodistribution
Biodistribution studies showed both renal and hepatobiliary excretion of 131 I-C19. The radiopharmaceutical study showed good specific tumor uptake of LoVo cells in athymic mice. In all other organs, the average percentage of injected radioactivity (%IA) of 131 I-C19 decreased in parallel after the administration of 5 MBq (135 µCi/mice) of 131 I-C19 ( Figure 9A,B). The 131 I-C19 radiopharmaceutical was cleared from the blood according to the ( ) = 89.8 . + 10.2 . fitted model, which showed that 89.8% of the activity was rapidly cleared from the blood within 0.45 h (ln2/1.524 = 0.45 h) and the remaining 10.2% was cleared within 21 h (ln2/0.033 = 21 h). I-C19 was also cleared rapidly from the blood as shown in the fitted biokinetic model: ( ) = 89.8 . + 0.260 . in which 89.8% of the activity was rapidly cleared from the blood within 0.46 h (ln2/1.520 = 0.46 h) and the remaining 10.2% was cleared within 24 h (ln2/0.029 = 24 h). The lung and liver were the organs that received the highest absorbed doses of radioactivity: 0.40 and 0.34 Gy, respectively; the colon received 0.14 Gy and the kidney received 0.12 Gy (Table  3). Because the reported maximum tolerated dose for the lungs and liver is 25-40 Gy, severe radiotoxicity was not expected after 131 I-C19 intraperitoneal administration. The radiation dose delivered to the colon was lower than that of the spleen (0.19 Gy) and pancreas (0.18 Gy). The highest absorbed dose was observed in the tumor (3.55 Gy/5 MBq) ( Table 3), which means that the administration of 74 MBq (2 mCi) would produce ablative radiation doses (52 Gy/74 MBq).

Genotoxicity Test (Micronucleus Assay)
To determine if the treatment could represent a genotoxic risk (i.e., increased damage to DNA), the structural chromosomal damage (i.e., clastogenicity) or lagging chromosomes in the erythrocytes of bone marrow (i.e., aneuploidogenesis) and the frequency of micronucleated polychromatic erythrocytes (MNPCEs) were evaluated [14]. A pronounced increase in the frequency of MNPCEs was observed with the treatment of 5-fluorouracil (5-Fu) (27 ± 4) compared to the vehicle (7 ± 2) ( Figure 10A,B; left panel). The mean values of the data obtained from 24 h of treatment showed an increase but not a significant number of MNPCEs in the presence of 30 mg/kg of C19 (12 ± 1) and I-C19 (11 ± 2) ( Figure 10A,B; right panel). These data suggest that exposure to C19 or I-C19 in vivo does not increase damage to DNA or the induction of aberrant chromosome distribution into the daughter cells.

Genotoxicity Test (Micronucleus Assay)
To determine if the treatment could represent a genotoxic risk (i.e., increased damage to DNA), the structural chromosomal damage (i.e., clastogenicity) or lagging chromosomes in the erythrocytes of bone marrow (i.e., aneuploidogenesis) and the frequency of micronucleated polychromatic erythrocytes (MNPCEs) were evaluated [14]. A pronounced increase in the frequency of MNPCEs was observed with the treatment of 5fluorouracil (5-Fu) (27 ± 4) compared to the vehicle (7 ± 2) ( Figure 10A,B; left panel). The mean values of the data obtained from 24 h of treatment showed an increase but not a significant number of MNPCEs in the presence of 30 mg/kg of C19 (12 ± 1) and I-C19 (11 ± 2) ( Figure 10A,B; right panel). These data suggest that exposure to C19 or I-C19 in vivo does not increase damage to DNA or the induction of aberrant chromosome distribution into the daughter cells.

