Untargeted Metabolomics Investigating the Intracellular Metabolic Alterations of HaCaT Cells Treated with Isotretinoin

Introduction

Isotretinoin (13-cis-retinoic acid) is a first-generation non-aromatic retinoid derived from β-carotene (provitamin A). This compound was first synthesized in 1955; however, it was not until 1973 that it was utilized for the treatment of psoriasis in Europe. Subsequently, in 1982, isotretinoin received approval for the treatment of acne. Beyond its well-established role in acne management, isotretinoin has increasingly been used off-label for various dermatological diseases, including rosacea,¹ seborrheic dermatitis,² periorificial dermatitis,³ granuloma annulare,⁴ and sarcoidosis.⁵ Despite its clinical efficacy, isotretinoin is associated with a range of adverse reactions including skin symptoms (dry skin, erythema, and cheilosis), teratogenic effects,⁶ mood alterations,⁷ hepatotoxicity,⁸ and dyslipidemia.⁹

The impact of isotretinoin on the skin is responsible for various mucocutaneous adverse effects. However, the mechanisms by which isotretinoin acts on the skin, particularly the epidermis, remain poorly understood. A previous study found that the dryness of the skin observed in patients with acne treated with isotretinoin is related to an increase in transepidermal water loss (TEWL) and a decrease in the hydration of the epidermal stratum corneum.¹⁰ Furthermore, a subsequent study indicated that the skin dryness caused by isotretinoin might be associated with its effects on aquaporin 3 (AQP3) expression, a water and glycerol channel protein.¹¹ Moreover, Tadini et al revealed that the topical application of isotretinoin on hairless mice epidermis leads to an enhancement in epidermis thickness and the number of epidermal cell layers.¹² However, in vitro studies show that isotretinoin inhibits the proliferation of HaCaT keratinocyte cells and decreases their survival.^13,14 Additionally, isotretinoin enhances the expression of Forkhead box protein O1 (FoxO1) in human primary keratinocytes (HPKs), promoting differentiation and apoptosis in these cells.¹⁵

Metabolomics is the qualitative and quantitative analysis of all endogenous low-molecular-weight metabolites (< 1.5 kDa) in the biological organism at a given time.Presently, several research techniques are commonly employed in metabolomics, such as ultra-high performance liquid chromatography coupled with time-of-flight mass spectrometry (UHPLC-Q-TOF/MS), liquid chromatography-mass spectrometry (LC-MS), gas chromatography-mass spectrometry (GC-MS), and nuclear magnetic resonance (NMR), and UHPLC-Q-TOF/MS has become a powerful tool for metabolomics research.¹⁶ In addition, non-targeted analysis has a wide application in drug research to reveal the potential efficacy and mechanisms of action of various drugs.^17–19

To date, no study has investigated the intracellular metabolic alterations in keratinocyte cells treated with isotretinoin. Although HaCaT cells are spontaneously immortalized through p53 mutation, this cell line closely resembles normal keratinocytes, making it a valuable experimental model. In this study, we employed a non-targeted metabolomics strategy utilizing UHPLC-Q-TOF/MS to investigate the intracellular metabolic alterations in HaCaT cells treated with isotretinoin, thereby providing a comprehensive perspective on the effects of isotretinoin on keratinocytes at the metabolomic level.

Materials and Methods

Cell Culture

A spontaneously immortalized skin keratinocyte cell line HaCaT (10 to 30 passage numbers), was obtained from iCell Bioscience Inc (Shanghai, China) and cultured in high glucose DMEM supplemented with 10% fetal bovine serum (FBS) and penicillin–streptomycin (Boster, China). The cells were maintained under standard cell culture conditions at 37°C, 5% CO₂ in a humid environment.

CCK8 Assay

Cell proliferation was evaluated using cell counting kit-8 (CCK8) (Solarbio, China) in accordance with the manufacturer’s instructions. Briefly, HaCaT cells were seeded at a density of 1×10⁴ per well (100 μL) in 96-well plates. After 24h incubation, the cells were treated with isotretinoin (MedChem Express,USA) at various concentrations (0, 1, 5, 10μM) for 48 h. Subsequently, 10 μL of CCK-8 was added to each well, and cells were incubated for 2 h at 37 °C. The absorbance value at 450 nm was read by the enzyme-labeled instrument. The experiments were performed in triplicate.

