PNAd-expressing vessels characterize the dermis of CD3+ T-cell-mediated cutaneous diseases

Abstract

T-cell recruitment to skin tissues is essential for inflammation in different cutaneous diseases; however, the mechanisms by which these T cells access the skin remain unclear. High endothelial venules expressing peripheral node address in (PNAd), an L-selectin ligand, are located in secondary lymphoid organs and are responsible for increasing T-cell influx into the lymphoid tissues. They are also found in non-lymphoid tissues during inflammation. However, their presence in different common inflammatory cutaneous diseases and their correlation with T-cell infiltration remain unclear. Herein, we explored the mechanisms underlying the access of T cells to the skin by investigating the presence of PNAd-expressing vessels in different cutaneous diseases, and its correlation with T cells’ presence. Skin sections of 43 patients with different diseases were subjected to immunohistochemical and immunofluorescence staining to examine the presence of PNAd-expressing vessels in the dermis. The correlation of the percentage of these vessels in the dermis of these patients with the severity/grade of CD3+ T-cell infiltration was assessed. PNAd-expressing vessels were commonly found in the skin of patients with different inflammatory diseases. A high percentage of these vessels in the dermis was associated with increased severity of CD3+ T-cell infiltration (P < 0.05). Additionally, CD3+ T cells were found both around the PNAd-expressing vessels and within the vessel lumen. PNAd-expressing vessels in cutaneous inflammatory diseases, characterized by CD3+ T-cell infiltration, could be a crucial entry point for T cells into the skin. Thus, selective targeting of these vessels could be beneficial in cutaneous inflammatory disease treatment.

In our study, we observed that PNAd-expressing vessels, identified as a crucial gateway for CD3+ T-cell entry into secondary lymph nodes, are a hallmark feature in the dermis across various CD3+ T-cell-mediated cutaneous diseases. This observation correlates with the extent of CD3+ T-cell infiltration within these conditions, suggesting that PNAd-expressing vessels may serve as a critical conduit for CD3+ T cells into the skin.

Graphical Abstract

Graphical Abstract.

Introduction

Inflammatory cutaneous diseases are a group of immune-mediated skin conditions characterized by increased levels of pro-inflammatory cytokines and chemokines [1] and histopathological infiltration of immunological cells, with T cells playing a major role in the adaptive immune system in these conditions [2].

Disorders dominated by T cells represent the largest group within chronic immune-mediated skin diseases [3], and clinical observations and studies have shown the essential role of these cells and their secreted cytokines in promoting inflammation [4, 5]. In psoriasis, for instance, which is a very common T-cell-mediated disease [6], the T cells secrete mediators and molecules (IL-17, IL-22, and TNF-α) in the disease lesion, leading to impaired keratinocyte cornification and disease development [7, 8]. Moreover, their effect on dermal endothelium cells (ECs) enhances T-cell infiltration from the bloodstream into the psoriatic lesion, resulting in disease progression [9]. In atopic dermatitis (AD), a common skin disease [10], the T cells through their secreted cytokines interfere with keratinocyte terminal differentiation, impairing the barrier function, a characteristic of AD [11]. These promote the production of IgE, which is strongly correlated with the severity of AD [12].

For T cells to maintain inflammation, they should be transmitted from the periphery to the dermis through their interaction with the vascular bed [13] and angiogenesis can promote and maintain immune cell influx during disease progression [14].

The role of high endothelial venules (HEVs), which are specialized blood vessels (BVs) located typically in secondary lymphoid organs (SLOs) [15] but can be formed in non-lymphoid structures under inflammatory conditions [16], in the amplification and maintenance of local immune responses in cutaneous diseases has gained research attention [17].

HEVs in SLOs are distinguished from other BVs by their characteristic plump, cuboidal EC morphology [18], and increased peripheral node addressin (PNAd) expression; PNAd is the ligand of L-selectin/CD62L (the classic homing receptor of naïve and central memory T- and B-lymphocytes which regulate entry into lymph nodes [LNs]) [19]. Under inflammatory conditions, these vessels can even recruit lymphocytes of activated/effector types that do not express L-selectin/CD62L [20], resulting in the generation of a structural and functional unit that supports enhanced lymphocyte recruitment from the bloodstream into lymphoid tissues [21].

