The L-lysine α-oxidase (LO) from Trichoderma harzianum is a flavoprotein from the family of L-amino acid oxidases (LAAOs) [32], an enzyme family extensively studied in medicine for their cytotoxic effects in different tumor cells. However, a challenge for the development of commercial recombinant LAAOs is the large-scale expression and isolation in prokaryotic hosts such as E. coli [33], mainly due to the formation of inclusion bodies [34], production of the inactive [35], and/or low yield [51]. Another difficulty in the heterologous expression is that they are usually expressed as propeptide precursors, rendering the host organism protection from the LAAO toxicity [36]. In this work, we employed several strategies to circumvent these challenges. We produced a recombinant LO from T. harzianum in an E. coli expression system, such as adding the coenzyme FAD to the expression-inducing culture medium and low temperatures during heterologous LO synthesis. The use of low temperatures in bacterial culture reduces system free energy and, therefore, protein synthesis rates, which favors the correct folding [37].
LO gene was cloned into the PGEX-4T1 expression vector, a system in which the enzyme is linked to a glutathione S-transferase (GST) tag, which increases protein solubility. Furthermore, as denaturing conditions are not necessary for any purification step, the expressed protein is more likely to maintain its structure and function. In most cases, the presence of GST does not affect its activity or antigenicity [38]. Belviso et al. produced GST-fused asparaginase and tested the enzyme's toxicity in K562, NALM-6, and MOLT-4 leukemic cells. Meanwhile, the fused enzyme showed high toxicity. The GST control had no impact on cell death [39]. Here, we show that GST had no significant effect on cell viability nor the promotion of apoptotic events, i.e., the toxicity of rLO can only be attributed to rLO activity.
Since the 1950s, it has been documented that cancer cells rely on glycolysis for energy production instead of oxidative phosphorylation, a more efficient process of producing ATP [40]. The amino acid uptake, steady-state levels, and catabolism are all elevated in the leukemia stem cell population. Still, they depend on oxidative phosphorylation and have lower glycolytic reserves than mature cancer cells [41]. Jones et al. demonstrate that LSCs isolated from de novo AML patients uniquely rely on amino acid metabolism for oxidative phosphorylation and survival. Pharmacological inhibition of amino acid metabolism reduces oxidative phosphorylation and induces cell death [42]. Here, we related that rLO was able to promote cell death by necrosis in CD34 + hematopoietic in the cells of leukemia patients.
On the other hand, healthy donor cells showed neither apoptosis nor necrosis following rLO treatment. Eradicating the malignant stem cell pool is the ultimate challenge in treating leukemia. Leukemic stem cells (LSC) hijack the normal hemopoietic niche by increasing the expression of efflux pumps, promoting quiescence [43] and the protection provided by the bone marrow microenvironment [44]. Michelozzi et al. explored the susceptibility of Leukemia Steam Cells to ASNase (L-asparaginase), which was effective against bone marrow cell types CD34 + CD38 + and CD34 + CD38-[45].
The antiproliferative effect of rLO was evaluated in the leukemic cell lines K562 and Jurkat through the PDL assay. K562 cells were more sensitive to rLO treatment as doses higher than 0.5 mU/mL promoted a significant reduction in the growth of treated cells. On the other hand, a significant effect was observed only from 2 mU/mL in Jurkat cells. The treatment of both strains with a concentration of 10 mU/mL completely rendered the cells unviable after 48 hours, preventing the progress of the assay. Our results corroborate the study developed by Zhukova et al. [46], in which the authors analyzed in more detail the effect of LO on the Burkitt lymphoma cells cycle, showing that 1 mU/mL LO prevented the transition from the S to the G2/M phase. At this stage, DNA duplication occurs in the cell nucleus. Several studies show the role of LAAOs in inducing apoptosis, necrosis, and autophagy [47], but their mechanisms of action and the signaling pathways involved remain elusive. In one of the few works on the matter, Pontes et al. (2016) reported the activation of the p38 MAPK and PI3K pathways during the induction of neutrophil death by the LAAO of Calloselasma rhodosthoma [48]. Here, we also evaluated the pathways related to cell death after LO treatment: JNK, ERK, p38, PKC, and PTK. In line with Pontes et al. (2016), we found that the rLO-promoted cell death depends on the activation of JNK and p38 since rLO effects were blunted by inhibitors of these pathways. JNK and p38 are members of the MAPK family of serine/threonine and tyrosine kinases that regulate diverse cellular activities related to cancer development, including proliferation, differentiation, apoptosis, autophagy, and inflammation [49]. In particular, JNK (N-terminal c-Jun kinase) and p38 are proapoptotic pathways in healthy cells, and failure in activating these cascades is strongly involved in carcinogenesis [50].