Discussion
In a previous work, the therapeutic effect of the C19 molecule against colorectal cancer was demonstrated [7]. That research suggested that C19 interacts with the K-Ras4B-PDE6δ protein complex by increasing the binding energy of the K-Ras4B-PDE6δ complex. The ΔG (binding) of the K-Ras4B-C19-PDE6δ system was −87.81 kcal/mol, blocking its dissociation and functions [7]. More significantly, this compound showed an increase in the binding energy and stabilization of the mutant K-Ras4B G13D -PDE6δ protein complex

Discussion
In a previous work, the therapeutic effect of the C19 molecule against colorectal cancer was demonstrated [7]. That research suggested that C19 interacts with the K-Ras4B-PDE6δ protein complex by increasing the binding energy of the K-Ras4B-PDE6δ complex. The ∆G (binding) of the K-Ras4B-C19-PDE6δ system was −87.81 kcal/mol, blocking its dissociation and functions [7]. More significantly, this compound showed an increase in the binding energy and stabilization of the mutant K-Ras4B G13D -PDE6δ protein complex with a ∆G (binding) of −95.82 kcal/mol, suggesting that this mechanism inhibits the activity of the mutated KRas4B protein. In this work, we demonstrated that the exchange of the chlorine for iodine in the C19 molecule improved the ∆G (binding) of the K-Ras4B-I-C19-PDE6δ system, which increased from −95.82 kcal/mol to −149.23 kcal/mol in the mutant K-Ras4B G13D -I-C19-PDE6δ system (Table 1). This result suggests that I-C19 strongly binds to the mutant K-Ras4B-PDE6δ complex, preventing its dissociation and activity. This highly selective inhibitor of PDE6δ presents greater affinity and produces less nonspecific cytotoxicity compared with other PDE6δ inhibitors, such as deltarasin [15]. In this work, I-C19 showed high specific cytotoxicity; the IC 50 value of the healthy CCD-18Co cell line after treatment with C19 was 60.8 µM [7], and a significantly increased IC 50 was measured (171.1 µM) with the I-C19 treatment (data not shown).
The lowest IC 50 value was obtained in the LoVo cell line (18.39 µM) and was one order of magnitude lower than the IC 50 obtained in healthy human colon cells (171.1 µM). The IC 50 in the SW620 cell line was 2.5 times lower than the IC 50 value obtained in the control cell line (68.28 vs. 171.1 µM). Compared to the IC 50 value in the K-Ras4B-independent HCT116 cell line, the obtained response in LoVo cells was five times higher, and the effect produced in the SW620 cell line was only 1.2 times higher than in HCT116.
The main drawback of cancer therapy (immunotherapy or small molecules) is the poor penetration of drugs into solid tumors, including colorectal cancer tumors. To circumvent this problem, the conjugation of drugs to αor β --emitting radionuclides can significantly improve their therapeutic effects [16]. In this work, I-C19 radiolabeled with 131 I exhibited a higher cytotoxicity at a concentration ten times lower than the IC 50 of unlabeled I-C19 in three different colorectal cancer cell lines ( Figure 4D). Consequently, in Na[ 125 I] uptake trials, it was observed that at 1 h posttreatment, three murine cell lines of CMT93 (colorectal cancer), EMT6 (breast cancer) and CMT64 (lung cancer) displayed significantly increased iodine-conjugate accumulation [17]. The radiolabeled compound 131 I-C19 showed an initial uptake of 3.28-3.6%, while after 3 h of treatment, colorectal cancer cell lines displayed a significantly increased uptake ( Figure 4E). The high uptake of 131 I-C19 at three hours of treatment suggests that the beta emission of [ 131 I] (β-, Emax = 606 keV) could decrease the viability of cancer cells through DNA double-stranded breaks (DSBs). Indeed, 131 I-C19 promoted DNA DSBs, which was evidenced by the significant accumulation of γH2AX foci in the colorectal cancer cell lines compared to the I-C19-treated and untreated cells ( Figure 5A,B). DNA DSBs are highly dangerous lesions with severe consequences for cell survival. DSBs in chromatin promptly initiate the phosphorylation of the histone H2A variant H2AX at serine 139 to generate γH2AX [18]. LoVo cells were more radioresistant than the SW620 KRAS-dependent cells [19], which may explain why the DSB concentrations were two times higher in LoVo than in SW620 cells.
Other studies demonstrate that a β-emitter can significantly decrease the proliferative capacity through the induction of DNA DSBs [20]. The high level of DSBs in HCT116 (six times higher than the control) due to the fact of exposure to 131 I beta radiation was expected, because the uptake did not show statistically significant differences among all evaluated cells. However, the radiation dose absorbed in LoVo cells was higher than that of the HCT116 and SW620 cells (i.e., 0.19, 0.29 and 0.43 Gy/Bq, respectively).
Targeting K-Ras4B by disrupting its cellular localization or posttranslational modifications has successfully harnessed K-Ras4B for clinical therapeutics [21]. The main objective of the I-C19 compound is to inhibit K-Ras4B activation, which occurs as shown in Figure 6A. The phosphorylation statuses of proteins involved in the downstream signaling pathway and other essential pathways in cancer, such as angiogenesis, survival and anti-apoptosis, proliferation, metabolism, chemoresistance and DNA repair, are important [7]. For example, Yang et al. (2021) describes the signaling pathway by which mutated KRAS can cause cells to acquire radioresistance. They demonstrated that KRAS mutations result in a more significant response to DNA damage and the positive regulation of 53BP1 (a protein associated to dsDNA repair) with increased associated nonhomologous end-junction repair (NHEJ). Moreover, KRAS mutations lead to the activation of NRF2 antioxidant signaling to increase the transcription of the 53BP1 gene [22]. Additionally, AKT-dependent WNK1 phosphorylation and enhanced WNK functions (such as increased phospholipase C-β signaling, angiogenesis and cell migration) have been reported. Here, after I-C19 treatment in K-Ras4B-dependent cell lines, AKT and WNK1 decreased phosphorylation ( Figure 6C,D, respectively). Moreover, one of the pathways that is upregulated in colorectal cancer is the mTOR pathway. It has been shown that RHEB (a Ras homolog enriched in the brain) silencing leads to a decrease in mTOR, P70S6K and 4EBP1 phosphorylation and the expression of RHEB, Ki67, mTOR, P70S6K, BCL-2 and PCNA as well as cell proliferation and differentiation [23]. Treatment with I-C19 showed a significant decrease in RAS-dependent pathway activity and, therefore, the phosphorylation of AKT1/2/3, PRAS40, P70S6K and ERK1/2 (essential kinases in anti-apoptosis and proliferation pathways) was diminished. Furthermore, after I-C19 treatment, the phosphorylation of AMP-activated protein kinases, AMPKa1 and AMPKa2, decreased only in the K-Ras4B-dependent cell line ( Figure 6D). Our results showed that I-C19 significantly inhibits the activation of K-ras4b and its signaling pathways in K-ras4b-dependent colorectal cancer cells ( Figure 6A,C,D). However, there is a possibility that I-C19 is inhibiting the binding of K-ras4B to GTP by a pathway other than stabilization of the K-ras4b-PDE6δ complex. Therefore, as future work, it is necessary to demonstrate the activity of the I-C19 compound at the molecular level, for example, by means of isothermal titration calorimetry (ITC) assays with recombinant proteins.
These findings suggest that treatment with I-C19 in combination with radiotherapy can be a powerful tool for treating certain types of K-Ras4B-dependent cancers. Furthermore, we analyzed the therapeutic potential of I-C19 in colorectal cancer in vivo. We used azoxymethane, a rodent colon-specific carcinogen that induces DNA damage and causes proto-oncogene K-Ras4B point mutations and subsequent tumor formation if DNA damage is not repaired or removed [24]. We were able to observe that a lower percentage of I-C19-treated mice developed colorectal adenocarcinomas compared to mice treated with the vehicle or C19 (Figure 7A), suggesting that I-C19 decreases the activation of K-Ras4B in vitro and in vivo and that the radiation delivery system through C19 to colon cancer cells is efficient, as it also reduces CEA ( Figure 7C; upper panel) and Ki67 staining ( Figure 7C; lower panel), which allowed us to confirm that this treatment decreased the malignancy and proliferative capacity of colorectal cancer with mutated K-Ras4B. This experimental evidence agrees with the theoretical results that suggested that I-C19 strongly binds to the mutant K-Ras4B-PDE6δ complex, avoiding its dissociation and activity. These findings are consistent with those reported by Shen and collaborators in 2017 [25], who identified miR-30a as a synthetic lethal agent in K-Ras4B-mutant CRC cells. miR-30a directly targeted ME1 and K-Ras4B and inhibited anchorage-independent growth and in vivo tumorigenesis by K-Ras4B-mutant CRC cells. These treatments showed reduced proliferation and increased apoptosis as indicated by the decrease in Ki67 and increase in cleavage caspase-3 staining [26]. This finding correlates with our results. After I-C19 treatment, we were able to observe a decrease in tumor growth in the LoVo xenograft model. I-C19 decreased malignancy and proliferation and increased apoptosis as demonstrated by the decrease in CEA and Ki67 and the increase in cleavage caspase-3 staining ( Figure 8D,E).
To confirm that I-C19 was taken up by the tumor cells and to evaluate the route of excretion, we performed biodistribution studies with 131 I-C19. We observed higher uptake rates of 131 I-C19 in the tumor and an accumulation in the liver and kidneys, suggesting that the compound is metabolized by the liver and eliminated through the urine. Similarly, biodistribution studies after injection of Na[ 125 I] in mice bearing CMT93 (mice colorectal cancer) allograft tumors resulted in 9.1% ID/g in tumor tissue at 30 min post-injection [17]. In this study, after 1 h, the 131 I-C19 activity concentration in the tumor tissue was 4.4% ID/g. These findings correlate with the reduction in tumor growth in mice treated with I-C19. The highest radiation dose was delivered to the tumor (3.55 Gy/5 MBq).
Finally, the increase in incidences of micronuclei formation is a useful biomarker of genotoxicity, either through the clastogenic mechanism (i.e., direct DNA damage) or through secondary interaction with DNA replication (i.e., indirect aneugenic mechanism) [14]. Cyclophosphamide (CP) is a potent antitumor agent used globally against many forms of human cancers. When mice were treated with 30 mg/kg of CP, the induction of micronuclei was strong [26]. Similarly, when we treated mice with 15 mg/kg of 5-Fu (a widely used chemotherapeutic), a significant increase in MNPCEs occurred ( Figure 10). Al-though this is expected from classical chemotherapeutic drugs, it is a secondary deleterious effect that reduces the quality of life of many patients and can lead to the production of other de novo tumors in patients over time. In contrast, the in vivo induction of micronuclei in PCEs treated with C19 or I-C19 was similar to the levels of micronuclei induction with the vehicle (Figure 10), indicating that the C19 or the I-C19 compounds did not represent an elevated genotoxic risk as do the classical chemotherapeutic agents.