When formed in non-lymphoid tissues, these BVs may resemble structurally distinct HEVs as in LNs [22] or may present as immature HEV-containing structures lined by flat ECs [17]. Moreover, they may be a part of an organized lymphoid aggregate (tertiary lymphoid organs [TLOs], which are similar to SLOs in cellular composition, structure, and chemokine production [15]), or may be found in the absence of histologically distinct TLOs [23]. Regardless of these structural differences, these PNAd-expressing vessels are the presumed entry sites of blood-borne B- and T-lymphocytes that express L-selectin and activated/effector type [22, 24].

PNAd-expressing vessels have been detected in TLO structures in the dermis of patients with various inflammatory diseases, such as syphilis and lupus erythematosus profundus [25]. However, in other common diseases that do not necessarily exhibit TLO elements in their dermis and are caused by the infiltration of T cells (CD3+ cells) [3], the presence of these vessels has not been extensively demonstrated despite their importance in T-cell trafficking. Therefore, we hypothesized that in common cutaneous inflammatory diseases such as psoriasis and AD, PNAd-expressing vessels may be formed, and along with regular BVs, they can maintain the inflammatory state through their interaction with T cells. We also extended our sample collection to involve other common T-cell-mediated inflammatory skin conditions [4] and accordingly, the presence of PNAd-expressing vessels was investigated and was found to be correlated with the severity/grading of CD3+ T-cell infiltrating into the dermis. The topographic arrangement of these vessels with CD3+ T cells was also studied.

Materials and methods

Ethics approval

This study was approved by the Ethics Committee of Kyoto University Hospital and the Institutional Review Board of Kyoto University Hospital. Informed consent was obtained from all participants as per the Declaration of Helsinki.

Patients

Skin sections of 43 patients were used in the study. The inclusion criteria were as follows: patients with a clinical history and those who presented with a clinical manifestation of different inflammatory cutaneous diseases. After collecting clinical history and performing clinical examinations of the patients, a provisional diagnosis was made. Subsequently, a punch biopsy was performed using the most active skin lesions (selected based on color, texture, and lesion type); in urticaria specifically, the sample was taken from the lesions that presented for 24 h or more, and before starting the patient on any treatment for dermatological diseases.

Formalin-fixed, paraffin-embedded, archived blocks of the biopsy specimens were used to prepare four consecutive sections for histology, hematoxylin and eosin (H&E) staining, immunohistochemistry (IHC), and immunofluorescence (IF) staining.

Patients for the diseases assessed had the following diagnosis: psoriasis, AD, lichen planus (LP), pityriasis rubra pilaris (PRP), discoid lupus erythematosus (DLE), dermatomyositis (DM), bullous pemphigoid (BP), mixed type (T- and B cell) pseudolymphoma, T-cell type pseudolymphoma, pseudolymphoma folliculitis, ordinary urticaria, insect bites, scabies, morbilliform drug eruption, and drug reaction with eosinophilia and systemic symptoms (DRESS).

Accordingly, the diseases were categorized into different inflammatory disease groups. Psoriasis, LP, AD, and PRP were categorized under the papulosquamous diseases group. DLE and DM were considered in the connective tissue diseases (CTD) group. BP cases were considered under the blistering disease group. Insect bites and scabies were categorized under the infectious disease group. Morbilliform drug eruption and DRESS were under the drug-induced disease group. Mixed-type (T- and B-cell) pseudolymphoma, T-cell type pseudolymphoma, and pseudolymphoma folliculitis were under the pseudolymphoma group, whereas cases of ordinary urticaria were retained in the urticaria group.

The diagnosis of diseases was reviewed and confirmed based on the clinical presentation, H&E staining, the presence of serum IgG targeting BP antigen (BP180) in BP cases, and direct IF for IgG and C3 deposition in the basement membrane in BP and DLE.

