Effects of Lithium on Mortality and Metabolite Profiles in Drosophila Lithium-Inducible SLC6 Transporter Mutants

Abstract

Lithium has long been the primary treatment for bipolar disorder and shows promise for managing other neurological and psychiatric conditions. We previously identified the Lithium-inducible SLC6 transporter (List) in Drosophila melanogaster as a gene significantly upregulated in response to lithium chloride supplementation. List encodes a putative amino acid transporter belonging to the Na⁺-dependent solute carrier family 6. Here, we show that List is expressed in the Malpighian tubules, glia, and hindgut. RNA interference-mediated List knockdown in the Malpighian tubules drastically increases lithium-induced mortality. Additionally, List loss-of-function mutants (List^TG4.2) accumulate six times more internal lithium than controls after lithium exposure. Metabolomic analysis revealed disrupted amino acid metabolism and a shift toward a more oxidized cellular redox state in lithium-treated List^TG4.2 mutants. Overall, our findings suggest that List protects flies from lithium toxicity by regulating internal lithium levels and maintaining metabolic and redox balance.

1. Introduction

Lithium is a simple alkali metal with chemical and physical properties similar to sodium. However, it has been a highly effective treatment for bipolar disorder and has remained the primary medication for this condition for over 70 years (Rybakowski, 2014). Lithium uniquely stabilizes mood swings and reduces suicidal ideation and behaviors in bipolar patients (Geddes and Miklowitz, 2013). Additionally, lithium has shown potential effectiveness in treating other psychiatric conditions, including autism spectrum disorders (ASD) and schizophrenia (Luo et al., 2020; Mintz and Hollenberg, 2019). Growing evidence also suggests that lithium can have protective and proliferative effects on neurons in both in vitro and in vivo systems (Puglisi-Allegra et al., 2021), raising the intriguing possibility of its use in medical interventions for neurodegenerative disorders such as Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS), and Parkinson’s disease (PD) (Forlenza et al., 2014; Lazzara and Kim, 2015; Singulani et al., 2024). Despite its known and potential therapeutic benefits, the clinical use of lithium is limited by its narrow therapeutic window and the risk of side effects, even at therapeutic doses (Medić et al., 2020). Unfortunately, the mechanisms underlying lithium’s beneficial and adverse effects remain poorly understood. To maximize lithium’s therapeutic benefits and minimize its toxic effects, it is essential to have a deep understanding of the biological responses to lithium at the molecular and cellular levels.

The fruit fly Drosophila melanogaster is a powerful experimental model for studying evolutionarily conserved biological processes, including development, physiology, and behavior (Hales et al., 2015; Rubin, 1988). It is also valuable in applied fields such as drug mechanism studies and toxicology (Huang and Lee, 2023). To investigate the molecular and genetic mechanisms underlying lithium’s biological actions, we used Drosophila to examine changes in transcriptional profiles following lithium treatment. We identified the Lithium-inducible solute carrier 6 (SLC6) transporter (List) as one of the most significantly upregulated genes in adult fly heads after 24 hours of feeding on a diet containing 50 mM lithium chloride (LiCl) (Kaas et al., 2016; Kasuya et al., 2009a; Kasuya et al., 2009b). List encodes integral membrane proteins belonging to the solute carrier 6 (SLC6) family of the Na⁺-dependent transporters (Thimgan et al., 2006). SLC6 transporters are responsible for transporting extracellular neurotransmitters, amino acids, and osmolytes across the plasma membrane into the cytosol (Pramod et al., 2013). In humans, the SLC6 family has 20 members (Bröer, 2008) classified into four subfamilies based on sequence similarity and substrate specificity: the GABA transporter subfamily, the monoamine transporter subfamily, and the amino acid transporter subfamilies I and II (Bröer and Gether, 2012).

List plays a crucial role in protecting Drosophila from the harmful effects of lithium. When List expression is reduced, either through a genomic deletion or RNA interference (RNAi)-mediated global knockdown, flies become highly susceptible to lithium toxicity, leading to significantly increased mortality after consuming lithium-containing food (Kasuya et al., 2009b). Other ions, such as sodium, potassium, and chloride, failed to induce List upregulation and did not affect the flies’ sensitivity to these ions’ adverse effects (Kasuya et al., 2009b), underscoring the unique interaction between List function and the biological response to lithium. In this study, we investigated the role of List in resistance to lithium toxicity by using a loss-of-function knock-in List mutant and List-RNAi, combined with cell-type specific GAL4 lines. Additionally, we performed a metabolomic analysis to examine how List dysfunction and lithium treatment influence cellular metabolic processes. Our results lay the groundwork for understanding how the SLC6 transporter List contributes to resistance against lithium’s adverse effects in Drosophila, providing valuable insights into similar mechanisms in vertebrate animals.

2. Materials and methods

2.1. Fly stocks and culture conditions.

Flies (Drosophila melanogaster) were reared at 25°C, 65% humidity in a 12-hour light/dark cycle on the diet developed by Edward Lewis (Lewis, 1960) and modified by Rodney Williamson (Beckman Research Institute of the City of Hope, Duarte, CA): a cornmeal/glucose/yeast/agar medium supplemented with the mold inhibitor methyl 4-hydroxybenzoate (0.05 %), propionic acid, and phosphoric acid. The exact composition of the diet was described in Kasuya et al. (2019). The following mutants and GAL4 lines were obtained from the Bloomington Stock Center (https://bdsc.indiana.edu/): yw, yw; TI[CRIMIC.TG4.2]List^{CR02506-TG4.2} (hereafter List^TG4.2), Df(2R)Exel7157, c42-GAL4, Myo31DF-GAL4, UAS-RedStinger, and UAS-List-RNAi. elav-GAL4 and repo-GAL4 were obtained from Dr. Chun-Fang Wu and Dr. John Manak, respectively, both affiliated with the University of Iowa.

2.2. Lithium treatment and survival assays.

Two- to four-day-old male and female flies were separately grouped into sets of approximately ten and transferred to vials containing fly food with either 10 mM or 50 mM LiCl (Sigma-Aldrich, St. Louis, MO). Under these conditions, internal lithium concentrations reached approximately 1 mM (Kasuya et al., 2009b), which is comparable to lithium’s therapeutic plasma concentration of 0.5–1.0 mM (Tondo et al., 2019). The lithium-containing food was prepared by mixing 9 parts of the regular cornmeal/glucose/yeast/agar diet with 1 part of 100 mM or 500 mM LiCl solution. Control food was prepared by replacing the LiCl solution with water. The vials with flies were maintained at 25°C, 65% humidity, under a 12-hour light/dark cycle. The number of live and dead flies was recorded daily.

