Kynurenine Metabolism in the Fat Body Non-autonomously Regulates Imaginal Disc Repair in Drosophila.

Tissue interactions are critical for maintaining homeostasis; however, little is known about how remote tissue regulates regeneration. Previously, we established a genetic dual system that induces cell ablation in Drosophila larval imaginal discs and simultaneously manipulates genes in non-damaged tissues. Using humoral metabolome analysis and a genetic damage system, we found that the Tryptophan (Trp)-Kynurenine (Kyn) pathway in the fat body is required for disc repair. Genetic manipulation of Trp-Kyn metabolism in the fat body impaired disc regeneration without affecting wing development. In particular, the fat body-derived humoral kynurenic acid (KynA) was required for disc repair. The impairment of S-adenosylmethionine (SAM) synthesis from methionine (Met) in the fat body hampers the maintenance of KynA levels in hemolymph at the early stage of disc repair, suggesting a connection between Met-SAM and Trp-Kyn metabolisms. Our data indicate KynA from the fat body acts as a permissive metabolite for tissue repair and regeneration.

Introduction

Tryptophan-kynurenine (Trp-Kyn) metabolites have shown colorful traits in various physiological functions throughout evolution. Trp is an essential amino acid, which serves as a substrate for protein synthesis as well as several bioactive compounds, such as serotonin and melatonin. However, in mammals, the vast majority of the catabolism of Trp happens through the Trp-Kyn pathway and generates a range of metabolites involved in inflammation, immune response, host-microbiome signaling, and neuronal excitability (Cervenka et al., 2017). Kyn is converted into two metabolites, 3-hydroxykynurenine (3-HK) and kynurenic acid (KynA). The metabolite 3-HK is known to exert neural toxicity effects linked to excitotoxicity (Zwilling et al., 2011). Additionally, 3-HK is finally converted to nicotinamide adenine dinucleotide (NAD+) via quinolinic acid in mammals, although 3-hydroxyanthranilic acid oxygenase or quinolinate phosphoribosyltransferase are not found in C. elegans or Drosophila melanogaster. On the other hand, KynA shows neuronal protection effects as an antagonist of N-methyl-D-aspartate receptor (NMDAR) and α-7 nicotinic acetylcholine receptor (α7nAChR) (Wirthgen et al., 2018). In addition to such neuronal contributions, Kyn and KynA bind to the transcription factor aryl hydrocarbon receptor (AhR) in immune cells and ameliorate immune activation (Wirthgen et al., 2018). KynA is also reported to bind to orphan G protein-coupled receptor 35 (GPR35) in the intestine and macrophages to modulate mucosal homeostasis or TNF-α secretion (Stone et al., 2013). Thus, the modulation of Trp-Kyn metabolism underlies the mechanisms of diseases.

Tissue damage causes several risks, including infection, tissue malfunction, and even lethality. Multicellular organisms have a homeostatic capacity against tissue damage. However, the underlying mechanisms are not confined to the damaged tissues; rather, they involve organismal regulation, including the immune, nervous, and endocrine systems, and even adipose tissues. Interesting examples of tissue interactions between damaged tissue and other tissues have been reported. When the leg of a mouse is locally injured, the contralateral leg muscle shows a high capacity to reenter the cell cycle in muscle stem cells (MuSCs) through the circulation of hepatocyte growth factor activator (HGFA) (Rodgers et al., 2017). A similar phenomenon is observed in appendage regeneration in axolotls. During blastema formation in a cut appendage, the contralateral uninjured appendage also shows cell proliferation via mTOR signaling (Johnson et al., 2018). We are focusing on such systemic damage responses (SDRs), including systemic wound responses (SWRs) (Kashio et al., 2017). When wounding by simple pricking was performed in adult Drosophila cuticle, an inter-tissue reaction was observed in the intestine and neurons, which are necessary for host survival (Lee and Miura, 2014). Interestingly, SDR also directly contributes to tissue repair processes. For example, reactive oxygen species (ROS) resulting from injuries induce the recruitment of leukocytes (white blood cells), which support fin repair in zebrafish (Yoo et al., 2011).