The exclusion criteria were as follows: any patient who was in the remission stage, those who presented a resolving lesion (e.g. post-inflammatory hyperpigmentation), and already received treatment for the dermatological inflammatory condition. The age of the patients, sex, and diagnosis of other chronic illnesses or treatments were not included in either criterion.

Four healthy skin tissue samples were obtained from patients who underwent other surgical procedures at the hospital and used as the negative control group. Four tissue samples from the tonsils and LNs of healthy individuals were used as positive controls for IHC and IF staining.

Tissue analysis

H&E staining

The first section from the archived blocks was used for H&E staining, which was performed as previously described [26]. The stained sections were examined by the same investigator under a light microscope for the severity/grade of inflammatory cell infiltration, which was represented as the density of inflammatory cells invading the skin. To determine the density of inflammatory cells, we determined the area with the highest proportion of inflammatory cells in the dermis (hot spot) by scanning the dermis at a 10× field. The inflammatory cell aggregation area was then measured (µm²) in a 20× field (Fig. 1A) and expressed as a proportion (%) of the slide area. The sections were then categorized into three grades of increasing severity (grades 1, 2, and 3) of inflammatory cell infiltration.

Figure 1.

Measurement of the severity or grade of inflammatory cell infiltration in psoriasis using H&E-stained skin sections. The severity or grade of inflammatory cell infiltration is represented by the density of inflammatory cells invading the skin. After the area with the highest proportion of inflammatory cells in the dermis (hot spot) was detected at 10×. Inflammatory cell aggregation area was measured in µm² in a 20× field and expressed as a proportion (%) of the slide area (A). Skin section of a patient with PRP stained with MECA-79 antibody (on the left) and CD31 antibody (on the right). The number of PNAd-positive and CD31-positive vessels were counted in 20× field, and the percentage of PNAd-positive vessels was calculated as the PNAd⁺/CD31⁺ vessel ratio (B).

Immunohistochemistry

Two consecutive sections were used to identify PNAd using a rat IgG anti-human MECA 79 antibody (BioLegend, San Diego, CA, USA) and BVs using a mouse IgG anti-human CD31 antibody [21] (Dako-Agilent Technologies, Santa Clara, CA, USA), respectively (Fig. 1B). The slides were subjected to heat-induced antigen retrieval using Target Retrieval Solution pH 9 (Agilent S2367; Dako, Glostrup, Denmark) for 40 min; thereafter, endogenous peroxidase activity was blocked with H₂O₂ for 10 min. Subsequently, the tissue samples were incubated with anti-human MECA-79 antibody or anti-human CD31 antibody, each diluted at 1:100 for 60 min at 24 °C, followed by incubation with Simple Stain MAX-PO (Nichirei Biosciences, Tokyo, Japan) at 1:100 dilution for 30 min at 24 °C. Finally, a premixed 3,3ʹ-diaminobenzidine (DAB) solution (ImmPACT DAB Peroxidase substrate; Vector Laboratories, San Francisco, CA, USA) was used to visualize the reactive dark brown reaction products, and the samples were counterstained with hematoxylin.

The percentage of PNAd-expressing vessels was calculated in each disease sample by first scanning the dermis area with an increased number of PNAd-positive vessels under a low-power field (10×). Next, the PNAd-positive vessels were counted under a higher power field (20×). The CD31-positive vessels were counted in the same way and same area as that of the positive PNAd-expressing vessels (Fig. 1B), and then the percentage of PNAd-positive vessels was calculated as the PNAd⁺/CD31⁺ vessel ratio.

Next, the statistical correlation between the percentage of PNAd-positive vessels and the severity/grading of inflammatory cell infiltration in each section was determined (see Statistical analysis section).

Immunofluorescence

We performed triple IF staining with CD34 antibody to detect EC [22], MECA-79 antibody, and CD3 antibody to assess the following: (a) severity/grading of CD3+ T-cell infiltration, which is represented as CD3+ T-cell density during inflammatory cell infiltration; this grading was similar to that of inflammatory cell infiltration in H&E staining, but a different microscope (confocal) type was used (A1R MP; Nikon, Tokyo, Japan); (b) the relationship of PNAd-positive vessels with CD3+ T cells with regard to their location and proximity.