Compounds and In Silico Molecular Analysis
The C19 compound [7] was modified by halogen exchange to synthesize I-C19 ((2S)-N-(2,5-diiodophenyl)-2-[(3,4-dimetoxyphenyl)-methylamine] propanamide), which was acquired from Enamine (Kyiv, Ukraine). In silico docking simulation, free binding energies calculation and molecular dynamics simulations for I-C19 and C19 that formed complexes with K-Ras4B/PDE6δ and K-Ras4B G13D /PDE6δ were performed as previously reported [6,7,27]. The coupling calculations were performed using 2.5 × 10 5 positions for the I-C19 compound on a specific target. Molecular docking was performed using the "matcher" function to produce the initial poses. The results from the London dG score were refined using energy minimization (MMFF94x force field) and rescored using Affinity dG scoring. The AMBER 16 package and the ff14SB force field were used to perform the protein-ligand molecular dynamics simulations. The free binding energies were carried out using the MMGBSA (Amber16 suite) [28], following the procedure detailed in our previous research [6,7,29]. The trajectories were analyzed using cpptraj tool in Amber16 to determine the time-dependence of the RMSD, radius of gyration (RG), and clustering analysis.

I-C19 Radiolabeling with 131 I
The electrophilic radioiodination of the I-C19 compound was performed using chloramine T as an oxidizing agent. Ten batches of 131 I-C19 were prepared using solutions containing 400 µg (700 µmol) of I-C19 to which 12 MBq (50 µL) of [ 131 I]-NaI and 50 µL of chloramine T (1 mg/mL) was added. After 2 min of vigorous agitation, 20 µL of a 15 µM Na 2 S 2 O 5 solution was added to interrupt the reaction. The final product, 131 I-C19, was purified by solid-phase extraction on a Sep-Pak C 18 .

