Ligand‐Independent Activation of Aryl Hydrocarbon Receptor and Attenuation of Glutamine Levels by Natural Deep Eutectic Solvent

Natural deep eutectic solvents (NADESs) are emerging sustainable alternatives to conventional organic solvents. Beyond their role as laboratory solvents, NADESs are increasingly explored in drug delivery and as therapeutics. Their increasing applications notwithstanding, our understanding of how they interact with biomolecules at multiple levels ‐ metabolome, proteome, and transcriptome ‐ within human cell remain poor. Here, we deploy integrated metabolomics, proteomics, and transcriptomics to probe how NADESs perturb the molecular landscape of human cells. In a human cell line model, we found that an archetypal NADES derived from choline and geranic acid (CAGE) significantly altered the metabolome, proteome, and transcriptome. CAGE upregulated indole‐3‐lactic acid and 4‐hydroxyphenyllactic acid levels, resulting in ligand‐independent activation of aryl hydrocarbon receptor to signal the transcription of genes with implications for inflammation, immunomodulation, cell development, and chemical detoxification. Further, treating the cell line with CAGE downregulated glutamine biosynthesis, a nutrient rapidly proliferating cancer cells require. CAGE's ability to attenuate glutamine levels is potentially relevant for cancer treatment. These findings suggest that NADESs, even when derived from natural components like choline, can indirectly modulate cell biology at multiple levels, expanding their applications beyond chemistry to biomedicine and biotechnology.


Introduction
Natural deep eutectic solvents (NADESs) are heralded as sustainable alternatives to petroleum-derived solvents. [1]Their unique and customizable properties spur the increasing attention NADESs receive in science, engineering, and biomedicine.Compositionally, NADESs are a eutectic mixture of natural metabolites, [1w] forming a homogeneous liquid at a specific molar ratio under ambient temperature and pressure.Classically, the synthesis of NADESs involved combining a hydrogen bond acceptor (HBA), such as betaine, with a hydrogen bond donor (HBD), such as a geranic acid, in different molar ratios. [2]1y] However, this eco-friendliness and biocompatibility are context-dependent, influenced by the concentration and chemistry of the HBA and HBD and the biology of the target organism. [3]While several studies support the contextdependent toxicity of NADESs, research on NADESs still remains very active, driven by their customizable properties and superior sustainability footprint compared with conventional solvents.
Despite the increasing investigation of NADESs, several knowledge gaps remain.For example, a molecularly informed understanding of how these solvents interact with biomolecules at the cellular level remains unclear.The variously reported toxicity studies are based on cell viability and cytotoxicity assays that do not report molecular-level perturbations at the transcriptomic, proteomic, and metabolomic levels.Considering the increasing interest in NADESs, a detailed dissection of their effect on multiple levels of human cells is critically needed to advance their applications beyond chemistry laboratories.To illustrate, choline-based NADESs, the most explored NADES, [4] feature low-to-moderate toxicity based on cell viability assays.However, how choline-based NADESs interfere with the transcriptome, proteome, and metabolome remains poorly understood, if not unknown.Bridging this gap will likely create the conceptual framework to advance the field.Further, considering that ionic liquids (ILs), another class of neoteric solvent like NADESs, [5] perturb the metabolome, proteome, and transcriptome, [6] it becomes more pertinent to understand how NADESs interact with cell biology.The present work aims to understand how NADESs, exemplified by a choline-based NADES (Choline And GEranic acid (CAGE)), impact human cell biology.We performed multiomic profiling of a CAGE-treated human cell model, the A431 cell line, to understand how the treatment impacts the transcriptome, proteome, and metabolome (Figure 1).Unlike previous studies that use mainly single omics or cell-based bioassays to evaluate cellular response to ILs and NADESs, a multiomics -integrated transcriptomics, proteomics, and metabolomics -enables a more comprehensive interrogation of molecular changes resulting from cells' exposure to NADESs across multiple levels of cell biology.This study will help us to understand how NADES-induced changes trigger biochemical crosstalk between metabolites, proteins, and genes, providing a holistic perspective to ascertain the biocompatibility of NADES and power novel discoveries to advance the application of NADESs.

