Epidermal proliferation, differentiation, and alteration of lipid composition: new key elements of the vitiligo puzzle

summary

Vitiligo is an acquired skin depigmentation disorder involving multiple pathogenic mechanisms that ultimately lead to cytotoxic CD8⁺ destruction of melanocyte cells

Brief introduction

Vitiligo is the most common acquired skin depigmentation disorder, affecting 0.

Over the years, most studies on vitiligo have tried to characterize the functional features associated with the disease, mainly lesions and surrounding skin

In skin cells, keratinocytes play a vital role in maintaining the balance of the environment within the skin, as keratinocytes form appropriate layers in the epidermis layer to ensure the composition

In lesions in patients with vitiligo, keratinocytes were found to be subject to many disturbances, including genes associated with differentiation and keratinization, as well as changes in epidermal thickness and structure (10, 11).

The differentiation of cutaneous keratinocytes and alterations in SC lipid profile in patients with vitiligo and their role as triggers for initiating the melanocyte immune response have not been studied

outcome

Non-damaging vitiligo keratinocytes exhibit changes in morphology and proliferation

The coordinated balance between keratinocyte proliferation and differentiation controls the correct alignment and integrity

Differentiation-related proteins and enzymes associated with skin barrier lipid biosynthesis and metabolism are deregulated in non-diseased vitiligo keratinocytes

After assessing the characteristics associated with cell morphology and growth potential, we analyzed the differentiation promise

Although NHKs undergo typical differentiation-related morphological changes at high calcium concentrations, manifesting as elongated, tightly arranged, and aggregated appearances, these changes are less pronounced in VHKs (Figure 1C).

At the same time as calcium rises, we also use cell density as an additional in vitro model to stimulate the differentiation of keratinocytes (25, 27).

Nonpathic vitiligo keratinocytes show impaired cell-cell junctions and actin cytoskeletal remodeling

Once they begin to differentiate, keratin forms contact between cells and tissue cells, which is essential

Nonpathic vitiligo keratinocytes with impaired energy metabolism

To determine whether the defects observed during VHK differentiation and stratification depend on impaired cellular energy and mitochondrial metabolic pathways, as previously demonstrated in vitiligo melanocytes (29) we examined levels of adenosine triphosphate (ATP) in NHKs and VHKs and observed low production of ATP within vitiligo keratinocytes (Figure 3A; **P<0.

Inhibition of intracellular ATP production has been reported to affect the differentiation of keratinocytes, which is reflected in a decrease in the expression of the differentiation marker total bracts when intracellular ATP synthesis is obstructed (34).
Therefore, our next step is to directly question the link
between impaired energy production and defective differentiation.
To do this, we use the immortalized human keratinocyte line kerct, as these cells show normal differentiation potential, assessed by the expression of basal differentiation markers and the ability to form differentiated epithelium (35–37).
We expose cells to calcium with or without the mitochondrial uncoupling agent 2,4-dinitrophenol (DNP), which alters the proton gradient of the mitochondria and hinders the production of cellular ATP (38, 39).
The effect of the compound on TFAM expression and differentiation processes was evaluated using K10 and total bractin protein levels as markers for early and late differentiation, respectively (Figure 3E).
Under high-calcium conditions, kerct cells exhibit increased expression levels of TFAM, K10, and total bracts compared to low-calcium cells, while in the presence of DNP, expression of these three proteins is dose-dependent decreased compared to untreated cells (Figure 3E).
Therefore, these findings suggest that an energy deficit may be responsible for the impaired ability of vitiligo keratinocytes to
differentiate normally and stratify.

Non-destabilizing vitiligo skin keratinocytes are rare and the distribution of keratin expression is altered

The composition of intact SC is strictly dependent on the proper differentiation of keratinocytes to form keratinocytes
.
Because our results show impaired differentiation of vitiligo keratinocytes, we wondered if this defect was also reflected in alterations in keratinocytes
.
SC specimens
of non-diseased skin of control skin and vitiligo patients were collected by tape stripping.
Measuring the peel zone area covered by keratinocytes by image analysis, the results showed that the number of keratinocytes in the non-diseased vitiligo skin sample was lower than that in the control skin sample, while the keratinocytes in the control skin were abundant and accumulated (Figure 4A; *P<0.
05).

.
The decrease in cell numbers was consistent with the results of total protein content, which is an indicator of SC cohesion (40, 41) in vitiligo patients, which was significantly lower than in the control group and therefore showed higher cohesion in vitiligo patients (Figure 4B; **P<0.
01).

Keratinous bridles are the main cell-cell adhesions in the uppermost layer of the epidermis (42); Taking into account defects in other cell-to-cell connections in the nonpathic vitiligo skin, we also performed immunofluorescence analysis of glycoprotein keratin in the core of the keratin bridge granule (43).
The analysis showed positive cell margins and occasional spot plasma staining
in the control group.
In most disease-free vitiligo specimens, keratin is less reactive and less prominent at the edges of keratinocytes; Conversely, diffuse positive lesions on the cell surface are prevalent (Figure 4C).
Thus, these results highlight changes
in non-damaging vitiligo keratinocytes.

The lipid composition of nonpathic vitiligo SC is unbalanced

To guarantee the correct integrity of the skin barrier, protein and lipid components are formed and work closely together
.
Therefore, we explore whether the results detected in vitiligo keratinocytes are reflected in changes in the lipid composition of the
keratinizing layer.
Since the lipid matrix of SC is mainly composed of ceramides (CERs), ffa, cholesterol (CH) and cholesteryl sulfate (CHS), we analyzed lipids
in normal vitiligo and healthy control skin SC with GC-MS and HPLC-MS.
There was a general increase in CH, CHS, and ffa in the vitiligo group compared to the control group, as shown by the heat map of a single sample and the mean of the vitiligo group and control group (Figure 4D).
Analysis of Variance (ANOVA) Simultaneous Component Analysis (ASCA) fractional plots showed a clear separation between the vitiligo group and the control group (Figure 4E).
Principal component analysis (PCA) analysis showed that 21 of the 24 assays in the vitiligo group were significantly higher compared to the control group, while the levels of two ffa (C19:0 and C26:0) were significantly lower (Figure 4F; *P<0.
05).

