Skip to main content
NCBI home page
Dermatology and Therapy logoLink to Dermatology and Therapy
. 2026 Apr 17;16(5):2541–2555. doi: 10.1007/s13555-026-01748-4

Early Proteomic Alterations in Non-segmental Vitiligo Skin Following Treatment with Topical Tacrolimus

Morten B Haulrig 1,2,3, Sonja Kabatnik 3,4, Lone Skov 1,2, Marianne B Løvendorf 1,3,5,✉,#, Beatrice Dyring-Andersen 2,3,4,5,6,#
PMCID: PMC13219589  PMID: 41996020

Abstract

Introduction

Vitiligo is a chronic depigmenting skin disease that can severely affect quality of life. First-line treatments include topical calcineurin inhibitors such as tacrolimus, but their molecular mechanisms of action in vitiligo remain incompletely understood. In this study, we investigated the early proteomic changes in patients with non-segmental vitiligo treated with topical tacrolimus.

Methods

We included eight patients and collected 2-mm punch biopsies from untreated lesional skin (LS) and nonlesional skin (NLS), as well as from LS skin after 2 and 8 weeks of treatment. Samples were analyzed with mass spectrometry-based proteomics. We used a pairwise comparison between untreated LS and NLS, and linear mixed models to compare untreated LS and tacrolimus-treated LS.

Results

In total, 4416 proteins were identified. Comparison between untreated LS and NLS revealed 32 differentially expressed proteins (DEPs; false discovery rate < 0.05), primarily associated with muscle contraction and melanocyte function. After 2 weeks of treatment, LS showed enrichment of 31 DEPs related to acute inflammatory responses and melanocyte adhesion, while LS treated for 8 weeks showed enrichment of 14 DEPs involved in melanogenesis, interferon-induced inflammation, and heat shock responses.

Conclusion

This study characterizes the early proteomic response of the skin to tacrolimus treatment and offers insights into disease-associated proteins in vitiligo.

Keywords: Depigmentation, Immunology, Mass spectrometry, Melanocytes, Proteomics, Topical calcineurin inhibitor

Key Summary Points

Why carry out this study?
Topical calcineurin inhibitors, such as tacrolimus, are first-line treatments for localized vitiligo; however, their molecular mechanisms of action in vitiligo remain incompletely understood.
In this study, we used mass spectrometry-based proteomics to analyse vitiligo skin before and during early treatment with topical tacrolimus.
What was learned from the study?
Untreated lesional vitiligo skin expressed decreased levels of proteins related to melanocyte function, while lesional skin treated with tacrolimus showed alteration of proteins involved in inflammation and melanocyte function.
The study reveals inflammatory and melanocyte-associated changes that provide insights into the effects and limitations of tacrolimus.
Open in a new tab

Introduction

Vitiligo is a chronic depigmenting disorder caused by an immune-mediated destruction of melanocytes. It affects approximately 0.4% of people worldwide and can severely reduce quality of life and self-esteem [1, 2]. The pathogenesis is multifactorial and involves a skewed type 1 immune response and melanocyte vulnerability to oxidative stress. However, several other aspects of the underlying pathophysiology remain unclear. The combined effects of genetic predisposition and environmental factors trigger innate and adaptive immune responses, ultimately leading to melanocyte apoptosis [3]. The recommended treatments for vitiligo include topical calcineurin inhibitors (TCIs) and phototherapy with narrow-band UVB [4]. In addition, topical ruxolitinib, a topical Janus kinase (JAK) inhibitor, has demonstrated clinical efficacy and is the first approved treatment for vitiligo [4]. However, vitiligo treatment is challenging and responses to current therapies are inconsistent, especially in nonfacial lesions [5].

TCIs such as tacrolimus are first-line treatments for vitiligo with limited involvement in both adults and children [4]. Tacrolimus is widely used because of its favorable safety profile and comparable efficacy to topical corticosteroids [6]. Additionally, a study of 30 patients with non-segmental vitiligo demonstrated comparable efficacy to the topical JAK inhibitor tofacitinib after 16 weeks of treatment [7].

Tacrolimus specifically targets the calcineurin–calmodulin complex, which blocks the nuclear factor of activated T-cells (NFAT) transcription factor, thereby reducing cytokine release and T-cell activation [8]. In lesional vitiligo skin, tacrolimus has been reported to increase regulatory cytokines while suppressing TNF-α [9, 10]. Other studies have reported increased melanocyte proliferation, migration, and melanin synthesis in healthy melanocytes in response to treatment [11].

While existing studies suggest possible mechanisms of action for tacrolimus, its overall effects in vitiligo remain unclear. Mass spectrometry (MS)-based proteomics can provide comprehensive quantification of proteins [12]. Although it has been used to characterize disease mechanisms, treatment-induced responses, and treatment targets in dermatological diseases, its application in vitiligo research is limited [13–15].

In this study, we used MS-based proteomics to characterize the early proteomic changes induced by tacrolimus in non-segmental vitiligo. We compared the differences between untreated lesional (LS) and nonlesional skin (NLS), and between untreated LS and LS after 2 and 8 weeks of topical tacrolimus treatment.

Methods

Patient Enrollment

Patients > 18 years with non-segmental vitiligo who were scheduled to receive treatment with tacrolimus were enrolled at the Department of Allergy and Dermatology, Herlev and Gentofte Hospital. The diagnosis of non-segmental vitiligo was based on the observation of symmetrical lesions during the clinical examination, supported by a Wood’s light examination. We excluded patients whose disease could not be clearly identified as vitiligo. Study inclusion required a minimum washout period of 4 weeks for systemic immunosuppressants, including phototherapy, and 2 weeks for topical treatments. We included patients with both progressive and stable disease. The study was approved by the Ethics Committee of the Danish Capital Region ​(H-19036270) and the Danish Data Protection Agency, and it followed the principles of the Declaration of Helsinki II. Informed, written consent was required before enrollment.