Perhaps the closest example of rLO is L-asparaginase, which functions similarly in amino acid depletion, although not an L-amino acid oxidase. L-asparaginase has been used for over 40 years in treating ALL and has a well-characterized mechanism of action. Studies show ERK and Akt/mTOR signaling pathways are the main pathways involved in asparaginase-induced apoptosis and autophagy [51]. The first study to evaluate the antitumor effect of LO was published in 1979 by Kusakabe et al., in which they observed that LO inhibits the growth of L5178Y mouse leukemic cells at low doses in vitro. The concentration that promotes 50% inhibition of cell growth (IC50) was 1 mU/mL, and complete inhibition was achieved with 3.3 mU/mL of LO [52]. The antiproliferative effect of LO was also evaluated in a cell cycle study that established the concentration of 1 mU/mL of LO as a preventive of Burkitt´s lymphoma cells' transition from the S phase to the G2 / M phase, in which DNA duplication occurs in the cell nucleus [53]. Besides, an in vitro study with radioactive isotopes showed that L5178Y leukemia cells CaOv human ovarian carcinoma cells and Burkitt lymphoma cells almost completely suppressed the DNA synthesis after being incubated for 40 minutes with LO [54]. The synthesis of RNA and proteins is also affected, with inhibition of approximately 70% [55].
Furthermore, LO from Trichoderma cf. aureoviride Rifai presented cytotoxicity in low concentrations against the following cell lines: K562, LS174T, HT29, SCOV3, PC3, and MCF7 with the IC50 ranging from 3.0×10–6 to 7.8×10–2 U/ml [56]. There are two events promoted by LO enzymatic reaction involved in its cytotoxicity: the depletion of L-lysine in the medium and the production of hydrogen peroxide. Some authors defend that hydrogen peroxide may be considered the primary mechanism of LO cytotoxic effect on neoplastic cells in vitro. They propose that the DNA breakage observed in cancer cells treated with LO is probably due to oxidative stress promoted by H2O2. The pre-incubation with catalase partially prevented the cell growth inhibition effect promoted by LO [57].
On the other hand, studies show that the growth-inhibitory effect of the enzyme on the cells is, at least in part, based on the decrease of L-lysine concentration in the culture medium. The L-lysine concentration is undetectable after 2 hours of incubation of LO. After the external addition of L-lysine, cell growth that was previously inhibited is restored, indicating that lysine depletion plays an important role in the cytotoxic effect of LO. Thus, it is unclear if hydrogen peroxide or amino acid depletion contributes most to the inhibition effect, and more studies must be done. The other sub-products of the enzymatic reaction, delta-piperidine-2-carboxylate, and ammonia, could not promote inhibition in cell growth in vitro (29).
The metabolic rewiring in cancer to support faster growth rates creates a window of intervention to differentiate normal and leukemic cells. This rewiring includes, for example, the glycolytic switch and alternative anaplerotic pathways [58]. Mitochondria are at the core of this adaptation, and the combination of mitochondrial-targeting drugs and radiotherapy has been proposed [59]. In this context, the reduction of mitochondrial function and the triggering of apoptotic pathways, centered in the release of cytochrome c following the decrease in mitochondrial membrane potential, are mediated by p38 phosphorylation and Bax accumulation in the mitochondria [60][61]. Indeed, evasive mechanisms that can maintain mitochondrial function are mediated by the master regulator of mitochondrial biogenesis PGC-1α [62][63](Fig. 7). In line with these findings, we demonstrate that rLO treatment impairs mitochondrial oxidative capacity and decreases PGC-1α expression in Jurkat cells, suggesting a potential accessory clinical intervention. The increase in mitochondrial superoxide production further supports an impairment of mitochondrial integrity, which could be an effect of rLO-produced peroxide. rLO is inactive after 4 hours in culture conditions (data not shown). Therefore, the increase in hydrogen peroxide measured after 24 hours of treatment is due to the surge in cell-generated reactive oxygen species rather than a direct measurement of the enzymatic product. Future work to dissect which outcome of rLO reaction is the major cause of the reduction in cell viability may help determine synergic drugs.