Based on these results, we further deepened the analysis of specific FFA changes to evaluate their distribution
in SC specimens in the control and vitiligo groups based on chain length.
Although the two groups are in the long chain (LC; There was no significant difference in percentages of C12–C20( Figure S3A) and VLC (C21–C25) FFA (Figure S3B), but the percentage of ULC (C26:0) FFA in SC specimens of vitiligo patients was significantly lower than in the control group (FigureS3C*P<0.
05).

In addition, the percentage of monounsaturated fatty acids (MUFAs) in the SC samples of the vitiligo group was significantly higher; Figures S3D*P<0.
05) and C18:2 fatty acids (FigureS3E**P<0.
01).

Overall, these results suggest that SC in vitiligo patients exhibits an imbalance
in cholesterol and FFA composition.
To dissect the distribution of all the major lipids involved in the assembly of the skin barrier, parallel to the assessment of cholesterol and FFA, we also analyzed the composition of ceramides
.
CERs are the main lipid components in SC and are divided into subgroups
according to molecular structure.
The most abundant CER subclass contains two fatty acid moieties, non-hydroxy [N] and α-hydroxy [A] moieties, combined with one of the four sphingosine base moieties: sphingosine [S], dihydrosphingosine [DS], phytosphingosine [P] and 6-hydroxysphingosphingosine [H].

We analyzed the α-hydroxy FA-CER (CER[ADS], CER[AS], CER[AP] and CER[AH]) and non-hydroxyFA-CER (CER[NDS], CER[NS], CER[NP] and CER[NH]) subclasses using predetermined multi-reaction monitoring (Table S1) techniques (Table S1).

Among the DETECTED CERs, we performed a comparative analysis of the 96 Cers, the most abundant of the two sets of SC samples
.
CER curves of SC specimens from the control group and vitiligo group are shown in the heat map (Figure 5A).
Although the individual differences in CER distributions were large, after averaging the control and vitiligo data grouping, the CER difference pattern between the two groups became apparent (Figure 5A).
The ASCA scoring plot of the analyzed CER showed a significant separation between the vitiligo group and the control group (P<0.
01; Figure 5B).
As shown in the load plot, of the 96 evaluated CERs, 61 of the two groups were in significantly different states (red bars): 57 vitiligo groups had lower levels and 4 were higher (Figure 5C; *P<0.
05).

In addition, in vitiligo SC samples, we observed a significant decrease in the level of the LC fatty acid (C22–C28) portion of the CER (Table S2).

This altered pattern may be related to a decrease in CERS activity, such as a high affinity for VLC or ULC fatty acids (e.
g.
, CERS3) in keratinocyte cultures (Figure 1F).
The distribution of sphingosine components of CER also differed between vitiligo patients and the control group, with a decrease in most sphingosine bases in vitiligo patient specimens, indicating a change
in the CER synthesis pathway.
All four CER increases in the vitiligo group contained the non-circulating component of C18, or a sphingosine group
.
CERs in the CERN (18) S(18) Vitiligo Group were the most significantly elevated, with fatty acids and sphingosine containing C18 (P<0.
01); This result may be related to the observed increase in C18 FFA and precursor C16 FFA (Table S2
).
All assessed CER subclasses in the vitiligo group showed a downward trend, and differences in CER[AP], CER[AS], CER[NDS] and CER[NP] were statistically significant, resulting in a significant separation between the two groups (Figures 5, D and E; *P<0.
05).

In addition, we calculated the ratio between CER[NP] and other CER subclasses as markers of skin barrier properties (44).
Compared with the control group, the ratios of CER[NP]/[NDS], CER[NP]/[NH], CER[NP]/[NH], CER[NP]/[AP], CER[NP]/[AH], and CER[NP]/[AS] in vitiligo patients were significantly reduced compared to the control group (Figure 5F; *P<0.
05).

ASCA scoring plots (Figure S4A) and load plots (Figure S4B) of the CER ratio expressed as a percentage of total content showed a greater
degree of separation between the control group and the vitiligo group.
The PCA plot, based on all lipid parameters analyzed, shows a clear separation between vitiligo patients and the control group, emphasizing significant changes in the composition of the epidermid barrier in vitiligo patients (Figure S4C
).
We analyzed possible correlations
between changes in a single lipid component and disease states measured by vitiligo degree score (VES).
There was no significant correlation between CER, FFA, CH and CHS content in SC and VES (mean) R= ? 0.
17, the average R=0.
03, R=0.
05, and R=? 0.
18 respectively).

Subsequently, we extended the analysis to the lesion area and selected another group of patients
with lesions in the forearm.
PCA confirmed the difference between the non-lesion area and the control region, while the lesion area and the non-lesion area showed similar abnormalities of the epidermid lipid spectrum (Figures 6, A and B).

Inflammatory mediators are upregulated in non-diseased vitiligo keratinocytes

Alterations in keratinocyte differentiation, barrier composition, and function are associated with dysregulation of the immune response in several skin diseases (16–20).
In addition, skin barrier defects weaken the skin's ability to properly resist trauma, thereby disrupting tissue homeostasis and possibly triggering local inflammation and making it persist for a
long time.
In this case, keratinocytes can trigger the Koebner phenomenon, in which mild and repeated mechanical damage can aggravate inflammatory skin diseases, including vitiligo (45, 46).
Therefore, based on the differentiation of epidermis and the imbalance of lipid components in patients with vitiligo, we evaluated the expression
of inflammatory mediators in VHKs.
We focused on the analysis of the chemokine CXCL10 because it is involved in the recruitment of T cells to the skin and the apoptosis of melanocytes (21, 47) interleukin-1β (interleukin-1β) and interleukin-6 (interleukin-1β) in the inflammatory network of vitiligo (48).
qRT-PCR analysis showed significantly higher than interleukin 1Β (**P<0.
01) and CXCL10 (*P< in the control group Gene expression in 0.
05) VHKs was higher than in NHKs, while no significant difference in interleukin 6 was observed in VHKs (Figure 7A).
At the protein level, enzyme-linked immunosorbent assay (ELISA) analysis of cultured supernatants showed that at 48 and 72 h, the yield of CXCL10 increased (**P<0.
01; Figure 7B) and the expression of IL-1β also increased after 72 h (**P <0.
01; Figure 7C) compared
to NHK culture in VHK culture.
Next, we assessed the release of chemokines and cytokines in the monolayer of scratch-injured cells, an in vitro model of The Cobner phenomenon (49).
From 24 h and 48 h post-injury, the release of CXCL10 in the VHK group gradually increased, and at these two time points, the level of CXCL10 in the culture supernatant of the VHK group was significantly higher than that in the NHK group (Figure 7B; **P<0.
01).