Despite many advances in the understanding of tissue regeneration, the molecular mechanisms of tissue non-autonomous regulation of repair processes are only beginning to be understood. This could be greatly enhanced by tissue-specific and comprehensive genetic analyses. Previously, we established a genetic cell ablation system for studying non-autonomous aspects of tissue repair in Drosophila (Kashio et al., 2016, 2017). Drosophila larvae possesses regenerative tissues, imaginal discs, which are known to show a remarkable ability to repair massive tissue damage with surgical ablation or genetic temporal ablation (Worley et al., 2012; Fox et al., 2020). In particular, wing discs are useful to investigate the extent of repair by observing adult wing phenotype. By combining a genetic temporal cell ablation system in imaginal discs with a genetic manipulation system in non-damaged tissues, we can perform functional analyses of systemic aspects of tissue repair. Drosophila has tissue-specific gene manipulation systems, such as the Gal4/UAS system and the Q system (Riabinina et al., 2015). We established a cell ablation system using a temperature-sensitive form of the diphtheria toxin A domain (DtAts) (Bellen et al., 1992). It has been demonstrated that DtAts is active at low temperatures (18°C) and induces cell death through nuclease activity but is inactivated at high temperatures (29°C) (Lee et al., 2005). Inducing DtAts expression using the Q system enabled us to manipulate Gal4/UAS-mediated gene expression in other organs, independent of the temporal tissue damage in wing discs. With this genetic “dual” system, we identified that Trp-Kyn metabolism in the fat body, an insect organ that is a functional counterpart of mammalian liver and adipose tissue, is required for disc regeneration without affecting normal wing development. In this study, we found that KynA, an end product of the Trp-Kyn metabolic pathway, is an essential hemolymph metabolite for non-autonomous tissue repair.

Results

Humoral Metabolomic Analysis Reveals that Trp-Kyn Metabolism Is Required for Disc Repair

First, we focused on the hemolymph as an important mediator. Hemolymph has been recognized as a homeostatic regulator at the systemic level (Droujinine and Perrimon, 2016; Leopold and Perrimon, 2007). We collected hemolymph from control and wing disc-damaged larvae at the early stage of disc repair (6 h after cell ablation; AA6) and performed metabolome analysis using capillary electrophoresis time-of-flight mass spectrometry (CE-TOF-MS) to explore the metabolic changes for tissue repair (Figure S1). Our previous data indicated that the levels of one of the essential amino acids, methionine (Met), and its downstream metabolite, S-adenosyl methionine (SAM), were lower in the ablated larval hemolymph at AA6 (Kashio et al., 2016). As expected, the level of Met was lower in disc-ablated larval hemolymph, which verified the metabolome analysis (Figure S1).

The metabolite with the lowest level was Met, whereas tryptophan (Trp) was the metabolite with the highest level in disc-damaged larval hemolymph (Figure S1). Drosophila has one rate-limiting enzyme in the Trp-Kyn pathway, Vermilion (V), which is a mammalian ortholog of Tryptophan 2,3-dioxygenase (TDO) (Figure 1A). Similarly to the strong expression of TDO in the mammalian liver (database from The Universal Protein Resource [UniProt]), vermilion is predominantly expressed in larval fat body (FB), a counterpart of the mammalian liver and white adipose tissue (data from FlyBase). Therefore, we checked the expression level of vermilion in the FB using quantitative RT-PCR (qRT-PCR) at 0 and 6 h after ablation (AA0, 6), and vermilion maintained its expression level in the FB of damaged larvae at AA6 (Figures 1A and 1B).

Figure 1

Involvement of Vermilion in Disc Repair

(A) Simplified diagram of Trp-Kyn metabolism. Kyn, Kynurenine; 3-HK, 3-hydroxykynurenine; KynA, Kynurenic acid; V, Vermilion; Cn, Cinnabar; Kyat, Kynurenine aminotransferase.

(B) Expression of v in the fat body. Control: +/UAS-DtA Ablation: WP-Gal4/UAS-DtA SEM was calculated from four independent samples. One-way ANOVA Tukey’s multiple comparison test was applied: NS, not significant; ∗p < 0.05.

