A total of 47 male C57BL/6 mice (8 weeks old; average weight, 22 g) were procured from the Shanghai SLAC Laboratory Animal Center (Shanghai, China) and housed at the Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (Hangzhou, China). The mice were maintained in a specific pathogen-free facility under controlled environmental conditions (temperature, 22±2°C; relative humidity, 50±10%; and a 12-h light/dark cycle) with ad libitum access to sterile water and their respective diets (standard chow or customized diets as detailed below). The animal study protocols were approved by the Animal Research Ethical Committee of the Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (approval no. SRRSH2025-0047). To address distinct experimental objectives, the mice were randomly assigned to specific cohorts as follows: i) For metabolomics profiling, mice were fed with standard chow and randomly divided into a normal control (NC) group and an imiquimod (IMQ)-induced model group (n=6 per group); ii) to evaluate the efficacy of phenylalanine (Phe) diets, mice were randomly assigned to three groups (n=5 per group) and fed diets containing either 0.25%, 1% (serving as the baseline control) or 2% Phe, which were manufactured by Hangzhou Hangsi Biotechnology Co., Ltd. (Table SI); and iii) For the in vivo L-type amino acid transporter 1 (LAT1) inhibition experiment, mice were allocated into four groups (n=5 per group). For the relevant experimental groups, the LAT1 inhibitor JPH203 (cat. no. s8667; Selleck Chemicals) was administered through subcutaneous injection at a dose of 50 mg/kg.

To establish mouse models of psoriasis, IMQ cream (Sichuan Med-Shine Pharmaceutical Co., Ltd.) was applied topically for 5 consecutive days. For the back skin model, 62.5 mg IMQ cream was applied daily to the shaved dorsal skin. For the ear inflammation model, to ensure consistent and complete absorption of the full dose and minimize the risk of medication loss due to animal grooming, 25 mg IMQ cream was applied evenly across both the dorsal and ventral surfaces of the right ear. All mice were weighed and recorded before the induction of psoriasis. Health status and behavioral patterns of all animals were monitored twice daily (morning and evening) during the 6-day study period to assess disease progression and treatment efficacy. An objective scoring system was used to score the severity of skin inflammation based on the clinical Psoriasis Area and Severity Index (PASI) (26). Next, erythema, thickening, and scaling were rated on a 0-4 scale, where 0, 1, 2, 3 and 4 indicated no symptoms, mild, moderate, severe or extremely severe symptoms, respectively. The cumulative scores served as the severity of inflammation index (scale 0-12). The researchers were blinded during the experiment and data analysis.

To monitor the dynamic progression of psoriasiform inflammation, the macroscopic clinical severity of the mouse skin was evaluated and recorded daily. The final PASI scores analyzed and presented in the present figures were recorded on day 6 of the experiment, prior to euthanasia. On day 7, all terminal procedures were performed in a strict sequence. Mice were euthanized through CO₂ asphyxiation with a controlled displacement rate of 30% of the chamber volume per min and mortality was strictly verified by determining the absence of heartbeat and respiration. Immediately following the confirmation of mortality, blood was collected through retro-orbital bleeding and lesional skin tissues and plasma were subsequently harvested for further histological and metabolic analyses. The excised spleens were immediately weighed and the spleen index was calculated as the ratio of spleen weight to body weight. The acanthosis thickness was analyzed using ImageJ software (version 1.53; National Institutes of Health) and the thickness of the ear skin was measured with a vernier caliper.

Mouse tissues and plasma were collected and analyzed by Wuhan Metware Biotechnology Co., Ltd. Specifically, 100 mg tissue or 100 µl plasma was homogenized in 500 µl 70% aqueous methanol solution. The mixture underwent a 3 min vortex-mixing step at 2,500 rpm, followed by centrifugation (13,800 × g, 10 min, 4°C). A total of 300 µl supernatant was isolated into fresh vials for liquid chromatography (LC)-mass spectrometry (MS) analysis. To ensure analytical stability, quality control samples were generated by combining equal aliquots from every individual specimen and handled identically.