VHKs also increased the amount of IL-1β produced at 24 hours and 48 h compared to NHKs (Figure 7C; **P<0.
01).

Thus, the high expression and release of inflammatory mediators in vitiligo keratinocytes and a further increase after the Koebner phenomenon abrasion model suggest that keratinocyte differentiation and alterations in the lipid composition of the skin barrier may be the initiating determinants of the occurrence of local inflammatory responses responsible for activating the immune response of melanocytes
.
To investigate whether the observed keratinocyte dysfunction is attributed to a pre-existing inflammatory environment, we treat normal primary keratinocytes with recombinant CXCL10 and analyze the effect
of chemokines on keratinocyte differentiation.
To do this, we first determined the mRNA expression levels of the CXCL10 receptor CXCR3 on NHKs and VHKs, and qRT-PCR analysis showed positive expression of the CXCR3 gene, while there was no difference between the control group and vitiligo keratinocytes (Figure 7D).
Next, we assessed the activity of keratinocytes with an increase in the concentration of CXCL10 with MTT, and no significant changes were observed at any dose analyzed compared to untreated cells (Figure 7E).
K10 and total bract protein levels were detected by westernblot and immunofluorescence analysis to explore the effect
of chemokines on cell differentiation.
Under high calcium conditions, NHKs exhibited an increase in K10 expression levels in the absence or presence of CXCL10 treatment compared to cells cultured with low calcium, mainly in the presence of K10 (Figures 7, F, and G) indicating that keratinocytes maintained the ability to
differentiate in the presence of chemokines.

Changes in differentiation markers and inflammatory mediator expression can be detected in the epidermis of nonpathic vitiligo

Next, we examine the structural arrangement of the entire epidermis by examining the skin biopsy collected from non-diseased areas of vitiligo patients, exploring whether the keratinocyte differentiation and assembly defects observed in vitro are obvious
.
Quantitative analysis of histological images showed that the thickness of all epidermal layers in non-diseased vitiligo skin, as well as SC alone, increased compared to control skin (Figures 8, A and B; *P<0.
05 and **P<0.
01).

There were no significant differences
in Ki67 (Figure S5A) and stress aging-related marker p16 (Figure S5B) staining between vitiligo patients and the control group.
K10 immunohistochemical analysis showed strong positives for all basal upper layers of the skin in the control group, but in most disease-free vitiligo specimens, the signal of K10 to the epidermis layer showed focal delay (Figure 8C).
Image analysis showed that the signal distribution of vitiligo patients was significantly reduced compared to the control group (Figure 8C; **P<0.
01).

Similarly, the total bracts of the upper epidermal layer of the control skin were positive, while in non-diseased vitiligo skin, the corresponding signal was significantly reduced, which was assessed by quantitative image analysis (Figure 8C; **P<0.
01).

Thus, weak expression of early and late differentiation markers suggests that changes detected by the differentiation process inducing keratinocytes in vitro reflect the skin characteristics of patients
with vitiligo.
As an additional indicator associated with skin barrier formation, we examined the expression of TGM1 to determine its critical role
in the formation of the keratinized envelope that makes up the skin barrier.
Image analysis showed that the intensity of TGM1 immunofluorescence in the outermost layer of lesion-free vitiligo skin was significantly reduced and the discontinuities were stronger compared to control skin (Figure 8C; **P<0.
01).

Immunohistochemical method to detect the expression of inflammatory mediators and chemokines, image analysis to detect positive areas; IFN-γ was expressed at sporadic low levels with no significant differences between groups (Figure 8D) while CXCL10 immunoreactivity was significantly enhanced in nonpathic vitiligo skin, although the degree of labeling differed between different specimens (Figure 8E; **P<0.
01).

Finally, assess whether changes in differentiation and expression of inflammatory markers are accompanied by the presence of CD8 ⁺ T lymphocyte infiltration, CD8 immunohistochemical analysis ⁺ T cells
.
The results showed that a small number of cells were scattered throughout the section, mainly in the perivascular region (Figure S5C
).

discuss

In this study, our aim was to assess the morphological and functional characteristics of pigmented skin in vitiligo patients, as well as the lipid components
.
In addition, we investigated possible links between these epidermal features and activation of the immune response destroyed by melanocytes
.
In vitro and in vitro evidence is obtained by analysis of keratinocyte culture and skin biopsy, and in vitro evidence is obtained by assessing keratinocyte morphology and epidermal lipid composition
.
Here we find epidermal defects and altered lipid components of the skin barrier in normal vitiligo skin
.
Non-injurious VHKs present with morphologically enlarged, low proliferation rates, and marked differentiation disorders, including poor
lamination and barrier assembly.
The increase in cell size is associated with the differentiation of keratinocytes (50, 51).
However, the cell size of aging keratinocytes also increases (52–54).
In addition, previously reported data show that the differentiation process of senescent keratinocytes is significantly slower than that of non-senescent keratinocytes (52).
Therefore, we observed an upward trend in the number of keratinocytes positive for the aging marker p16 in non-damaging vitiligo skin slices, which is consistent with the presence of stress and aging-related markers in similar samples previously reported (55).