Results and Discussion
Despite several reports on NADESs, many questions remain unanswered.An overlooked but relevant question is how NADESs interact with biomolecules at multiple levels -metabolome, proteome, and transcriptome -of the human cell.Answering this question is pertinent because many natural components of NADES, including choline, interact with proteins and nucleic acids, influencing protein translation and gene transcription. [7]To address this question, we treated a model human cell, the A431 cell line, with CAGE and employed multiomics to profile intracellular levels of metabolites, proteins, and mRNAs.We opted for CAGE because it is an archetypal choline-based NADESs, the most widely explored NADESs.Also, CAGE is increasingly being investigated as a drug delivery system, [8] enzyme inhibitor, [8a,c] anticancer agent, [9] antidiabetic agent, [10] and antimicrobial agent. [11]As previously reported, [8c,9,11,12] we used a well-established general protocol to synthesize CAGE from one mole of choline bicarbonate and three moles of geranic acid. [9]In our prior work, [9] we show that CAGE reduces the viability of the A431 cell line in a concentration-dependent manner.Therefore, we supplemented the culture medium of the cell line for 7 days with 6.25 μM CAGE (Supporting Information (SI)), a concentration far below the IC 50 of CAGE. [9]We opted for this concentration because it has no impact on cell viability and proliferation, [9] allowing insight into how non-lethal doses of NADESs impact the cell's molecular landscape over an extended time.Notably, a concentration below the IC 50 mitigates cell death, which introduces dead cells' metabolites, proteins, and nucleic acids into the culture medium, [13] confounding the effect of CAGE.Metabolomic, proteomic, and transcriptomic analyses uncovered differential levels of intracellular metabolites, proteins, and mRNAs in CAGE-treated (treated) versus non-treated (control) cells (Figure 2a-c).The metabolomic analysis identified 87 metabolites in the cells, with the intracellular levels of several metabolites differentially elevated or depleted between the treated and control (Figure 2a).Further, proteomic analysis detected 2547 proteins in treated and control, with 19 and 9 proteins exclusive to control and treated, respectively (Figure 2d).Again, we detected 14779 mRNAs in treated and control cells, with 1726 and 817 mRNAs unique to control and treated cells, respectively (Figure 2d).Also, 2265 of the detected mRNAs were translated into proteins in both treated and control cells (Figure 2d).The difference between the number of detected proteins and mRNAs is not uncommon because mRNA levels alone do not determine protein levels. [14]14a] Enrichment analysis using MetaboAnalyst 5.0 [15] shows that depleted metabolites in the treated cells are associated with ammonia recycling, nicotinate and nicotinamide metabolism, and glutamate metabolism (Figure 2e).Depleting metabolites related to ammonia recycling and glutamate metabolism has implications for rapidly proliferating cancer cells.For instance, cancer cells recycle ammonia into the central amino acid metabolic pathway for efficient nitrogen utilization. [16]Therefore, CAGE's ability to deplete L-glutamine (Figure 2a) that generates ammonia has therapeutic relevance.Further, the elevated metabolites in the treated cells were associated with pathways such as purine, methionine, and betaine metabolism (Figure 2f).Purine metabolism involves guanine, guanosine, hypoxanthine, inosine, xanthine, and uric acid that are elevated, although not significantly, (Table S2), suggesting a bias toward purine catabolism in treated cells.Purine catabolism plays a vital role in nitrogen metabolism, providing a carbon and nitrogen source for the biosynthesis of biomolecules.Perhaps, purine catabolism increased to compensate for the depleting levels of L-glutamine, another critical source of carbon and nitrogen in proliferating cells.At the transcriptomic level, gene set enrichment analysis in Reactome [17] shows an over-representation of genes associated with Phase 1 functionalization of compounds (Figure 2g).The Phase 1 functionalization of compounds pathway functionalizes biomolecules with reactive groups to enable their detoxification through excretion or enhance their subsequent reaction with other biomolecules.
Whereas some genes related to Phase 1 functionalization were downregulated, others, such as CYP1A1, which regulates chemical detoxification [18] were upregulated in treated cells (Figure 2g).
An important question is how CAGE components, choline and geranic acid, perturb the metabolome, proteome, and transcriptome.Specifically, we expected perturbations from choline since it is metabolizable, generating metabolites capable of activating receptors at the proteomic level to regulate gene transcription or increasing metabolic flux to shift the equilibrium of biosynthetic pathways.Unlike choline, little is known about geranic acid metabolism; however, it is established that geranic acid inhibits tyrosinase, [19] an enzyme that catalyzes tyrosine oxidation.Therefore, we also expect geranic acid, an enzyme inhibitor, to interfere with cellular metabolism.Metabolomic analysis shows a statistically significant 2-fold increase (p-value < 0.05) of intracellular levels of betaine and Lmethionine in treated cells (Figure 3).This increase was expected, considering that choline is a metabolic precursor of betaine, which methylates homocysteine to generate Lmethionine. [20]The proteomic analysis detected three choline metabolism-regulating enzymes -choline transporter-like protein 1, choline transporter-like protein 2, and choline/ethanolamine kinase -with no statistically significant difference between the intracellular levels of these enzymes in treated and control cells (Table S3 in Supporting Information).This suggests that the elevated level of betaine and L-methionine likely resulted from increased metabolic flux from choline into the choline-betaine-L-methionine conversion pathway rather than from changes in enzyme levels.
As mentioned above, geranic acid inhibition of tyrosinase is documented. [19]Ideally, tyrosinase inhibition should increase tyrosine intracellular levels, increasing tyrosine metabolic flux.Proteomic analysis identified no tyrosinase enzyme (Table S3), and metabolomics detected a statistically significant depletion in tyrosine levels in treated cells (Figure 3c), suggesting no interaction between geranic acid and tyrosinase.Further, geranic acid is a fatty acid and a terpene; therefore, we expected alteration in the lipid profile because terpenes such as geraniol, an analog of geranic acid, modulate lipid metabolism through enzyme inhibition. [21]However, we found no treatment-induced alterations in the levels of several lipids and lipids derivatives, including arachidonic acid, taurocholic acid, and glyceryl palmitate (Figure 3d-f and Table S2).While the effect of choline on the metabolome is evidenced by the enhanced level of betaine and methionine (Figure 3), the impact of geranic acid, especially on lipid metabolism, requires further investigation.