Study Design and Sample Collection

Participants were clinically assessed at three scheduled visits: at baseline and after 2 and 8 weeks of treatment. At baseline, vitiligo diagnosis was confirmed, Vitiligo Extent (VES) and Vitiligo Disease Activity (VIDA) scores were recorded, and a nonfacial vitiligo lesion measuring > 2 cm in diameter was selected as the treatment site. Preference was given to areas showing recent disease progression, as these areas would show the highest inflammatory activity. Clinical photographs were obtained at all visits. The 2 mm punch biopsies were collected from LS selected for the study treatment < 1 cm from the lesion border, and from NLS > 10 cm from the nearest lesion. All biopsies were immediately frozen in liquid nitrogen and stored at −80 °C. Whenever possible, biopsies from LS and NLS were obtained from the same anatomical region. Patients applied topical tacrolimus (0.1%) twice daily to the selected treatment area, starting the day after baseline. Treatment adherence and response were evaluated at week 2 and 8, and a repeated lesional biopsy was obtained near the initial biopsy site.

Sample Preparation

All samples were suspended in a 200 µL 50 mM Tris–HCL (pH8) digestion buffer. The samples underwent mechanical lysis while kept on ice with sterile scissors and probe sonication (Model 505, Fisher Scientific, Bremen, Germany) for 8 × 30 s each. We added 200 µL of 2,2,2-trifluorethanol (TFE) and 20 µL of 1 M DL-dithiothreitol (DTT) and heated the samples in a thermomixer. Each sample was lysed again using cup horn sonication for 30 × 30 s. To prevent disulfide bond reformation, 20 µL of 550 mM chloroacetamide (CAA) was added for 20 min at room temperature. A vacuum concentrator (Concentrator Plus, Eppendorf, VWR, Søborg, Denmark) was used to reduce sample volume (60 °C; 30 min.). We then added 200 µL of 50 mM Tris–HCL (pH 8) along with digestion enzymes (4 µL of trypsin and 4 µL of Lys-C) and left the samples for digestion overnight. To stop digestion, we added trifluoroacetic acid (TFA) 1%. Approximately 200 µL of each peptide mixture were loaded into cartridges (iST kit, PreOmics, Planegg-Martinsried, Germany) and washed twice with 200 µL of 0.2% TFA. Each sample was eluted with a 60 µL buffer (80% acetonitrile and 1% ammonium ammonia [v/v]) in a 96-well plate and dried in a vacuum concentrator (45 °C; 60 min). We resuspended the peptides with 40 µL loading buffer (2% acetonitrile in 0.2% TFA [v/v]) and measured the peptide concentration with an LVis Plate (SPECTROstar Nano, BMG Labtech, Ortenberg, Germany). The peptide mixtures (1.5 µg/µL concentration) were loaded onto two MS plates in a randomized order (#60,180-P217B, Thermo Fisher Scientific, Waltham, MA, USA).

Liquid Chromatography Tandem Mass Spectrometry

We loaded 700 ng of each sample onto Evotips and ran them on an OrbitrapTM AstralTM mass spectrometer (Thermo Fisher Scientific) coupled to an Evosep One liquid chromatography system (https://www.evosep.com/). Peptides were separated using the 60 Samples Per Day (SPD) method on an Evosep Performance column (8 cm × 150 µm, 1.5 µm particle size). All samples were acquired in data-independent acquisition (DIA) mode.

Full MS1 scans were performed in the Orbitrap analyzer at a resolution of 240,000 over an m/z range of 380–980, with the automatic gain control (AGC) target set to 500%. MS/MS scans were acquired using an isolation window of 2 Th, an ion injection time (IIT) of 3 ms, and a scan range of 150–2000 m/z. Fragmentation was carried out using higher-energy collisional dissociation (HCD) with a normalized collision energy (NCE) of 25%.

Bioinformatics Analysis

MS files were analyzed in Spectronaut (v19.0.240606), using a spectral library-free approach against the human protein sequence (20,654 entries, UP000005640, UniProt, 2024). Default settings were applied with trypsin and LysC digestion, CAA as a variable modification, and “cross-run model normalization” disabled. Protein quantification outputs (.tsv) were processed in Python and R. Proteins with > 50% valid values across all experimental groups were retained following an evaluation of data completeness and missing value distribution. Data were normalized using the VSN package (Bioconductor, R) [16], and remaining missing values were imputed within each group using K-nearest neighbor (KNN) imputation using the scikit-learn library in Python (k = 5). Differential abundance between lesional and nonlesional samples at baseline (week 0) was assessed using two-sided Student’s t-test with the Benjamini–Hochberg FDR procedure, considering adjusted p-values < 0.05 statistically significant.

Significant changes across treatment time points (weeks 2 and 8 relative to baseline) were analyzed using per-protein linear mixed-effects model (LMM). We used Python (statsmodels), with time point as a fixed effect and patient ID as a random intercept to account for correlation between repeated measures within individuals. From each fitted model, log2 fold changes and p-values were extracted and corrected for multiple hypothesis testing using the Benjamini–Hochberg FDR procedure (adjusted p-values < 0.05).

Gene set enrichment analysis (GSEA) was performed using the GSEApy package (v1.0.4). Functional network analysis was conducted using the STRING database network [17]. No original code was generated for this analysis.

Results

Patient Characteristics and Treatment Outcomes

Nine patients diagnosed with vitiligo were enrolled in the study. Each patient was enrolled upon planned treatment with tacrolimus (baseline) and completed visits after 2 and 8 weeks of treatment with tacrolimus. At the baseline visit, we assessed the extent and disease activity of vitiligo and obtained a 2 mm punch biopsy from LS and NLS. The biopsies were obtained from nonfacial areas, prioritizing sites with the most recent disease activity. All patients-initiated treatment with topical tacrolimus 0.1%, applied twice daily to the lesional skin area from which the biopsy were taken, starting the day after their baseline visit. At the visits after 2 and 8 weeks of treatment, we obtained punch biopsies and assessed the treated lesional skin area for repigmentation and disease activity.

