Phthalimide dithioate derivatives were synthesized according to the procedure reported in our previous report (Zahran et al., 2017), which resulted in two TH analogs including analog 1 [2-(1, 3-Dioxoisoindolin-2-yl) ethyl cyclohexylcarbamodithioate] and analog 2 [2-(1, 3-Dioxoisoindolin-2-yl) ethyl piperidine-1-carbodithioate]. Their chemical structures are shown in Figure 1.

Swiss albino mice (Age: 10 weeks; weight: 26 ± 1.8 g; sex: male) were obtained from the National Research Centre Animal Facility, Cairo, Egypt. The mice were divided into groups (n = 10), housed under standard conditions, and given standard food/water ad libitum. The mice experiments have been approved by the Ethical Committee Board. Tested analogs and CP were dissolved in sterile saline solution. The mice were injected intraperitoneally (IP) with 15 mg/kg body weight of the tested analogs. This dose proves safety in acute toxicity studies (data are not mentioned). Mice groups were injected with tested analogs, 4 h before IP-injection with CP (25 mg/kg body weight). After 1 day from treatment, blood and bone marrow specimens have been collected from mice. Control mice were administered saline by IP-injection, and CP-treated mice (CP-group) were administered IP-injected with CP alone (25 mg/kg body weight). To analyze chromosomal aberration, other groups under the same treatment settings were additionally IP-injected with colchicine (10 mg/kg body weight), 2 h before specimen collection.

Chromosome aberrations analysis has been performed based on the method of Choudhury et al. (2000). Briefly, bone-marrow cells (BMCs) have been harvested from femurs in hypotonic solution. The cell suspension was incubated at 37°C for 20 min, before being centrifuged (1,000 rpm; 10 min). The cells were transferred to cold fixative (acetic acid: methanol; 1:3; V/V) and re-centrifuged before being submitted to chromosome staining and analysis. A hundred spread metaphases were examined per animal. Metaphases with gaps, numerical aberrations (polyploidy), fragments, chromatid breakage, and deletions have been noted.

The same slides that have been stained for the analysis of chromosomal aberrations were utilized to analyze the mitotic index (MI), which depends on counting 1,000 cells per mouse. The number of dividing cells including prophases and metaphases was recorded. MI was calculated as the number of dividing BMCs/1,000 cells.

Both femurs of the mice were separated and the epiphyses were used to isolate BMCs (Valette et al., 2002). BMCs have been suspended in fetal calf serum and spread onto glass slides. BMC smears were set in six replicates for fixation (methanol; 10 min) and staining (May-Grunwald/Giemsa, pH 6). The number of nucleated BMCs and the micronucleated BMCs (MN-BMCs) have been scored according to the nuclei status (mono, bi, and poly) by scoring their count in 1,000 BMCs/mouse (D’Souza et al., 2002).

The comet assay was used to analyze the DNA damage in murine BMCs, based on the protocol of Collins (2004). The image analysis software (comet imager measurements software V2.2, Metasystems, Germany) connected to the fluorescent microscope (Zeiss, Germany; 40×) has been utilized to investigate and evaluate DNA damage/comet. A hundred BMCs/slide were examined. The quantification of tail moment (tail DNAl and mean migration distance of tail) has been performed in comparison to that of BMCs from control and CP-group using the nonparametric Mann–Whitney U test.

The lymphocytes were isolated from the murine peripheral blood according to the method of Chi and Harris (1978). The lymphocytes have been stained with pimonidazole as an anoxic indicator (Challapalli et al., 2017), and then the cells were washed three times to remove excess pimonidazole. The cells have been digested using the cell lysis solution (#LSKCLS500; Merck, United States) supplemented with protease inhibitor cocktail (#P8340; Merck, United States), and the pimonidazole adducts were measured using a microplate fluorometer. In another experiment, the separated lymphocytes of different groups were lysed by the same buffer and the lysates were investigated for hypoxia-inducible factor-1 α (HIF-1α) and HIF-2α concentrations using the HIF-1α ELISA fluorescent kit (#ab229433; Abcam, Germany) and HIF-2-alpha ELISA kit (ab227898; Abcam, Germany).