Briefly, the sections were incubated with mouse anti-human CD34 antibody (Abcam, Cambridge, UK [1:500 dilution]); rat IgG anti-human MECA 79 antibody (1:75 dilution); and rabbit IgG anti-human CD3 antibody (Abcam [1:100 dilution]) overnight at 4 °C, followed by incubation with Alexa Fluor 647-conjugated anti-rabbit antibody, Alexa Fluor 555-conjugated anti-mouse antibody, and Alexa Fluor 488-conjugated anti-rat antibody (Thermo Fisher Scientific, Osaka, Japan) for 60 min at 24 °C. Images were captured using a confocal microscope (Nikon) at 20× magnification.

Statistical analysis

When calculating the severity/grade of inflammatory cell infiltration, 25% and 75% percentiles were obtained and used for cutoffs of grade 1 (≤4.38%) and 3 (≥21.10%), respectively. The values in between the cutoffs were assigned to grade 2. The median value was 9.25% (mean = 17.28%; Supplementary Table S1).

For severity/grade of CD3+ T-cell infiltration: given the strong right-hand skew in the data distribution, the 25% and 75% percentiles were obtained and these were used for cutoffs of grade 1 (≤1.73%) and grade 3 (≥12.68%), respectively. The values in between the cutoffs were assigned to grade 2. The median value was 3.16% (mean = 8.13%; Supplementary Table S2).

To compare the disease group’s PNAd+ status (vessel percentage), inflammatory severity (grade 1–3), and CD3+ severity (grade 1–3) across groups, separate unbalanced one-way ANOVA was used. In the case of a significant difference found among groups (P < 0.05 for the F-ratio; i.e. null hypothesis rejected), the means were compared pairwise for differences using Fisher’s least significant difference or Tukey’s honestly significant difference test (for further details see Supplementary Tables S3–S6).

Results

Detection of PNAd-expressing vessels in the dermis of different inflammatory cutaneous diseases

Profile of the 43 samples is summarized in Table 1. Immunohistochemical staining with MECA-79 ab showed the expression of PNAd on BVs in all inflammatory disease groups; normal skin sections did not show PNAd expression on BVs.

Table 1.

Sample profiles and MECA-79 ab staining results

Inflammatory disease groups	Number of samples	Number of samples with +PNAD expressing vessels
Papulosquamous diseases
Psoriasis	8	6
Pityriasis rubra pilaris	2	2
Atopic dermatitis	5	5
Lichen planus	6	6
Connective tissue diseases
Discoid lupus erythematosus	3	3
Dermatomyositis	2	1
Infectious diseases
Scabies	3	2
Insect bites	2	0
Drug-induced diseases
Morbilliform eruption	2	2
Drug reaction with eosinophilia and systemic symptoms	1	1
Blistering diseases
Bullous pemphigoid	2	2
Pseudolymphomas
Mixed type pseudolymphoma (T and B cells)	2	2
T-cell type pseudolymphoma	1	1
Pseudolymphoma folliculitis	1	1
Urticaria
Ordinary urticaria	3	1

In the slides having the positive PNAd-expressing BVs, they were found in the dermis of the skin and were located mainly in areas that contain inflammatory cells (Fig. 2A and B).

Figure 2.

PNAd-expressing vessels in the dermis of different inflammatory cutaneous diseases. Immunohistochemical staining of a skin section of a patient with lichen planus (A) and pseudolymphoma (B) shows the PNAd-expressing vessels in areas containing inflammatory cells.

In certain diseases, we observed the absence of PNAd-expressing vessels in the dermis; in insect bite-related diseases and some others, PNAd-expressing vessels were found in only one-third of the corresponding tested slides, such as in urticaria (Table 1). This finding could explain the role of chronicity in inducing the PNAd-expressing vessel formation.