2.3. Quantitative reverse transcription polymerase chain reaction (qRT-PCR).

Total RNA was extracted from three 1- to 2-day-old male flies of the indicated genotype using the Qiagen RNeasy Kit (Qiagen, Germantown, MD), following the manufacturer’s instructions. The purified RNA was treated with DNase I Set (Cat. # E1010, Zymo Research, Irvine, CA) and further purified using the Zymo RNA Clean & Concentrator-25 Kit (Cat. # R1017, Zymo Research). Next, 1 μg of DNase-treated RNA was reverse-transcribed into cDNA using the iScript cDNA Synthesis Kit (Cat. # 1708890, Bio-Rad, Hercules, CA). Quantitative PCR (qPCR) was performed using SYBR^™ Green Universal Master Mix (Cat. # 4309155, Thermo Fisher, Waltham, MA) in four technical replicates across three biological replicates for each genotype. The primer sets used for amplifying List and ribosomal protein 49 (rp49) cDNA were: List (TTTCTTCCTCATGCTATTCGTCTTAGGCAT and GAGCGGACATATCGAGCAGCATGT) and rp49 (GATCGATATGCTAAGCTGTCGC and CGACCACGTTACAAGAACTCT). The Δ cycle threshold (ΔCT) values were calculated by normalizing the CT values of List to those of rp49. The relative fold change in List gene expression was determined using the formula: Fold Change = 2^−ΔΔCT.

2.4. Immunohistochemistry and confocal microscopy.

Immunostaining was performed essentially as described in (Ishimoto et al., 2013). Briefly, brains were dissected from the adult progeny of a cross between List^TG4.2 and UAS-RedStinger. They were fixed, blocked, and stained with 8D12 anti-Repo IgG (1:200; Developmental Studies Hybridoma Bank, University of Iowa), followed by goat Alexa Fluor 488-conjugated anti-mouse IgG (1:500; Invitrogen). A confocal microscope (Zeiss 510) was used to collect z-sections of the adult brain, gut, or Malpighian tubules, expressing fluorescent reporters (RedStinger) or stained with antibodies, at 1 or 2 μm intervals with a 10× objective lens. ImageJ (Schindelin et al., 2012) was used to pseudocolor and create maximum intensity z-projections.

2.5. Quantification of lithium levels.

2.5.1. Sample preparation.

Two- to four-day-old adult males (yw/Y and yw/Y; List^TG4.2/List^TG4.2) were collected and transferred to vials containing a standard diet supplemented with 10 mM LiCl, with approximately 20 flies per vial. After two days of LiCl treatment, the flies were flash-frozen using liquid nitrogen and stored at −80°C until metal analysis. Prior to metal analysis, the fly and diet samples were completely dissolved using a microwave digestion system (Milestone Ethos UP MA182, Sorisole, Italy). Before use, the digestion vessels were cleaned with 50% (v/v) HCl. The vessels were then rinsed three times with deionized water and allowed to air dry. Approximately 50 mg of each sample was added to a digestion vessel. The samples were digested in a microwave digestion system in a total of 5 mL solution containing 10% trace metal grade nitric acid, 50% trace metal grade hydrogen peroxide, and 40% Milli-Q water (v/v). After digestion, the samples were diluted with deionized water to a final volume of 25 mL, resulting in a final nitric acid concentration of 2% (v/v). The standard fly food was processed in the same way to measure its metal content.

2.5.2. Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) analysis.

The concentrations of 14 metals (Al, Ag, As, Cd, Co, Cr, Cu, Fe, Li, Mn, Ni, Se, V, and Zn) in the prepared solution were quantified using inductively coupled plasma-mass spectrometry (ICP-MS) (Agilent 7900 ICP-MS, Agilent Technologies, Inc., Santa Clara, USA). Calibration of the ICP-MS was conducted using multi-element standards (Inorganic Ventures, Inc., Christiansburg, USA). High-purity helium (>99.999%) was employed as a collision gas to eliminate polyatomic interference during the analysis of As, Fe, Se, and V. For the remaining metals, argon was used as the carrier gas in low matrix mode. ⁸⁹Y and ¹⁵⁹Tb were used as the internal standard at a concentration of 20 μg/L.

2.6. Metabolomic analysis.

2.6.1. Sample preparation.

Two- to four-day-old adult males (yw/Y and yw/Y; List^TG4.2) raised on a standard diet, were collected and transferred to vials containing a diet either with or without 10 mM LiCl (approximately 20 males per vial). After two days of feeding a control diet or a LiCl-supplemented diet, the flies were flash-frozen with liquid nitrogen and stored at −80°C until metabolomic analysis using gas chromatography–mass spectrometry (GC-MS) and liquid chromatography–mass spectrometry (LC-MS).

2.6.2. GC-MS method.

Whole fly samples were lyophilized for 2 hours prior to bead mill homogenization in 18:1 (μL:mg wet tissue weight) ice-cold 2:2:1 methanol/acetonitrile/water extraction buffer containing a mixture of internal standards (D4-citric acid, D4-succinic acid, D8-valine, and U13C-labeled glutamine, glutamic acid, lysine, methionine, serine, and tryptophan; Cambridge Isotope Laboratories). Homogenates were rotated for 1 hour at −20°C. Homogenates were centrifuged for 10 minutes at 21,000 × g, and 150 μL of the cleared metabolite extracts were transferred to autosampler vials and dried using a SpeedVac vacuum concentrator (Thermo). Dried metabolite extracts were reconstituted in 30 μL of 11.4 mg/mL methoxyamine (MOX) in anhydrous pyridine, vortexed for 5 minutes, and heated for 1 hour at 60°C. Next, to each sample 20 μL of N,O-Bis(trimethylsilyl)trifluoroacetamide (TMS) was added to each sample, then samples were vortexed for 1 minute and heated for 30 minutes at 60°C. Derivatized samples were analyzed by GC-MS. 1 μL of derivatized sample was injected into a Trace 1300 GC (Thermo) fitted with a TraceGold TG-5SilMS column (Thermo) operating under the following conditions: split ratio = 20:1, split flow = 24 μL/minute, purge flow = 5 mL/minute, carrier mode = Constant Flow, and carrier flow rate = 1.2 mL/minute. The GC oven temperature gradient was as follows: 80°C for 3 minutes, increasing at a rate of 20°C/minute to 280°C, and holding at a temperature at 280°C for 8 minutes. Ion detection was performed by an ISQ 7000 mass spectrometer (Thermo) operated from 3.90 to 21.00 minutes in EI mode (−70eV) using select ion monitoring (SIM).