For the statistical analysis of chromosome aberrations, Chi-square test (2X² contingency table) was carried out. The Student’s t-test was used to analyze normally distributed data of mitotic index, micronucleus test, comet tail migration, and comet head diameter. ANOVA followed by a Tukey’s post hoc (at 99% confidence interval) was used to analyze HIF-1α and HIF-2α results. The differences between groups are considered significant at p < 0.05.

In the current work, the administration of analogs alone has led to a significant disturbance in the chromosomal integrity in murine BMCs. The percentages of chromosomal gaps have been elevated significantly (p < 0.05) by both analogs. The percentages of total chromosomal aberrations, excluding gaps, have increased compared to control by analog 1 (p < 0.05) and analog 2 (p < 0.01). This significant induction is suggested to be due to a cumulative significant induction in fragment breaks, deletions, multiple aberrations, and tetraploidy, (Table 1), while TH treatment has resulted in no chromosomal aberrations. Additionally, the administration of CP alone has resulted in a high increase in the total chromosomal aberrations (p < 0.001) including fragment breaks (p < 0.05), deletions (p < 0.05), multiple aberrations (p < 0.001), and tetraploidy (p < 0.01), compared to the control (Table 1).

The administration of TH, analog 1, or analog 2 to mice before CP resulted in a dramatic induction in the aberration percentages, without gaps, compared to the CP-group. This induction (p < 0.001) of the total chromosomal aberrations has been detected as 168.2, 177.8, and 197.9% in TH/CP, analog 1/CP ,and analog 2/CP groups, respectively (Table 1). It is clear that analog 1 and analog 2 strongly induced the same aberrations types of CP-like gaps, fragments, breaks, deletions, multiple aberrations, and tetraploidy (p < 0.05–p < 0.001). Conclusively, both analogs 1 and 2 enhance the genotoxicity of CP.

The mitotic index has been dramatically diminished in the CP-group (p < 0.01) compared to the control. The BMC proliferation rate has been depressed (p < 0.05) in analog 1 and analog 2 treated groups, compared to the control mice, (Table 1). The rate was remarkably inhibited in TH/CP (p < 0.01), analog 1/CP (p < 0.001) and analog 2/CP groups (p < 0.001), compared to the CP-group, as shown in Table 1. These results suggest that the analogs, to variable degrees, induced CP cytotoxicity and inhibited the mitosis, and that TH exhibited the lowest mitosis suppression, while analog 1 and analog 2 possessed potential inhibitory affinity for the mitotic activity (Table 1).

The MN assay affords a sensitive recognition of structural chromosomal damages in BMCs (Fenech, 2000). MN are minor chromatin bodies, which occur after the fragment condensation of acentric chromosome in the cytoplasm, usually encouraged by clastogenics or spindle-poison in dividing cells (Fenech, 2000). MN incidences have been recognized as a consistent index for chromosome loss and breaks (Lajmanovich et al., 2005).

Table 2 demonstrates that the nuclear impairment was remarkable in the CP-group. The total MN-BMC number has increased significantly (p < 0.001), when the animals were treated with TH, analog 1, and analog 2. MN-BMCs existed together with the variable types of BMCs including mononucleated, binucleated, and polynucleated cells. All of the three types of MN-BMCs have elevated dramatically (p < 0.001) in the CP-group, compared to the control mice. The treatment with TH/Cp, analog 1/Cp, and analog 2/Cp has remarkably increased this mononucleated MN-BMC induction (p < 0.05, p < 0.05 and p < 0.001, respectively). Likewise, binucleated MN-BMCs have elevated significantly (p < 0.05, p < 0.05 and p < 0.01, respectively), while analog 1/Cp and analog 2/Cp showed a remarkable inhibition (p < 0.001) in polynucleated MN-BMCs compared to the CP-group.