Of the nine patients enrolled, one was excluded due to concomitant use of systemic corticosteroids to treat an unrelated skin condition. The majority of the included patients were male (n = 7) and had generalized vitiligo (n = 6; Table 1). The median age was 33 years (interquartile range [IQR]: 26–38), and the median disease duration was 4 years (IQR: 1–8). Four patients had clinical signs of activity at the baseline visit (Table 2). After 8 weeks of treatment, two patients had developed perifollicular repigmentation. There was no clinical sign of treatment effect in the rest of the patients. Adherence was high throughout the study, with the number of missed treatments ranging from 1 to 12 out of 112 planned treatments.

Table 1.

Study participant characteristics (n = 8)

Sex, Male 7 (87.5)
Age, years, median (IQR) 33 (26–38)
Fitzpatrick skin type
II 2 (25.0)
III 3 (37.5)
IV 2 (25.0)
V 1 (12.5)
Age at debut, years, median (IQR) 29 (23–34)
Disease duration, years, median (IQR) 4 (1–8)
Type of non-segmental vitiligo
Acrofacial 2 (25.0)
Generalised (vulgaris) 6 (75.0)
Vitiligo extent, median % (IQR)* 1.0 (0.6–1.9)
Previous treatments
No 6 (75.0)
Yes 2 (25.0)
Topical corticosteroids 1 (50.0)
Topical calcineurin inhibitors 2 (100)
NB-UVB phototherapy 1 (50.0)
Open in a new tab

Data are n (%) unless otherwise stated

IQR Interquartile range, NB-UVB Narrow-band Ultraviolet B

*Vitiligo extent was assessed with the Vitiligo Extent Score (VES)

Table 2.

Treatment specific outcomes

Patient ID VIDA score Clinical signs of activity* Anatomical site of LS biopsies Anatomical site of NLS biopsy Missed treatments, total† Repigmentation at week 8
1  +4 Yes Elbow Elbow 12 No
2  +4 Yes Back Back 1 No
3  +2 Yes Back Thigh 7 No
4  +3 No Elbow Upper arm 2 No
5 0 No Hip Hip 5 No
6 0 No Hip Hip 4 No
7  + 3 Yes Wrist Hip 5 Yes
8 -1 No Back Hip 9 Yes
Open in a new tab

VIDA Vitiligo Disease Activity Score

*Clinical signs of activity were determined at the designated treatment area and included trichrome vitiligo and/or a confetti-like pattern

†The total number of possible missed treatments was 112

To investigate the proteomic differences in vitiligo skin, proteins extracted from the biopsies were analyzed using MS with data-independent acquisition. We conducted a paired comparison between untreated LS and NLS at baseline, followed by a comparison between untreated LS and LS after 2 and 8 weeks of treatment. We identified a total of 4416 proteins across all sample types. Protein quantifications and identifications were comparable between the groups (Fig. 1A and B).

Fig. 1.

Fig. 1

Open in a new tab

Proteomic depth and overlap. A Number of quantified proteins. B Overlap of quantified proteins between untreated nonlesional skin (NL0) and untreated lesional skin (L0) at baseline, and lesional skin after 2 (L2) and 8 weeks (L8) of topical tacrolimus treatment

Lesional Skin Showed Proteomic Changes Related to Muscle Contraction, MAPK Signaling, and Melanocyte Function Prior to Treatment

A baseline comparison between untreated LS and NLS was performed by using a pairwise proteomic analysis to identify differences between patients prior to treatment. The analysis revealed 32 differentially expressed proteins (DEPs; false discovery rate [FDR] < 0.05; Fig. 2A).

Fig. 2.

Fig. 2

Open in a new tab

Proteomic comparison between untreated lesional and nonlesional vitiligo skin at baseline. A Pairwise proteomic comparison between untreated nonlesional (NL0) and lesional (L0) vitiligo skin at baseline (two-sided t-test, Benjamini–Hochberg multiple testing correction, FDR < 0.05). Proteins significantly enriched in untreated lesional skin are highlighted in red, and proteins enriched in untreated nonlesional skin are highlighted in blue. B Gene set enrichment analysis (GSEA) using the Gene Ontology (GO) term Biological Process (2025), performed on the comparison between untreated nonlesional and lesional skin at baseline. Pie charts indicate the percentage of matched genes from our dataset to the respective GO term. Blue bars represent significantly enriched pathways with a negative enrichment score, and red bars indicate significant enrichment of positively regulated pathways. C Functional network of the proteins that were significantly decreased in the differential proteomics analysis, generated using the STRING database. Proteins shown in dark blue represent the functional cluster annotated as “muscle proteins”; those in mid-blue represent “MAK16, TEX10,” and those in light blue represent “muscle alpha-actinin binding” proteins. The thickness of the connecting lines indicates the degree of evidential support for the associations. Two proteins could not be mapped and are not illustrated (ALDH1B4 and POTEKP). D Log2 intensity levels of proteins showing significant enrichment between untreated nonlesional and lesional skin at baseline

Functional enrichment analysis of log2 fold changes for all identified proteins demonstrated the strongest enrichment related to muscle contraction, which showed decreased expression in LS (Fig. 2B). To investigate the functional relationship, we conducted a network pattern analysis of all proteins that showed decreased expression in LS and identified three clusters (Fig. 2C). The largest cluster contained 11 proteins, including calponin-1 (CNN1), transgelin (TAGLN), actin alpha 2 (ACTA2), and myosin light chain kinase (MYLK), which are all related to smooth muscle cell contraction and have been observed in skin fibroblast subtypes (Fig. 2D) [18, 19].

Among the most strongly enriched proteins in LS were mitogen-activated protein kinase kinase 4 (MAP2K4). In response to oxidative stress, MAP2K4 activates c-Jun N-terminal kinase (JNK) and p38 MAPK, which are associated with melanocyte senescence and regulation of melanogenesis [20]. Additionally, we observed increased expression of protein phosphatase 6 catalytic subunit (PPP6C), which negatively regulates MAP3K7, an upstream kinase of MAP2K4 [21].