(C) Vermilion mutant background caused severe wing phenotype after ablation. Statistical analysis was conducted using the Chi-squared test to compare the control (+/+;;WP) with v mutant (v/v;;WP); ∗∗∗p < 0.001. Temperature treatment was described in Materials and Methods.

We measured the metabolites of Trp-Kyn metabolism at AA0 and 6 in both disc-damaged and non-damaged larval hemolymph and tissues, including the FB, gut, cuticle (including larval muscle), and brain (Figures S2A–S2E). The level of Trp was higher in disc-damaged larval hemolymph at AA6 (Figure S2A), as observed in metabolome analysis (Figure S1). As for Kyn, the level tended to be higher in damaged larval hemolymph (Figure S2A), which was similar to the level in the FB (Figure S2B) and cuticle at AA0 (Figure S2D). KynA level was higher in almost all samples at AA6 other than brain (Figure S2E).

We also investigated metabolic change during the third larval development. The level of Trp-Kyn metabolites decreased at a whole-body level from the early to late third larval stage (Figure S2F), implying that the Trp-Kyn metabolism decreased toward the pupation stage and may be correlated with the decrement of the regenerative capacity of the imaginal disc as development proceeds.

FB-Specific Knockdown of Vermilion Affects Disc Repair Non-autonomously

To confirm the involvement of Vermilion in disc repair, we performed DtAts cell ablation with a vermilion mutant and it showed a severe wing phenotype (Figure 1C). Previously, oral administration of TDO inhibitor 680C91 was utilized for repressing Vermilion in Drosophila and ameliorated neurodegeneration (Breda et al., 2016). Oral administration of 100 μM TDO inhibitor was performed from 24 h before AA6, which resulted in the impairment of disc repair (Figures S3A and S3B). Additionally, TDO inhibitor caused a decrease of the levels of downstream metabolites, 3-HK and KynA (Figure S3C). These results support the contribution of Trp-Kyn metabolism to disc repair.

To investigate the tissue-specific involvement of Vermilion in the FB in disc repair, vermilion (v) was knockdown in the FB. Knockdown of vermilion did not affect normal wing development but worsened the adult wing phenotype after disc ablation (Figures 2A, 2F, 2G and S5A). We then examined the expression of morphogen Wingless (Wg) and proliferative cells marked with phospho-histone 3 (PH3). At AA0 (following 38 h of cell ablation treatment), there were no significant differences between the control larval discs and vermilion knockdown larval discs (Figures 2B, 2C, and 2H). Damaged discs showed a weakened signal of Wg in the wing pouch (WP) region, and the number of PH3-positive cells in the WP region was lower compared with other disc regions. Thirty hours after the shift back to 29°C (condition for cell ablation stop; AA30), cells in the WP were proliferating and the typical Wg staining pattern was restored during the recovery phase in control discs (Figures 2D, 2D′, and 2H), indicating that the repair process had been activated. On the other hand, wing discs in FB-specific vermilion knockdown larvae indicated a hampered repatterning of Wg without substantial differences in cell proliferation (Figures 2E, 2 E′, and 2H). Additionally, the knockdown of vermilion in the WP region did not worsen wing normal development and disc repair (Figure S3D, S5C and S5D), which supports the idea that Vermilion acts non-autonomously in disc repair.

Figure 2

Genetic Manipulation of Vermilion in the Fat Body Causes Impairment of Wing Disc Repair

(A) Comparison of adult wing sizes. From top to bottom: WP>+, FB>lacZ-RNAi (n = 154); WP>+, FB>v-RNAi 1 (n = 144); WP>+, FB>v-RNAi 2 (n = 216); WP> DtA, FB>lacZ-RNAi (n = 60); WP> DtA, FB>v-RNAi 1 (n = 80); and WP> DtA, FB>v-RNAi 2 (n = 112). Statistical analysis was conducted using the chi-square test to compare the control (WP>+, DtA, FB>+) with the treatment larvae; ∗∗∗, p < 0.001.

(B-E') Representative examples of wing discs developed within the indicated time course: WP (B, B', D, and D') and WP (C, C', E, and E').

