Construction of Salmonella Gallinarum expressing ADI
In the initial phase of this study, we constructed a plasmid expressing and secreting the arcA gene encoding Arginine Deiminase (ADI) from genomic DNA of Lactococcus lactis ssp. lactis (ATCC 7962). The ADI in the Lactococcus lactis was chosen since it has been shown to be highly suppressive to SNU-1 human gastric adenocarcinoma [27]. The arcA gene was placed under TAC promoter. Additionally, we incorporated a 3xFLAG tag immediately after the arcA gene to facilitate the detection and quantification of ADI using a specific antibody [28]. The TAC promoter in S. Gallinarum lacking Lac repressor expresses constitutively any gene cloned downstream (data not shown). For ADI to operate in TME, it must be secreted from tumor-targeted bacteria. We exploited the 22-amino acid PelB leader sequence from Erwinia chrysanthem that allows secretion of the downstream cargo protein [29]. The ADI gene was first inserted into pET20b, which contains the PelB signaling sequence, using the BamHI and XbaI site of the plasmid, generated PelB-ADI. This was subsequently cloned into the KpnI/XbaI sites of the pTAC at the downstream of TAC promoter, generating a plasmid named pTAC_PelB::ADI (Fig. 1A).
We used ΔppGpp Salmonella Gallinarum SGKS1004, as it is attenuated in a mouse model such that i.v. injection of 5x108 Colony Forming Units (CFU) caused no harm at all [30]. The ppGpp-defective mutant was created by disrupting ppGpp synthetase I and II, encoded by relA and spoT genes, respectively [31]. To maintain the plasmid in the absence of selection pressure, the plasmid carried a balanced-lethal host-vector system in which the glmS gene of S. Gallinarum was incorporated into the plasmid pTAC_PelB::ADI while the host carried a mutation in the chromosomal glmS gene locus, as the GlmS− mutant undergoes lysis unless complemented by the glmS gene on the plasmid [30, 32]. For subsequent studies, ΔppGpp S. Gallinarum transformed with the pTAC_PelB::ADI carrying glmS gene was deployed.
Firstly, this strain was grown in LB overnight, and the bacteria were separated from the medium after centrifugation. The bacterial pellet was sonicated, and the soluble fraction was taken for analysis together with the medium as well as the total cell lysate (Fig. 1B). These were subjected to SDS-PAGE gel electrophoresis and probed for detection of ADI by Western analysis using specific antibody (Fig. 1B). ADI was detected in the medium as well as in the total cell lysate and the soluble fraction of the bacterial pellet. Apparently, ADI was expressed from pTAC as a soluble protein and secreted out of bacteria via the PelB leader sequence. To determine the functionality of ADI, the enzyme activity was assessed using a standard protocol [33]. ADI activities in total cell lysate and the soluble fraction of bacterial pellet of 10 ml bacterial culture grown overnight in LB were 660 IU, while that in the bacterial medium was 73 IU. The PelB mediated secretion of ADI seemed to be 11% (Fig. S1). The effect of expressing and secreting PelB-ADI on the growth of S. Gallinarum was evaluated through a comparison of bacterial growth curves between strains carrying the empty plasmid and those expressing PelB-ADI. The expression and secretion of PelB-ADI did not exert any notable impact on bacterial growth kinetics, indicating that the presence of PelB-ADI did not interfere with bacterial proliferation under the experimental conditions employed in this study (Fig. 1C). This was a favorable characteristic for its potential application as a therapeutic agent in cancer treatment.