2.6.3. LC-MS method.

Whole fly samples were lyophilized and transferred to ceramic bead tubes. 18-fold (w/v) extraction solvent (with 9 heavy internal standards) was added to each sample and homogenized. After homogenization, samples were rotated in −20°C for 1 hr and then centrifuged at 21,000 × g for 10 min. 200 μL of the supernatant extracts was transferred to microcentrifuge tubes, and the extracts were dried using a speed vac apparatus. Dried extracts were reconstituted in 20 μL acetonitrile/water (1:1 v/v) and vortexed well, kept in −20°C overnight. The following day, samples were centrifuged, and the supernatant was transferred to the LC-MS autosampler vials for analysis. LC-MS data were acquired on a Thermo Q Exactive hybrid quadrupole Orbitrap mass spectrometer with a Vanquish Flex UHPLC system or Vanquish Horizon UHPLC system. The LC column used was a Millipore SeQuant ZIC-pHILIC (2.1 × 150 mm, 5 μm particle size) with a ZIC-pHILIC guard column (20 × 2.1 mm). The injection volume was 2 μL. Mobile phase is as follows: Solvent A (20 mM ammonium carbonate [(NH₄)₂CO₃] and 0.1% Ammonium Hydroxide (v/v) [NH₄OH]) and Solvent B (Acetonitrile). The method was run at a flow rate of 0.150 mL/min. The gradient starts at 80% B and decreasing to 20% B over 20 minutes; returning to 80% B in 0.5 minutes; and held there for 7 minutes (Cantor et al., 2017). The mass spectrometer was operated in full-scan, polarity-switching mode from 1 to 20 minutes, with the spray voltage set to 3.0 kV, the heated capillary held at 275°C, and the HESI probe held at 350 °C. The sheath gas flow was set to 40 units, the auxiliary gas flow was set to 15 units, and the sweep gas flow was set to 1 unit. MS data acquisition was performed in a range of m/z 70–1,000, with the resolution set at 70,000, the AGC target at 1 × 106, and the maximum injection time at 200 ms (Cantor et al., 2017).

2.6.4. Data analysis.

Raw data were analyzed using TraceFinder 5.1 (Thermo). Metabolite identification and annotation required at least two ions (target + confirming) and a unique retention time that corresponded to the ions and retention time of a reference standard previously determined in-house. A pooled-sample generated prior to derivatization was analyzed at the beginning, at a set interval during, and the end the analytical run to correct peak intensities using the NOREVA tool (Li et al., 2017). NOREVA corrected data were then normalized to the sum total of signal per sample to control for extraction, derivatization, and/or loading effects. Metabolomics datasets were analyzed using MetaboAnalyst 5.0 (Pang et al., 2021).

2.7. Statistical analysis.

Statistical analysis was conducted using GraphPad Prism 10 (GraphPad Software, Inc., Boston, MA). To evaluate the effects of genetic manipulations on the survival curves of adult flies (Figures 2 and 5), we performed log-rank (Mantel-Cox) tests, comparing each experimental group to the control group. When appropriate, P-values were adjusted for multiple comparisons using the Bonferroni correction. The results of qRT-PCR experiments were analyzed using one-way ANOVA on ΔCT values, followed by Turkey’s multiple comparison test. A comparisons of lithium content between control flies and List^TG4.2 mutants was performed using a two-tailed student’s t-test, assuming unequal variance. Data from metabolic analyses were processed and statistically evaluated using MetaboAnalyst 5.0 (Pang et al., 2021).

Figure 2. Effects of List^TG4.2 on lithium-induced mortality.

Proportion of surviving adult flies after consuming a diet supplemented with 10 mM or 50 mM LiCl. Viability was assessed every 24 hours. (A) Males on a 10 mM LiCl diet. Total number of flies tested: N = 65 (+/+), 61 (List^TG4.2/+), 87 (UAS-List-RNAi/Y; List^TG4.2/+), and 55 (List^TG4.2/List^TG4.2). (B) Females on a 10 mM LiCl diet. N = 70 (+/+), 70 (List^TG4.2/+), 101 (UAS-List-RNAi/+; List^TG4.2/+), and 65 (List^TG4.2/List^TG4.2). (C) Males on a 50 mM LiCl diet. N = 70 (+/+), 70 (List^TG4.2/+), 75 (UAS-List-RNAi/Y; List^TG4.2/+), and 70 (List^TG4.2/List^TG4.2). (D) Females on a 50 mM diet. N = 67 (+/+), 70 (List^TG4.2/+), 108 (UAS-List-RNAi/+; List^TG4.2/+), and 68 (List^TG4.2/ List^TG4.2). (E) Males on a 10 mM LiCl diet. N = 60 (List^TG4.2/Df(2R)Exel7157; +/+) and 60 (List^TG4.2/Df(2R)Exel7157; UAS-List-cDNA/+). (F) Females on a 10 mM diet. N = 80 (List^TG4.2/Df(2R)Exel7157; +/+) and 60 (List^TG4.2/Df(2R)Exel7157; UAS-List-cDNA/+). Statistical analyses were conducted using the log-rank (Mantel-Cox) test, followed by Bonferroni correction. Statistically significant differences compared to the control group (+/+ or List^TG4.2/Df(2R)Exel7157; +/+) are indicated with adjusted P-values. ****P < 0.0001.

Figure 5. Effects of cell type-specific List knock down on lithium-induced mortality.

Proportion of surviving adult flies after consuming a diet supplemented with 50 mM LiCl. Viability was assessed every 24 hours for flies carrying GAL4 only (black) or GAL4 as well as UAS-List-RNAi (red) to knockdown List expression in a cell type-specific manner. (A) Males and (E) females carrying c42-GAL4 (Malpighian tubule). Total number of flies tested: N = 73 (male, c42-GAL4/UAS-List-RNAi), 79 (male, c42-GAL4/+), 82 (female, c42-GAL4/UAS-List-RNAi), and 94 (female, c42-GAL4/+). (B) Males and (F) females carrying nSyb-GAL4 (neuron). N = 79 (male, nSyb-GAL4/UAS-List-RNAi), 32 (male, nSyb-GAL4/+), 87 (female, nSyb-GAL4/UAS-List-RNAi), and 48 (female, nSyb-GAL4/+). (C) Males and (G) females (N=) carrying repo-GAL4 (glia). N = 56 (male, repo-GAL4/UAS-List-RNAi), 29 (male, repo-GAL4/+), 58 (female, repo-GAL4/UAS-List-RNAi), and 26 (female, repo-GAL4/+). (D) Males and (H) females carrying Myo31DF-GAL4 (gut). N = 31 (male, Myo31DF-GAL4/UAS-List-RNAi), 38 (male, Myo31DF-GAL4/+), 41 (female, Myo31DF-GAL4/UAS-List-RNAi), and 47 (female, Myo31DF-GAL4/+). Statistical analyses were conducted using the log-rank (Mantel-Cox) test. **P < 0.01; ****P < 0.0001.

3. Results

3.1. A knock-in List mutation leads to increased sensitivity to lithium toxicity.

The Drosophila Lithium-inducible SLC6 transporter (List) gene (CG15088, FBgn0034381) is localized on the right arm of the second chromosome (2R: 18727827..18730281, Flybase: https://flybase.org/) and consists of seven exons (Figure 1). TI[CRIMIC.TG4.2]List^{CR02506-TG4.2} (List^TG4.2) is a knock-in mutant allele of List. This allele carries the Swappable Integration Cassette (SIC) containing an artificial exon with attP-FRT-Splice Acceptor (SA)-T2AGAL4-polyA-3XP3EGFP-polyA-FRT-attP (T2A-GAL4 or Trojan-GAL4) (Kanca et al., 2022) (Figure 1). List^TG4.2 is presumed to be a null or strongly hypomorphic allele of List because the inserted element contains the artificial splice acceptor and the viral T2A sequence (Lee et al., 2018), which leads to the truncation of the List polypeptide.