Sample Preparation for untargeted Metabolic Analysis

After 48 hours of culture, the cells reached an 80% adherent density in the T-75 culture flask. The cells were divided into three groups: Group A (control group), Group B(treated with 10 μM isotretinoin for 48 h), and Group C (treated with 10 μM isotretinoin for 7 d). In Group C, the medium was changed every two days with 10 μM isotretinoin.Consistency in cell culture conditions was meticulously monitored to ensure that the cells did not exceed 90% confluence.The culture medium harvested from the HaCaT cells was aspirated using a pipette. Subsequently, the cells were rinsed with PBS at 37°C.

Eight biological replicates were used for each group for metabolomics analysis.The samples were thawed at 4°C and mixed with 1 mL of cold methanol/acetonitrile/H₂O (2:2:1, v/v/v). The homogenate was sonicated at low temperature for 30 min, repeated twice. The mixture was then centrifuged for 20 min at 14,000 g and 4°C. The supernatant was dried using a vacuum centrifuge. For LC-MS analysis, the samples were re-dissolved in 100 μL acetonitrile/water (1:1, v/v) solvent.To ensure the reliability and stability of the instrument analysis, quality control (QC) samples were created by pooling 10 μL from each sample, which were then analyzed alongside the other samples. The QC samples were inserted regularly and analyzed every five samples.

UHPLC-Q-TOF MS

Analysis was performed using an UHPLC (1290 Infinity LC, Agilent Technologies) coupled to a quadrupole time-of-flight (AB SciexTriple TOF 6600) in Shanghai Applied Protein Technology Co., Ltd.

For hydrophilic interaction liquid chromatography (HILIC) separation, samples were analyzed using a 2.1 mm × 100 mm ACQUIY UPLC BEH Amide 1.7 µm column (waters, Ireland). In both ESI positive and negative modes, the mobile phase contained A=25 mM ammonium acetate and 25 mM ammonium hydroxide in water and B= acetonitrile. The gradient was 95% B for 0.5 min and was linearly reduced to 65% in 6.5 min, and then was reduced to 40% in 1 min and kept for 1 min, and then increased to 95% in 0.1 min, with a 3 min re-equilibration period employed.

The ESI source conditions were set as follows: Ion Source Gas1 (Gas1) as 60, Ion Source Gas2 (Gas2) as 60, curtain gas (CUR) as 30, source temperature: 600°C, IonSpray Voltage Floating (ISVF) ± 5500 V. In MS only acquisition, the instrument was set to acquire over the m/z range 60–1000 Da, and the accumulation time for TOF MS scan was set at 0.20 s/spectra. In auto MS/MS acquisition, the instrument was set to acquire over the m/z range 25–1000 Da, and the accumulation time for product ion scan was set at 0.05 s/spectra. The product ion scan is acquired using information dependent acquisition (IDA) with high sensitivity mode selected. The parameters were set as follows: the collision energy (CE) was fixed at 35 V with ± 15 eV; declustering potential (DP), 60 V (+) and −60 V (−); exclude isotopes within 4 Da, candidate ions to monitor per cycle: 10.

Data Processing

The raw MS data were converted to MzXML files using ProteoWizard MSConvert before importing into freely available XCMS software. For peak picking, the following parameters were used: centWave m/z = 10 ppm, peakwidth = c (10, 60), prefilter = c (10, 100). For peak grouping, bw = 5, mzwid = 0.025, minfrac = 0.5 were used. CAMERA (Collection of Algorithms of MEtabolite pRofile Annotation) was sued for annotation of isotopes and adducts. In the extracted ion features, only the variables having more than 50% of the nonzero measurement values in at least one group were kept. Compound identification of metabolites was performed by comparing of accuracy m/z value (<10 ppm), and MS/MS spectra with an in-house database established with available authentic standards.

Metabolites were compared with free online databases KEGG (http://www.genome.jp/kegg/) and HMDB (http://www.hmdb.ca/), and the corresponding KEGG pathways were extracted. KEGG enrichment analysis was performed using MetaboAnalyst (www.metaboanalyst.ca).