Sample extracts were analyzed using an LC-electrospray ionization (ESI)-MS/MS system (ExionLC™ AD UPLC coupled with a QTRAP^® 6500+ System; SCIEX). The instrument operated under both positive (5,500 V) and negative (-4,500 V) ion spray modes utilizing an ESI Turbo Ion-Spray (SCIEX) interface (source temperature: 550°C and curtain gas: 35.0 psi), managed using Analyst^® (version 1.6; SCIEX) software. Multiple reaction monitoring transitions were continuously optimized for each elution window of target amino acids.

Raw mass spectrometry outputs were transformed into MzXML files utilizing ProteoWizard MSConvert (version 3.0; http://proteowizard.sourceforge.net) prior to XCMS integration (using the 'xcms' package on the R platform; version 4.1.2; https://bioconductor.org/packages/xcms). Strict thresholds [variable importance in projection (VIP) >1 and P<0.05] were applied within the Partial Least Squares Discriminant Analysis model to filter differential metabolites. Pathway impact assessments were conducted using MetaboAnalyst (version 5.0; https://www.metaboanalyst.ca), where pathways exhibiting an impact factor >0.1 were deemed highly significant.

Mouse skin samples were immersed in 4% paraformaldehyde for optimal fixation at room temperature for 24 h. Tissues were subsequently dehydrated, cast into paraffin blocks and microtomed into 5-µm cross-sections. H&E staining was carried out following the manufacturer's instructions (Beyotime Biotechnology). For immunofluorescence assays, the sections were first deparaffinized in xylene and rehydrated using a descending ethanol gradients. Heat-mediated antigen unmasking was performed at 95°C for 15 min in a pH 6.0 citrate buffer. To prevent non-specific binding, slides were treated with a 5% BSA solution (Beyotime Biotechnology) enriched with 0.1% Triton X-100 for 1 h at room temperature. Primary antibody incubation with anti-p-p65 (1:200 dilution; cat. no. 3033S; Cell Signaling Technology, Inc.) occurred overnight at 4°C. After rigorous washing with PBS. specimens were exposed to CoraLite^® Plus 488-conjugated AffiniPure goat anti-rabbit IgG (H+L) (1:200 dilution; cat. no. RGAR002; Proteintech Group, Inc.) for detection of p-p65, for 1 h at room temperature in the dark. Subsequently, specimens were incubated with anti-CD11c (Alexa Fluor^® 555 conjugate) (1:200 dilution; cat. no. 64675S; Cell Signaling Technology, Inc.) overnight at 4°C. CD11c was visualized directly via its Alexa Fluor^® 555 conjugate without requiring a secondary antibody. All images were eventually captured utilizing a digital slide scanning system (Konfoong Bioinformation Tech Co., Ltd.).

Single cells from the ear were generated according to the methods of a previous study (27). To prevent non-specific Fc receptor binding, single-cell suspensions were pre-treated with the anti-mouse CD16/32 antibody (1:500 dilution; cat. no. 553141; BD Biosciences) on ice for 20 min prior to the surface marker labeling. The antibodies (all used at 1:500 dilution) for flow cytometry included allophycocyanin (APC)-A700-CD45 (56-0454081; eBioscience; Thermo Fisher Scientific, Inc.), FITC-CD3 (cat. no. 11-0032-82; eBioscience; Thermo Fisher Scientific, Inc.), BV650-CD11b (cat. no. 416-0112-80; eBioscience; Thermo Fisher Scientific, Inc.) and BV421-LY6G (cat. no. 404-9668-80; eBioscience; Thermo Fisher Scientific, Inc.). The Zombie Aqua™ Fixable Viability Kit (cat. no. 423101; BioLegend, Inc.) was also used. After washing, the cells were assayed with a CytoFLEX LX flow cytometer (Beckman Coulter, Inc.) and the data were analyzed using FlowJo software (version 10; BD Biosciences).