Radiochemical Purity Evaluation
The radiochemical purity of 131 I-C19 was evaluated on a reversed-phase HPLC system (Waters Corporation, Milford, Massachusetts, USA) with a UV photodiode array detector and a radioactivity detector. To separate the components of the mixture (i.e., unlabeled I-C19 and 131 I-C19), a µBondapak C18 column (3.9 mm × 300 mm, 5 µm) was used. A gradient at 1 mL/min of TFA:water (0.1%, v/v, phase A) and TFA:acetonitrile (0.1%, v/v, phase B) was used. The gradient started at 100% of A and was maintained for 3 min before being changed to 50% of A (3-23 min) and, finally, 100% of A was maintained for 27-30 min.

Cell Uptake
The HCT116, LoVo and SW620 cells were seeded in 48-well tissue culture plates at a density of 10 5 cells/well. After 24 h, the cells were exposed to 1.44 Bq/cell of 131 I-C19 (200 µL) at 37 • C for 1, 3 and 5 h. After treatment, two washes (1 mL of ice-cold PBS) were performed. For the calculation of the cell uptake, the cells in each well were treated (at room temperature for 5 min) with 500 µL of glycine buffer (50 mM, pH 2.8). Then, this fraction was removed, and the radioactivity was measured in a NaI (Tl) detector (BCL, San Rafael CA, USA). To evaluate the 131 I-C19 internalization, the cells were incubated with 1 mL of NaOH 0.3 M (at room temperature for 10 min). Once more, the radioactivity was measured in a NaI (Tl) detector (BCL, San Rafael CA, USA). A 100% standard solution (200 µL) of 131 I-C19 was also measured (n = 5). The uptake (%) (plasmatic membrane) and internalization (%) (cytoplasm) were used to obtain the radiation absorbed dose.

I-C19 Cell viability
The CCD-18Co, SW620, LoVo and HCT116 cell lines were plated in 96-well plates at a density of 10 4 cells/well for 24 h to allow for adherence. Then, the cells were treated with the I-C19 compound for 72 h. Cell viability was evaluated by the XTT (2,3-bis[2-methoxy-4nitro-5-sulfophenyl]-2H-tetrazolium-5-carboxyanilide) assay kit (Roche, Germany), based on to the manufacturer's instructions. The required I-C19 concentration, which produced fifty percent cell death at 72 h, was calculated by fitting a model with Prism software (GraphPad Software, San Diego, CA, USA).

131 I-C19 Cell Viability
For the assessment of dual therapy, beta radiation emission from 131 I-labeled C19 was also evaluated. The mitochondrial dehydrogenase activity in living HCT116, LoVo and SW620 cells was considered to measure the contribution of 131 I beta radiation to cell death, and the cell viability analysis was carried out after exposure to concentrations ten times lower than the IC 50 obtained for each cancer cell line (8.8, 1.8 and 6.8 µM, respectively). Cell viability was evaluated by the XTT (2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]-2Htetrazolium-5-carboxyanilide) assay kit (Roche, Germany), based on to the manufacturer's instructions. Briefly, colorectal cancer cell lines at a density of 10 4 cells/well were seeded in 96-well microtiter plates. The viability after treatment was measured 72 h (incubation in 5% CO 2, 85% humidity and 37 • C). The absorbance in each well was recorded in a microplate reader (Epoch; BioTek Instruments; Winooski, VT, USA) at 450 nm. The cells without treatment were considered the control group (100% viability).

Radiation Absorbed Dose to the Colon Cancer Cell Nucleus
For estimation of the radiation absorbed doses to the HCT116, LoVo and SW620 cells' nuclei, Equation (1) was used: where D N←source denotes the mean absorbed dose to the nucleus from source regions (i.e., cell surface-CS; cytoplasm-Cy); N CS and N Cy are the total nuclear transformations occurring in the membrane and cytoplasm, respectively, calculated from the mathematical integration (from t = 0 to t = ∞) of the activity estimated in the internalization and cell uptake experiments; DF N←CS and DF N←Cy represent the dose factors specific for [ 131 I] from the cell surface (plasmatic membrane) and cytoplasm regions to the nucleus configuration. The dose factor geometries were obtained from S values [30] (HCT116 cell radius = 10 µm, nucleus radius = 5 µm; LoVo and SW620 cell radius = 10 µm, nucleus radius = 6 µm).

