Methods

Results

To visualise and quantify immune biomarkers in formalin-fixed paraffin-embedded (FFPE) abdominoplasty skin sections, we established 4 immunofluorescence biomarker staining panels to allow analysis using multiplex immunohistochemistry. These were as follows: 1) CD3, CD4, and FoxP3 to identify CD3⁺CD4⁺ helper T cells and CD4⁺FoxP3⁺ regulatory T cells; 2) CD3, CD8, and CD69 to identify CD3⁺CD8⁺ cytotoxic CD8 T cells and CD8⁺CD69⁺ resident memory CD8 T cells; 3) HLA-DR, CD14, and neutrophil elastase to identify general antigen-presenting cells including macrophages, monocytes, and neutrophils, respectively; and 4) CD11c, CD1a, and Langerin to identify dermal dendritic cells and Langerhans cells. Sections were subsequently scanned using the Vectra 3 automated quantitative pathology imaging system and analysed with inForm software. A representative example of staining and automated phenotype and score mapping is presented in Figure 1 A (in each image, epidermis on the left, dermis on the right). Except for Langerin, which was mainly expressed in the epidermis (data not shown), the majority of immune markers were found to be expressed in the dermal layers. Thus, our analysis approach allowed multispectral and automated imaging detection and measurement of weakly expressed and overlapping biomarkers in abdominoplasty skin within a familiar digital pathology workflow.

Figure 1

A – Representative automated image analysis of multiparameter immunofluorescence panels for CD3, CD4, and FoxP3 in the skin. Phenotype map: each dot represents the phenotype assigned to a DAPI-positive cell by the custom inForm algorithm. Score map: scoring function classifies cells according to co-expression into 4 classes (double negative, single positive for marker 1 or 2, and double positive). Red – single positive CD3; Green – single positive FoxP3; Yellow – single positive CD4; Pink – double positive; Blue – double negative. B – Enumeration of immune populations across 5 patient samples as determined by inForm analysis. Each dot within a graph represents a separate biopsy. Bars represent mean + SD from 3 biopsies per patient, Kruskal-Wallis one-way ANOVA with post hoc Dunn’s multiple comparisons test, *p < 0.05. Y-axis – numbers of cells, X-axis – patient number

As shown in Figure 1 B, the inForm software was able to successfully quantitate numbers of immune cells in 5-mm abdominoplasty biopsies, even when less than 50 cells per section were present. The fold difference in average immune cell subset numbers between patients was 7.8 for CD3⁺CD4⁺ helper T cells, 5.2 for CD4⁺FoxP3⁺ regulatory T cells, 5.8 for CD3⁺CD8⁺ cytotoxic CD8 T cells, 6.1 for CD8⁺CD69⁺ resident memory CD8 T cells, 19.0 for HLA-DR⁺ antigen presenting cells, 4.7 for CD14⁺ monocytes, 4.0 for neutrophil elastase⁺ neutrophils, 6.3 for CD11c⁺ dermal dendritic cells, 2.5 for CD1a⁺ dermal dendritic cells, and 4.2 for Langerin⁺ Langerhans cells (Table I). Intra-patient variability was relatively conserved for some subsets across patients, e.g. 1- to 1.9-fold for CD14⁺ monocytes, 3.6- to 4.7-fold for CD4⁺FoxP3⁺ regulatory T cells, and 1.3- to 2.9-fold for Langerin⁺ Langerhans cells; however, it differed more markedly for other subsets, e.g. CD11c⁺ dermal dendritic cell counts varied 1.6 fold for patients 1 and 4, but 5.6 fold for patient 5.

Discussion

Abdominoplasty skin explants are increasingly used to make translational dermatological research discoveries in a broad range of disciplines including skin healing [11], compound testing [12], and the role of immune cells in disease [13, 14]. However, there is growing evidence to suggest that the skin at different anatomical sites differs in multiple immunological contexts, including in immune cell prevalence, cytokine and chemokine production, and the production of anti-microbial peptides [15]. Furthermore, it has been reported that white adipose tissue associated with abdominoplasty skin from obese individuals contains unique subsets of innate lymphoid cells and distinct subsets of dendritic cells and macrophages not found in abdominoplasty skin from lean individuals [16]. Hence, both the site of skin harvest and the abdominoplasty donor themselves may have significant impacts upon the prevalence of immune cells within abdominoplasty skin explants.

Multiplex immunohistochemistry is a relatively new platform that has been reported to enable quantitative and qualitative assessment of immune cells, including Th17 cells, in formalin-fixed paraffin embedded melanoma tissue [17, 18]. Using this approach, we established 4 staining panels that permitted the detection markers associated with helper T cells, regulatory T cells, cytotoxic CD8 T cells, resident memory CD8 T cells, HLA-DR⁺ antigen-presenting cells, monocytes, neutrophils, CD11c⁺ and CD1a⁺ dermal dendritic cells, and Langerhans cells in abdominoplasty skin explants. Our data indicate that there is considerable variation in immune cell numbers between patients and between biopsies from similar healthy skin regions of the same patient. The impact of this variation on the usefulness of abdominoplasty skin explants for research will depend on the experimental purpose and design. For example, a large influx of immune cells might be expected to enter immune-mediated diseased and/or inflamed skin [19, 20], in which case the 2- to 7-fold differences in immune cell numbers we detected in otherwise healthy skin may have less of an impact if being used for comparative controls. However, the levels of immune cell variation would certainly be impactful if, for example, the focus was on biology connected to a specific subset, such as HLA-DR⁺ cells, which varied 19-fold between patients. Although small scale, this study provides useful baseline data detailing the abundance and variance of immune cell subsets in abdominoplasty skin explants.