Figure 1. Schematic representation of the lithium-inducible SLC6 transporter gene (List) and the insertional mutation List^TG4.2.

The List gene, previously known as CG15088, is located at cytogenetic position 55E10 on the right arm of the second chromosome (2R: 18,727,817..18,730,281). It consists of seven exons (represented as boxes, with coding regions shown in black) and six introns. The direction of transcription is indicated by a red arrow. List^TG4.2 is a loss-of-function allele of List that incorporates the Swappable Integration Cassette (SIC), which includes the T2A-GAL4 (Trojan-GAL4) element (see text for details).

To assess the effect of List^TG4.2 on flies’ sensitivity to lithium toxicity, adult flies homozygous or heterozygous for List^TG4.2 were fed a diet containing 10 mM or 50 mM LiCl, and their survival rate was examined every 24 hours. For both males and females, all or nearly all List^TG4.2 homozygotes died within 72 hours (3 days) on a 10 mM LiCl diet (Figure 2A, B) and within 48 hours (2 days) on a 50 mM LiCl diet (Figure 2C, D). Treatment with 10 mM or 50 mM LiCl for 120 hours (5 days) had little effect on the viability of control yw flies (Figure 2A–D). This heightened sensitivity to LiCl is a recessive trait associated with List^TG4.2, as List^TG4.2 heterozygotes showed almost no mortality on lithium-infused food under these conditions, displaying survival rates indistinguishable from those of control flies (Figure 2A–D).

We previously demonstrated that lithium sensitivity is significantly increased when the X-linked UAS-List-RNAi transgene (GD2072, VDRC) (Dietzl et al., 2007) is expressed under the control of the ubiquitous da-GAL4 driver (Kasuya et al., 2009b). Similarly, expression of List-RNAi using T2A-GAL4 activity associated with List^TG4.2 also led to increased lithium sensitivity in both males and females, although to a lesser extent than in List^TG4.2 homozygotes. Notably, heightened sensitivity to 10 mM LiCl was observed in List-knockdown females but not in males (Figure 2A–D).

To further confirm the specific role of List in mediating sensitivity to lithium toxicity, we used the UAS-List-cDNA transgene (Kasuya et al., 2009b). Df(2R)Exel7157 is a molecularly defined genetic deficiency that deletes a 109 kb DNA segment (2R: 18,621,522..18,730,771), including the entire List gene (2R: 18,727,817..18,730,281) (FlyBase, https://flybase.org/reports/FBab0038052). Flies heterozygous for List^TG4.2 and Df(2R)Exel7157 were highly sensitive to lithium toxicity. However, expression of wild-type List cDNA, driven by T2A-GAL4 activity associated with List^TG4.2 (List^TG4.2/Df(2R)Excel7157; UAS-List-cDNA/+), completely rescued the phenotype (Figure 2E, F).

3.2. Quantification of transcript levels with quantitative RT-PCR.

We examined List transcript levels in List mutants and transgenic flies expressing either List-RNAi or List-cDNA using quantitative RT-PCR (qRT-PCR) (Figure 3A). List^TG4.2 is a presumptive null allele (Figure 1). Consistently, List transcript levels in List^TG4.2 homozygotes and List^TG4.2/Df(2R)Exel7157 trans-heterozygotes were estimated to be less than 1% of those of List wild-type flies (Figure 3B). Expression of List-RNAi under the control of T2A-GAL4, which is associated with List^TG4.2 (UAS-List-RNAi/Y; List^TG4.2/+), resulted in an over 80% reduction in List transcript levels. In contrast, expression of List-cDNA restored List transcript levels in List^TG4.2/Df(2R)Exel7157 trans-heterozygotes (List^TG4.2/Df(2R)Exel7157; UAS-List-cDNA/+). Notably, lithium resistance was also restored in these flies, as shown in Figure 2E and 2F. We statistically analyzed List transcript levels across five different genotypes using one-way ANOVA on ΔCT values. The results of the statistical analyses, including adjusted P-values, are presented in Supplementary Table 1.

Figure 3. Effects of List mutations and transgene expression on List transcript levels.

Whole-body List transcript levels were measured in five genotypes using qRT-PCR, with rp49 as an endogenous control for gene expression: +/+, List^TG4.2/List^TG4.2, UAS-List-RNAi/Y; List^TG4.2/+, List^TG4.2/Df(2R)Exel7157, and List^TG4.2/Df(2R)Exel7157; UAS-List-cDNA/+. (A) ΔCT values. (B) Fold changes (2^−ΔΔCT) relative to the control genotype (+/+). Means ± SD are shown. Statistical analyses were performed using one-way ANOVA, followed by Turkey’s multiple comparison test. *P < 0.05, ****P < 0.0001.

3.3. Visualization of List expression in adult tissues.

FlyAtlas 2 (https://motif.mvls.gla.ac.uk/FlyAtlas2/) is a publicly accessible database that utilizes RNA-seq data to provide information on gene expression across various tissues in Drosophila melanogaster (Krause et al., 2022). According to FlyAtlas 2, List is expressed in the adult eye, brain, thoracicoabdominal ganglion, hindgut, Malpighian tubules, and the rectal pads. To visualize List expression in adult flies, we drove expression of the nuclear-targeted fluorescent marker, RedStinger (Barolo et al., 2000), using the T2A-GAL4 activity associated with List^TG4.2 (Figure 1). The presence of a splice acceptor site directly upstream of the T2A-GAL4 sequence in List^TG4.2 ensures that the GAL4 protein is translated independently, reflecting the endogenous spatial and temporal expression pattern of the List protein (Lee et al., 2018).

Prominent expression of the List reporter was observed in the nervous system, the gut, and the Malpighian tubules. A significant portion of the nuclear-directed reporter expression in the brain (Figure 4A) coincided with the expression pattern of the pan-glial nuclei marker, anti-Repo (Figure 4B), suggesting that List is predominantly expressed in most, if not all, glial cells within the adult brain. Malpighian tubules are genetically and functionally segmented into a distal initial and transitional segment, a main segment, a proximal lower segment, and an upper and lower ureter (Sözen et al., 1997). The List reporter was expressed in the main and lower segments, but not in the initial/transitional segment or the ureter (Figure 4C, D). The expression of the reporter gene in the gut was restricted to the ileum and rectal pads within the hindgut region (Figure 4C–F).

Figure 4. Expression of the List reporter gene in adult flies.

(A) An adult brain showing expression of the nuclear-targeted fluorescent reporter gene UAS-RedStinger (red), driven by T2A-GAL4 activity associated with List^TG4.2 (See Figure. 1). (B) The same brain as in (A), immunostained with anti-Repo antibodies, a pan-glial nuclear marker (green). (C) Malpighian tubules and the ileum in the hindgut showing RedStinger expression (red). (D) Bright-field image of the same tissue as in (C). (E) A female abdomen showing RedStinger expression in the ileum and rectal pads in the hindgut as well as the Malpighian tubules (red). (F) Bright-field image of the same sample as in (E).