Statistical Analysis

For non-targeted metabolomics,after sum-normalization, the processed data were analyzed by R package (ropls), where it was subjected to multivariate data analysis, including Pareto-scaled principal component analysis (PCA) and orthogonal partial least-squares discriminant analysis (OPLS-DA). The 7-fold cross-validation and response permutation testing were used to evaluate the robustness of the model. The variable importance in the projection (VIP) value of each variable in the OPLS-DA model was calculated to indicate its contribution to the classification. Student’s t test was applied to determine the significance of differences between two groups of independent samples. VIP > 1 and P value < 0.05 were used to screen significant changed metabolites.One-way analysis of variance (ANOVA) method (GraphPad Prism 10.1.2) was applied for statistical analysis of the CCK-8 assay. P < 0.05 indicates a statistically significant difference.

Results

Effects of Isotretinoin on Proliferation of HaCaT Cells

The effect of isotretinoin on proliferation of HaCaT cells was determined by CCK8 assay. The results indicated that isotretinoin inhibited the proliferation of HaCaT cells by 21.3 ± 6.32% at a concentration of 1 μM (P = 0.652), 29.57 ± 5.15% at 5 μM (P = 0.002), and 40.14 ± 2.88% at 10 μM (P < 0.001).These results indicate that isotretinoin may block the proliferation of HaCaT cells in a dose-dependent manner (Figure 1).

Figure 1 The effect of isotretinoin on proliferation of HaCaT cells. Cells were exposed to various concentrations of isotretinoin (0–10 μmol/L) and CCK8 assay was used to analyze cell proliferation.Data are expressed as means ± SD from three independent experiments, **P < 0.01, ***P < 0.001 compared to control by one-way ANOVA.

Multivariate Analysis

Unsupervised principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were utilized to discriminate the overall distribution among samples. As shown in the PCA score plot, PC1 and PC2 accounted for 26.3 and 16.2% of the total variability, respectively. A clear cluster of the pooled QC samples was exhibited, indicating that the sample analysis sequence had satisfactory stability and repeatability (Figure 2). Furthermore, PCA revealed distinct differences among these groups, suggesting significant metabolic differences between them. Similarly, the OPLS-DA model also displayed clear segmentation between Group B and Group A (Figure 3A), as well as between Group C and Group A (Figure 3B). The model was evaluated by monitoring the model R2 (goodness of fit) and Q2 (predictive ability) values and permutation tests (n = 200) were performed on R2 and Q2. The OPLS-DA parameters of R2Y and Q2 for each comparison were as follows:Group B and Group A (R2Y = 0.972, Q2 = 0.88), Group C and Group A (R2Y = 0.994, Q2 =0.949), suggesting a good fit and predictability of these models.Moreover,the permutations plots showed that the permuted R2 (in green) and Q2 (in blue) values were lower than the original ones, and Q2 regression line intersected the vertical axis below zero, indicating all the OPLS-DA models were valid (Figure 3C,3D).

Figure 2 The score plots of principal component analysis.

Figure 3 OPLS scores and 200 permutation tests for OPLS-DA models. The score plots of OPLS-DA: (A) Group A and Group B; (B): Group A and Group C. The permutation tests for OPLS-DA: (C) Group A and Group B; (D): Group A and Group C.