As expected, LS showed decreased expression of proteins associated with melanocyte functions (Fig. 2A and 2D). These proteins included glycoprotein nonmetastatic melanoma protein B (GPNMB) and zinc finger DHHC-type palmitoyltransferase 13 (ZDHHC13). GPNMB positively regulates melanosome formation, oxidative stress protection in melanocytes, and keratinocyte-melanocyte adhesion [22]. GPNMB was previously reported to be undetectable in LS [23]. ZDHHC13 contributes to skin-barrier maintenance and regulates melanogenesis by promoting melanocortin-1 receptor (MC1R) [24]. LS also showed higher intensity of tight junction protein 2 (TJP2), involved in epithelial barrier maintenance [25], and secreted LY6/PLAUR domain containing 1 (SLURP1), which promotes keratinocyte apoptosis [26].

Tacrolimus Induces Acute Inflammatory Responses, Altered Melanocyte Adhesion, and ECM Remodeling After 2 weeks of Treatment

Having characterized the differences between untreated LS and NLS, we next investigated the tacrolimus-induced response in LS after 2 and 8 weeks of treatment. A comparison of untreated LS and LS after 2 weeks of treatment revealed 31 DEPs (Fig. 3A and B).

Fig. 3.

Fig. 3

Open in a new tab

Proteomic changes in lesional vitiligo skin after 2 and 8 weeks of topical tacrolimus treatment. A Pairwise proteomic comparison between untreated lesional skin at baseline (L0) and lesional skin after 2 (L2) and 8 weeks (L8) of tacrolimus treatment. A linear mixed-effects model (LMM) was applied, and for each fitted model, log2 fold changes and p-values were extracted (L0 versus L2 and L0 versus L8) and corrected for multiple hypothesis testing (Benjamini–Hochberg, FDR < 0.05). Both time points are visualized in the volcano plot, where the direction of the arrows indicates whether proteins become more enriched or depleted over the course of treatment. Red highlights proteins enriched along the L0–L2–L8 axis, while blue indicates the opposite trend. B Clustered heatmap of significantly enriched proteins, sorted by log2 fold change and adjusted p-value. C Log2 fold change of a subset of significantly enriched proteins. The red line represents the median log2 fold change across patients, and the red shading indicates the standard deviation

LS treated with tacrolimus showed an enrichment of proteins related to keratinocyte stress responses and inflammation, with keratin 6A (KRT6A), and S100 calcium binding proteins A7/A8/A9 (S100A7/A8/A9) showing the strongest enrichment. KRT6A promotes keratinocyte inflammation, while S100A7/A8/A9 function as damage associated molecular patterns (DAMPs) in response to reactive oxygen species [27, 28]. In addition, LS showed enrichment of six transmembrane epithelial antigen of prostate family member 4 (STEAP4), nicotinamide phosphoribosyltransferase (NAMPT), and cyclin-dependent kinase 9 (CDK9). STEAP4 and NAMPT are both involved in cytokine-mediated chemokine production in keratinocytes, while CDK9 promotes keratinocyte differentiation [29, 30].

Moreover, LS treated with tacrolimus showed protein changes linked to melanocyte attachment and migration. We observed increased expression of aldo–keto reductase family 1 member B10 (AKR1B10) and SAM and SH3 domain containing 1 (SASH1). AKR1B10 is mainly expressed in keratinocytes and has negative regulatory effects on transforming growth factors (TGFs) and E-cadherin [31]. In contrast, SASH1 positively regulates melanocyte E-cadherin expression and melanocyte migration [32]. LS treated with tacrolimus also showed increased expression of coiled-coil domain containing 80 (CCDC80), involved in fibroblast-mediated collagen production [33].

LS Shows Increased Melanosome Maturation and Proinflammatory Stress Responses After 8 Weeks of Tacrolimus Treatment

Proteomic comparison of untreated LS to LS after 8 weeks of treatment revealed 14 DEPs. All DEPs were enriched in tacrolimus-treated LS, including three that were shared with LS after 2 weeks of treatment: CDK9, S100A8, and CCDC80 (Fig. 3B).

Tacrolimus-treated LS showed increased expression of proteins related to inflammation, which included ring finger protein 114 (RNF114) and mitochondrial ribosomal protein L18 (MRPL18). RNF114 increases production of chemokines in response to type I interferons [34], while MRPL18 promotes expression of the heat shock protein 70 (HSP70) [35]. Additionally, LS treated with tacrolimus showed increased expression of proteins involved in melanocyte functions. We observed increased expression of mediator complex subunit 23 (MED23), a positive regulator of melanocyte inducing transcription factor (MITF) expression [36], and dystrobrevin binding protein 1 (DTNBP1; dysbindin), a component of the BLOC-1 complex, which is associated with melanosome maturation [37].

We compared the samples from LS from the two patients who showed perifollicular repigmentation after 8 weeks of treatment with the rest of the cohort but found no statistically significant differences.

Discussion

In this study, we investigated the cutaneous molecular changes in eight patients treated with topical tacrolimus. We used MS-based proteomic analyses of skin biopsies collected from untreated LS and NLS, and from untreated LS and LS after 2 and 8 weeks of treatment. We quantified a total of 4416 proteins. A total of 32 DEPs were observed in the comparison between LS and NLS, while 42 were observed in LS during treatment.

The comparison between untreated LS and NLS showed decreased expression of CNN1, TAGLN, ACTA2, and MYLK in LS, which could suggest decreased contractility of smooth muscle cells or fibroblasts in LS. All of these proteins have previously been observed in distinct fibroblast subtypes, including myofibroblasts, which express TAGLN, MYLK, and ACTA2 [18, 19, 38]. Interestingly, we also observed decreased levels of CNN1 in perilesional vitiligo skin compared to NLS in our previous study (manuscript submitted). Fibroblasts in vitiligo are implicated in IFN-γ-mediated inflammation while possibly also promoting melanocyte detachment [39]. While fibroblasts in LS and NLS both share a senescent phenotype, previous studies have also observed functional differences [39]. Fibroblasts in LS show increased expression of ACTA2 compared with NLS, which may contrast with our findings [40]. However, fibroblasts in NLS also show increased expression of ACTA2 compared with healthy skin [41]. Taken together, our results could indicate loss of contractile properties in the dermal layer in LS.