Total RNA was isolated from harvested skins or DCs using RNAiso Plus (Takara Bio, Inc.). Reverse transcription for cDNA synthesis was conducted utilizing the PrimeScript™ RT Reagent Kit (cat. no. RR037A; Takara Bio, Inc.) following the provided technical guidelines. RT-qPCR was conducted using the Hieff™ qPCR SYBR Green Master Mix (cat. no. 11201ES08; Shanghai Yeasen Biotechnology Co., Ltd.) on a LightCycler^® 480 system (Roche Diagnostics GmbH). The thermocycling program commenced with a 5 min pre-denaturation at 95°C, sequentially followed by 40 cycles (95°C for 10 sec and 60°C for 30 sec). Target gene expression levels were normalized against the housekeeping gene β-actin leveraging the 2^−ΔΔCq method (28). Primer sequences are listed in Table SII.

To isolate primary cells for both BMDC and naïve CD4⁺ T cell experiments, healthy and untreated male C57BL/6 donor mice (aged 6-8 weeks) obtained from the Shanghai SLAC Laboratory Animal Center, (Shanghai, China) were utilized. A total of 10 donor mice were euthanized using CO₂ asphyxiation (at a displacement rate of 30% of the chamber volume per min) and mortality was verified by cervical dislocation. To obtain sufficient cell yields, biological replicates for the following in vitro assays were generated using pooled samples from these donor mice.

To prepare BMDCs, femurs and tibias were flushed with RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) to collect bone marrow. The cell pellet was subsequently resuspended in red cell lysis buffer (Beyotime Biotechnology) for 3 min. The primary culture system consisted of RPMI-1640 fortified with 10% FBS (CellMax Technologies AB), 1% penicillin-streptomycin, 50 µM β-mercaptoethanol (cat. no. abs9592; Absin Bioscience, Inc.), 20 ng/ml GM-CSF (cat. no. 315-03; PeproTech^®; Thermo Fisher Scientific, Inc.) and 10 ng/ml recombinant mouse IL-4 (cat. no. 214-14; PeproTech^®; Thermo Fisher Scientific, Inc.). On days 3 and 5, half of the old culture medium was carefully removed and replaced with an equal volume of fresh medium. All cells were cultured at 37°C in a 5% CO₂ incubator. On day 6, BMDCs were used for relevant experiments. All in vitro pharmacological treatments were conducted at 37°C in a 5% CO₂ incubator. For in vitro stimulation, IMQ powder (cat. no. R837; InvivoGen) was dissolved in sterile water and added to the culture medium at a final concentration of 10 µg/ml. For inhibitor treatments, cells were incubated with the NF-κB inhibitor BAY11-7082 (10 µM; MedChemExpress) or the LAT1 inhibitor JPH203 (100 µM; Selleck Chemicals) for 2 h.

To detect maturation of BMDCs, expression of CD80, CD86 and major histocompatibility Complex (MHC)-II was analyzed on CD11c-positive cells using APC-labelled anti-CD80 (cat. no. 104713; BioLegend, Inc.), primary binding-labelled anti-CD86 (cat. no. 105021; BioLegend, Inc.), phycoerythrin (PE)-labelled anti-MHCII (cat. no. 12-5321-81, eBioscience; Thermo Fisher Scientific, Inc.) and PE-Cy7-labelled anti-CD11c (cat. no. 117317, BioLegend, Inc.) antibodies (all used at a 1:500 dilution). The incubation was performed on ice for 20 min in the dark. After antibody staining, cells were washed and analyzed with a CytoFLEX LX flow cytometer (Beckman Coulter, Inc.).

On day 6, BMDCs were pre-treated with phenylalanine (1 mM) for 48 h at 37°C in a humidified incubator containing 5% CO₂, followed by incubation with IMQ 10 µg/ml for 24 h to harvest the supernatant. Naïve CD4⁺ T cells were isolated from the spleens of C57BL/6 mice using an isolation kit (cat. no. 19765A, Stemcell Technologies, Inc.). A total of 200 µl BMDC supernatant was used to resuspend naïve CD4⁺ T cells for co-culture for 72 h. Differentiated T cells were stimulated with 750 ng/ml ionomycin (cat. no. 5608212; PeproTech^®; Thermo Fisher Scientific, Inc.), 50 ng/ml phorbol 12-myristate 13-acetate (cat. no. S1819; Beyotime Biotechnology) and brefeldin A (cat. no. 00-4506-51; eBioscience; Thermo Fisher Scientific, Inc.) for another 4 h at 37°C. The cells were then collected, stained with PE-labelled anti-CD4 (cat. 17-0041-82; eBioscience; Thermo Fisher Scientific, Inc.) and APC-labelled anti-IL-17A (cat. 17-7177-81; eBioscience, Thermo Fisher Scientific, Inc.) antibodies (both at a 1:500 dilution) on ice for 20 min in the dark. The stained cells were then analyzed on a CytoFLEX LX flow cytometer (Beckman Coulter, Inc.).