PNAd-expressing vessel percentages in different cutaneous inflammatory disease groups

The percentage of PNAd-positive vessels among different inflammatory disease groups was highly variable, and there were clear statistical differences among them (Fig. 3A; Supplementary Table S3). In general, the pseudolymphoma group had a higher average percentage of these BVs in the dermis, followed by the CTD group, whereas the infectious disease and urticaria groups had the lowest average percentage of PNAd-expressing vessels in their dermis.

Figure 3.

PNAd-expressing vessel percentage among different inflammatory cutaneous disease groups. PNAd-expressing vessel percentages among different disease groups were highly variable; the highest percentage was found in the skin of patients in the pseudolymphoma group followed by the CTD group (P < 0.05). The middle horizontal bar in each box plot corresponds to the mean *P < 0.05.

PNAd-expressing vessels in the dermis increase with an increase in CD3+ T-cell infiltration

To test whether the PNAd-expressing vessels contribute to maintaining the inflammatory process in the different disease groups, we first assessed if the severity of inflammatory cell infiltration and CD3+ T cells’ infiltration varied among disease groups as in the case of PNAd %, and found that the severity of inflammatory cells infiltration was largely similar across the disease groups (Fig. 4A; Supplementary Table S4). Contrarily, more profound differences were found in CD3+ T-cell severity among disease groups, whereby pseudolymphoma and CTD groups showed the greatest severity due to CD3+ T cells in the dermis (Fig. 4B), same as the disease groups with higher PNAd % in the dermis (Supplementary Tables S3 and S5).

Figure 4.

Grading of inflammatory cell infiltrates and CD3+ T-cell infiltrates among different inflammatory cutaneous disease groups and their relation to the PNAd-expressing vessel percentages. The grading of inflammatory cell infiltration among different disease groups was similar (A). The grading of CD3+ T-cell infiltration among different inflammatory cutaneous disease groups was highly variable; pseudolymphoma and CTD groups showed the highest grading of CD3+ T-cell infiltration in the dermis (B). The relationship between inflammatory cell infiltration grading and PNAd-expressing vessels percentage in the tested slides showed that slides with grade 3 inflammatory cell infiltration (represented by the letter b) had higher PNAd percentage than grade 1 (represented by the letter a) but not grade 2 (represented by letters ab). Grade 2 did not show a significant difference compared to grade 1 for the PNAD percentage (C). The relationship between CD3+ T-cell infiltration grading and PNAd-expressing vessel percentage, grade 3 (represented by the letter b) shows a significantly higher percentage of PNAd vessels compared to that in both grades 1 and 2 (represented by the letter a; D). The middle horizontal bar in each box plot corresponds to the mean. *P < 0.05.

We then explored whether the PNAd % changed by the severity of inflammatory cell infiltration and CD3+ T-cell infiltration and found in the former case that PNAd % did not differ significantly between the consequent grading of inflammatory infiltration (Fig. 4C). This was in contrast in the case of CD3+ T-cell infiltration, whereby slides with grade 3 CD3+ T-cell infiltration had a significantly higher percentage of PNAd-expressing vessels than slides of grade 1 or 2 (Fig. 4D; Supplementary Table S6), indicating an association of high PNAD % with the most severe case of (grade 3) CD3+ T-cell infiltration.

PNAd-expressing vessels contained and surrounded by CD3+ T cells

While determining the topographical arrangement of the CD3+ T cells and PNAd-expressing vessels, we observed that in each tested section with positive PNAd expression, the CD3+ T-cell aggregation areas contained PNAd-expressing vessels within them (Fig. 5A–C), suggesting that these vessels are located in T-cell zones as in SLOs. Moreover, some of the PNAd-expressing vessels within these areas with CD3 cell aggregates showed that CD3+ T cells were located inside the vessels (Fig. 5D), suggesting that PNAd-positive vessels contribute to the formation of the T-cell zone and aggregation in the dermis of patients with cutaneous diseases.

Figure 5.