(F and G) Representative example of adult wings of Group A of WP> DtA, FB>lacZ-RNAi, and Group D of WP> DtA, FB>v-RNAi 1. Black scale bar, 1 mm.

(H) PH3 positive cell numbers in the wing pouch region was quantified. Samples were same as Figure 2B’, C’, D’ and E’. n = 4, 4, 7, and 6 for each sample. Error bars indicate standard error of the mean.

(I) Timing of pupation for “Control + FB” larvae (WP+, FB), “Control + FB” larvae (WP+, FB), “Damage + FB” larvae (WP), and “Damage + FB” larvae (WP). n = 60, 73, 38, and 35, respectively. SEM was calculated from four repeated experiments. Developmental timing is represented as both the fraction of total larvae pupated and hours after larval hatch. Temperature treatment was described in Materials and Methods.

Previous studies indicated that damage in discs causes a developmental delay, which is required for modulating organ size control (Colombani et al., 2012; Garelli et al., 2012; Halme et al., 2010). The influence of Vermilion on developmental timing was not observed in either damaged or non-damaged conditions (Figure 2I), indicating that Vermilion regulates disc repair by affecting the repair processes and not the developmental timing.

We also examined tissue remodeling and dying cell clearance with phalloidin (F-actin probe) and active executioner caspase, Dcp1, antibody staining. Recent studies indicated that dying cell clearance and tissue remodeling are involved in disc repair (Cosolo et al., 2019; Iida et al., 2019; Yoo et al., 2016). Notably, vermilion knockdown in the FB affected tissue structure and dying cells clearance (Figures S4A and S4B). At AA30, regenerative discs indicated the basal extrusion of dying cells; dying cells were accumulated throughout the discs, and the tissue structure was winding, establishing excess folds in non-regenerative discs (Figures S4A and S4B). We also quantified the Dcp1 signals in a projected XZ cross section of wing discs and found that dying cells or debris exist on relatively apical sides of discs compared with lacZ knockdown larvae (Figure S4A’, S4B′, and S4C). Additionally, the intensity of Dcp1 in the WP region did not show a significant difference between regenerative and non-regenerative discs (Figure S4D), implying that the amount of dying cells was not significantly increased by Trp-Kyn metabolism impairment.

Humoral KynA Derived from the FB Is Required for Disc Repair

To narrow down the metabolites responsible for disc repair, we measured Trp-Kyn metabolites in the hemolymph at the early stage of disc repair (AA6) using LC-MS/MS. As for Trp, the amount of Trp at AA6 in cell-ablated larval hemolymph was the same given by the metabolome analysis (Figure 1A), and v-RNAi caused a higher amount of Trp in the hemolymph (Figure 3A). Importantly, the KynA level, and not that of 3-HK, was higher in the cell-ablated condition at AA6 and significantly decreased in v-RNAi larval hemolymph (Figure 3A). Furthermore, FB-specific knockdown of Kynurenine aminotransferase (Kyat), a gene required for KynA synthesis (Figure 1A), caused a severe wing disc phenotype (Figure 3B), and adults showed a worse wing phenotype compared with the knockdowns of cinnabar (cn), a kynurenine 3-monooxygenase (KMO) producing 3-HK (Figures 3C and S5B). Therefore, these results indicated that KynA synthesis in the FB was required for disc repair.

Figure 3

Humoral KynA from the FB Is Required for Disc Repair

(A) The levels of Trp, KynA, and 3-HK in the disc-cell-ablated and non-ablated larval hemolymph with FB-specific RNAi of lacZ or v at AA6. The metabolites were measured with UPLC-MS/MS. SEM was calculated from four to five independent samples. One-way ANOVA Tukey's multiple comparison test was applied: NS, not significant; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

(B) Representative examples of wing discs developed within the indicated time course. The discs were stained with anti-Wg. White scale bar, 100 μm.

(C) Comparison of adult wing sizes when Trp-Kyn enzymes are manipulated with the FB-Gal4 driver. n = 78, 88, 58, 94, 46, and 40 for each sample from top to bottom. Statistical analysis was conducted using the chi-square test to compare the control (WP> DtA, FB>lacZ-RNAi) with the treatment larvae; NS, not significant; ∗, p < 0.05; ∗∗∗, p < 0.001.