Unexpectedly, our findings revealed an enhanced inflammatory response in keratinocytes in LS after 2 weeks of treatment, which is suggestive of a paradoxical inflammatory response to tacrolimus. S100A8/A9 and KRT6A showed the strongest enrichment in LS after 2 weeks of treatment. S100A8/9 forms a complex (calprotectin) that functions as a DAMP in response to TNF-α and IFN-γ signaling [28], while KRT6A increases after skin injury and promotes keratinocyte migration and inflammatory responses via JAK1/STAT3 signaling [27]. All of these proteins have previously been observed in lesional keratinocytes [42]. We also observed increased expression of keratinocyte-associated proteins involved in chemokine production, which have not previously been described in vitiligo. NAMPT increases the production of the chemokines CXCL8/10 in response to TNF-α [30]. Similarly, STEAP4 promotes CXCL8 production and is induced by multiple cytokines, including IL-1β, IL-36, and TNF-α [29]. These chemokines are implicated in both innate and adaptive responses, with CXCL8 primarily promoting neutrophil chemotaxis, whereas CXCL10 is a well-known promoter of cytotoxic T-cell activation in vitiligo [43]. The increased inflammatory response may be caused by a compensatory activation of other proinflammatory pathways such as TNF-α in response to the tacrolimus-induced inhibitory effects on cytokine production [8]. Our results indicate that this response is transient, as nearly all DEPs were differentially expressed in LS only after 2 weeks of treatment. Consistently, decreased levels of TNF-α have been observed in LS after 24 weeks of treatment with tacrolimus [10]. The increased inflammatory response seen after 2 weeks of treatment may also reflect application-site reactions, such as erythema and burning, which are common during the first few weeks of treatment with tacrolimus [44]. However, these reactions have been sparsely investigated and only within the first day of treatment. One study observed mast cell-induced allergic reactions, while TNF-α-levels remained unchanged [45]. In contrast, another study observed vascular inflammation with increased TNF-α in response to systemic treatment [46].

In support of previous findings, LS showed increased expression of proteins associated with melanocyte adhesion and migration after 2 weeks of treatment [11]. These proteins included AKR1B10 and SASH1, which have not previously been described in vitiligo. AKR1B10 decreases keratinocyte expression of TGF-β, which normally keeps melanocytes in an immature state and downregulates MITF, a key pigmentation transcription factor [31]. AKR1B10 also decreases gene expression of E-cadherin [31]. In contrast, SASH1 positively regulates E-cadherin, while promoting melanocyte development and migration [32]. E-cadherin is an important melanocyte–keratinocyte adhesion protein. Disruption of E-cadherin levels weaken cell adhesion, which causes melanocytes to detach [39]. Hence, the opposing effects of AKR1B10 and SASH1 on E-cadherin expression could reflect the complex regulation of melanocyte adhesion during their migration. Interestingly, these proteins were differentially expressed only after 2 weeks but not after 8 weeks of treatment.

After 8 weeks of treatment, LS showed enrichment of proteins associated with melanosome maturation. We observed increased expression of DTNBP1, a component of the BLOC-1 complex, which colocalizes with tyrosinase related protein 1 (TYRP-1), oculocutaneous albinism type 2 (OCA2), and tyrosinase into melanosomes to promote their maturation [37]. Additionally, we observed increased expression of MED23, a positive regulator of MITF expression [36]. Our findings are consistent with previous reports of increased melanosome maturation and MITF transcription in response to tacrolimus treatment in melanocytes [11].

LS showed increased expression of proteins associated with a persistent inflammatory response after 8 weeks of treatment. We observed increased expression of RNF114, which enhances NF-κB transcription and production of chemokines CCL5 and CXCL10 in keratinocytes in response to IFN-α and IFN-γ [34]. The release of CXCL10 induced by IFN-γ through JAK/STAT1 signaling is recognized as an important driver of T-cell activation and recruitment during inflammation in vitiligo [3]. IFN-α is mainly produced by plasmacytoid dendritic cells in response to HSP70 [3]. Interestingly, we also observed increased expression of MRPL18, which promotes HSP70 synthesis in response to heat shock stress [35]. Although these findings were suggestive of increased STAT1 activity, it was not differentially expressed. Since tacrolimus suppresses IFN-γ gene expression via inhibition of the NFAT pathway, our findings may indicate that the inhibitory effects of tacrolimus on the NFAT pathway are insufficient to fully suppress IFN-γ transcription [8]. This is supported by previous reports, which have observed unchanged IFN-γ expression in LS after 24 weeks of treatment [10]. These factors may contribute to the inconsistent treatment responses observed in vitiligo. Consistent with previous untargeted proteomic studies, we did not identify cytokines TNF-α, IFN-γ, and some chemokines, such as CXCL10 [47]. This could reflect their rapid turnover or short half-lives [48].

We compared the results with those of our previous proteomic study to identify any shared vitiligo proteomic signatures across experiments (manuscript submitted). In our previous study, we observed increased expression of proteins related to keratinocyte function, proinflammatory signaling and immune regulation in LS compared with NLS. However, we observed no shared DEPs, except for CNN1. This limited overlap may reflect methodological differences, including mass spectrometers, and sample sizes. A recent study investigated the proteomic skin alterations in untreated LS and NLS, along with LS after 12 weeks of treatment with TCIs, topical corticosteroids, and narrow-band UVB in 30 patients [15]. This study observed increased expression of proteins related to inflammatory signaling and triglyceride metabolism. We compared our results but observed no shared DEPs. The study used suction-blister biopsies, a limited proteomic panel, and combination therapy, which could explain the differences.

