4. Discussion
IL-18 has been proposed as a candidate for cancer treatment for a long time. Many studies have revealed the role of this cytokine as inducing factor for IFN-γ, which plays an important role in anti-cancer immunity (Castro et al., 2018). However, the limitation of using this cytokine was demonstrated by the low impact of this cytokine in clinical trials (Robertson et al., 2006). This may be due to the presence of natural inhibitor, IL-18 binding protein, that increases in several types of cancers (Srivastava et al., 2010). Interestingly, many research tried to improve both efficiency and stability of this protein, mainly by protein engineering. The popular technique that has been applied to modify this protein is site-directed mutagenesis based on the knowledge of binding mode of IL-18 to its receptor (Kato et al., 2003; S. H. Kim et al., 2001; S.-H. Kim et al., 2002; Saetang et al., 2016; Yamamoto et al., 2004). Most of studies showed the successful development of engineered IL-18 with enhancing activity and bioavailability, especially the development of the decoy-resistant IL-18 (DR-18) that displayed the low affinity to IL-18 binding protein while maintaining the signaling cascade when binding to its receptor (T. Zhou et al., 2020). These all emphasized the role of this cytokine in cancer immunotherapy.
In this work, we tried to improve the aggregation phenomenon of our modified IL-18 found at the high concentration (data not shown). Although the improvement of efficiency of our IL-18 could enhance the satisfactory result from the study, the aggregation has still been the problem because of the unpredictable activity and immunogenicity results. This was revealed by several types of FDA-approved alpha-helical cytokines, such as IL-2 (Fatima et al., 2012), IFN-β-1b (Lipiäinen et al., 2015) or granulocyte colony stimulating factor (G-CMS) (Krishnan et al., 2002) that were found to form multimer precipitation in some condition, especially physiological condition. The aggregation may result in loss of activity or induction of unfavorable immune response. For IL-18, Yamamoto and colleagues reported that the free surface cysteines, including C38, C68, C76 and C127 may contribute to IL-18 aggregation based on the computational structure prediction (Yamamoto et al., 2004). After the replacement of these cysteine by serine and the exposure of oxidative stress, the C-to-S mutations tend to form aggregation lower than wild-type IL-18, especially when all four surface cysteines were replaced (Yamamoto et al., 2004). This is in accordance with our study that when applied all these modifications to IL-18 DM, the fluorescent signal corresponding to protein aggregation was lower compared to IL-18 DM and wild-type. This suggested the contribution of surface cysteine on intermolecular bonding which also found in other cytokines (Karpusas et al., 1997; Lipiäinen et al., 2015). For example, the cysteine to serine substitution was performed for IFN-β-1b, which has been commercially used to treat relapsing forms of multiple sclerosis (MS) to increase the stability of this cytokine (Lin, 1998). The aggregation due to cysteine residue was also found in IL-31 which the formation of disulfide bond occurs at the intracellular levels (Shen et al., 2011). The replacement of these cysteine residues by serine also helped to improve the aggregation phenomenon (Shen et al., 2011). Interestingly, human α-synuclein is another case that emphasized the role of cysteine in protein stability. The aggregation of α-synuclein in Lewy bodies in midbrain dopamine neurons is usually associated with Parkinson’s disease. Importantly, amino acid replacement of tyrosine by cysteine showed the increasing aggregation rate under the oxidative stress led to the cellular toxicity which may be associated with Parkinson’s disease (W. Zhou & Freed, 2004).
In addition to the decrease of aggregation rate under oxidative condition, IL-18 DM1234 showed the higher ability to induce IFN-γ from NK-92MI cell compared to IL-18 DM. This may be caused by the structural changes of IL-18 DM1234 that led to the more suitable form for receptor binding. As our MD simulation revealed, D35 of IL-18 DM1234 changed its direction closer to V125 and S127 of IL-18 receptor compared to IL-18 WT and IL-18 DM. This probably led to the formation of stronger hydrogen bonds between these molecules. Importantly, this interaction was reported to played an important role for stabilizing α-helix structure of IL-18 that mediated the interaction between IL-18 and IL-18 receptor α (Tsutsumi et al., 2014). One another explanation is that the replacement of cysteine by serine may increase the surface polarity since serine showed that less hydrophilic index compared to cysteine (Kyte & Doolittle, 1982) and this scenario may improve IL-18-receptor binding. The reduction of aggregation rate might be another factor that improve the efficacy of this cytokine since it may increase the active form of IL-18 DM1234 in the system.