To investigate the requirement of humoral KynA for disc repair, an oral KynA feeding experiment was conducted (Figure 4A). KynA administration from AA0 rescued the level of humoral KynA of vermilion knockdown larvae (Figure 4B). Furthermore, the severe wing phenotype was also recovered via KynA feeding (Figures 4D and S5E), and KynA treatment also decreased the population of severe Wg phenotype (Class C) of the FB-specific vermilion-RNAi larval disc (Figure 4F). Because KynA is an end product of the Trp-Kyn pathway, these results suggested the requirement of humoral KynA, rather than in the FB tissue, for disc repair. On the other hand, KynA treatment did not enhance the regenerating phenotype of FB-specific lacZ-RNAi larval discs and adult wings (Figures 4C, 4E, 4G and S5F), possibly because humoral KynA level in damaged control larvae was saturated to supporting disc repair. As for PH3 staining, KynA treatment did not increase the number of PH3-positive cells in the WP region of either FB-specific lacZ-RNAi or vermilion-RNAi larvae (Figure 4H), which was accompanied with the results of no apparent effects of vermilion-RNAi in the FB on PH3 (Figures 2B–2E and 2H).

Figure 4

Oral KynA Administration Rescues Disc Repair in Trp-Kyn Metabolism-Impaired Larvae

(A) Schematic view of KynA administration.

(B) The levels of Trp and KynA in ablated and v knockdown (v-RNAi 1) larval hemolymph at AA6 under the KynA administration condition. Error bars represent the SEM, and p-values were calculated using the two-tailed student’s t-tests. SEM was calculated from six independent samples; NS, not significant; ∗∗p < 0.01.

(C) The levels of Trp and KynA in ablated and lacZ knockdown larval hemolymph at AA6 under the KynA administration condition. Error bars represent the SEM, and p-values were calculated using the two-tailed student’s t-tests. SEM was calculated from six independent samples; NS, not significant; ∗∗p < 0.01.

(D) Comparison of adult wing sizes in the condition of v knockdown in the fat body with or without KynA administration. n = 56 and 70. Chi-squared test to compare control with KynA treated larvae; ∗∗p < 0.01.

(E) Comparison of adult wing sizes in the condition of lacZ knockdown in the fat body with or without KynA administration. n = 54 and 32. Chi-squared test to compare control with KynA treated larvae; NS, not significant.

(F and G) Representative example of wing discs at AA30: WP (F) and WP (G).

(H) PH3 positive cell numbers in the wing pouch region were quantified. n = 15, 19, 15, and 17. Error bars indicate standard error of the mean.

Impairment of SAM Synthesis in the FB Represses the Level of Kyn and KynA at the Early Stage of Disc Repair

In addition to the Trp-Kyn pathway, another major metabolic pathway in the FB is the Met-SAM pathway (Kashio et al., 2017). Our previous study indicated that the expression level of SAM synthase (sams) was upregulated in the FB of disc-cell-ablated larvae at AA6 and that sams in the FB was required for disc regeneration (Kashio et al., 2016). Therefore, we wondered whether there were any interrelationships between these two major metabolic pathways in the FB. We compared the metabolic changes of FB-specific vermilion or sams knockdown larval hemolymph at AA0, 6, and 12 (Figure 5A). Vermilion knockdown resulted in the highest level of Trp compared with the control, but sams knockdown did not present the same results as the vermilion knockdown. On the other hand, the downregulation of KynA in both the vermilion and sams knockdowns was observed at AA6 (a similar tendency was also observed at AA12). As for Kyn and 3-HK, vermilion knockdown tended to repress the level of Kyn or 3-HK at all time points, whereas the sams knockdown showed almost the same levels as the control, besides a decreasing tendency of Kyn at AA6. From these results, we assumed that Sams is required in the FB for the maintenance of KynA in the early stage of disc repair (Figure 5B).