DNA Double-Stranded Breaks Assay
For the evaluation of DNA double-stranded breaks (DSBs), the phosphorylation of the Ser-139 gamma histone variant H2AX (γH2AX) was used as a biomarker. Cells, diluted to a suitable concentration, were grown on a glass slide after exposure to I-C19 or 131 I-labeled C19 (1.8, 6.8 and 8.8 µM for the cell lines LoVo, SW620 and HCT116, respectively) for 3 h. The anti-phospho-histone H2A.X (ser139) (1:200; CST, Boston, MA, USA) antibody was used as the primary antibody. The Alexa-488-conjugated anti-rabbit IgG was used for the visualization of γH2AX. Afterwards, the slides were stained with DAPI, and the foci were observed with a Leica confocal microscope. To quantify the foci, clear and easily distinguishable dots were counted. At least 30 cells were imaged per slide, and the average number of foci per cell was calculated.

RAS Activation Level Assay
For the evaluation of the RAS activation level after treatment with the I-C19 compound, the RAS-GTP pull-down assay was carried out via Western blotting using the RAS Activation Assay Biochem Kit (BK008; Cystoskeleton, Inc., Denver, CO, USA), according to the standard procedure. Briefly, HCT116, LoVo and SW620 cells were grown in to obtain 3 × 10 6 cells and then lysed in ice-cold lysis buffer (400 µL) supplemented with cOmplete ™ EDTA-free Ultra Protease Inhibitor Cocktail and PhosSTOP ™ (Sigma-Aldrich, St. Louis, MO, USA). Lysates were centrifuged, and the protein (300 µg) was collected. Lysates were incubated by end-over-end rotation with 100 µg of Raf-RBD-conjugated beads for 1 h. The supernatant was removed, and the beads were washed and boiled in Laemmli sample buffer, followed by Western blot analysis using the pan-RAS antibody.

Phospho-Kinase Downstream Activation of K-Ras4B Signaling
To evaluate the status of phospho-kinase downstream activation of K-Ras4B signaling after I-C19 treatment, cells were rinsed with cold PBS and immediately lysed in a buffer supplemented with Complete ™ EDTA-free Ultra Protease Inhibitor Cocktail (Sigma-Aldrich) and PhosSTOP ™ (Sigma-Aldrich) at 4 • C for 30 min. After centrifugation (14,000 g, 5 min), supernatants were transferred to a clean tube, and protein concentrations were quantified using the Precision Red Advanced Protein Assay (Cytoskeleton, Inc., ADV02-A). The lysates were diluted and analyzed using the Human Phospho-Kinase Antibody Array (Proteome Profiler, R&D Systems, Minneapolis, MN, USA). Nitrocellulose membranes were scanned using ChemiDoc TM Imaging Systems (BIO-RAD Laboratories, Inc., Hercules, CA, USA).

Experimental Animals
All mice were supplied by the Laboratory Animals Production and Experimentation Unit of CINVESTAV-Zacatenco. The Institutional Animal Care and Use Committee of CINVESTAV (Protocol No. 0238-17) previously approved the animal trials. The mice were housed in the CINVESTAV animal facility under standard conditions (i.e., temperature 21 • ± 2, humidity of 50% ± 10% and a 12:12 light:dark cycle). At the end of all treatments, the mice were euthanized by CO 2 asphyxiation.

Colorectal Cancer Mouse Model (Azoxymethane/Dextran Sulfate Sodium)
Mice were treated as previously described [24]. Six-week-old C57BL/6 male mice, with a weight of approximately 20 g, were intraperitoneally injected with a single dose of azoxymethane (AOM) (10 mg/kg). One week later, the mice were given a course of 1.5% dextran sulfate sodium (DSS) in sterile drinking water for 7 days, followed by pure drinking water for two weeks for a total of three courses. Mice were then randomly divided into three equal groups: vehicle (10% DMSO in PBS); I-C19 (30 mg/kg); C19 (30 mg/kg). For the control group, the same procedure was performed with normal intraperitoneal saline and pure drinking water instead of the AOM/DSS treatment.