Metabolites can interact electrostatically or bind covalently to other biomolecules, modifying the structure and function of these biomolecules.These modifications can trigger a cascade of signals that could culminate in regulating gene transcription downstream.For instance, betaine suppresses transcription factor nuclear-kB, downregulating the expression of genes encoding pro-inflammatory cytokines such as tumor necrosis factor-alpha (TNF-α), interleukin 1 beta (IL1B), and interleukin 23 (IL23). [20]Further, terpenes, a metabolite class that includes geranic acid, attenuate pro-inflammatory cytokine expression in cell lines and animal models. [22]Therefore, the elevated levels of betaine (Figure 3), coupled with geranic acid, should suppress pro-inflammatory cytokine expression in treated cells.On the contrary, transcriptomics reveals that CAGE upregulated genes encoding several interleukins (IL1A, IL1B, and IL11) (Figure 4a-c).This seemingly unexpected finding (discussed in the next paragraph) is counterintuitive because betaine and geranic acid have been shown to suppress pro-inflammatory cytokines expression.However, the result highlights the complexity inherent in biological systems and context-dependence of biological experiments.
Because biochemical crosstalk between the metabolome and transcriptome regulate mRNA expression, we examine the metabolomic data to understand why CAGE treatment bolstered IL11, IL1A, and IL1B expression.In addition to betaine, CAGE treatment significantly elevated the intracellular levels of other metabolites, including 4-hydroxyphenyllactic acid and indole-3-lactic acid, by more than two folds (Figure 5).The Human Metabolome Database (HMDB) [23] describes indole-3lactic acid as tryptophan metabolites.The elevation of indole-3lactic acid contrasted with kynurenine, another tryptophan metabolite, [23] depleted in the treated cell line (Figure 5).Tryptophan metabolism occurs via the kynurenine pathway, producing kynurenine, or indole pyruvate, and L-amino acid oxidase mediated pathways, producing indole-3-lactic acid. [24]he metabolomic data (Figure 5) shows that the treatment depleted kynurenine and elevated indole-3-lactic acid.Therefore, it is logical to posit that CAGE concurrently downregulated the kynurenine pathway, depleting kynurenine levels, and upregulated the indole pyruvate or L-amino acid oxidasemediated pathway, elevating indole-3-lactic acid levels.The Lamino acid oxidase-mediated pathway is more likely than the indole pyruvate pathway to bolster the levels of indole-3-lactic acid for three reasons.First, literature precedence suggests that the indole pyruvate pathway requires mammalian intestinal flora, [24,25] which is absent in our experiment.In contrast, the Lamino acid oxidase pathway is feasible in mammalian cells. [24]econd, the L-amino acid oxidase pathway catalyzes other Laromatic amino acids, such as tyrosine. [24]Here, we found elevated levels of 4-hydroxyphenyllactic acid, a tyrosine metabolite, in the treated cells (Figure 5).Third, unlike the indole pyruvate pathway, the L-amino acid oxidase pathway generates hydrogen peroxide (H 2 O 2 ). [26]The third reason is noteworthy because H 2 O 2 , upregulated in CAGE-treated A431, [9] can inhibit indoleamine-2,3-dioxygenase [27] to downregulate the conversion of tryptophan to kynurenine.In this study, we infer that H 2 O 2 downregulated the kynurenine pathway, biasing tryptophan metabolism towards the L-amino acid oxidase pathway and eventually favoring the biosynthesis of indole-3-lactic acid against kynurenine.
Next, we ask if elevated indole-3-lactic acid and 4-hydroxyphenyllactic acid perturbs gene expression, including genes encoding pro-inflammatory cytokines.25a,28] Again, treatment of a human squamous cell carcinoma cell line with the 2,3,7,8-tetrachlorodibenzo-p-dioxin, a high-affinity AhR ligand, increases the mRNA levels of pro-inflammatory IL1B. [29]Here, we found that CAGE treatment increased the mRNA levels of IL1A, IL1B, IL11, CYP1A1, CYP1B1, ALDH3A1, and NQO1 (Figure 4).Therefore, we attributed the counterintuitive upregulation of some genes encoding interleukins (discussed above) to elevated levels of indole-3-lactic acid and 4-hydroxyphenyllactic acid, known AhR ligands.Choline and geranic acid are unlikely high-affinity AhR ligands and are more likely to indirectly activate AhR through their metabolites or by modulating metabolic pathways.Indeed, in a protein-ligand docking studies, we found that indole-3-lactic acid and 4-hydroxyphenyllactic acid were better AhR ligands than choline and geranic acid (Supporting Information).
CAGE's ability to indirectly activate the expression of AhRregulating genes is a significant finding because it advances our understanding of ligand-independent activation of AhR.28a] The ligand-independent activation theory postulates that molecules such as carotenoids indirectly activate AhR through their metabolites that are ligands, or by modulating pathways that generate ligands.28a] The multiomics approach in this study suggests that CAGE indirectly activates AhR by upregulating the biosynthesis of indole-3-lactic acid and 4hydroxyphenyllactic acid, which are proven AhR ligands.Various studies support the ligand-independent activation of AhR. [30]AGE indirect activation of AhR is also evidenced at the proteome level, where altered expression of some enzymes regulated by AhR activity was detected by proteomics (Figure 4g and h).
Additionally, the metabolomic data show that CAGE treatment significantly depleted the intracellular levels of several metabolites, including glutamine (Figure 6).From a cell biology perspective, glutamine is vital, being a source of cellular energy and a critical nutrient required by rapidly proliferating cancer cells. [31]Consequently, glutamine deprivation is intensively investigated as an anticancer therapy.Thus, CAGE's ability to deplete intracellular glutamine levels has implications for cancer chemotherapy and explain the previously reported antiproliferative property of CAGE in the A431 cell line. [9]ecause glutamine plays a pivotal role in proliferating cells' bioenergetics, [31e] we ask if cellular bioenergetic changes with CAGE treatment.The ADP/ATP ratio, an indicator of cellular bioenergetics, [32] was impervious to the CAGE treatment, as evidenced by the non-significant difference (p-value > 0.05) between the ratio of control and treated cell lines (Figure 7).31e] Indeed, metabolomic analysis of the control cell line, which had elevated glutamine levels, reveals elevated levels of glutamic acid and aspartic acid (Table S2) that are biosynthesized from glutamine and used for proteins and nucleic acid biosynthesis.This finding suggests that glutamine depletion in the treated cells more likely resulted from impeded biosynthesis glutamine or uptake from cell culture medium rather than increased metabolic conversion to other metabolites.The culture medium in this study was not supplemented with glutamine, implicating de novo biosynthesis as the source of the difference in glutamine level between treated and control cells.LC-MS/MS-based proteomics shows no statistically significant difference in the intracellular levels of glutamine synthetase (GS) (Figure 4i), which catalyzes glutamine biosynthesis from glutamate and ammonia.We, therefore, infer that the treatment-induced depletion of glutamine probably results from the inhibited activity of GS rather than a differential expression of the enzyme.The present study did not elucidate whether CAGE inhibited the activity of GS, but previous studies demonstrated CAGE's ability to inhibit the activity of other enzymes. [10]More specifically, several studies showed that methionine sulfoxide, a metabolite derivable via the choline-betaine-methionine metabolic pathway and elevated in the treated cells (Table S2), inhibits the activity of GS. [33] It is, therefore, rational to assume that methionine sulfoxide inhibited GS activity to attenuate glutamine levels (Figure 6).