Figure 5

Met-SAM and Trp-Kyn Pathways in the FB Remotely Regulate Disc Repair via Humoral KynA

(A) The levels of Trp, Kyn, KynA, and 3-HK in disc-ablated larval hemolymph with FB-specific RNAi of lacZ, v (v-RNAi 1), or sams at AA0, 6, and 12. The metabolites were measured using UPLC-MS/MS. SEM was calculated from five to six independent samples. One-way ANOVA Tukey's multiple comparison test was applied: ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

(B) Scheme of the tissue nonautonomous effects of Met-SAM and Trp-Kyn metabolisms in the fat body on repairing disc via KynA.

Discussion

In the context of organ communications, peptides or secreted proteins are mainly studied as signaling molecules (Droujinine and Perrimon, 2016), even though small molecules like metabolites have several other advantages, such as the possibility of rapid diffusion or transcellular transport. In this paper, we focused on metabolites as inter-tissue mediators. Trp-derived metabolites function in diverse situations. As for tissue regeneration, a recent planarian study implied the involvement of indole, a Trp catabolite via the microbiome, impaired regeneration (Lee et al., 2018). Ischemia by cardiac arrest is reported to change the plasma KynA level in rats, pigs, and humans (Ristagno et al., 2013), and a study reported that KynA has a protective effect against ischemic tissue damage (Olenchock et al., 2016), which supports our idea that KynA regulates tissue repair non-autonomously. Even though the concrete mechanisms are still under investigation, such functions imply the multifactorial mechanisms of Trp metabolites on tissue repair and regeneration.

The reason why humoral Trp level was higher in AA6 of disc damaged larvae was also unclear. During normal development of larvae, both Met and Trp metabolites gradually decrease from mid-larval stage at a whole-body level (Figure S2F) (Kashio et al., 2016). Considering that developmental progression decreases the regenerative capacity of discs autonomously (Narbonne-Reveau and Maurange, 2019), this implies that a damage-induced developmental delay maintained the level of humoral Trp in regenerating larvae. Besides the effects of developmental delay, it also indicated that plasma free Trp is increased under fatigue conditions in mammals. For example, exercise increased the plasma free Trp in rats, which is associated with an increase of plasma nonesterified fatty acid (NEFA) competing for binding to albumin with Trp (Fernstrom and Fernstrom, 2006). Considering this point, we can also assume that disc injury causes some stresses like fatigue, which increases the amount of Trp in larval hemolymph during disc repair. Additionally, the gut in disc-damaged larva showed a lower level of Trp at AA6 (Figure S2C), which implied that the intake of Trp from gut to hemolymph possibly affected humoral Trp level. Furthermore, considering that Trp is the lowest-frequency amino acid in eukaryote proteins (Id et al., 2019), the lower usage of Trp for protein synthesis at the early stage of disc repair is also one of the possible causes of the humoral level of Trp. Future studies are required for the trigger of Trp-Kyn pathway changes during disc repair.

Several targets of KynA were studied in various biological contexts. As for NMDAR, α7nAChR, or glutamate receptor (GluR), the neural activation of these receptors is repressed by the allosteric binding of KynA (Schwarcz et al., 2012; Wirthgen et al., 2018). Several papers indicated that GPR35 is a target of KynA, even though a paper also showed an undetectable binding capacity of KynA to GPR35 (Inoue et al., 2012). For instance, muscle-derived KynA increased energy expenditure by activating GPR35 (Agudelo et al., 2018). KynA was also reported as a potent AhR endogenous ligand that could induce interleukin-6 (IL-6) production and xenobiotic metabolism (DiNatale et al., 2010). For transportation of KynA into cells, human organic anion transporters hOAT1 (SLC22A6) and hOAT3 (SLC22A8) mediate the transport of KynA using a Xenopus laevis oocyte expression system (Uwai et al., 2012). So far, there are no reports about the obvious way to use KynA in Drosophila. Furthermore, it is still unclear if KynA acts in disc repair directly or indirectly; thus, further studies seeking the KynA target are needed.