3.4. Effects of cell type-specific List knockdown on susceptibility to lithium toxicity.

Global reduction of List function in List^TG4.2 homozygotes led to markedly decreased resistance to lithium toxicity (Figure 2). To investigate critical sites of List-dependent resistance to lithium, we assessed the impact of cell type-specific List knockdown on the survival rates of adult flies exposed to a diet containing 50 mM LiCl. We utilized the UAS-List-RNAi transgene to suppress List expression in a cell type-specific manner via four different GAL4 drivers: c42-GAL4, Myo31DF-GAL4, nSyb-GAL4, and repo-GAL4. c42-GAL4 targets expression to Malpighian tubules (Rosay et al., 1997), and a subset of neurons in the ellipsoid body, fan-shaped body, and pars intercerebralis of the adult brain (Renn et al., 1999). Myo31DF-GAL4 drives gene expression in gut enterocytes (Jiang et al., 2009). nSyb-GAL4 and repo-GAL4 are widely used as pan-neuronal (Riabinina et al., 2015) and pan-glial (Sepp et al., 2001) drivers, respectively.

Adult progeny from crosses between UAS-List-RNAi females and GAL4 males (GAL4/List-RNAi) were fed a diet containing 50 mM LiCl, and their viability was examined every 24 hours. GAL4 heterozygotes (GAL4/+) were tested under the same conditions as controls. The most significant effect on lithium sensitivity was observed when List expression was knocked down with c42-GAL4. Almost all c42-GAL4/List-RNAi flies, regardless of sex, died after 72 hours (3 days) on a lithium-containing diet (Figure 5A, E). In sharp contrast, pan-neuronal knockdown of List using nSyb-GAL4 had no obvious effect on survival rates (Figure 5B, F). This indicated that, while c42-GAL4 directs gene expression in both Malpighian tubules and neuronal subsets, List knockdown in the Malpighian tubules, not in the neuronal subsets, is responsible for the increased lithium sensitivity in c42-GAL4/List-RNAi flies. Additionally, repo-GAL4/List-RNAi flies (Figure 5C, G) were more susceptible to lithium toxicity than the corresponding GAL4/+ flies, indicating that List knockdown in glial cells had a measurable impact on the flies’ sensitivity to lithium toxicity.

3.5. Internal lithium levels in lithium-treated control flies and List mutants.

Malpighian tubules are functionally analogous to vertebrate kidneys, playing a key role in waste excretion and osmoregulation of body fluids (Rodan, 2019). Given the critical function of List in the Malpighian tubules for resistance to lithium toxicity (Figure 5), it is possible that List may regulate the bioavailability or clearance of lithium. To investigate this, we measured lithium levels in control flies (yw/Y; +/+) and List mutants (yw/Y; LIst^TG4.2/LIst^TG4.2) after feeding them a diet supplemented with 10 mM LiCl for two days. Among the metals we examined, trace amounts of metals such as iron (Fe) and zinc (Zn) were found in the control food, but lithium was not detected (Supplementary Table 2). Analysis of lithium-fed flies showed that List mutants retained lithium at levels six times higher than those in control flies (24.7 ng vs. 3.68 ng per mg dry weight; unpaired t-test, P < 0.0001) (Figure 6).

Figure 6. Internal lithium levels in control flies and List^TG4.2 mutants.

The lithium content (ng/mg of dry weight) in male control flies (Ctrl) and List^TG4.2 mutants (List) after consuming a diet supplemented with 10 mM LiCl for two days. Unpaired student’s t-test. N = 5, ****P < 0.0001.

3.6. Metabolomic analysis of control flies and List mutants.

3.6.1. Overall changes in metabolite levels due to the List mutation and lithium treatment.

To investigate the physiological role of List and the mechanisms underlying the increased mortality observed in List mutants following lithium treatment, we performed metabolomic analysis using GC-MS and LC-MS. This analysis aimed to determine how List dysfunction and lithium exposure affect cellular metabolism. We analyzed four groups of flies: control male flies (yw/Y) raised on a regular diet (Ctrl-R) and a diet supplemented with 10 mM LiCl (Ctrl-L), as well as homozygous List^TG4.2 mutants (yw/Y; List^TG4.2/List^TG4.2) raised on a regular diet (List-R) and a 10 mM LiCl-supplemented diet (List-L) (Figure 7A). Each group consisted of six biological replicates, and a total of 143 metabolites were analyzed (Supplementary Table 3). A sparse partial least squares discriminant analysis (sPLS-DA) (Lê Cao et al., 2011) was performed on the combined data set obtained from GC-MS and LC-MS. The analysis revealed that samples within each group clustered closely together, while distinct separation was observed between groups (Figure 7B). These results indicate that both the List mutation and lithium treatment induce distinct and reproducible metabolomic changes. Heatmaps of the top 25 most significantly altered metabolites illustrate the primary metabolic effects of the List^TG4.2 mutation and lithium treatment (Figure 8A–D). Supplementary Table 4 provides a comprehensive list of metabolites that exhibited significant changes with false discovery rate (FDR) < 0.05 and a fold change (FC) > 1.2 in response to the List^TG4.2 mutation and lithium treatment. The number of such metabolites was 27, 89, 16, and 67 for the comparisons between Ctrl-R and List-R (effects of List^TG4.2 on a regular diet), Ctrl-L and List-L (effects of List^TG4.2 on a lithium diet), Ctrl-R and Ctrl-L (effects of lithium in control flies), and List-R and List-L (effects of lithium in List^TG4.2 mutants), respectively.

Figure 7. Metabolomic analysis of control flies and List^TG4.2 mutants with or without lithium treatment.

(A) Schematic overview of the metabolomic analysis workflow. (B) Sparse partial least squares discriminant analysis (sPLS-DA) plots showing metabolite comparisons among the following groups: Ctrl-R (control flies on a regular diet), List-R (List^TG4.2 mutants on a regular diet), Ctrl-L (control flies on a lithium diet), and List-L (List^TG4.2 mutants on a lithium diet).

Figure 8. Heatmaps illustrating metabolites altered by List^TG4.2 and lithium treatment.

The top 25 most significantly altered metabolites in response to List^TG4.2 on a regular diet (A) and a diet supplemented with 10 mM LiCl (B). The top 25 most significantly altered metabolites in response to lithium treatment in control flies (C) and List^TG4.2 mutants (D). Figures were generated using MetaboAnalyst 5.0 (Pang et al., 2021).