Moreover, the intact BMC type (mononucleated cells), plus binucleated and polynucleated types were detected without MN (Table 3). In the CP-group, binucleated and polynucleated BMCs, without MN, showed a noticeable induction (p < 0.001), while the regular mononucleated cells showed a remarkable inhibition (p < 0.05), compared to the control. Mononucleated BMCs, without MN, have significantly declined (p < 0.05) in analogs groups. The binucleated BMCs, without MN, have noticeably elevated in TH/CP (p < 0.05), analog 1/CP (p < 0.05), and analog 2/CP (p < 0.01) groups. Polynucleated BMCs are elevated in the analog 2/CP group, compared to the CP-group. The analogs have enhanced the tetraploid BMC number in the chromosomal aberration analysis and binucleated and polynucleated BMC numbers in CP-stimulated with MN, which suggests that diploid cells with aberrations have impaired cell division as a consequence of interstrand crosslinks (Krishnaswamy and Dewey, 1993; Martínez et al., 2005). These findings indicate the mutagenic efficacy of the analogs.

The comet assay provides a tool to detect DNA damage and multiple DNA impairment types (e.g., single and double-strand breaks, alkali-labile sites, incomplete repair loci, cross-links, and fragmentation) (Azqueta et al., 2020). In the current study, a dramatic induction in DNA damage (p < 0.001) was observed in the CP-group as designated by the comet tail length/tail moment compared to their corresponding control (Table 4). Mice treated with analogs alone showed dramatic changes in the comet pattern. The findings demonstrated that the analogs are strong inducers of the pre-induced-DNA damage that lead to elevated tail lengths and tail moments. Analog 1 possesses the highest damaging affinity to DNA, as shown in Table 4. However, in analogs/CP-group, the results indicated that the analogs stimulate a further elevation of the CP-induced DNA damage that leads to the stimulation of the pre-induced-DNA damage, as concluded from the elevated tail lengths and moments (Table 4).

The previous report suggests that low CP dose interacts with the variable types of tumor microenvironment (TME) cells, including tumor-infiltrating lymphoid cells, endothelial cells, pericytes, and recruited circulating endothelial cells (Picoli et al., 2021). The anti-angiogenic mechanisms produced by CP have been identified by previous experimental studies, resulted in a declined microvessel density and then the enhancement of hypoxia (Shahrzad et al., 2008; Laheurte et al., 2020). The pathways of hypoxic stress response, largely controlled by HIF, are highly involved in the regulation of immune cells function, as a major regulator of immune cell metabolic function. HIF expression and stabilization in immune cells can be triggered by hypoxia among other factors (Chen and Gaber, 2021). Hypoxia status in TME is firmly associated with increased resistance to chemotherapy. Hypoxia induces HIFs that upregulate the expression of many drug-resistant-related genes that ultimately led to tumor resistance to chemotherapy (McAleese et al., 2021). The HIF function is determined by HIF-1a and HIF-2a (Hsu et al., 2020). Therefore, we studied the CP influence on hypoxia status in lymphocytes. In the current study, the treatment with CP resulted in a high hypoxia status as indicated by the formation of pimonidazole adducts and the remarkable induction of HIF-1α and HIF-2α concentrations (p < 0.001), in mice lymphocytes (Table 4). On the other hand, TH and the analogs showed a non-significant change in hypoxia degree compared to control mice, while in analogs/CP-treated mice, the results revealed that only the analogs have inhibited the CP-induced-hypoxia as concluded from the low pimonidazole adducts (p < 0.001), (Table 4). The further investigation of the analogs’ effect on HIF-1α and HIF-2α concentrations indicated that both analogs have a non-significant effect on HIF-2α, on contrary both analog 1 and analog 2 significantly have inhibited HIF-1α concentration (p < 0.01 and p < 0.05, respectively), (Table 4). These findings suggest that inhibition of hypoxia by both analogs is HIF-1α-dependent and that analog 1 is a more potent anti-hypoxic agent.