Anti-cancer effect of ADI on cancer cell lines in vitro
It is widely acknowledged that ADI exerts its inhibitory effects on de novo protein synthesis predominantly in cells with low ASS1 activity while showing a diminished impact on cells with elevated ASS1 activity. To develop this phenomenon for our approach, we meticulously investigated the effects of ADI expressed and secreted by ΔppGpp S. Gallinarum carrying pTAC_PelB::ADI on four distinct cancer cell lines in vitro (Fig. 1D). The chosen cell lines encompassed ASS1-expressing entities, such as B16F10 mouse melanoma cells and HEK-293 human embryonic kidney cells, alongside ASS1-deficient counterparts, including 4T1 mouse breast cancer cells and CT26 murine colorectal carcinoma cells (Fig. 1E). Employing a systematic approach, cells were cultured in 96-well plates and subjected to sequentially diluted bacterial media grown by the ADI-secreting S. Gallinarum. After a 24-hour incubation period, MTT staining was performed, enabling the quantification of the percentage of viable cells. Notably, ASS1-positive cells exhibited a sustained high percentage of viability across all concentrations of tested ADI, underscoring the protective role of ASS1 in mitigating ADI-induced cytotoxic effects. Conversely, ASS1-negative cells experienced a concentration-dependent increase in ADI-induced cytotoxicity, further corroborating the differential impact of ADI on cells with varying levels of ASS1 expression. While treating ASS1-negative cancers with ADI, there was a theoretical possibility for ASS1 to become reactivated and begin producing arginine. To investigate whether ADI treatment could potentially induce the reactivation of ASS1 expression in CT26 mouse cancer cells, we conducted in vitro experiments to assess ASS1 levels following ADI treatment. Our results revealed no significant activation of ASS1 expression in CT26 cells treated with ADI, as demonstrated by Western blot analysis (Fig. 1F). These findings suggest that ADI treatment does not lead to the restoration of ASS1 expression in ASS1-deficient CT26 cancer cells. This provides valuable insight into the notable response of cancer cells to ADI based on their ASS1 status, shedding light on potential avenues for targeted cancer therapies with ADI-expressing bacteria.
Anti-cancer effect of ADI expressed by ΔppGpp S. Gallinarum on grafted tumor in mouse model.
Subsequently, ΔppGpp S. Gallinarum carrying the pTAC_PelB::ADI construct at a dose of 1×108 CFU/mouse was administered to BALB/c mice bearing CT26 colon carcinoma grafts through the intravenous route when tumor volumes reached ~ 100 mm3. As part of the experimental design, a mock control was included, involving the injection of ΔppGpp S. Gallinarum carrying the parental plasmid lacking PelB::ADI, denoted as pTAC. After 5 days post-injection (dpi), tumor tissues were excised and subjected to Western analysis to assess the expression of ADI. A robust presence of ADI was observed in the soluble fraction of tumor tissue derived from mice treated with ΔppGpp S. Gallinarum carrying pTAC_PelB::ADI, in stark contrast to no ADI expression in the mock control mice (Fig. 2A).
Concurrently, at this 5-day juncture, not only were tumor tissues collected but also vital organs like the liver and spleen. Bacterial colony-forming units (CFU) were enumerated, revealing a striking count of over 109 CFU in the tumor tissues of mice treated with ΔppGpp S. Gallinarum carrying pTAC_PelB::ADI or pTAC (mock) (Fig. 2B). In contrast, ~ 5×103 CFU and ~ 1×103 CFU were counted in the spleens and livers, respectively, of the animals treated with ΔppGpp S. Gallinarum carrying either type of plasmid. These results strongly suggested that the expression and secretion of ADI from ΔppGpp S. Gallinarum did not compromise the efficacy of tumor targeting or impede bacterial proliferation within the tumor microenvironment.
Antitumor efficacies were also examined in these animals (Fig. 2C). Notably, mice treated with ΔppGpp S. Gallinarum carrying pTAC_PelB::ADI displayed a significant reduction in tumor size compared to those treated with the bacteria carrying the empty plasmid (mock), the latter still demonstrating notable retardation of tumor growth compared to the PBS control (Fig. 2C, E). Treatment with ΔppGpp S. Gallinarum containing pTAC_PelB::ADI or empty plasmid (mock) did not cause any significant weight loss in the CT26 mice (Fig. 2D). Importantly, this distinct divergence in tumor dynamics translated into a remarkable survival advantage for the group treated with ΔppGpp S. Gallinarum expressing ADI compared to those treated with ΔppGpp S. Gallinarum (mock) or PBS (Fig. 2F). The median overall survival for mice subjected to ΔppGpp S. Gallinarum expressing ADI extended significantly to ~ 34 days, a noteworthy contrast to the ~ 23 days observed in mice treated with the bacteria carrying pTAC (mock). Moreover, in comparison with the PBS control group, which exhibited a median overall survival of ~ 16 days, treatment with ΔppGpp S. Gallinarum expressing ADI conferred an approximately two-fold extension in median overall survival. This result underscored the profound antitumor efficacy of ADI-expressing S. Gallinarum, suggesting its potential as a therapeutic intervention with significant implications for extending survival in CT26 colon carcinoma-bearing mice.