Conclusions
In conclusion, NADESs, even when composed of seemingly non-toxic and natural motifs, can generate metabolites to alter  the molecular landscape of cells.In this study, CAGE, composed of choline and geranic acid, upregulated the levels of indole-3lactic acid and 4-hydroxyphenyllactic acid, ligands that activate the AhR to signal the transcription of genes with implications for inflammation, immunomodulation, chemical detoxification, and cell development.Indeed, transcriptomic data show upregulated levels of AhR-regulated genes, CYP1A1, CYP1B1, ALDH3A1, NQO1, IL1A, IL1B, and IL11, in CAGE-treated cell line, suggesting ligand-independent activation of AhR.Further, CAGE-treated cell lines exhibited attenuated glutamine levels, a vital nutrient for proliferating cells.While it remains unknown whether CAGE inhibit enzymes related to glutamine biosynthesis, metabolomic data shows upregulated levels of methionine sulfoxide, a metabolite derivable from choline, and demonstrated to inhibit glutamine biosynthesis.The results show that CAGE indirectly activated AhR and attenuated glutamine levels.It is important to note that because the properties of NADES depend on the molar ratios of the HBA and HBD, [1] the effects of CAGE on the transcriptome, proteome, and metabolome will depend on the molar ratios of choline and geranic acid.Overall, the findings provide a conceptual framework to reconsider the effect of NADESs on cell biology and expand NADES's application beyond chemistry to biomedicine and biotechnology.