Total RNA was isolated from harvested tissues or DCs leveraging the RNAiso Plus reagent (Takara Bio, Inc.) as directed by the manufacturer. RNA purity and integrity were verified using a NanoDrop 2000 spectrophotometer and an Agilent 2100 Bioanalyzer (RIN >7.0). Transcriptomic libraries were constructed, verified using an Agilent 2100 Bioanalyzer, and subsequently subjected to 2×150 bp paired-end sequencing. The sequencing runs were executed on the Illumina NovaSeq™ X Plus system (Illumina, Inc.) at a final loading concentration of 1.5 nM by LC-Bio Technologies (Hangzhou) Co., Ltd. The 'DESeq2' package (29) was utilized within the R platform (version 4.1.2) for transcriptomic comparisons between specific groups, whereas the 'edgeR' algorithm (30) was employed for evaluating individual sample variations. Differentially expressed genes (DEGs) were defined by a false discovery rate (FDR) <0.05 and an absolute fold change ≥2. DEGs were then subjected to enrichment analysis of Gene Ontology (GO) functions and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

BMDCs were lysed in standard RIPA buffer (Beyotime Biotechnology) to liberate total cellular proteins. Lysates were subjected to low-temperature centrifugation (10,000 × g, 4°C, 20 min) to clear cellular debris. Protein quantification was executed using a BCA reagent kit (cat. no. FD2001; Fdbio Science). Equivalent protein masses (30 µg per lane) were resolved using 12% SDS-PAGE and electrically blotted onto 0.22 µm PVDF transfer membranes. The membranes were blocked with 5% non-fat dairy milk for 1 h at room temperature prior to overnight incubation at 4°C with distinct primary antibodies: anti-NF-κB p65 (1:1,000 dilution; cat. no. 8242; Cell Signaling Technology, Inc.), phosphorylated (p)-NF-κB p65 (Ser536; 1:1,000 dilution; cat. no. 3033; Cell Signaling Technology, Inc.), anti-IκBα (1:1,000 dilution; cat. no. 4814; Cell Signaling Technology, Inc.) and anti-p-IκBα (Ser32; 1:1,000 dilution; cat. no. 2859; Cell Signaling Technology, Inc.). An anti-β-tubulin antibody (1:1,000 dilution; cat. no. AF2839; Beyotime Biotechnology) was used as the loading control. The subsequent day, blots were probed with an HRP-conjugated goat anti-rabbit secondary antibody (1:5,000 dilution; cat. no. A0208; Beyotime Biotechnology) for 1 h at room temperature. For the detection of phosphorylated and total proteins (such as p-p65/p65 and p-IκBα/IκBα) on the same membrane, the PVDF membranes were first probed for the phosphorylated proteins, then stripped using a commercial stripping buffer (cat. no. FD0050; Fdbio Science), blocked again and reprobed for the respective total proteins. Immunoreactive bands were optically captured using a VILBER Fusion FX7 imaging system (Vilber Lourmat) supported by a specialized chemiluminescent substrate (cat. no. FD8030; Fdbio Science).

All quantitative datasets were processed and visualized utilizing GraphPad Prism (version 9.0; Dotmatics). For statistical comparisons involving two independent cohorts, a two-tailed Student's t-test was executed. When evaluating datasets containing >2 experimental groups, either one-way or two-way ANOVAs were applied, followed by Tukey's post-hoc test for multiple pairwise comparisons. P<0.05 was considered to indicate a statistically significant difference. Data are presented as the mean ± SEM.