In vitiligo cutaneous keratinocyte cultures, a damaged aging process and attempts to control it have been described (56).
Although the study focused on in vitro behavioral changes in diseased keratinocytes, p16 expression levels that were already evident at the time of first culture passage have been reported in uninfected cells compared to their associated
counterparts.

Thus, the features shown suggest an aging predisposition behavior of non-diseased keratinocytes, as reported by vitiligo melanocytes and fibroblasts (5, 55).

Studies on the mechanisms of differentiation regulation release have shown intrinsic metabolic defects in nonpathic vitiligo melanocytes and fibroblasts (6, 29).
We found that the basal ATP level in patients with VHK was lower than in patients with
NHK.
Decreased intracellular ATP synthesis due to inhibition of F1F0-ATP synthase (also known as mitochondrial complex V) has been reported, resulting in downregulation of the expression of keratinocyte differentiation markers, thereby providing a link between energy metabolism and cellular behavior such as differentiation (34).
A decrease in the parallel expression of the mitochondrial transcription factor TFAM in vitiligo cells indicates impaired mitochondrial-associated
energy.
TFAM plays a vital regulatory role in mitochondrial function, including the expression of subsections of respiratory chains encoded by mitochondria, which are essential for energy production and transcription and replication of mitochondrial DNA (30–32).
After treatment with mitochondrial uncoupling agent DNP, differentiation-related abnormalities developed in the kerct cell line, further suggesting that mitochondria are the root cause
of impairing the function and energy exhaustion of keratinocyte differentiation.
In vitiligo melanocytes, low production of ATP is accompanied by simultaneous upregulation of enzymes that control glycolytic metabolism (29).
This induction of anaerobic glycolysis may be a compensatory strategy for reduced mitochondrial ATP production, but it also impairs differentiation, as abnormal expression of key regulators of glycolytic metabolism such as fructose-2-phosphate/fructose-2,6-diphosphatase 3 (PFKFB3) has been shown to inhibit differentiation of keratinocytes (57).

When keratinocytes differentiate, they form cell-cell contact and begin to aggregate
in the epidermal multilayer membrane.
E-cadherin is a key determinant
of proper cell stratification after differentiation.
In addition, E-cadherin is the main adhesion molecule that regulates the interaction between keratinocytes and melanocytes
.
In normally pigmented cutaneous melanocytes, defects in the expression and distribution of E-cadherin have been highlighted
.
There have also been studies that show that this adhesion damage affects the resistance of melanocytes to oxidative and mechanical stress, suggesting that the presence of preclinical primary skin defects favors the shedding and loss of melanocytes (14).
Our results show a reduced ability of keratinocytes to properly tissue cell contact during differentiation and stratification, thus demonstrating that impaired E-cadherin expression reported in nonpathic melanocytes extends to the entire epidermis
.
The uneven distribution of E-cadherin in VHKs compared to NHKs may be due to lower ATP levels, as its depletion interferes with the localization of E-cadherin and degrades it (58).
Similarly, abnormal ATP concentrations can also serve as the basis for the recombination of the backbone of actin cells that induce differentiation in VHKs, as previously reported in Hailey-Hailey keratinocytes (59).
As a result, these inherent metabolic damage prevents VHKs from producing enough energy for proper differentiation and stratification
.

Parallel to the changes in the differentiation of keratinocytes, the integrity of the skin barrier is also sensitive
to the correct balance of lipid components.
In the current study, we noticed a change in the episebric component of the nonpathic vitiligo skin, a significant decrease in CER levels, and an increase in
FFA, CH, and CHS levels.
The synthesis, release, localization, and binding of these lipids play a key role in forming a complete and well-maintained SC and a fully competent osmotic barrier (60).
Alterations in total CER levels and subclass ratios are associated with skin barrier damage (44) glycosylated short chains and LC-CERs promote differentiation of keratinocytes (61).

In patients with vitiligo, a decrease in CER can be used as an indicator
of changes in proliferation and differentiation balance within the epidermis layer.
After in vitro exposure to prodigative stimuli, VHKs show impaired expression of skin barrier lipid-related enzymes such as CERS3 and ELFL4, which are consistent
with the reduction in CER associated with VLC and ULC-FAs detected in vitro and in vitro.
There was no difference in FAS expression between post-differentiation VHKs and NHKs, a finding that supports the hypothesis that the increase in FFA in SC stems from deregulation of the CER synthesis pathway, rather than the biosynthesis
of FFA.
The increase in CH and CHS is associated with inhibition of epidermal shedding and SC thickening (62) which may explain the reason for thickening of the epidermis in patients with
vitiligo.
These results also reflect the presence of higher cohesion in vitiligo keratinocytes than in the control group, and a decrease in protein content removed by tape stripping was assessed, a reliable method to measure scliance cohesion (40, 41).
Therefore, differences in lipid content and keratinocyte adhesion between vitiligo patients and the control group suggest a change
in the structure of the skin barrier in vitiligo patients.
We observed that the content of FFA in diseased and non-diseased keratinocytes showed similar differences
compared to the control group.
To our knowledge, in vitiligo patients, the comprehensive liposome analysis of SC associated with the expression profile of enzymes that regulate lipid barrier synthesis has not been performed
before.
Decreased expression of keratinocyte differentiation and keratinization genes stratifin, keratin, envoplakin, periplakin, and transglutaminase 1 in the skin lesions and non-lesion epidermis was associated with no change in the expression of filament proteins and total bracts (11).
However, unlike our study, there was no correlation
with normal control skin.
There was no difference in lipid composition between lesions and non-lesions observed in our samples, which is consistent with previous data, i.
e.
there was no significant difference between basal TEWL (percutaneous dehydration) at vitiligo-affected sites and non-lesion skin (12).
On the other hand, the restoration of the barrier at the site of injury is significantly delayed compared to non-injured sites (12).
Based on these data, we can speculate that when mechanical damage acts on non-damaging skin, it may lead to an additional but transient deterioration of the intrinsically altered lipid barrier component, triggering an immune response against melanocytes
.
Once the lesion has stabilized, the skin barrier may partially recover
.

