Transcriptome-wide analysis of blood vessels laser captured from human skin and chronic wound-edge tissue

Abstract

Chronic wounds represent a substantial public health problem. The development of tools that would enable sophisticated scrutiny of clinical wound tissue material is highly desirable. This work presents evidence enabling rapid specific identification and laser capture of blood vessels from human tissue in a manner which lends itself to successful high-density (U133A) microarray analysis. Such screening of transcriptome followed by real-time PCR and immunohistochemical verification of candidate genes and their corresponding products were performed by using 3 mm biopsies. Of the 18,400 transcripts and variants screened, a focused set of 53 up-regulated and 24 down-regulated genes were noted in wound-derived blood vessels compared with blood vessels from intact human skin. The mean abundance of periostin in wound-site blood vessels was 96-fold higher. Periostin is known to be induced in response to vascular injury and its expression is associated with smooth muscle cell differentiation in vitro and promotes cell migration. Forty-fold higher expression of heparan sulfate 6-O-endosulfatase1 (Sulf1) was noted in wound-site vessels. Sulf1 has been recently recognized to be anti-angiogenic. During embryonic vasculogenesis, CD24 expression is down-regulated in human embryonic stem cells. Wound-site vessels had lower CD24 expression. The findings of this work provide a unique opportunity to appreciate the striking contrast in the transcriptome composition in blood vessels collected from the intact skin and from the wound-edge tissue. Sets of genes with known vascular functions but never connected to wound healing were identified to be differentially expressed in wound-derived blood vessels paving the way for innovative clinically relevant hypotheses.

Fig. 1.

Rapid identification of blood vessel elements from human wound tissue. (a) Human wound sections stained with UEA I lectin (green) and anti-human VWF antibody (red) with DAPI nuclear stain (blue). (i) UEA I lectin and DAPI. (ii) VWF and DAPI. (iii) Colocalization of UEA I lectin with VWF. (iv) UEA lectin, VWF and DAPI. (Scale bars: 20 μm.) (b) Human wound section stained with UEA I lectin. (i) A vessel is identified and marked for capture and catapult (shown with a white arrow). (ii) Laser-assisted cutting and separation of the identified vessel. The vessel is ready to be catapulted. (iii) The tissue section after the cut vessel has been catapulted. (iv) Isolated vessel captured in chaotropic solution.

Fig. 2.

Optimization of vessel staining protocol: fixation and dehydration. (a) Human wound sections were subjected to standard fixation methods, i.e., RNALater, acetone, formalin (neutral buffered formalin), or ethanol (95% vol/vol ethanol) as shown. After fixation, sections were stained with UEA I lectin (green). The bar graph shows relative β-actin mRNA levels quantified by using real-time PCR. RNA was extracted from 400,000 μm² of vessel elements captured after laser microdissection after specified fixation and UEA I lectin staining. *, P < 0.05 lower compared with the RNALater treated group. (b) UEA I stained dehydrated versus nondehydrated wound tissue sections. The bar graph shows relative β-actin mRNA levels quantified from tissue elements captured from RNALater fixed and dehydrated or nondehydrated. *, P < 0.05 lower compared with the dehydrated group. (c) Stability of β-actin transcript in tissue sections as a function of time after RNALater fixation, staining, and dehydration. Vessel area (400,000 μm²) was processed, and β-actin expression was quantified as described. *, P < 0.05 lower compared with 5-min group. (Scale bars: 200 μm.)

Fig. 3.

Characterization of captured vessel and nonvessel elements. (a) Images of vessel and nonvessel elements are marked by using the LMPC system. (b) Quantification of vessel-specific mRNA from 400,000 μm² of vessel elements (VE) or nonvessel elements (non-VE) captured following laser microdissection from RNALater treated, UEA I stained and dehydrated human wound tissue. RNA was extracted from the captured tissue, amplified, and reverse transcribed into cDNA. Gene expression was quantified by using real-time PCR (normalized to β-actin). *, P < 0.05 compared with vessel element. SMA, smooth muscle actin; KRT14, keratin 14; VIM, vimentin.

Fig. 4.

Hierarchical cluster images illustrating genes up- or down-regulated in vessels from human wound edge compared with that from intact skin. Gene expression data obtained by using GeneChip were subjected to t test analysis with false discovery rate correction (see SI Fig. 8). For comparative visualization, those genes that were expressed at significantly different levels in the wound-edge skin versus the intact skin were subjected to hierarchical clustering as shown. Line graphs at Right illustrate the average pattern of gene expression in the corresponding cluster. Red to green gradation in color represent higher to lower expression signal. A scale representing fold change is indicated at the bottom.

Fig. 5.

Real-time PCR validation of GeneChip microarray expression analysis. Expression levels of selected genes identified by using GeneChip analysis were independently determined by using real-time PCR. For comparison, the real-time PCR data (normalized to β-actin, a housekeeping gene) were proportionately adjusted to fit to the scale with GeneChip expression values (normalized by using global scaling approach). *, P < 0.05 compared with corresponding wound vessels. Solid bars, wound-edge skin vessel elements; open bars, intact skin vessel elements. Shown are wound-edge skin vessels versus vessels in intact skin for up-regulated genes (a) and down-regulated genes (b). Sulf1, sulfatase 1; COL4A1, collagen 4A1; HLA-DRB4, major histocompatibility complex, class II, DR beta 4; THY1, Thy-1 T cell antigen; COL5A2, collagen 5A2; COL3A1, collagen 3A1; TIMP1, tissue inhibitor of metalloproteinases; CYR61, cysteine-rich, angiogenic inducer, 61; MMP1, matrix metalloproteinase-1; ANGPT2, angiopoietin 2; KRT14, keratin 14; SOX9, SRY (sex determining region Y)-box 9; KRT15, keratin 15; PROM1, CD133 antigen, or prominin 1.

Fig. 6.

Immunofluorescent staining of differentially expressed proteins (identified by using microarray analysis) in blood vessels from human wound edge and intact skin tissue. Human wound and skin tissue sections were stained with UEA I lectin (green) and coimmunostained with Cyr61 (red) (a), MMP1 (red) (b), ANGPT2 (red) (c), or Thy1 (red) (d) as shown. (Scale bars: 50 μm.) CYR61, cysteine-rich angiogenic inducer 61; MMP1, matrix metalloproteinase-1; ANGPT2, angiopoietin 2; THY1, Thy-1 T cell antigen.