To investigate the metabolic landscape associated with psoriasis progression, untargeted metabolomics profiling was performed on skin samples from IMQ-induced psoriasis-like mouse models and NCs (Fig. 1A). PCA revealed distinct metabolic clustering, indicating a shift in the metabolic profile following IMQ treatment (Fig. 1B). Specifically, out of 14,887 detected features, 3,506 significantly upregulated and 1,160 significantly downregulated metabolites were identified in the psoriatic skin (Fig. 1C). To interpret the biological importance of these shifts, KEGG pathway enrichment analysis was performed. The results demonstrated that the altered metabolites were predominantly enriched in amino acid metabolism pathways, including 'arginine biosynthesis', 'histidine metabolism' and notably, 'phenylalanine metabolism' (Fig. 1D). Given the high enrichment ratio and statistical significance associated with phenylalanine metabolism, this pathway was selected for further validation.

These untargeted findings were next corroborated using targeted quantitative metabolomics to measure absolute concentrations of phenylalanine and its downstream metabolite, tyrosine. Consistent with the present initial metabolomics screening, phenylalanine and tyrosine levels were found to be significantly elevated in the skin lesions of IMQ-treated mice compared with NC group (Fig. 1E). Furthermore, plasma analysis also revealed a significant systemic increase in circulating phenylalanine and tyrosine levels (Fig. 1F). These findings collectively suggested that the progression of psoriasis was closely associated with aberrant phenylalanine metabolism (Fig. S1), leading to the accumulation of both phenylalanine and tyrosine systemically and locally in skin lesions.

Having identified phenylalanine accumulation as a metabolic hallmark of psoriasis, the concentration of phenylalanine was adjusted in the diet of mice during psoriasis modeling to determine whether altering systemic phenylalanine levels functionally modulates disease outcome (19). C57BL/6 mice were acclimated to diets containing varying concentrations of phenylalanine (0.25, 1 or 2%) for 1 week prior to and during the course of IMQ treatment (Fig. 2A). Notably, these dietary modifications did not significantly affect general health or body weight before IMQ treatment (Fig. 2B). However, after IMQ application, phenotypic assessment revealed a significant dose-dependent effect of phenylalanine on skin inflammation. Mice fed a high-phenylalanine diet (2%) exhibited exacerbated psoriasis-like lesions characterized by severe erythema, heavy scaling and skin thickening. Conversely, dietary restriction of phenylalanine (0.25%) significantly alleviated these clinical manifestations compared with the standard diet (1%) group (Fig. 2C). This observation was quantified by PASI scores, which exhibited a clear dose-dependent trajectory (Figs. 2D and S2). Histological analysis of the dorsal skin further demonstrated these findings: High dietary phenylalanine induced pronounced epidermal hyperplasia (acanthosis) and inflammatory cell infiltration, whereas phenylalanine restriction effectively attenuated these pathological changes (Fig. 2E and F). To characterize the inflammatory milieu driven by phenylalanine, the gene expression of key psoriasis-associated cytokines in the skin were analyzed. It was found that elevated dietary phenylalanine significantly promoted the gene expression levels of Il1b, Il6 and C-X-C motif chemokine ligand (Cxcl-)-1, as well as the key psoriasis drivers Il12b, Il23a and Il17a (Fig. 2G). Consistent results were observed in the ear skin, whereby high phenylalanine aggravated ear swelling and histological severity, accompanied by a significant increase in ear thickness (Fig. 2H-J). To further establish the pathological severity and terminal phenotype of the disease model, the infiltration of neutrophils was evaluated, which serve as a well-established marker of psoriasis severity (31). Flow cytometry analysis of single-cell suspensions from the ear skin revealed that dietary phenylalanine significantly dose-dependently increased the infiltration of neutrophils (Fig. 2K and L). In summary, these results indicated that phenylalanine acted as an aggravating factor in psoriasis, likely by amplifying pro-inflammatory cytokine production and promoting immune cell recruitment to skin lesions.