We then designed an experiment to delineate the mechanism underlying the anti-tumor effect of ADI expressed by the bacteria in the tumor microenvironment (TME). Purified ADI (2 IU) was injected directly into tumor tissue every two days, with or without co-injection of ΔppGpp S. Gallinarum carrying pTAC through the intravenous route once at 0 dpi (Fig. 3A). Control experiments included injections of PBS, ΔppGpp S. Gallinarum carrying pTAC, and ΔppGpp S. Gallinarum carrying pTAC-PelB-ADI as before. Tumor sizes in each group of mice were recorded every two days after the treatments. Injection of the purified ADI demonstrated a limited effect only at earlier time points, attributed to the protein's inherent characteristics, including a short serum half-life and marked immunogenicity, as reported previously [22]. Intriguingly, the combined administration of ΔppGpp S. Gallinarum carrying pTAC and purified ADI exhibited a better effect compared to bacteria alone (mock). The most compelling results emerged from the profound impact of ADI expressed by ΔppGpp S. Gallinarum, underscoring its efficacy within the cancer tissue (Fig. 3B). Combination therapy with bacteria as well as treatment with purified ADI did not result in significant weight loss in the treated mice groups (Fig. 3C). As a conclusive measure, on 10 dpi, tumor tissues from all experimental groups were dissected, and their weights were quantified, thereby providing tangible data that substantiates the outcomes of this intricate experiment (Fig. 3D). The results clearly suggested that the observed effect was due to the expression of the ADI by S. Gallinarum on-site in tumor tissue, presumably allowing maintenance of this protein, which otherwise would vanish in a short period.
Enhancement of therapeutic effect of ADI depletion in situ by co-treatment with chloroquine
It has been reported that the depletion of intracellular arginine by ADI induces metabolic stress and subsequent cellular autophagy [34, 35]. The role of autophagy is to protect cancer cells from death, thus playing a cytoprotective role in ADI-based cancer therapy [36]. Chloroquine diphosphate (CQ), renowned for its antimalarial properties, has demonstrated diverse biological effects, including inhibition of cell growth and induction of cell death across various cancer cells [37–41]. Notably, CQ blocks the induction of autophagy [26], thereby disrupting mechanisms that promote tumor survival and growth, potentially enhancing the effectiveness of cancer therapies by amino acid deprivation [42, 43]. We evaluated the therapeutic potential of combined treatment with S. Gallinarum carrying a PelB-ADI-expressing plasmid and CQ in a murine tumor model (Fig. 4).
Following the establishment of CT26 tumors, mice were treated with PBS, CQ alone, bacteria carrying the mock vector (mock), mock plus CQ, bacteria expressing PelB-ADI, or the bacteria expressing PelB-ADI plus CQ. CQ was administered at 50 mg/kg per day, as suggested in previous studies [44], concurrent with bacterial injection (Fig. 4A). While CQ treatment alone or in combination with mock vector-carrying S. Gallinarum revealed some effect due to its anti-cancer activity, the combination of CQ and PelB-ADI-expressing bacteria exhibited greatly enhanced tumor growth inhibition compared to the bacteria expressing PelB-ADI alone (Fig. 4B). Daily chloroquine administration throughout the treatment period did not adversely affect animal weight (Fig. 4C). These findings suggest the potential of combined bacterial therapy and CQ to enhance therapeutic efficacy in cancer treatment.