Several studies had shown that CP exposure enhances intracellular reactive oxygen species (ROS) generation, which consumes the cellular oxygen (Manda and Bhatia, 2003; Emmenegger et al., 2006). CP had been reported to significantly increase HIF-1α protein levels, which was accompanied by an induction in macrophage infiltration and elevated iNOS levels (Viola et al., 2008). These facts are linked to that after radiation treatment whereby nitric oxide induction was responsible for aerobic stabilization of HIF-1α via nitrosylation of a cysteine residue in the oxygen-dependent degradation domain, which prohibited the recognition of the protein by the von Hippel–Lindau complex (Li et al., 2007). As reviewed in Penel et al. (2012), multiple previous studies have reported that the combination of CP with other effective agents markedly enhanced the CP anti-cancer effect through multiple mechanisms in these tumor-bearing mice models including classic cytotoxic drugs (paclitaxel, doxorubicin, and cisplatin), immunostimulant agents (thalidomide (Zhao et al., 2005), lenalidomide, and interleukin-2), molecule-targeting hypoxic cells (tirapazime and hyperthermia) (Emmenegger et al., 2006), and molecular-targeting agents (sunitinib, imatinib, and trastuzumab). Similarly, the current study is an in-line trial to investigate the derivatives of phthalimide (as effective part of TH), which are dithiocarbamate analogs connected through either methylene or ethylene bridges to phthalimide pharmacophoric core, as cytogenotoxic and anti-hypoxic agents to potentiate CP anti-cancer activity. The study findings suggest that both analogs have a high potential to increase CP-genotoxicity and to inhibit CP-hypoxia via the HIF-1α-dependent mechanism. Analog 1 is a potent anti-hypoxic agent, and it is suggested as a promising adjacent CP-complementary agent to induce CP-genotoxicity and to suppress CP-associated hypoxia.

Taken together, the study findings suggest that both analogs have a higher potential to induce CP-genotoxicity than TH and that both analogs inhibit the CP-hypoxia via HIF-1α-dependent mechanism, in which analog 1 is a more potent anti-hypoxic agent than analog 2. Analog 1 is suggested as an adjacent CP-complementary agent to induce CP-genotoxicity and to inhibit CP-associated hypoxia. In conclusion, both analogs have increased CP-stimulated chromosomal aberrations than those induced by TH. The analogs have elevated the cytotoxic effect of CP by inhibiting the mitotic activity, in which analog 2 showed higher inhibition. In the absence of MN, CP has induced bi- and poly-nucleated BMCs. TH and analogs have elevated the CP-stimulated bi-nucleated BMCs, while only analogs have increased the CP-induced poly-nucleated BMCs and inhibited the mono-nucleated BMCs. CP-induced MN-BMCs were accompanied with mono-, bi-, and poly-nucleated cells. Both analogs remarkably have elevated mono- and poly-nucleated MN-BMCs. In the presence of CP, TH and analogs have enhanced mono- and bi-nucleated MN-BMCs. The analogs significantly induce DNA fragmentation, in which analog 1 is the strongest inducer. CP treatment has resulted in a high hypoxia status as indicated by high pimonidazole adducts and HIF-1α/HIF-2α concentrations in lymphocytes. Analogs/CP-treated mice showed low pimonidazole adducts. Both analogs have inhibited HIF-1α concentration but not HIF-2α. Comparing the results of both analogs indicated that analog 2 was a higher inducer of chromosomal aberrations and inhibitor of mitosis than analog 1. On the other hand, analog 1 was a higher inducer of DNA-damage and MN formation and a higher inhibitor of hypoxia than analog 2.