Our results also suggested interactions between Met-SAM and Trp-Kyn metabolism in the early stage of disc regeneration. Although the nature of the connection between SAM synthesis and tryptophan catabolism during disc repair is still unclear, the enzymatic activity in the Trp-Kyn pathway could be affected by Met or SAM level. Mouse kynurenine aminotransferase III (KAT III) that catalyzes the transamination of Kyn to KynA is effectively inhibited by Met in vitro (Han et al., 2009). We previously showed that sams knockdown in larval fat body significantly increased Met level (Kashio et al., 2016); thus, Drosophila Kyat activity might be suppressed and the KynA production could be reduced in sams knockdown condition.

Previous studies indicated the enzymatic regulatory mechanisms by methylation, for example, it has been reported that methylation of the catalytic subunit in protein phosphatase 2 A (PP2A) keeps the phosphatase activity by hampering the binding of TOR signaling pathway regulator (TIPRL), which binds to the demethylated PP2A and disassembles holoenzyme (Wu et al., 2017). Additionally, Lys 260 methylation in MAPK kinase kinase 2 (MAP3K2) blocks the interaction with its negative regulator PP2A, leading to an increase in the activity of MAP/ERK kinase 1 (MEK1) and MEK2 (Mazur et al., 2014). Even though it has not been investigated whether enzymatic activity or gene expression in the Trp-Kyn pathway is regulated by SAM, including the involvement of PP2A and MAP3K2, further biochemical experiments will be needed for the requirement of methylation on the regulation of enzymes in Trp-Kyn pathway.

Correlation of Met-SAM and Trp-Kyn metabolism changes in several biological processes has been reported. As for the regulation of pluripotency in human embryonic stem cells (hESCs), both the Kyn/Trp ratio and SAM level is higher in primed than in naive hESCs (Sperber et al., 2015). It was also indicated that Trp-Kyn metabolites affect histone modification. Although the mechanisms have not been studied, the supplementation of Kyn, 3-HK, and Anthranilate during the neural differentiation from committed human iPSCs induced histone H3 lysine 4 trimethylation (H3K4me3) for maintenance of neural gene expression (Hayakawa et al., 2019). On the other hand, SAM is also required for maintenance of undifferentiated ESC and iPSC and SAM depletion enhanced differentiation (Shiraki et al., 2014), and Kyn catabolism by KMO and KAT II is enhanced during ectodermal differentiation (Yamamoto et al., 2019). Thus, both Trp-Kyn and Met-SAM pathways act in the same direction for regulation of differentiation status. Additionally, glioblastoma shows abnormal Met and Trp-Kyn metabolic changes (Palanichamy et al., 2016). Notably, in the context of aging and the reduction of dietary amino acids, tryptophan and methionine can extend the lifespan in animal models (Fontana and Partridge, 2015). Alzheimer's disease and mild cognitive impairment are associated with an overlapping pattern of perturbations in Trp, tyrosine, Met, and purine pathways (Kaddurah-Daouk et al., 2013). Additionally, both Trp and Met restrictions exert protective effects in ischemia/reperfusion damage (López-Otín et al., 2016). Therefore, further studies on the interconnection of Met and Trp-Kyn metabolisms would be worthy of note in various biological contexts. Since our data suggested the metabolic connection between Met-SAM and Trp-Kyn pathway in the fat body during disc repair, it is the next important research topics to know the regulatory mechanisms of these two pathways for repairing discs.

Limitations of the Study

Compared with the Met-SAM pathway, the Trp-Kyn metabolic pathway is rather simple, and its catabolic pathway proceeds linearly to produce KynA as an end product. We conducted fat body-specific gene manipulation and tissue-specific measurements of metabolites for the Trp-Kyn pathway in this study. Although we found the production of KynA in the FB is required for systemic control of disc repair and regeneration, we could not exactly reveal the dynamics of KynA and whether KynA directly acts on wing imaginal disc or on other tissues. In mammals, KynA can exert its function through binding to NMDAR, a7nAChR, GluR, GPR35, or AhR. Therefore, it is necessary to identify KynA-binding receptors required for disc repair. It is also intriguing to know whether KynA functions in tissue regeneration in other organisms.

Resource Availability

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Masayuki Miura (miura@mol.f.u-tokyo.ac.jp).

Materials Availability

Materials and protocols used in this study are available from the authors upon request.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.