Xenograft Model: Subcutaneous Injection and Preclinical Drug Testing
Six-week-old nu/nu male mice, with weights of approximately 20 g, were used. A suspension of LoVo cells (5 × 10 6 /200 µL PBS) was subcutaneously injected into the right flanks of nude mice. When tumors reached an average volume of 150 mm 3 , the mice were randomly assigned to two groups. The first group was administered with a vehicle (i.e., 10% DMSO in PBS) and the second was administered with I-C19 (i.e., 30 mg/kg). Treatments were administered intraperitoneally, in a final volume of 300 µL, once daily, for 12 days. Tumor volumes were calculated using the formula mm 3 = ( 1 2 d)(D 2 ), with d and D being the shortest and longest diameters, respectively.

Biodistribution and Biokinetic Models
131 I-C19 was intraperitoneally injected (200 µL, 5 MBq (135 µCi/mice)) into nu/nu mice. The mice (n = 3) were euthanized at 0.5, 1, 3, 5, 24, 48, 92 and 144 h post-injection. Whole tumor, lung, liver, pancreas, spleen, kidney, large intestine (colon) and blood were rinsed and placed into pre-weighed plastic test tubes. The activity was determined in a NaI(Tl) detector (Canberra) along with three 0.5 mL aliquots of the diluted standard, representing 100% of the injected activity and used for radioactive decay correction. Mean activity values in each organ were correlated with those of the standard (representing the initial injected activity) to obtain the percentage of injected dose (% ID) at different times. The % ID values of each organ or tumor were input into the Olinda/EXM 2.0 software and adjusted to exponential functions (first-order kinetics) to obtain biokinetic models corrected by radioactive decay (q h (t)) (Equation (2)), which corresponded to the I-C19 compound. Likewise, the A h (t) functions, which corresponded to the 131 I-C19 biokinetic models, were obtained considering λ e f f as = λ B + λ R , where the biological constant is represented by λ B and the 131 I−radioactive decay constant is λ R (Equation (3)).

Radiation Absorbed Dose Estimation
The mathematical integration of A h (t) functions yielded the total nuclear transformations (MBq . s) that occurred in each source organ (Equation (4)).
The absorbed doses delivered by 131 I-C19 to selected organs were estimated based on the following Equation (5): where D target←source is the mean absorbed dose to the target organ from a source organ; DF target←source is a dose factor that considers the fraction of absorbed energy for each nuclear emission of iodine-131 in the different geometries and chemical compositions of the organs. Iodine-131 DF values (mGy/MBq·s) for a 25 g mouse model were obtained from the Olinda code, version 2.2. For the tumor dose calculation, the sphere model was used considering an average tumor mass of 0.1 g (Olinda 1.2 software).

Genotoxicity Assay
To evaluate the genotoxic effect of I-C19, a micronucleus assay with bone marrow cells was carried out according to the method previously described [31]. The test compounds were intraperitoneally administered once, as a solution, at a concentration of 15 mg/kg of 5-fluorouracil (5-Fu) and 30 mg/kg of C19 or I-C19. Bone marrow cells were obtained 24 h after treatment and stained with Giemsa-Wright (Diff-Quick; Harleco; Gibbstown, NJ, USA). Two-thousand polychromatic erythrocytes were counted per animal using an optical microscope at 100× magnification to determine the number of micronucleated polychromatic erythrocytes.

Statistical Analysis
Statistical comparisons were performed with one-way analysis of variance (ANOVA) followed by Dunnett's multiple comparisons test, using GraphPad Prism 5.0 software. The data are shown as the mean ± SEM. A p-value < 0.0001 represent as (****) in the graphs and p-value < 0.01 represent as (*) in the graphs, was considered statistically significant.

Conclusions
Based on in vitro and in vivo evaluations, it is conceivable to design a formulation and continue preclinically and clinically assessing I-C19 and 131 I-C19 as potential molecules for the treatment of K-Ras4B-dependent or K-Ras4B-independent colon cancers.