3.6.2. Effects of the List mutation and lithium treatment on amino acid levels.

List encodes a membrane transporter protein that belongs to the SLC6 family (Kasuya et al., 2009b). Using the Drosophila Integrative Ortholog Prediction Tool (DIOPT, http://www.flyrnai.org/diopt), which integrates data from multiple ortholog prediction algorithms (Hu et al., 2017), we identified SLC6A7 as the putative human ortholog of List protein, with the highest weighted DIOPT score of 4.87. SLC6A5 and SLC6A14 were the next closest candidate orthologs, each with a DIOPT score of 3.92. The SLC6A7, SLC6A5, and SLC6A14 proteins correspond to the high-affinity brain L-proline transporter (PROT) (Shafqat et al., 1995), glycine transporter 2 (GlyT2) (Liu et al., 1993), and the neutral and cationic amino acid transporter B^0,+ (ATB^0,+) (Karunakaran et al., 2011), respectively. These three proteins are members of the mammalian SLC6A amino acid transporter subfamily I. The List protein shares 36% amino acid identity and 55% similarity with mouse PROT and 35% identity and 54% similarity with human PROT. Given these significant overlap between List and these amino acid transporters, we hypothesize that amino acids are critical to both the function of List protein and the mechanisms driving increased mortality in lithium-treated List mutants.

Our metabolomics analysis encompassed all 20 proteogenic amino acids. Under a regular diet, the List^TG4.2 mutation resulted in elevated asparagine (Asn) levels and reduced concentrations of six amino acids: histidine (His), methionine (Met), phenylalanine (Phe), proline (Pro), threonine (Thr), and tryptophan (Trp) (Figure 9A: light blue vs. pink, 9B: effects of List mutation, regular diet). In contrast, when exposed to 10 mM LiCl, the List^TG4.2 mutation led to increased levels of 10 amino acids, while alanine (Ala) and proline (Pro) levels decreased by 21.4% and 56.4%, respectively (Figure 9A: dark blue vs. red, 9B: effects of List mutation, lithium diet).

Figure 9. Effects of List^TG4.2 and lithium treatment on amino acid levels.

(A) Levels of each amino acid in flies from four groups: Ctrl-R (light blue), Ctrl-L (dark blue), List-R (pink), and List-L (red). Average values are shown relative to the levels in the Ctrl-R group. N = 6. (B) Amino acids with levels significantly altered by (1) List^TG4.2 on a regular or lithium diet and (2) lithium treatment in control flies or List^TG4.2 mutants. The direction of changes (green: up, red: down) and the levels of statistical significance (adjusted P value) are indicated by arrows and color coding, respectively. A Tukey’s multiple comparisons test was used in conjunction with an ANOVA analysis.

Lithium treatment had markedly different effects on amino acid levels between control flies and List^TG4.2 mutants. In control flies, lithium exposure decreased the levels of six amino acids: isoleucine (Ile), leucine (Leu), valine (Val), phenylalanine (Phe), threonine (Thr), and tryptophan (Trp), with no observed increases in amino acid levels (Figure 9A: light blue vs. dark blue, 9B: effects of lithium treatment, control flies). Conversely, in List^TG4.2 mutants, lithium treatment led to significant increases in eight amino acids, notably asparagine (Asn) and lysine (Lys), which rose by 98.8% and 61.3%, respectively. However, levels of alanine (Ala), phenylalanine (Phe), and proline (Pro) decreased by 19.6%, 18.1%, and 52.1%, respectively (Figure 9A: pink vs. red, 9B: effects of lithium treatment, List mutant).

3.6.3. Metabolic pathways significantly affected by the List mutation and lithium treatment.

To better understand the changes in metabolite profiles caused by the List^TG4.2 mutation and lithium treatment, we performed a pathway analysis using the pathway analysis module in MetaboAnalyst (Pang et al., 2021). Under a regular diet without lithium treatment (Table 1A), four of the top ten pathways most significantly affected by the List^TG4.2 mutation pertained to amino acid metabolism: histidine metabolism (FDR = 0.0105), taurine and hypotaurine metabolism (FDR = 0.0147), tyrosine metabolism (FDR = 0.0202), and alanine, aspartate, and glutamate metabolism (FDR = 0.0309). Two top pathways are directly involved with antioxidant response, including the top hit, ascorbate and aldarate metabolism (FDR = 0.00102), and nicotinate and nicotinamide metabolism (FDR = 0.0105). Two of the top pathways are associated with cellular membrane composition and lipid signaling: sphingolipid metabolism (FDR = 0.0105) and glycerophospholipid metabolism (FDR = 0.0112). The second most significant hit, pyruvate metabolism (FDR = 0.00102) links the List^TG4.2 mutation to central carbon metabolism and the TCA cycle. Taken together, these data suggest that, on a regular diet, the List^TG4.2 mutation primarily affects amino acid metabolism and may have additional effects on cellular oxidation state, carbohydrate metabolism, and lipid signaling pathways.

Table 1.

Summary of metabolic pathways significantly affected by List^TG4.2 and lithium treatment^*

A
Ctrl-R vs List-R
Metabolic pathways affected by ListTG4.2	FDR
(Regular diet)
Ascorbate and aldarate metabolism	0.00102
Pyruvate metabolism	0.0105
Sphingolipid metabolism	0.0105
Nicotinate and nicotinamide metabolism	0.0105
Histidine metabolism	0.0105
Glycerophospholipid metabolism	0.0112
Taurine and hypotaurine metabolism	0.0147
Tyrosine metabolism	0.0202
Alanine, aspartate, and glutamate metabolism	0.0309
Pyrimidine metabolism	0.0445
B
Ctrl-L vs List-L
Metabolic pathways affected by ListTG4.2	FDR
(Lithium diet)
Arginine and proline metabolism	8.67 × 10−9
Alanine, aspartate and glutamate metabolism	5.85 × 10−8
Glutathione metabolism	6.24 × 10−8
Glyoxylate and dicarboxylate metabolism	1.24 × 10−7
Pentose phosphate pathway	2.75 × 10−7
Terpenoid backbone biosynthesis	2.75 × 10−7
Lysine degradation	4.44 × 10−7
Sphingolipid metabolism	4.44 × 10−7
Glycerophospholipid metabolism	7.30 × 10−7
Glycolysis / Gluconeogenesis	7.30 × 10−7
C
Ctrl-R vs Ctrl-L
Metabolic pathways affected by lithium treatment	FDR
(Control flies)
Valine, leucine, and isoleucine biosynthesis	0.0123
Phenylalanine, tyrosine, and tryptophan biosynthesis	0.0123
Phenylalanine metabolism	0.0123
Riboflavin metabolism	0.0286
Steroid biosynthesis	0.0286
Insect hormone biosynthesis	0.0286
Nicotinate and nicotinamide metabolism	0.0680
alpha-Linolenic acid metabolism	0.0915
Ascorbate and aldarate metabolism	0.228
Pyrimidine metabolism	0.278
D
List-R vs List-L
Metabolic pathways affected by lithium treatment	FDR
(ListTG4.2 mutants)
Arginine and proline metabolism	6.27 × 10−7
Pentose phosphate pathway	1.06 × 10−6
Terpenoid backbone biosynthesis	1.42 × 10−6
Alanine, aspartate, and glutamate metabolism	1.94 × 10−6
Fatty acid biosynthesis	2.72 × 10−6
Butanoate metabolism	3.97 × 10−6
Arginine biosynthesis	4.38 × 10−6
Glutathione metabolism	4.38 × 10−6
Sphingolipid metabolism	5.15 × 10−6
Lysine degradation	5.15 × 10−6

