Reversing the aging stromal phenotype prevents carcinoma initiation

Abstract

The accumulation of senescent stromal cells in aging tissue changes the local microenvironment from normal to a state similar to chronic inflammation. This inflammatory microenvironment can stimulate the proliferation of epithelial cells containing DNA mutations which can ultimately lead to cancer. Using geriatric skin as a model, we demonstrated that senescent fibroblasts also alter how epithelial keratinocytes respond to genotoxic stress, due to the silencing of IGF-1 expression in geriatric fibroblasts. These data indicate that in addition to promoting epithelial tumor growth, senescent fibroblasts also can promote carcinogenic initiation. We hypothesized that commonly used therapeutic stromal wounding therapies can reduce the percentage of senescent fibroblasts and consequently prevent the formation of keratinocytes proliferating with DNA mutations following acute genotoxic (UVB) stress. Sun-protected skin on the lower back of geriatric human volunteers was wounded by dermabrasion and the skin was allowed to heal for three months. In geriatric skin, we found that dermabrasion wounding decreases the proportion of senescent fibroblasts found in geriatric dermis, increases the expression of IGF-1, and restores the appropriate UVB response to epidermal keratinocytes in geriatric skin. Therefore, dermal rejuvenation therapies may play a significant role in preventing the initiation of skin cancer in geriatric patients.

Figure 1.. Senescent human fibroblasts contain markers of DNA damage response in vitro

(A) Low passage neonatal normal human fibroblasts (PD-5), stress-induced senescent fibroblasts (PD-5, H₂O₂), and replicatively senescent fibroblasts (PD-40) were stained for the presence of senescence-associated β-galactosidase activity (blue), with α-53BP1 antibodies (multiple punctate nuclear staining), or α--p21 antibodies (bar = 20 µm). (B) The percentage of senescent cells were determined for senescence-associated β-galactosidase and 53BP1 staining. 53BP1-positive cells contained at least four individual fluorescent pin-point spots per nucleus (asterisks indicate significant difference from PD-5 cells, p <0.001, two-tailed t-test).

Figure 2.. Senescent fibroblasts accumulate in geriatric dermis in vivo

Biopsies of sun-protected skin were obtained from six young adult (20-28 years old) and six geriatric (>65 years old) volunteers. (A) Sections of skin were stained with antibodies to 53BP1 and γH2AX. Positive nuclei are indicated by white arrows; dashed line specifies the location of the basement membrane; bar = 25 µm. (B) Quantitative PCR analysis of mRNA isolated from skin biopsies, normalized to actin expression. Asterisk indicate statistical significance from young adult values (IGF-1 p = 0.005, COL1 p = 0.091; two-tailed t-test). (C) The number of senescent fibroblasts (based on circular or elliptical nuclear morphology as determined using Nikon Elements Image Analysis software) was counted in the papillary dermis. Asterisk indicates statistical significance from young adult values (p = 0.001; two-tailed t-test). (D) The area of the epidermis and papillary dermis were calculated from 3mm punch biopsies using Nikon Elements image analysis software. Asterisk indicates statistical significance from young adult values (Epidermis p = 0.0577, Papillary dermis p = 0.022; two-tailed t-test).

Figure 3.. Dermabrasion restores young adult fibroblast function in geriatric dermis

A 5 cm² area of sun-protected skin on geriatric (≥65 years old) volunteers was dermabraded. After a healing period of three months, biopsies of untreated and dermabraded skin were obtained. (A) Representative H&E sections from untreated and dermabraded geriatric skin. Panels i and iv are higher magnification images of dark boxes indicated in panels ii and v. Panels iii and vi are higher magnification images of light boxes indicated in panels ii and v. Panels i and iv, bar = 10 µm; panels ii and v, bar = 50 µm; panels iii and vi, bar = 12.5 µm. (B) The area of epidermis and papillary dermis were calculated as described in Fig. 2. Asterisk indicates statistical significance from geriatric control values (Epidermis p = 0.287, Papillary dermis p = 0.013; two-tailed t-test). (C) The number of senescent fibroblasts in the papillary dermis was determined as described in Fig. 2. Asterisk indicates statistical significance from control values (p = 0.018, two-tailed t-test).

Figure 4.. Dermabrasion stimulates fibroblast replication and suppresses senescence in geriatric dermis

(A) Skin biopsies described in Fig. 3B were stained with antibodies to 53BP1 and γH2AX. Senescent nuclei are indicated by white arrows, the basement membrane is designated by a dashed grey line, bar = 25 µm. (B) The density of fibroblasts in the papillary dermis was determined using the Nikon Elements Image Analysis software. Asterisk indicates statistical significance from control values (p = 0.0048), two-tailed t-test). (C) Sections of biopsies were stained with antibodies to Ki67. The percentage of Ki67(+) fibroblasts in the papillary dermis and the percentage of Ki67(+) keratinocytes in the basal layer of the epidermis were calculated using Nikon Elements image analysis software. Asterisks indicate statistical significance from control values (papillary dermis, p = 0.039; epidermis p = 0.058, two-tailed t-test).

Figure 5.. Dermabrasion upregulates IGF-1 expression and restores the appropriate UVB response to geriatric skin

Small areas of dermabraded and untreated skin on geriatric volunteers described in Fig. 2 were irradiated with UVB (dose of 350 J/m²). Twenty-four hours post-UVB, the irradiated skin was biopsied. (A) Total mRNA was isolated and the relative level of IGF-1 expression was determined by quantitative PCR (normalized to actin expression). Asterisk indicates statistically significant differences from control values (p <0.006, two-tailed t-test). (B) Sections of biopsies were stained with antibodies to Ki67 and thymine dimers (TD). The number of dual Ki67(+):TD(+) basal keratinocytes were determined for both sets of biopsies. Asterisk indicates statistically significant differences from control values (p <0.05, two-tailed t-test).