This study has some limitations. The study was designed to investigate early treatment-induced changes and did not include an assessment of long-term clinical efficacy or repigmentation patterns. Repigmentation in vitiligo typically requires sustained therapy over months to years and would therefore not be expected within the relatively short observation period of this study. Because nonlesional vitiligo skin is not equivalent to healthy skin, comparisons between LS and NLS reflect differences within vitiligo rather than changes toward a normal skin state. The small sample size of eight patients limits the statistical power, and LS and NLS samples were not always paired, which may have affected comparability between groups. Additionally, samples were obtained from anatomical sites with different skin thickness and UV exposure, which may also have influenced the results. The use of bulk proteomic comparisons limits interpretations of findings related to cell-specific changes.

The strengths of this study included the use of a proteomic approach to evaluate tacrolimus monotherapy in full-thickness skin punch biopsies, which provided direct identification of previously undescribed proteins across all skin compartments in vitiligo. There is a need for future studies to validate the functional aspects of the DEPs and their cellular origin.

Conclusions

This study characterized the proteomic differences between untreated LS and NLS, as well as the early treatment response to tacrolimus in LS. Our study reveals inflammatory and melanocyte-associated changes that provide insights into the effects and limitations of tacrolimus.

Acknowledgements

We thank the healthcare staff at Hareskov Skin Clinic and dermatologists Gitte Strauss, Anne Danielsen, and Luise Idorn for their assistance in recruiting study patients. We thank the Jesper V. Olsen Group (Novo Nordisk Foundation Center for Protein Research, Department of Cellular and Molecular Medicine) for their technical support, and the clinical staff at The Copenhagen Translational Immunology Biobank Research Programme (BIOSKIN), Copenhagen University Hospital, Herlev and Gentofte Hospital, Denmark. We also thank all the participants in the study.

Author Contributions

Morten Bahrt Haulrig: Conceptualization, methodology, investigation, project administration, data curation, writing—original draft, writing—review and editing, visualization, and funding acquisition. Sonja Kabatnik: Conceptualization, methodology, validation, formal analysis, data curation, writing—original draft, writing—review and editing, and visualization. Lone Skov: Conceptualization, methodology, resources, writing—review and editing, supervision, project administration, and funding acquisition. Marianne B. Løvendorf: Conceptualization, methodology, resources, writing—review and editing, supervision, project administration, and funding acquisition. Beatrice Dyring-Andersen: Conceptualization, methodology, resources, writing—review and editing, supervision, project administration, and funding acquisition.

Funding

This study was funded by the Novo Nordisk Foundation (NNF21OC0066694), Herlev and Gentofte Hospital Research Council, Hofbuntmager Aage Bang Foundation, Carl and Ellen Hertz’ grant to Danish Medical and Natural Sciences, King Christian X Foundation, Else and Mogens Wedell-Wedellsborg Foundation, The Medical Sciences Part of the Faculty of Health and Medical Sciences’ Foundation for Scientifically Employed Candidates and Students at the University of Copenhagen, The A.P. Møller Foundation for the Advancement of Medical Science, the Leo Foundation (LF-ST-21–500002), and Copenhagen University Hospital, Herlev and Gentofte Hospital, Denmark. The funders had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript. The journal’s Rapid Service Fee is covered by research funds from Copenhagen University Hospital—Herlev and Gentofte Hospital and Zealand University Hospital, Roskilde.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of Interest

Lone Skov has been an advisor, investigator, and speaker for Abbvie, Eli Lilly, Novartis, Sanofi, LEO Pharma, BMS, UCB, Incyte, Stada, Takeda, Pfizer and Almirall, unrelated to the submitted work. Lone Skov reports nonfinancial support from grants from Almirall, Novartis, Janssen, BMS, Sanofi and UCB. Morten Bahrt Haulrig, Sonja Kabatnik, Beatrice Dyring-Andersen and Marianne B. Løvendorf have nothing to disclose.

Ethical Approval

The study was approved by the Ethics Committee of the Danish Capital Region ​(H-19036270) and the Danish Data Protection Agency, and it followed the principles of the Declaration of Helsinki II.

Footnotes

Prior publication: This work has not been previously presented or published in any form.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Marianne B. Løvendorf and Beatrice Dyring-Andersen Shared last authorship.

Change history

5/5/2026

Equal contribution updated.