^*Pathway analysis, which combines enrichment analysis and pathway topology analysis, was conducted using the pathway analysis module in MetaboAnalyst 5.0 (Pang et al., 2021)

When flies were fed lithium, the ten pathways most significantly impacted by the List^TG4.2 mutation (Table 1B) were predominantly related to amino acid metabolism, central carbon metabolism, and antioxidant activity. Three pathways—two of which are the top two hits—are related to amino acid metabolism: arginine and proline metabolism (FDR = 8.67 × 10⁻⁹), alanine, aspartate, and glutamate metabolism (FDR = 5.85 × 10⁻⁸), and lysine degradation (FDR = 4.44 × 10⁻⁷). These results reinforce our hypothesis that the List^TG4.2 mutation interferes with amino acid transport. Three pathways were directly connected to central carbon metabolism: glyoxylate and dicarboxylate metabolism (FDR = 1.24 × 10⁻⁷), the pentose phosphate pathway (PPP; FDR = 2.75 × 10⁻⁷), and glycolysis/gluconeogenesis (FDR = 7.30 × 10⁻⁷). Notably, of the top ten pathway hits, two are critical pathways involved with antioxidant response, including the PPP, which generates the NADPH necessary for recharging spent antioxidants, and glutathione metabolism (FDR = 6.24 × 10⁻⁸), which is the pathway responsible for generating the most important antioxidant in the cell, glutathione. Additionally, arginine, proline, and glutamate are all connected to antioxidant metabolism through regulation of NADPH levels (proline), through their effect on glutathione biosynthesis (glutamate and arginine), or through direct electron scavenging (lysine) (Egbujor et al., 2024). In further support of an oxidative stress-based hypothesis, NADP⁺ increased 38% in List^TG4.2 mutants when fed lithium, whereas NADPH remained unchanged, resulting in a decreased NADPH/NADP⁺ ratio. The NADPH/NADP⁺ ratio is a critical regulator of cellular redox balance, and decreases in this ratio indicate an increase in the oxidation state of the cell. Decreases in the NADPH/NADP+ ratio also promote flux through the PPP as an adaptation to increase the reductive potential of the cell by increasing the NADPH pool (Holten et al., 1976). Taken together, these results strongly suggest that one possible source of toxicity from lithium in List^TG4.2 mutants is elevated oxidative stress.

In addition to the effects observed on key antioxidant pathways, nearly all glycolysis and TCA cycle metabolites were significantly reduced in lithium-fed List^TG4.2 mutants compared to those on a regular diet (see Supplementary Table 4, part D). Both glycolysis and the TCA cycle are essential for supplying electrons (in the form of NADH and FADH₂) to the electron transport chain for mitochondrial ATP production. Consequently, reduced flux through these pathways is often associated with ATP depletion, which could be another source of toxicity in List^TG4.2 mutants. Indeed, we found that ADP levels increased in lithium-fed List^TG4.2 mutants, while ATP remained unchanged, leading to a decreased ATP/ADP ratio. NAD⁺ decreased by 31%, whereas NADH levels also appeared to decline but were highly variable and did not quite reach statistical significance (46% decrease, P = 0.06). As a result, we cannot draw firm conclusions about the NADH/NAD⁺ ratio, though the total NAD(H) pool is likely affected and appears to be significantly reduced overall. Notably, β-hydroxybutyrate (BHB) increased by 53%, a common response when reducing equivalents from glycolysis and the TCA cycle are insufficient to meet energy demands. BHB, generated via fatty acid oxidation, provides reductive potential by generating NADH during its conversion to acetoacetate. Acetoacetate is subsequently converted to acetyl-CoA, which replenishes the TCA cycle and facilitates further NADH production. Taken together, these data suggest that a reduction in electron donors may drive toxicity in lithium-fed List^TG4.2 mutants through ATP depletion.

As shown by the drastic lithium-induced changes in amino acid levels described above (Figure 9), lithium treatment had a significant impact on metabolic pathways associated with amino acid metabolism in both control flies and List^TG4.2 mutants. In control flies, lithium treatment markedly altered pathways such as valine, leucine, and isoleucine biosynthesis (FDR = 0.0123), the phenylalanine, tyrosine, and tryptophan biosynthesis (FDR = 0.0123), and phenylalanine metabolism pathways (FDR = 0.0123) (Table 1C). Notably, in List^TG4.2 mutants, lithium treatment had a pronounced effect on the arginine and proline metabolism pathway (FDR = 6.27 × 10⁻⁷) as well as other amino acid related pathways, including alanine, aspartate, and glutamate metabolism (FDR = 1.94 × 10⁻⁶), arginine biosynthesis (FDR = 4.38 × 10⁻⁶), and lysine degradation (FDR = 5.15 × 10⁻⁶) (Table 1D).

4. Discussion

In this study, we used a loss-of-function knock-in mutant, List^TG4.2, and List RNAi to investigate the role of the Drosophila SLC6 transporter List in resisting lithium toxicity. As shown in Figure 4C and 4E, List is expressed in the Malpighian tubules, which serve as the renal epithelium of insects. These tubules are essential for maintaining water and electrolyte balance in the hemolymph (insect blood) and for eliminating metabolic waste and toxins (Rodan, 2019). Notably, knocking down List in the Malpighian tubules significantly increased lithium-induced mortality (Figures 5A, E). Additionally, after lithium treatment, internal lithium levels were significantly higher in List^TG4.2 mutants compared to control flies (Figure 6). These findings highlight the critical role of List protein in the Malpighian tubules in reducing lithium levels in the treated animals, thereby protecting lithium-sensitive cells throughout the body.

Since List loss-of-function mutants show increased internal lithium levels, it is likely that the wild-type List protein either inhibits lithium absorption or facilitates its excretion. Considering that List is part of the SLC6 family of transporters, which mediate the movement of solutes across membranes, the latter explanation seems more plausible. SLC6 transporters typically rely on a sodium ion (Na⁺) concentration gradient to transport substrates across cell membranes. However, lithium ions (Li⁺), due to their similar charge and size, can substitute for Na⁺, albeit with varying efficiency depending on the transporter. For example, substituting sodium with lithium alters the transport-associated currents of the neutral amino acid transporter B₀AT1 (SLC6A19), suggesting that lithium can permeate through B₀AT1, although less efficiently than sodium (Margheritis et al., 2013). Similarly, studies on glycine transporters GlyT1 (SLC6A9) and GlyT2 (SLC6A5) have shown differing effects of lithium. In COS7 cells expressing GlyT2, lithium was found to have a stimulatory effect, whereas it exerted a noncompetitive inhibitory effect on GlyT1 transport (Benito-Muñoz et al., 2018). To better understand how the List protein prevents lithium accumulation and reduces lithium-induced mortality, future molecular and physiological studies are needed. These studies should investigate whether List directly mediates the transport of lithium across cellular membranes. If it does, its specificity and efficiency should be evaluated.