To validate the therapeutic potential of restricting cellular phenylalanine availability, LAT1, the primary transporter for large neutral amino acids, was targeted. JPH203, a selective high-affinity LAT1 inhibitor, was utilized in mice subjected to the high-phenylalanine diet and IMQ challenge (Fig. 3A). Notably, administration of JPH203 alone to healthy mice did not induce any cutaneous abnormalities or toxicity. Aligning with the present hypothesis, JPH203 exerted a potent protective effect. Treatment with this inhibitor markedly abrogated the exacerbating impact of dietary phenylalanine. Clinically, this was evidenced by a marked reduction in erythema and scaling compared with the vehicle-treated IMQ group (Fig. 3B). Consistent with these macroscopic observations, histological assessment of the dorsal skin revealed that JPH203 effectively attenuated pathological epidermal thickening (Fig. 3C). Quantitative analyses further demonstrated these protective effects, demonstrating significant reductions in PASI scores (Figs. 3D and S3) and epidermal acanthosis thickness (Fig. 3E). Similarly, in the ear skin model, JPH203 treatment alleviated tissue swelling and restored epidermal architecture, characterized by reduced hyperkeratosis and fewer pustules (Fig. 3F), as well as a significant reduction in ear thickness (Fig. 3G). At the cellular level, flow cytometry analysis revealed that JPH203 treatment effectively suppressed the immune response, leading to a significant reduction in the frequency of infiltrating neutrophils in skin lesions (Fig. 3H and I). Furthermore, systemic inflammation was assessed by examining the spleen. While IMQ treatment induced significant splenomegaly, JPH203 co-administration normalized both spleen size and the spleen index (Fig. 3J and K). Collectively, these findings demonstrated that blocking phenylalanine transport through JPH203 effectively disrupted the phenylalanine-dependent inflammatory loop, providing therapeutic benefits both locally and systemically.

DCs serve as key orchestrators in the initiation of psoriasis, primarily by driving the IL-23/IL-17 axis (32). The present in vivo findings demonstrated that systemic phenylalanine markedly aggravated IMQ-induced psoriasis-like dermatitis. To elucidate the molecular basis of this effect, RNA-seq was performed on BMDCs exposed to phenylalanine in the presence or absence of IMQ stimulation. Transcriptomics analysis revealed that phenylalanine exposure alone was sufficient to notably alter the DC landscape (Fig. 4A). Compared with vehicle controls, phenylalanine treatment significantly upregulated 116 genes (Fig. 4B). KEGG enrichment analysis showed phenylalanine significantly regulated genes associated with the 'cytokine-cytokine receptor interaction', 'NOD-like receptor signaling pathway' and 'NF-κB signaling pathway' (Fig. 4C). Notably, in the context of inflammatory challenge (IMQ stimulation), phenylalanine acted as a potent amplifier. Co-treatment with phenylalanine and IMQ induced a distinct transcriptional profile compared with IMQ alone. Differential expression analysis highlighted further upregulation of 245 genes, including key pro-inflammatory mediators such as Il1a, Il1b, Il6 and Cxcl1-3. Functional enrichment analysis determined that phenylalanine potentiated pathways key in psoriasis pathogenesis, including the 'IL-17 signaling pathway' and 'cytokine-cytokine receptor interaction' (Fig. 4D).

To validate these transcriptomic signatures, qPCR was performed on independent samples. Consistent with the RNA-seq data, phenylalanine treatment alone significantly increased the basal expression of Il6 and Il12b, while only slightly elevating Il1b and Il23a levels without statistical significance. Notably, it significantly and synergistically enhanced the IMQ-induced expression of these cytokines, as well as the neutrophil-attracting chemokines Cxcl1, Cxcl2 and Cxcl3 (Fig. 4E). To determine whether this effect was specific to phenylalanine or shared by its downstream metabolites, the impact of tyrosine supplementation was evaluated. While tyrosine also exacerbated the IMQ-induced expression of Il1b, Il6 and Il12b, its effect was notably less potent compared with that of phenylalanine in combination with IMQ (Fig. 4F). This suggested that while tyrosine contributes to the phenotype, phenylalanine itself acted as the primary driver of the hyper-inflammatory state in DCs. Collectively, these data suggest that phenylalanine reprogramed the DC transcriptome, lowering the threshold for activation and amplifying the magnitude of the inflammatory response.