Topographic arrangement of PNAd-expressing vessels and CD3+ T cells. IF staining for CD34, MECA-79, and CD3 in inflammatory cutaneous disease slides showing the presence of PNAd-expressing vessels located among the CD3+ cell aggregates in DLE (A); bullous pemphigoid (B); pseudolymphoma (C); PNAd-expressing vessels showing CD3+ cells located within them, indicating that they are a site of entry to these cells into the dermis in pseudolymphoma slide (white arrow, D); PNAd-expressing vessels surrounded by CD3+ cells in psoriasis (E); CD3+ cells located merely adjacent to the areas that express PNAd in DLE (F); PNAd-expressing vessels located in areas distant from the CD3+ cell aggregates and are devoid of CD3+ cells, and 4ʹ,6-diamidino-2-phenylindole-stained cells surrounded them in PRP (G).

Additionally, some PNAd-expressing vessels were located in areas distant from CD3 aggregates but were still surrounded by CD3+ T cells (Fig. 5E), moreover, in some slides, the CD3+ T cells were merely located adjacent to areas of PNAd expression in BVs (Fig. 5F). This finding indicates that either CD3+ T cells promote PNAd expression in the BVs or PNAd expression in these BVs facilitates the presence of CD3+ T cells around BVs and the dermis in patients with these disorders.

In contrast, in two of the above-mentioned skin sections, PNAd-expressing vessels were located in areas distant from CD3+ T-cell aggregation and were devoid of CD3+ T cells and even 4ʹ,6-diamidino-2-phenylindole-stained cells surrounding them (Fig. 5G)—we called them naked PNAd-expressing vessels. However, this finding could not be explained in terms of the association of these vessels with immunological cells infiltrating the skin and, therefore, warrant further studies.

Discussion

Our study showed that PNAd-expressing vessels were a common feature in the slides of the chronic inflammatory cutaneous diseases, unlike slides of acute dermatological conditions, such as insect bites and urticaria, where these vessels were totally absent in all the tested slides. This finding could clearly explain the crucial role of chronicity in inducing and maintaining PNAd-expressing vessel formation in the dermis [16].

Interestingly, among the different disease groups, some were more permissive to the formation of these specialized vessels than others; for instance, the pseudolymphoma group had a significantly higher percentage of PNAd-expressing vessels in its dermis than other groups, and the CTD group had second highest PNAd-expressing vessel percentage. This finding could be attributed to several factors, including the effect of treatments that the patients may be receiving at the time of biopsy to test for the presence of PNAd-expressing vessels [27] and the genetic signature of the microenvironment itself that is associated with the diseases included in these groups. For instance, in SLE that is associated with DLE in 20% of the cases [28], the gene expression of TLO structures observed in the kidneys of patients is comparable to that in patients with LN [29]. However, no previous study has checked the presence of TLO structures or their gene expression in the skin of patients with SLE and DLE; it is generally checked in the kidneys. Hence, the genetic signature should be monitored when studying factors that control PNAd-expressing vessel development during chronic disease progression.

The formation of these specialized BVs may contribute to the maintenance of the inflammatory process (along with the PNAd-negative BVs) in these diseases [14]. Previous studies on non-cutaneous sites have shown that both CD3+ T cells and PNAd-expressing vessels are present in the tissues of patients with chronic inflammatory conditions [30, 31]; however, our study revealed that these specialized vessels contributed to increased T-cell invasion into the skin tissues based on the following findings: (a) these specialized vessels within lymphocyte-rich areas frequently contained luminally attached or extravasating CD3+ T cells similar to their homologs in LNs, (b) CD3+ T cells were merely located adjacent to the areas that expressed PNAd in some PNAd-expressing vessels, (c) the disease groups that were having significantly higher percentage of PNAd-expressing vessels in the dermis were the same as those showing the greatest severity of CD3+ T-cell infiltration in the dermis, and (d) the significant relationship between high grade of CD3+ T-cell infiltration and high PNAd-expressing vessel percentage.

However, the CD3+ T-cell infiltration in cutaneous diseases is not merely a PNAd-expressing vessel-dependent process, as PNAd-expressing vessel-negative slides showed CD3+ T-cell infiltration (and some with grade 2 CD3+ T-cell infiltration). Despite this, the possibility of CD3+ T cells using PNAd-expressing vessels, if they exist, to enhance their access to the skin further leading to severe CD3+ T-cell infiltration cannot be underestimated.