Figure 1 .
Figure 1.(a) Schematic representation of Choline And GEranic acid (CAGE) NADES synthesized from one mole of choline bicarbonate and three moles of geranic acid.The water by-product was removed under vacuum.(b) Schematic illustration of the multiomics workflow.(i) The human A431 cell line was treated with 6.25 μM of CAGE for 7 days; (ii) cells were harvested and RNAs extracted, then (iii) sequenced on an Illumina platform; (iv) cells were harvested; (v) metabolites extracted, then (vi) quantified using a UHPLC-MS/MS platform; (vii) protein separated from metabolites, then (viii) digested to peptides, and (ix) quantified using an LC-MS/MS platform.For details, please see Supporting Information.

Figure 2 .
Figure 2. Volcano plots showing differentially expressed (a) metabolites, (b) proteins, and (c) mRNAs in non-treated vs. CAGE-treated A431 cells.The spheres indicating individual biomolecules -metabolites, proteins, or RNAs -and the colored (blue: upregulated in treated or red: downregulated in treated) indicates biomolecules that statistically change (p-value = 0.05) over two folds.(d) Venn diagram showing the number of proteins and mRNA in control (À C) and treated (À T) cells.Dot plots showing association of depleted (e) and (f) elevated metabolites with metabolic pathways.The size and color of the spheres correspond to enrichment ratio and p-value, respectively.(g) Heat map showing the detected mRNAs associated with Phase 1 functionalization of compounds pathway, the enriched pathway with the highest Q-value.

Figure 3 .
Figure 3. LC-MS/MS-based metabolomics show that CAGE significantly altered the levels of metabolites (a) betaine, (b) methionine, (c) tyrosine but did not significantly affect the levels of (d) arachidonic acid, (e) taurocholic acid, and (f) glyceryl palmitate.

Figure 5 .
Figure 5. LC-MS/MS-based proteomics shows that CAGE elevated the levels of (a) indole-3-lactic acid and (b) 4-hydroxylphenyllactic acid but depleted the levels of (c) kynurenine in treated cells.

Figure 6 .
Figure 6.LC-MS/MS-based metabolomics shows that CAGE depleted the levels of (a) glutamine acid in treated cells without significantly altering (b) the ADP/ATP ratio.