Epidermal cell differentiation and alterations in the composition of the skin barrier can stimulate keratinocytes to recruit immune cells and release inflammatory mediators, thus acting as
immune sensors.
Damage to the skin barrier is associated
with dysregulation of the immune response.
Inflammatory signals in vitiligo herald the activation of epidermal dendritic cells and natural killer cells (NK), which release IFN-γ, stimulate keratinocytes to produce chemokines, which in turn acquire and activate cytotoxic CD8⁺ T cells
.
Increased expression of inflammatory mediators including IFN-γ, tumor necrosis factor-α, IL-6 and IL-1 cytokines, and chemokines CXCL9, CXCL10 and CXCL16 in the lesion and surrounding skin has been widely reported (48, 63–65).
Some studies have attempted to define some of these messengers as potentially reliable biomarkers of disease activity and severity, both at serum levels and at tissue levels (21, 64, 65).
In particular, CXCL10 levels detected in aspirated vesicle fluid have been reported to be significantly higher than stable disease
in active vitiligo.
Conversely, stable lesions exhibit higher levels of chemokines than non-lesions (64).
However, exactly what causes the recruitment and maintenance of inflammatory cells in the skin is not fully understood
.
Elucidating these mechanisms may provide viable targets for inhibiting or preventing the emergence and spread of clinical manifestations
.
Recently, a single-cell RNA sequencing study of aspirated blisters obtained from lesions and non-lesions found intercellular signals that had not previously been reported in the lesion area, and the researchers created a comprehensive set
of intercellular communication interactions within the epidermis of vitiligo.
In addition, these authors demonstrated that non-diseased skin is in a constant state of subclinical inflammation (24).
In this study, we did not detect differences in epidermal IFN-γ expression between vitiligo and control skin samples
.
Thus, in vitiligo, deficiencies in proper epidermal tissue and barrier composition may trigger stimulation of local pro-inflammatory signaling (66) first a congenital immune response and then an adaptive immune response, leading to loss
of melanocytes.
Therefore, we found that VHKs are more spontaneously secreted IL-1β, especially CXCL10
, than NHKs.
In a model that simulated mechanical stress, the production of in vitro abrasions Koebnerization models of VHKs, IL-1β and CXCL10 was further upregulated
.
Defective keratinocyte differentiation and alteration of barrier components may represent a priming determinant of the Koebner phenomenon, defined as the occurrence of new lesions in unaffected skin areas after trauma in patients with dermatological diseases (45, 46).
This hypothesis is further supported by positive CXCL10 detected in non-diseased vitiligo skin samples, consistent with previous studies (23).
Thus, defects in the differentiation of keratinocytes and abnormal lipid barrier components may be the functional basis of the observed persistent inflammation and the cause of
the Koebner phenomenon.
The identification of similar changes in the lipid composition of the lesion area and the non-lesion area, as well as the lack of correlation between CER, FFA, CH, or CHS levels and the degree or duration of the disease, support the idea that the unbalanced structure of the skin barrier is not the result of inflammatory processes, but the constituent features
of pigmented vitiligo skin.
Consistent with this hypothesis, our findings suggest that treatment with recombinant CXCL10 does not impair the differentiation ability of keratinocytes, further suggesting that the cause of changes in keratinocyte behavior is an intrinsic defect rather than a pre-existing inflammatory environment
.
After mechanical or stress damage, the activation of keratinocytes may be exacerbated, leading to a gradual increase in CD8 ⁺ T cell infiltration
.
Thus, epidermal changes detected in non-diseased skin are accompanied by the presence of a small amount of CD8 ⁺ T cells
.
Once stimulated, keratinocytes may release more pro-inflammatory mediators, as well as other factors such as matrix metalloproteinase (MMP)–9(67) that cause imbalances in the melanocyte microenvironment, promoting melanocyte bleeding and percutaneous melanocyte loss
.
The limited consistency of vitiligo symptoms in monozygotic twins further underscores the critical role of environmental factors in the disease (68).

One limitation of our study is that disease activity
is defined primarily based on clinical symptoms and subjective assessment by the doctor.
In addition, we did not use objective measurements to assess the function of the skin barrier, nor did we evaluate the composition and function after treatment, such as light therapy
.
A survey of samples collected from patients who responded to treatment will help validate our findings and may provide a good clinical translation
.
Overall, our data identified changes in the structural and functional properties of non-diseased keratinocytes, suggesting that the epidermis is an unknown key factor in the multifaceted pathogenesis of vitiligo
.
Therefore, targeted treatments, aimed at improving the inherent epidermal defects, can provide effective protection against the occurrence and development of vitiligo pigmentation, even in areas
of the skin where the lesions have not yet developed.

Materials and methods

Research design

The objectives of this study were: (1) to explore the differentiation process of keratinocytes and the lipid composition of the skin barrier in vitiligo patients; 2.
Identify the underlying mechanisms for possible regulatory and regulatory imbalances; And (iii) to study whether keratinocyte-related alterations constitute factors that promote the development of local inflammation, which is responsible for activating immune infiltrates
.
The subject of the study was IFO-IRCCS
of the Vitiligo Family at the San Galicano Institute of Dermatology.
Fondazione Bietti, the Ifo Institutional Research Ethics Committee (Instituti Regina Elena e San Gallicano), approved the study and obtained informed written consent during the registration process for patients and healthy volunteers
.
The study was conducted
in accordance with the Principles of the Helsinki Declaration.
We established primary cultures of non-pathological vitiligo keratinocytes and control keratinocytes, we collected SC samples and examined paraffin-embedded skin biopsies
.
Real-time quantitative RT-PCR technology was used to analyze the gene expression of the sample; (ii) protein expression using Western blot, ELISA, immunofluorescence and immunohistochemical analysis; and (iii) lipid component methods
using GC-MS and HPLC-MS.
For each experiment, at least three biological replicates are used, and the data analysis is not blind
.
Sample size and P are reported values
in each figure legend.