In contrast, PNAd-expressing vessel existence in the dermis was probably unrelated to the severity or grading of inflammatory cells invading the skin despite the potential role of these vessels in transmitting different types of immune cells into non-lymphoid tissues and vice versa; innate and adaptive immune cells are involved in PNAd-expressing vessel formation [32, 33]. This finding has several interpretations. The grading of these infiltrates in the dermis did not reflect the actual immune activity that occurred in the skin as prolonged chronic stimulation may lower the presence of some types of immune cells [34] and the opposite effects of different cells that may be present within inflammatory cell infiltrates on PNAd-expressing vessel development because some immune cells can regulate the suppressive function of PNAd expression [35]. Therefore, in many cutaneous inflammatory diseases, the density of inflammatory cells invading the skin could not be used as an indicator of the presence of PNAd-expressing vessels.

In addition to the above findings, we observed that some of the PNAd-expressing vessels were surrounded by scarce cells and located in areas distant from lymphocyte aggregates. This could be because PNAd-expressing vessels are formed before lymphocyte infiltration under the influence of transient expression of chemokines secreted by the stromal fibroblast cells. This phenomenon occurs in the tumor microenvironment [36], and thus, verifying it in chronic inflammatory conditions warrants further investigation.

This study has some limitations. This was a descriptive and observational study and we did not obtain evidence through functional experiments. Additionally, we did not study other factors that might have affected the formation of these specialized BVs for example, medications. Additionally, we did not determine whether inflammation at the time of sampling was in the initial stage or final stage; this greatly affected PNAd-expressing vessel formation and/or disappearance.

Lastly, fundamental questions remain unanswered as follows: (a) what mechanisms operate in the regulation of leukocytes that infiltrate through PNAd-positive or negative BVs to enter the affected skin in chronic diseases? (b) do CD3+ T cells accompanying PNAd-expressing vessels possess special protective or damaging functions? and (c) is a TLO-like structure in dermatological diseases a potential therapeutic target? Functional experiments to determine the extent and role of PNAd-expressing vessels in dermatological diseases can address the above questions.

Glossary

Abbreviations

AD

atopic dermatitis

BP

bullous pemphigoid

BVs

blood vessels

CTD

connective tissue diseases

DAB

diaminobenzidine

DLE

discoid lupus erythematosus

DM

dermatomyositis

DRESS

drug reaction with eosinophilia and systemic symptoms

EC

endothelium cell

H&E

hematoxylin and eosin

HEVs

high endothelial venules

IF

immunofluorescence

IHC

immunohistochemistry

LNs

lymph nodes

LP

lichen planus

PRP

pityriasis rubra pilaris

SLOs

secondary lymphoid organs

TLOs

tertiary lymphoid organs

Ethical approval

This study was approved by the Ethics Committee of Kyoto University Hospital and the Institutional Review Board of Kyoto University Hospital. Informed consent was obtained from all participants.

Conflict of interests

None declared.

Funding

This work was supported by the Advanced Research and Development Programs for Medical Innovations of the Core Research for Evolutional Science and Technology (AMED-CREST), Grant-in-Aid for Scientific Research (KAKENHI) from the Ministry of Health, Labor and Welfare (Project for Research of Intractable Disease; grant number JP20K08649), and the Takeda Science Foundation. The author was granted funds by the Japan Society for the Promotion of Science (JSPS) and this work was conducted during this period. No role of the research funder(s) or sponsor(s) in the research design, execution, analysis, interpretation, and reporting.

Data availability

The data underlying this article are available in the article and its online supplementary material.

Author Contributions

F. Budair: data curation, formal analysis, investigation, methodology, writing—original draft preparation, software. Takashi Nomura: conceptualization, supervision, validation, visualization, writing—reviewing and editing, funding acquisition, project administration, resources. M. Hirata: writing—reviewing and editing. K. Kabashima: supervision.