Bioinformatic analysis suggests that List encodes a putative amino acid transporter (Kasuya et al., 2009b), most likely a proline transporter. As shown in Figure 9, the List^TG4.2 mutation and lithium treatment cause complex alterations in the levels of various amino acids. Since the metabolism of amino acids is highly interconnected through shared pathways, intermediates, and regulatory mechanisms (Ling et al., 2023), it is challenging to distinguish primary metabolic defects caused directly by List^TG4.2 and lithium treatment from secondary or tertiary changes. However, our results show that proline levels were the most significantly reduced in lithium-treated List^TG4.2 mutants (Figure 9), supporting the hypothesis that List serves as a proline transporter.

Other notable changes in amino acid profiles observed in lithium-treated List^TG4.2 mutants include a significant increase in asparagine and lysine levels (Figure 9). Both asparagine and lysine are derived from oxaloacetate, a metabolic intermediate involved in several critical biochemical pathways, including gluconeogenesis, the urea cycle, the glyoxylate cycle, fatty acid synthesis, the citric acid cycle, and amino acid synthesis (Arnold and Finley, 2023; Koendjbiharie et al., 2021). Additionally, we observed a concurrent decrease in the branched-chain amino acids isoleucine, leucine, and valine in lithium-treated control flies (Figure 9). A previous study investigating the effects of LiCl on gene expression profiles provided evidence that lithium significantly impacts branched-chain amino acid metabolism (Kasuya et al., 2009a). The findings from our current metabolomic analysis further confirm that lithium treatment strongly influences the metabolism of these amino acids.

In addition to the above changes in amino acid levels, our metabolomics data indicate that the List^TG4.2 mutation affects cellular redox pathways on a normal diet, but the effects are much more pronounced when the flies are fed lithium (Table 1). We observed changes in pentose phosphate metabolites, as well as in many metabolites that have antioxidant capacity themselves. However, perhaps most strikingly, the top two most significant metabolite hits between control and List^TG4.2 mutants upon feeding with lithium (Ctrl-L vs List-L, Supplementary Table 4B) were proline (FDR = 2.87 × 10⁻⁸, FC = 0.438) and ornithine (FDR = 4.42 × 10⁻⁸, FC = 2.83), which can be interconverted through the intermediate 1-pyrroline-5-carboxylate (P5C). This interconversion reaction plays a significant role in regulation of cellular redox balance through electron shuttling between the key redox couples NAD⁺/NADH and FADH/FADH₂ (Chalecka et al., 2021). Furthermore, P5C can also be reversibly converted to glutamate, which uses NADP⁺/NADPH as a cofactor. The altered proline levels we observed in List^TG4.2 mutants could therefore be altering cellular redox balance through changing the ratios between proline, ornithine, and glutamate and their respective cofactors. Since ornithine is a key metabolite in the urea cycle, the P5C reactions could also help explain why we observed significant changes in flux through the urea cycle in List^TG4.2 mutants.

In addition to the effects observed on redox pathways, lithium-fed List^TG4.2 mutants also exhibited signs of ATP depletion, including elevated ADP levels and a reduced ATP/ADP ratio. Energy depletion often coincides with increased oxidative stress, as cells divert electrons from classical NADH-producing pathways (glycolysis and the TCA cycle) toward NADPH-producing pathways to support antioxidant systems, primarily the pentose phosphate pathway (PPP). The decreases observed in glycolytic and TCA intermediates, along with increases in β-hydroxybutyrate (BHB) and PPP intermediates, suggest that List^TG4.2 mutants may engage in pentose cycling to boost NADPH production while compensating for energy deficits by generating NADH, FADH₂, and TCA intermediates through fatty acid oxidation, BHB metabolism, and anaplerotic amino acid pathways. Notably, elevated BHB is most commonly associated with high energy demand or fasting (Newman and Verdin, 2014), further supporting the hypothesis that lithium-fed List^TG4.2 mutants may experience toxicity through energy depletion.

Another interesting feature in our data is that lithium treatment led to significantly elevated ribose levels in List^TG4.2 mutants (FDR = 0.0157, FC = 1.70, Supplementary Table 4D), which could be a result of increased flux through the PPP. Elevated ribose has been linked to cognitive impairment in type 2 diabetes (Yu et al., 2019), and it is possible that increases in ribose induced by lithium administration could lead to deleterious effects on the nervous system. In addition, ribose glycates proteins faster than glucose (Wei et al., 2012), and increased ribose levels have been shown to increase advanced glycation end product generation, which is linked to endoplasmic reticulum (ER) stress. Furthermore, elevated ribose has also been shown to induce severe nephropathy in mice (Hong et al., 2018), suggesting ribose may be a source of lithium-induced renal impairment. Interestingly, inhibition of NF-κB phosphorylation significantly reduced this effect (Hong et al., 2018), raising the possibility that inhibition of this pathway could treat or prevent lithium-induced renal damage, which is a major issue for patients taking lithium for bipolar disorder and other psychiatric conditions.

5. Conclusion

This study, using the Drosophila model, demonstrates that knocking down the SLC6 transporter gene List in the Malpighian tubules significantly increases mortality caused by a diet supplemented with lithium chloride. Based on our lithium and metabolomic analyses, we hypothesize that List functions as a proline transporter, mitigating lithium toxicity by reducing internal lithium levels and modulating amino acid profiles in a way that protects against oxidative stress and energy depletion. However, further experiments are needed to confirm this hypothesis, including assays using selective inhibitors known to target the transport of specific amino acids, such as proline. Additionally, mechanistic studies are required to determine whether lithium toxicity in List^TG4.2 mutants can be rescued through metabolic interventions aimed at counteracting oxidative stress and/or ATP depletion.

Highlight.

• Lithium remains a key treatment for bipolar disorder.

• The mechanisms behind lithium’s beneficial and adverse effects are unclear.

• The Drosophila SLC6 transporter List may function as a proline transporter.

• List knockdown in the Malpighian tubules increases lithium-induced mortality.

• List dysfunction alters amino acid metabolism and redox signaling pathways.

Abbreviations:

AD

Alzheimer’s disease

ALS

amyotrophic lateral sclerosis

PD

and Parkinson’s disease

ASD

autism spectrum disorders

DIOPT

Drosophila integrative ortholog prediction tool

ER

endoplasmic reticulum

FDR

false discovery rate

FC

fold change

GABA

gamma-amino butyric acid

GC-MS

gas chromatography-mass spectrometry

PROT

high-affinity brain L-proline transporter

ICP-MS

inductively coupled plasma-mass spectrometry

LC-MS

liquid chromatography-mass spectrometry

List

Lithium-inducible SLC6 transporter

PPP

pentose phosphate pathway

RNAi

RNA interference

S7P

sedoheptulose 7-phosphate

qRT-PCR

quantitative reverse transcription polymerase chain reaction

rp49

ribosomal protein 49

SLC6

solute carrier 6

sPLS-DA

sparse partial least squares discriminant analysis

SIC

swappable integration cassette