Study participants and approvals

The study was conducted
on patients in the Vitiligo Department (IFO-IRCC) of the San Gallicano Deatological Institute.
There were no eligibility criteria for a specific disease, and any vitiligo patient who was willing to follow the study procedure was included in the study
.
A total of 86 vitiligo patients and 49 healthy volunteers
were enrolled in the study.
Fondazione Bietti, the Ifo Institutional Research Ethics Committee (Instituti Regina Elena e San Gallicano), approved the study and obtained informed written consent during the registration process for patients and healthy volunteers
.
The study was conducted
in accordance with the Principles of the Helsinki Declaration.
Exclusion criteria included any skin conditions that might alter the clinical evaluation, use of previous treatments such as topical treatments (corticosteroids, topical calcineurin inhibitors, cosmetics, and camouflages), prior light therapy within 12 weeks, and any systemic immunomodulation
within 8 weeks after examination.
During the study visit, a detailed medical background, including a history of
vitiligo, was obtained for each patient.
VES measurements are performed
by a dermatologist experienced in the diagnosis and treatment of vitiligo.
In addition, SC samples
from all patients involved in the study were collected.

Patient characteristics

The main demographic and disease characteristics are shown in
Tables S3 and S4.
Of the 86 patients, 47 were female and 39 were male, with a mean age of 43.
6 years (range 22 to 71 years).

Vitiligo course 2 to 48 years, the average of 15.
1 years
.
Overall, all patients were affected by non-segmental vitiligo as defined by the Global Consensus Meeting on Vitiligo (VGICC) and were included in the assessment
.
Table S4 shows the different subtypes
of vitiligo within the patient group.
Facial vitiligo is the most common subtype, followed by diffuse vitiligo
.
Overall, the average VES score was 2.
65 (ranging from 0.
135 to 15.
3875
).
No patients showed markers of clinical activity, such as confetti-like decolorization, Koebner phenomenon, and trichromatic lesions, while patient-reported activity was considered unreliable and therefore excluded from
any evaluation.
Each registered patient was followed up at our institute, emphasizing that there had been no recent appearance of vitiligo and that the duration of the disease was at least 2 years
.

Cell culture and treatment

As previously described, primary cultures of human keratinocytes (NHK and VHK) were isolated from control skin and non-destabilizing vitiligo skin and cultured in 154 medium (Monza Life Technologies Italia, Italy), supplemented by human keratinocyte growth supplements (HKGS) (Life Technologies Italy), as well as antibiotics and ^Ca2+ (0.
07 mm).

Cells are passaged once a week and experiments
are performed between generations 2 and 4.
Keratinocyte differentiation
is induced by changing the calcium concentration in the medium from low (0.
07 mm) to high (1.
8 mm).
Alternatively, allow keratinocytes to achieve density-mediated differentiation and maintain cells in low calcium and HKGS-supplemented medium for 7, 10, and 14 days
.
The immortalized human keratinocyte line Ker-CT (American type culture specimen set CRL-4048) is stored in 154 medium containing HKGS (LifeTech Of Italy), antibiotics and 0.
1 mM Ca^.
In DNP (Sigma-Aldrich) experiment, Kerct cells are cultured in 154 medium, The addition of antibiotics and Ca ^di+ in the presence or absence of DNP (0.
05 to 0.
1 mm) is sustained at low or high concentrations for 48 h
.
For experiments with CXCL10 (Peprotech, London, UK), NHK is stored in medium with high Ca ^{and treated} with chemokine concentration increased for 48 h
.

Immunofluorescence analysis

Fix keratinocytes with 4% paraformaldehyde and 0.
1% Triton X-100, allowing them to penetrate, or fix in low-temperature methanol? 20 degrees Celsius
.
Cells are then cultured with the following major antibodies: anti-Ki67 rabbit polyclonal antibody (1:500; Abcam, Cambridge Science Park, Cambridge, UK), anticellular keratin 10 (K10) rabbit polyclonal antibody (1:500; Abcam), anti-foreskin rabbit polyclonal antibody (1:100; Abcam), Anti-E-cadherin monoclonal antibodies (1:200; Dako Corp.
, Carpentia, CA, USA), anti-ELOVL4 rabbit polyclonal antibodies (1:100; Abcam), anti-LAS3/CERS3 rabbit polyclonal antibodies (1:100; Abcam), anti-TGM1 (1:200) rabbit polyclonal antibodies (Abcam), and anti-TFAM rabbit monoclonal antibodies (1:100; Cell Signaling Technology，MA，USA）
。 The primary antibody
was observed using goat anti-rabbit Alexa-fluor546 conjugate and goat anti-mouse Alexa-fluo488 binding antibody (1:800; Thermo-Fisher-Scientific, Italy).
Using a 4′,6-diamino-2-phenylindole (DAPI; Invitrogen) long-acting gold anti-fading reagent mounting coverslips
.
The fluorescence signal is analyzed by recording stained images using charge-coupled device cameras (ZEISS, Ober Cochin, Germany
).
The percentage of Ki67-positive cells was assessed by counting at least 1800 cells of NHKs=5 and VHKs=5, and the results represented the percentage
of positive cells/total cells in each sample.
To measure cell area, at least 1000 cells (NHKs=6 and VHKs=6) were measured, and the results were expressed in mean area/cell (μm) ²) for each sample
.
Quantitative analysis of TFAM fluorescence intensity of at least 700 cells with NHKs=4 and VHKs=4 using Zen 2.
6 (blue edition) software (ZEISS
).
Results are expressed
as the average of the intensity/cells per sample.
The stripped keratin cells are immunolabeled with antikeratotic complexion polyclonal antibodies (1:100; Abcam) and then immunelabeled with goat anti-rabbit Alexa Fluor546 conjugate (1:800; Thermo Fisher Scientific, Italy
).

Determination of proteins by sandwich ELISA method

CXCL10 and IL-1β in the control group and supernatant of vitiligo keratinocytes were analyzed
using a commercially available ELISA kit (Biotech Co.
Ltd.
) as per the manufacturer's protocol.
Results obtained from primary cells are normalized to represent 6 cells in picograms every 1×10.