References

  • 1.Haulrig MB, Al-Sofi R, Baskaran S, Bergmann MS, Løvendorf M, Dyring-Andersen B, et al. The global epidemiology of vitiligo: a systematic review and meta-analysis of the incidence and prevalence. JEADV Clin Pract. 2024;3:1–10. [Google Scholar]
  • 2.Ezzedine K, Eleftheriadou V, Jones H, Bibeau K, Kuo FI, Sturm D, et al. Psychosocial effects of vitiligo: a systematic literature review. Am J Clin Dermatol. 2021;22(6):757–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Seneschal J, Boniface K, D’Arino A, Picardo M. An update on vitiligo pathogenesis. Pigment Cell Melanoma Res. 2021;34(2):236–43. [DOI] [PubMed] [Google Scholar]
  • 4.Seneschal J, Speeckaert R, Taïeb A, Wolkerstorfer A, Passeron T, Pandya AG, et al. Worldwide expert recommendations for the diagnosis and management of vitiligo: Position statement from the international Vitiligo Task Force—Part 2: specific treatment recommendations. J Eur Acad Dermatology Venereol. 2023;37(11):2185–95. [DOI] [PubMed] [Google Scholar]
  • 5.Seneschal J, Boniface K. Vitiligo: current therapies and future treatments. Dermatol Pract Concept. 2023;13(4):2–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Duplaine A, Tannous J, Seneschal J, Ezzedine K, Passeron T, Dupin N, et al. Value of tacrolimus 0.1% in the treatment of vitiligo in the era of targeted therapy. Ann Dermatol Venereol. 2025;152(2):103352. 10.1016/j.annder.2025.103352. [DOI] [PubMed] [Google Scholar]
  • 7.Mehta H, Bishnoi A, Parsad D, Kumaran MS, Vinay K. Comparative efficacy of topical tofacitinib versus topical tacrolimus in the treatment of localized vitiligo: a randomized, investigator-blinded, intraindividual trial. Br J Dermatol. 2025;4:1–8. [DOI] [PubMed] [Google Scholar]
  • 8.Tsuda K, Yamanaka K, Kitagawa H, Akeda T, Naka M, Niwa K, et al. Calcineurin inhibitors suppress cytokine production from memory T cells and differentiation of naïve T cells into cytokine-producing mature T cells. PLoS ONE. 2012. 10.1371/journal.pone.0031465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Taher ZA, Lauzon G, Maguiness S, Dytoc MT. Analysis of interleukin-10 levels in lesions of vitiligo following treatment with topical tacrolimus. Br J Dermatol. 2009;161(3):654–9. [DOI] [PubMed] [Google Scholar]
  • 10.Grimes PE, Morris R, Avaniss-Aghajani E, Soriano T, Meraz M, Metzger A. Topical tacrolimus therapy for vitiligo: therapeutic responses and skin messenger RNA expression of proinflammatory cytokines. J Am Acad Dermatol. 2004;51(1):52–61. [DOI] [PubMed] [Google Scholar]
  • 11.Lee JH, Kwon HS, Jung HM, Lee H, Kim GM, Yim HW, et al. Treatment outcomes of topical calcineurin inhibitor therapy for patients with vitiligo: a systematic review and meta-analysis. JAMA Dermatol. 2019;155(8):929–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Aebersold R, Mann M. Mass-spectrometric exploration of proteome structure and function. Nature. 2016;537:347–55. [DOI] [PubMed] [Google Scholar]
  • 13.Goleva E, Berdyshev E, Kreimer S, Reisz JA, D’Alessandro A, Bronova I, et al. Longitudinal integrated proteomic and metabolomic skin changes in patients with atopic dermatitis treated with dupilumab. J Allergy Clin Immunol. 2025;155(5):1536–46. 10.1016/j.jaci.2025.01.020. [DOI] [PubMed] [Google Scholar]
  • 14.Nordmann TM, Anderton H, Hasegawa A, Schweizer L, Zhang P, Stadler PC, et al. Spatial proteomics identifies JAKi as treatment for a lethal skin disease. Vol. 635, Nature. Springer US; 2024. 1001–1009 p. 10.1038/s41586-024-08061-0. [DOI] [PMC free article] [PubMed]
  • 15.Ouwerkerk W, Narayan VS, Chielie S, van Uden NOP, Schneider MA, Hillenbrand R, et al. Immune and biological changes during treatment in patients with non-segmental vitiligo and their relation to repigmentation. J Invest Dermatol. 2025;1–12. [DOI] [PubMed]
  • 16.Huber W, von Heydebreck A, Sültmann H, Poustka A, Vingron M. Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics. 2002;18(Suppl 1):S96-104. [DOI] [PubMed] [Google Scholar]
  • 17.Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRING v11: Protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47(D1):D607–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Janson D, Saintigny G, Mahé C, El Ghalbzouri A. Papillary fibroblasts differentiate into reticular fibroblasts after prolonged in vitro culture. Exp Dermatol. 2013;22(1):48–53. [DOI] [PubMed] [Google Scholar]
  • 19.Janson DG, Saintigny G, Van Adrichem A, Mahé C, El Ghalbzouri A. Different gene expression patterns in human papillary and reticular fibroblasts. J Invest Dermatol. 2012;132(11):2565–72. [DOI] [PubMed] [Google Scholar]
  • 20.Bellei B, Maresca V, Flori E, Pitisci A, Larue L, Picardo M. P38 regulates pigmentation via proteasomal degradation of tyrosinase. J Biol Chem. 2010;285(10):7288–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kajino T, Ren H, Iemura SI, Natsume T, Stefansson B, Brautigan DL, et al. Protein phosphatase 6 down-regulates TAK1 kinase activation in the IL-1 signaling pathway. J Biol Chem. 2006;281(52):39891–6. 10.1074/jbc.M608155200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Nishida N, Otsu M, Mizutani Y, Ishitsuka A, Mizukami Y, Inoue S. The glycoprotein GPNMB protects against oxidative stress through enhanced PI3K/AKT signaling in epidermal keratinocytes. J Biol Chem. 2025;301(3):108299. 10.1016/j.jbc.2025.108299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Biswas KB, Takahashi A, Mizutani Y, Takayama S, Ishitsuka A, Yang L, et al. GPNMB is expressed in human epidermal keratinocytes but disappears in the vitiligo lesional skin. Sci Rep. 2020;10(1):1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chen LY, Lin KR, Chen YJ, Chiang YJ, Ho KC, Shen LF, et al. Palmitoyl acyltransferase activity of ZDHHC13 regulates skin barrier development partly by controlling PADi3 and TGM1 protein stability. J Invest Dermatol. 2020;140(5):959-970.e3. 10.1016/j.jid.2019.09.017. [DOI] [PubMed] [Google Scholar]