To mechanistically validate the present transcriptomic signatures, the NF-κB signaling axis, a central regulator of inflammatory cytokine production (33), was investigated. The activation status of DCs was first examined in vivo. Immunofluorescence staining of skin lesions from mice fed different diets revealed that while the total density of infiltrating DCs (CD11c⁺) remained comparable across dietary groups (Fig. 5A and B), the activation state was markedly altered. Specifically, the fluorescence intensity of p-p65⁺ within DCs exhibited an upward trend, though a statistically significant increase was only observed between the 1 and 2% phenylalanine groups (Fig. 5A and C). The signaling kinetics were next dissected in vitro using BMDCs. Western blotting analysis demonstrated that phenylalanine treatment independently elevated the basal phosphorylation levels of IκBα and p65. Following the addition of IMQ, the expression of phosphorylated IκBα and p65 were further upregulated (Fig. 5D). Quantification showed that the phosphorylation levels of IκBα and p65 were significantly amplified at numerous time points compared with the controls (Fig. 5E and F). To prove causality, cells were treated with BAY11-7082, a specific NF-κB inhibitor. While phenylalanine strongly potentiated the IMQ-induced expression of Il1b, Il6, Il12b, Il23a and Cxcl1, pharmacological blockade of NF-κB abrogated this amplification effect (Fig. 5G), determining that the pro-inflammatory action of phenylalanine was dependent on the NF-κB axis. Finally, to determine an association between intracellular signaling to amino acid transport, the LAT1 inhibitor JPH203 was utilized. Western blotting revealed that blocking phenylalanine uptake effectively suppressed hyperactivation of the pathway, evidenced by reduced phosphorylation of IκBα and p65 in JPH203-treated cells (Fig. 5H-J). Consistent with this, JPH203 treatment significantly attenuated gene expression of downstream cytokines and chemokines in phenylalanine-exposed DCs (Fig. 5K). In summary, these data indicated that LAT1-mediated phenylalanine uptake fueled the hyperactivation of NF-κB signaling, thereby drove the excessive production of psoriasis-associated cytokines in DCs.

DCs serve an important role in immature T cell differentiation and the present data showed that phenylalanine amplified the production of Th17-polarizing cytokines (IL-6, IL-23 and IL-1β) by DCs. The present study hypothesized that this metabolic cue would ultimately exacerbate psoriasis by fueling a Th17-dominant response. To test this in vivo, the systemic immune profile of mice fed diets with varying phenylalanine concentrations were analyzed. Flow cytometry analysis of drained lymph nodes revealed a notable dose-dependent expansion of Th17 cells. Mice on the 2% phenylalanine diet exhibited the highest frequency and absolute numbers of Th17 cells among all diet groups (Fig. 6A-C). To determine that this Th17 expansion was mechanistically driven by DC-derived signals rather than a direct effect on T cells, an in vitro differentiation assay was performed using DC-conditioned medium. Naïve CD4⁺ T cells were cultured in supernatants collected from BMDCs that had been subjected to different treatments. Results indicated that the conditioned medium from phenylalanine-primed, IMQ-stimulated DCs possessed a potent capacity to drive Th17 polarization. T cells cultured in this medium exhibited significantly upregulated gene expression of the master transcription factor RAR-related orphan receptor C (Rorc) and the effector cytokines Il17a, Il17f and Il22 compared with controls (Fig. 6D). This was further corroborated by flow cytometry, which exhibited a robust increase in the population of CD4⁺ IL-17A⁺ T cells (Fig. 6E-G).

Finally, to demonstrate that blocking phenylalanine transport in DCs could reverse this pathogenic axis, LAT1 inhibitor JPH203 was utilized. Notably, T cells cultured in supernatants from DCs co-treated with JPH203 failed to promote a strong Th17 response, as evidenced by the normalized expression of Rorc, Il17a, Il17f and Il22 (Fig. 6H). In conclusion, these findings demonstrated that phenylalanine acted as a metabolic checkpoint that instructed DCs to create a pro-Th17 microenvironment, thereby sustaining the chronic inflammation characteristic of psoriasis.