RNA extraction with real-time quantitative PCR

Total RNA was isolated using the Aurum Total RNA
Miniature Kit (Bio-Rad Laboratories Srl, Milan, Italy).
By OD260/280 absorbance measurement, the yield, purity, and quality
of the RNA were determined.
According to the manufacturer's instructions, cDNA was synthesized from 1 μg of total RNA using the Reverted First Strand cDNA
Synthesis Kit (Thermo Fisher Scientific, Monza, Italy).
Quantitative real-time RT-PCR in a reaction mixture containing 2X ChamQ universal SYBR-qPCR
masterbatch (Vazyme-Biotech) and 25pmol forward and reverse primers.
The sequence of all primers is shown in
Table S5.
Three reactions
were performed using the CFX96 real-time system (Bio-Rad Laboratories Srl).
Melt curve analysis is performed on each gene to ensure the specificity
of the amplification product.
Gene expression levels with 2^{? ΔΔCT} Methods: Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an endogenous control and expressed
relative to untreated control cells.

Western blot analysis

Keratinocyte extracts were prepared with radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitors, and cell lysate concentration
was determined with Bradford protein analysis reagent (Bio-Rad).
Equal amounts of protein are dissolved on an SDS-polyacrylamide gel, transferred to a nitrocellulose membrane (Amersham Biosciences, Milan, Italy), and then treated with the following main antibodies: anticellular keratin 10 rabbit polyclonal antibody (1:1000; Abcam), anti-foreskin rabbit polyclonal antibody (1:500; Abcam).
Anti-EVOL4 rabbit polyclonal antibodies (1:500; Abcam), anti-LAS3/CERS3 rabbit polyclonal antibodies (1:500; Abcam), anti-TGM1 rabbit polyclonal antibodies (1:500; Abcam) and anti-TFAM rabbit monoclonal antibodies (1:1000; cell signaling technology).

Anti-Gapdh rabbit polyclonal antibody (Santa Cruz Biotechnology Company, California, USA) was used as a load control
.
Secondary anti-rabbit immunoglobulin G (IgG) horseradish peroxidase (HRP) binding antibody and anti-mouse IgG-HRP binding antibody
were used.
Detection of antibody complexes
by chemiluminescence.
Imaging and density analysis
are performed by measuring optical density in specific wavelength bands using the GS-800 Calibrated Image Density Meter (Bio-Rad Laboratories Srl, Milan, Italy) or the ULITEC Mini HD9 Acquisition System (Alliance UVITEC Ltd.
, Cambridge).

Immunohistochemistry

Dewaxing from formalin-fixed and paraffin-embedded continuous sections (3 μm) and rehydrating
by fractional ethanol to phosphate-buffered saline.
After antigen recovery, tissue sections are cultured with the following main antibodies: anti-Ki67 monoclonal antibodies (MIB-1, Dako, Agilent, Santa Clara, CA, USA), anti-p16 mouse monoclonal antibodies (Thermo Fisher Scientific, Italy), anticellular keratin 10 rabbit polyclonal antibodies (1:500; Abcam), Anti-foreskin rabbit polyclonal antibody (1:300; Abcam), anti-IP-10 polyclonal antibody (1:200), anti-IFN-γ polyclonal antibody (1:400), anti-CD8 monoclonal antibody (clone C8/144B, Dako).

Staining
was observed with HRP, a thermal ultrasound quantum detection system, using 3-amino-9-ethylcarbazole or 3,3'-diaminobiphenylamine as substrates.
All slices are counterstained with hematoxylin
.
Antigen recovery is achieved by heating the portion of the anti-IP10 antibody at pH9 and pH6
.
Negative controls are obtained
by omitting primary antibodies.
During TGM1 immunofluorescence staining, tissue sections are dewaxed to extract antigens and then incubated
with anti-TGM1 polyclonal antibodies (1:200) (Abcam).
The primary antibody
was then observed using the goat anti-rabbit Alexa-fluor546 conjugate (1:800; Thermo-Fisher Scientific).
Slides are made with a long-acting gold anti-fading reagent
containing DAPI (Invitrogen).
Positive area/total area was measured using Zen2.
6 (blue edition) software (Zeiss) and the results were expressed
as the average of each sample.
The epidermis thickness was measured with zen2.
6 (blueedition) software, and the results were expressed
in the average thickness value ± SD.

Human SC sampling

As previously described, sc(CuDerm Corporation) of the patient's forearm is continuously stripped with D-scales (CuDerm Corporation, Dallas, Texas, USA) (69).
Simply put, remove very shallow suspension particles with a single D-square patch stripping three times the determined sampling area and treat them as clean discs
.
After marking the initial position of the tape with a fine-tipped pen, four patches were prepared for sampling and three peels
were performed in each area without surface SC.
For vitiligo subjects, SC was collected from clearly unaffected areas
.
The collected patches are stored in ? 80 °C until processing
.
The same assessment was conducted in a control group of hospital staff whose sex and age matched those of the
vitiligo group.

ATP assay

Intracellular ATP levels are measured using commercial ATP colorimetric/fluorescence assay kits (BioVision) according to the manufacturer's instructions
.
Results were reported as ATP content/1×¹⁰⁶ cells
.

MTT law

NHK treated with CXCL10 is then incubated at 37 °C with 3-(4,5-dimethyl-2-thiazolidinyl)-2,5-diphenyl-2H-tetrazolammonium bromide (MTT) (1 mg/ml) for 2 h and lysed
in dimethyl sulfoxide.
The absorbance at 570 nm was measured with the spectrophotometer DTX880 multimode
detector (Beckman Coulter Srl, Milan, Italy).
The results were manifested as a multiple change in the value of untreated cells grown in a low-calcium environment ^{by two+}, defined as 1
.