Identification of Differentially Expressed Metabolites

VIP reflected the significance of those variables in the OPLS-DA model and was used to select the significant variables.To further identify the differentially metabolites in the samples, we screened the metabolites using variable importance projection (VIP)>1 and p-value<0.05 in the OPLS-DA model. Compared to Group A, we identified 115 significantly differentially metabolites in Group B,including glycerophosphocholine, adenosine, acetylcarnitine, creatine,Dl-lactate, hypoxanthine,uracil,LPC 18:1,creatine phosphate,5′-S-Methyl-5′-thioadenosine(43 upregulated, 72 downregulated)(Table 1). The chemical classification of these metabolites primarily consists of 35.96% lipids and lipid-like molecules, 21.93% organic acids and derivatives, 13.16% nucleosides nucleotides and analogues, 11.40% Organoheterocyclic compounds, and other metabolites (Figure 4A). In total, 140 significantly differential metabolites (81 upregulated and 59 downregulated) were identified in Group C compared to Group A. These metabolites include glycerophosphocholine, niacinamide, creatine, phytosphingosine, oleic acid, hypoxanthine, 1-Palmitoyl-sn-glycero-3-phosphocholine, Dl-lactate, 5’-S-Methyl-5’-thioadenosine and NG,NG-Dimethyl-L-arginine (Table 2). The highest proportion of differential metabolites was observed in lipids and lipid-like molecules (32.86%), followed by organic acids and derivatives (20.00%), organoheterocyclic compounds (13.57%), and nucleosides, nucleotides, and analogues (10.00%) (Figure 4B). Furthermore, to visualize the distribution of relative levels of the differentially metabolites in each group, the Euclideandistance matrix and complete linkage method were applied for the hierarchical clustering.The heatmap illustrated the top 50 differential metabolites and indicated that the metabolites in Group A were significantly separated from those in both Group B and Group C (Figure 5).

Table 1 The Top10 Significant Differential Metabolites Among Group A and B

Table 2 The Top10 Significant Differential Metabolites Among Group A and C

Figure 4 Proportion of identified metabolites in each chemical classification. (A) Differential metabolites between Group A and Group B. (B) Differential metabolites between Group A and Group C.

Figure 5 Hierarchical clustering heat map. (A) The top 50 differential metabolites between Group A and Group B. (B) The top 50 differential metabolites between Group A and Group C. Color scale on the right of the heat map represented normalized values of each identified metabolites content. Red represents significant upregulation; blue represents significant downregulation.

Pathway Enrichment Analysis

Fisher’s exact test was employed to analyze and calculate the significance level of metabolite enrichment for each pathway.The p-value was used to identify the metabolic pathways that are significantly affected:the smaller the p-value, the more significant the difference in the metabolic pathway. Compared with the Group A,30 metabolic pathways were significantly changed (p < 0.05) in Group B. The most significant metabolic pathways included purine metabolism, nucleotide metabolism, ABC transporters, taurine and hypotaurine metabolism, and cGMP-PKG signaling pathway (Figure 6A). In total, 46 metabolic pathways were significantly changed (p < 0.05) in Group C compared with Group A. The enriched pathways primarily involved nucleotide metabolism, ABC transporters, purine metabolism, glycerophospholipid metabolism, and choline metabolism in cancer (Figure 6B).

Figure 6 Statistics of KEGG enrichment. (A)The top 10 enriched pathways between Group A and Group B. (B) The top 10 enriched pathways between Group A and Group C. The x axis indicates the rich factor corresponding to each pathway, and the y axis indicates name of the KEGG metabolic pathway. The size and color of bubbles represent the number and degree of enrichment of different metabolites, respectively.

Discussion

In this study, we investigated the therapeutic mechanism of isotretinoin on HaCaT cells at two different time points: short-term (48 h) or long-term (7 d), utilizing a metabolomics approach. Our findings revealed that, in comparison to Group A, there were 115 significantly different metabolites identified in Group B and 140 in Group C. Furthermore, when compared to Group A, 30 metabolic pathways were significantly altered in Group B, while 46 pathways were perturbed in Group C.

Nucleotide Metabolism and Purine Metabolism

Our data demonstrated that nucleotide metabolism and purine metabolism were significantly perturbed pathways at all time points (48 h and 7 d), implicating their important roles in the mechanism of action of isotretinoin on keratinocytes in both short-term and long-term contexts. Nucleotides are vital metabolites in organisms and participate in almost all biochemical processes within cells.²⁰ Nucleotide metabolism fuels DNA replication and RNA synthesis, which are indispensable for cell proliferation. In our study, isotretinoin inhibited the proliferation of HaCaT cells in a dose-dependent manner. Previous evidence has demonstrated that isotretinoin regulates keratinocyte proliferation by binding to the nuclear receptors all-trans retinoic acid receptor (RAR) and all-trans retinoid X receptor (RXR).²¹ Our results suggest that the inhibition of keratinocyte proliferation by isotretinoin may be associated with its influence on the nucleotide metabolism pathway through the activation of RAR/RXR.Furthermore, seven metabolites related to the purine metabolic pathway changed significantly, including adenosine monophosphate, hypoxanthine, uric acid, inosine, 2′,3′-cyclic adenosine monophosphate, 3′-adenosine monophosphate, and 2′-deoxyinosine. Adenosine monophosphate is not only a synthesis product of purine nucleotides, but also a product of their oxidative decomposition.²² Both hypoxanthine and inosine serve as key intermediates in the degradation of purines, while uric acid is the final product of purine catabolism in humans.²³ In our study, the treatment with isotretinoin disrupted the metabolism of these substances. Hence,the effect of isotretinoin on keratinocytes is associated with the regulation of purine synthesis and degradation.