  • 25.Roy S, Elgharably H, Sinha M, Ganesh K, Chaney S, Mann E, et al. Mixed-species biofilm compromises wound healing by disrupting epidermal barrier function. J Pathol. 2014;233(4):331–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Arredondo J, Chernyavsky AI, Webber RJ, Grando SA. Biological effects of SLURP-1 on human keratinocytes. J Invest Dermatol. 2005;125(6):1236–41. 10.1111/j.0022-202X.2005.23973.x. [DOI] [PubMed] [Google Scholar]
  • 27.Chen M, Wang Y, Wang M, Xu S, Tan Z, Cai Y, et al. Keratin 6A promotes skin inflammation through JAK1-STAT3 activation in keratinocytes. J Biomed Sci. 2025;32(1):1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Vogl T, Stratis A, Wixler V, Völler T, Thurainayagam S, Jorch SK, et al. Autoinhibitory regulation of S100A8/S100A9 alarmin activity locally restricts sterile inflammation. J Clin Invest. 2018;128(5):1852–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Liang Y, Xing X, Beamer MA, Swindell WR, Sarkar MK, Roberts LW, et al. Six-transmembrane epithelial antigens of the prostate comprise a novel inflammatory nexus in patients with pustular skin disorders. J Allergy Clin Immunol. 2017;139(4):1217–27. 10.1016/j.jaci.2016.10.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kanda N, Hau CS, Tada Y, Tatsuta A, Sato S, Watanabe S. Visfatin enhances CXCL8, CXCL10, and CCL20 production in human keratinocytes. Endocrinology. 2011;152(8):3155–64. [DOI] [PubMed] [Google Scholar]
  • 31.Jumper N, Hodgkinson T, Arscott G, Har-Shai Y, Paus R, Bayat A. The aldo-keto reductase AKR1B10 is up-regulated in keloid epidermis, implicating retinoic acid pathway dysregulation in the pathogenesis of keloid disease. J Invest Dermatol. 2016;136(7):1500–12. 10.1016/j.jid.2016.03.022. [DOI] [PubMed] [Google Scholar]
  • 32.Bishnoi A, Arunima A, Vinay K, Kumaran MS, Parsad D. SASH1 Mutations and hereditary disorders of pigmentation: review of literature. Pigment Cell Melanoma Res. 2025. 10.1111/pcmr.70032. [DOI] [PubMed] [Google Scholar]
  • 33.Jin SP, Suh JH, Kim CE, Oh IG, Seo EY, Kim MK, et al. Functionally similar genes exhibit comparable/similar time-course expression kinetics in the UV-induced photoaged mouse model. PLoS ONE. 2023;18(11 NOVEMBER):1–17. 10.1371/journal.pone.0290358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Bijlmakers MJ, Kanneganti SK, Barker JN, Trembath RC, Capon F. Functional analysis of the RNF114 psoriasis susceptibility gene implicates innate immune responses to double-stranded RNA in disease pathogenesis. Hum Mol Genet. 2011;20(16):3129–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zhang X, Gao X, Coots RA, Conn CS, Liu B, Qian SB. Translational control of the cytosolic stress response by mitochondrial ribosomal protein L18. Nat Struct Mol Biol. 2015;22(5):404–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Xia M, Chen K, Yao X, Xu Y, Yao J, Yan J, et al. Mediator MED23 links pigmentation and DNA repair through the transcription factor MITF. Cell Rep. 2017;20(8):1794–804. 10.1016/j.celrep.2017.07.056. [DOI] [PubMed] [Google Scholar]
  • 37.Goff PS, Patel S, Harper DC, Carter T, Marks MS, Sviderskaya EV. Reprogramming of endolysosomes for melanogenesis in BLOC-1-deficient melanocytes. Curr Biol. 2025;35(15):3570-3586.e7. 10.1016/j.cub.2025.06.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.McAndrews KM, Miyake T, Ehsanipour EA, Kelly PJ, Becker LM, McGrail DJ, et al. Dermal αSMA + myofibroblasts orchestrate skin wound repair via β1 integrin and independent of type I collagen production. EMBO J. 2022;41(7):1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Wu Y, Yang Y, Lin Y, Ding Y, Liu Z, Xiang L, et al. Emerging role of fibroblasts in vitiligo: a formerly underestimated rising star. J Invest Dermatol. 2024;144(8):1696–706. 10.1016/j.jid.2024.02.007. [DOI] [PubMed] [Google Scholar]
  • 40.Gupta R, Misri R, Gupta A, Chowdhary M, Singh A. Genome-wide profiling reveals pervasive transcriptional alterations in fibroblasts derived from lesional skin in vitiligo including a reduced potential to proliferate. Exp Dermatol. 2023;32(4):331–40. [DOI] [PubMed] [Google Scholar]
  • 41.Kovacs D, Bastonini E, Ottaviani M, Cota C, Migliano E, Dell’Anna ML, et al. Vitiligo skin: exploring the dermal compartment. J Invest Dermatol. 2018;138(2):394–404. 10.1016/j.jid.2017.06.033. [DOI] [PubMed] [Google Scholar]
  • 42.Shiu J, Zhang L, Lentsch G, Flesher JL, Jin S, Polleys C, et al. Multimodal analyses of vitiligo skin identifies tissue characteristics of stable disease. JCI Insight. 2022;7(13):1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Speeckaert R, Belpaire A, Speeckaert MM, van Geel N. A meta-analysis of chemokines in vitiligo: recruiting immune cells towards melanocytes. Front Immunol. 2023;14(February):1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Sisti A, Sisti G, Oranges CM. Effectiveness and safety of topical tacrolimus monotherapy for repigmentation in vitiligo: a comprehensive literature review. An Bras Dermatol. 2016;91(2):187–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Du X, Che D, Peng B, Zheng Y, Hao Y, Jia T, et al. Dual effect of tacrolimus on mast cell–mediated allergy and inflammation through Mas-related G protein-coupled receptor X2. J Dermatol Sci. 2023;112(3):128–37. 10.1016/j.jdermsci.2023.10.003. [DOI] [PubMed] [Google Scholar]
  • 46.Rodrigues-Diez R, González-Guerrero C, Ocaña-Salceda C, Rodrigues-Diez RR, Egido J, Ortiz A, et al. Calcineurin inhibitors cyclosporine A and tacrolimus induce vascular inflammation and endothelial activation through TLR4 signaling. Sci Rep. 2016;6(March):1–16. 10.1038/srep27915. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Møller LBP, Kromann B, Kabatnik S, Hjortlund JH, Haulrig MB, Sølberg JBK, et al. Spatial proteomic profiling reveals increased levels of cholesterol synthesis proteins in psoriasis vulgaris. J Invest Dermatol. 2025. 10.1016/j.jid.2025.05.002. [DOI] [PubMed] [Google Scholar]
  • 48.Zhou X, Fragala MS, McElhaney JE, Kuchel GA. Conceptual and methodological issues relevant to cytokine and inflammatory marker measurements in clinical research. Curr Opin Clin Nutr Metab Care. 2010;13(5):541–7. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Articles from Dermatology and Therapy are provided here courtesy of Springer

RESOURCES