Extraction and determination of SC proteins

SC protein content is used to quantify SC for tape stripping removal
.
Protein extraction as described earlier by Raj and so on
.
(70) Minor modifications
.
Simply put, incubate two D-scales in a 1.
5 ml Eppendorf tube with 750 μl of 1 M sodium hydroxide solution and shake at 37 °C at 1400 rpm using a hot mixer (Eppendorf, Hamburg, Germany) for 1 h
.
Then add 1 M hydrochloric acid neutralizing SC protein solution
.
The protein content
was determined using the Pierce BCA Protein Analysis Kit according to the manufacturer's instructions (Thermo Fisher Scientific, MO, Italy).
Before measuring absorbance at 562 nm, incubate the microplate at 37 °C for 2 h
.
The protein content of each sample is analyzed into two parts, with the average value considered the correct value
.
For each microplate analyzed, a new calibration curve is prepared and the same extraction procedure described above is performed on two blank D-scales to minimize the background signal
.

D-Scale lipid extraction

With 0.
025% butylhydroxytoluene (BHT; Sigma-Aldrich, Milan, Italy) of 2 mL ethanol (Darmstadt Merck, Germany) extracts SC lipids from two D-Squames to prevent oxidation
.
Samples were extracted after adding 200 μl of deuterated internal standard, the deuterated internal standard containing 100 μM d6 cholesterol (d6CH), 25 μM d7 cholesterol sulfate (d7CHS), 50 μM d17 palmitic acid (d17PA), 10 μMN- palmitoyl-d31-d- Erythosphingosine (d31CER16:0) and 0.
01% BHT, acetone/methanol/isopropanol (40/40/20).

Transfer the ethanol used for extraction to a 2-ml Eppendorf tube, centrifuge at 18,000 rpm for 10 min at 4 °C, and then filter through a filter plate with a 0.
2-μm PTFE membrane (Captiva, Filter Plate, Agilent Technologies, California, USA
).
After evaporation under nitrogen until dry, the sample extract is stored at ? Before analysis, dissolve it in 250 μl acetone/methanol/isopropanol (40/40/20
).

GC-MS lipid analysis

Direct silanation of N,O-bis-(trimethylsilyl)-trifluoroacetamide containing 1% trimethylchlorosilane was used
as a catalyst (Sigma-Aldrich, Milan, Italy) after direct silanization and cholesterol (CH).
After 30 min at 60 °C, samples were analyzed
using the GC 7890A and MS 5975 VL analyzers (Agilent Technologies, Santa Clara, California, USA).
Chromatographic separation
was performed on HP-5MS (Agilent Technologies, Santa Clara, Ca.
) capillary columns (30 m×250 μm×0.
25 μm) using helium as the carrier gas.
Oven temperature gradients are 80°C to 200°C (8°C/min) and 250°C (10°C/min).

The syringe and gas chromatography (GC)-MS transfer lines are maintained at 260 °C and 280 °C
, respectively.
The samples were scanned using electronic impact mass spectrometry (ELECTRONIC IMPACT MS) to verify the consistency
of the detected ffa by comparison with real standards and matching with library spectral data.
Quantitative analysis was performed with calibration curves for different ffa, and the results were reported
in the form of micrograms of SC protein per milligram.

Ultra-high performance liquid chromatography/electrospray ionization-MS/MS lipid analysis

CER and CHS analysis using ultra-high performance liquid chromatography (1200 HPLC liquid chromatography system; Agilent Technologies, Inc.
, Palo Alto, Calif.
, USA) performed CER and CHS analysis with a triple quadrupole mass spectrometer (6400 triple quadrupole LC/MS, Agilent Technologies
, Palo Alto, Ca.
, USA) with a galvanized spray.
Use C8 columns (Zorbax SB-C8 fast resolution HT, 1.
8 μm, 2.
1×100 mm; Agilent Technologies, Palo Alto, California, USA) and gradient separation of solvent a (water containing 0.
1% formic acid) and solvent B (methanol and 0.
1% formic acid) as follows: 0.
2 ml/min 30% B (0-1 min), 0.
2 ml/min 30-70% B (1-2.
5 min), as follows 0.
2 ml/min 70 to 80% B (2.
5 to 4 minutes), 0.
3 ml/min 80 to 90% B (4 to 8 minutes), 0.
3 ml/min 90% B (8 to 50 minutes), 0.
2 ml/min 90% B (50 to 52 minutes) and 0.
2 ml/min 90 to 30% B (52 to 60 minutes).

The column temperature is maintained at 40 °C
.
The CER is broken using helium as the collision gas and monitored by MRM in positive ion mode (see Table S1 for
parameters).
Relative quantification of analytes using class-specific internal standard d31CER16:0, results reported in the form of micrograms of SC protein per milligram
.

CHS separation
was performed using c18 columns (symmetrical, 3.
5 μm, 2.
1×100 mm1, Waters, Wilmslow, UK) and solvent a (acetonitrile-water-formic acid 20:80:0.
1, v/v/v) and solvent B (acetonitrile and 0.
1% formic acid).
Mobile phase B increases from 0 to 100% in a linear gradient within 6 min and remains at 100% until 10 min
.
Then mobile phase B is reduced from 10 to 11 min to 0% and remains at 0% until 22 min
.
The total running time is 25 minutes
.
The column temperature is maintained at 40 °C
.
ChS is lysed in negative ion mode with magnetic resonance spectrometer (MRM) and the following transitions are monitored: chs465→97 and d7 CHS (for CHS) 472→ For CHS quantification, a linear curve is generated where the ratio of the analyte standard peak area to the peak area of the internal standard to the amount of analyte standard
。 Results were reported
in the form of micrograms of SC protein per milligram.

Statistical analysis

The data was analyzed
using statistical and data analysis from the Python libraries Scipy and Statsmodels and MatLab (9.
4.
0 version R2018a; Mathworks, Natick, MA).

Students' t-test and analysis of variance and Tukey's post-mortem test are used to assess differences
in continuous variables between two or more groups, respectively.
Pearson's coefficient (R) is used to measure the correlation
between two quantitative variables.
The difference and correlation were considered statistically significant P<0.
05
.
For lipidomics analysis, the absolute values of FFA, CH, CHS, and CER (measured in micrograms or nanograms per milligram of SC protein) are converted and normalized; In addition, the significant effect of control factors on chromatogram was assessed using ASCA, taking into account the multivariate nature of the experiment (71).