ABC Transporters

ATP-binding cassette (ABC) transporters are a class of transmembrane proteins that are extensively expressed in various tissues, including keratinocytes, hepatocytes, intestinal epithelial cells, and adipocytes.²⁴ They play multiple key functions in keratinocytes, including lipid transport, barrier formation, drug metabolism, defense responses, and responses to environmental stress such as ultraviolet radiation.^25,26 In our study, ABC transporters was another significantly perturbed pathway in HaCaT cells treated with isotretinoin for 48 h and 7 d. Previous research has indicated that ABC transporters was the most altered in patients with moderate-to-severe acne.²⁷ Similarly, in a rat acne model, ABC transporters were identified as a significantly perturbed pathway.²⁸ Therefore, the regulation of the ABC pathway in keratinocytes by isotretinoin may contribute to its therapeutic efficacy in treating acne.

Glycerophospholipid Metabolism

Glycerophospholipids, commonly referred to as phospholipids, are essential components of the lipid bilayer of cellular membranes and play a pivotal role in cell signaling and metabolism.²⁹ Our study revealed that after isotretinoin treatment, the glycerophospholipid metabolism pathways in HaCaT cells were significantly altered at 7 d, rather than at 48 h.This indicates the differences in the mechanism of action of isotretinoin on keratinocytes in short-term and long-term contexts. Furthermore, seven metabolites associated with the glycerophospholipid metabolic pathway exhibited significant changes, including glycerol 3-phosphate, O-phosphoethanolamine, cytidine 5′-diphosphate ethanolamine, phosphorylcholine, glycerophosphocholine, sn-glycerol 3-phosphoethanolamine, and 1-palmitoyl-sn-glycero-3-phosphocholine. Notably, phosphorylcholine is the most abundant phospholipid in biological membranes, contributing to the characteristic bilayer structure of cells and regulating membrane integrity.³⁰ Glycerol 3-phosphate serves as a key precursor for the synthesis of membrane lipids,³¹ while phosphoethanolamine is involved in phospholipid turnover, acting as a substrate for various phospholipids in cell membranes. Glycerophosphocholine is a product of the breakdown of phospholipid components in cell membranes.³² In our study, a 7-day isotretinoin treatment led to a decrease in the levels of these substances. This finding suggests that long-term isotretinoin treatment may influence the synthesis and function of cell membranes by inhibiting the production of these metabolites.

The limitations of this study include the following: the use of immortalized cell lines instead of primary keratinocyte cell lines; the lack of targeted metabolomics validation; the absence of functional assays; and the limited translational relevance.

Conclusion

This study employed UHPLC-qTOF-MS to investigate the intracellular metabolic alterations induced by isotretinoin in HaCaT keratinocytes. Through the analysis of metabolic profiles, we identified 115 metabolites that were significantly differentially expressed after 48 h of isotretinoin treatment and 140 metabolites after 7 d of treatment.Isotretinoin modulates multiple metabolic pathways, including nucleotide metabolism,purine metabolism, ABC transporters,and glycerophospholipid metabolism.Our findings will help clarify the underlying mechanism of action of isotretinoin on keratinocytes at the metabolite level and the mucocutaneous adverse effects associated with isotretinoin in the treatment of dermatological diseases.However, this study has certain limitations. Further validation studies, involving targeted metabolomics and functional assays, are required to confirm and validate these initial discovery results.