1. Elevated Local Senescence in Diabetic Wound Healing Is Linked to Pathological Repair via CXCR2 
Holly N. Wilkinson, Christopher Clowes, Kayleigh L. Banyard, Paolo Matteuci, Kimberly A. Mace, Matthew J. Hardman 
Journal of Investigative Dermatology 
Volume 139, Issue 5, Pages e6 (May 2019)
DOI: /j.jid
Copyright © 2019 The Authors Terms and Conditions

2. Figure 1
SA-βGAL staining elucidates pathological accumulation of senescent cells in murine skin and wounds. (a) SA-βGAL staining illustrates increased senescent dermal cells in aged skin and in Db and aged wounds (compared with young). (b) Representative images of SA-βGAL+ (blue) cells, depicted by black arrows (n = 5–6 per group). Scale bars = 100 μm. (c–f) Further profiling of senescence markers via qRT-PCR showed an increase in (c) Cdkn1a, (d) Cdkn1b, (e) Cdkn2a, and (f) Trp53 in aged skin, whereas Cdkn1a was up-regulated only in Db wounds. n=3–4 per group. Data represent mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < D0, normal skin; D7, postinjury day 7 wound; Db, diabetic; qRT-PCR, quantitative real-time reverse transcriptase–PCR; SA-βGAL, senescence-associated β-d-galactosidase.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

3. Figure 2
Db wounds show heightened recruitment of Mφs, which are intrinsically susceptible to senescence. (a) Masson’s trichrome staining comparing Db and NDb wounds at day 7 after injury. Wound edges are indicated by black arrows. Scale bars = 500 μm. (b) Quantification of a. (c) Db wounds contain highly senescent Mφs, shown via (d) representative immunofluorescence (white arrows). Scale bar = 20 μm. n = 4–6 mice per group. (e, f) Bone marrow-derived Db Mφs showed significantly more SA-βGAL+ staining (blue cells). Scale bar = 50 μm. (g) Polarized Db Mφs had heightened Cdkn1a and Cdkn2a mRNA 72 hours after cytokine stimulation with less Nos2 and Ym1 than NDb Mφs. (h) Immunofluorescence further shows reduced polarization of Db Mφs (merged images). Scale bar = 20 μm. n = 5 animals per group. Data show mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < Db, diabetic; Mφ, macrophage; NDb, nondiabetic; SA-βGAL, senescence-associated β-d-galactosidase.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

4. Figure 3
Db Mφs produce a SASP governed by the CXCR2 receptor that alters human dermal fibroblast gene expression. (a, b) Cytokine array of CM from M1 and M2 NDb and Db Mφs (two independent experiments, representative M2 blots). CXCL1 was not expressed in NDb M2 Mφs (∞). (c) qRT-PCR of Cxcl1, Cxcl2, and Cxcr2 in Db Mφs. (d) Zymography of Mφ CM protease and quantification. (e–g) qRT-PCR for (e) Cxcl1, (f) Cxcl2, and (g) Cxcr2, performed on skin (D0) and wounds (D7) from mice. (h) Mφ CM stimulated changes in human dermal fibroblast gene expression (Db vs. NDb fold changes). n = 3 human donors per group. Fold changes above 4 and below –4 are shown with saturated color. Data show mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < CM, conditioned media; D, day; Db, diabetic; Mφ, macrophage; NDb, nondiabetic; qRT-PCR, quantitative real-time reverse transcriptase–PCR; SA-βGAL, senescence-associated β-d-galactosidase; SASP, senescence-associated secretory phenotype.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

5. Figure 4
Ectopic expression of CXCL2 induces paracrine activity that influences nuclear localization of p21. (a) HDFs transfected with a CXCL2-containing plasmid show heightened levels of CXCL2 (green fluorescence) at day 2 after transfection that are further up-regulated by day 6. (b) Quantification of a. (c, d) CXCL2-transfected fibroblasts also displayed elevated nuclear p21 (red fluorescence) by day 6 after transfection. DAPI indicates nuclei. Scale bars = 50 μm. (e) YFP-positive staining. (f, g) CXCL2-expressing HDFs also produced a SASP-rich secretome. (h) Zymography shows changes in MMP2 activity (representative of three gels). (i) Quantification of h. (j) qRT-PCR showed changes in MMP2, COL3A1, and COL1A1 between control-transfected and CXCL2-transfected HDFs. Data show mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < CTCF, corrected total cell fluorescence; D, day; FC, fold change; HDF, human dermal fibroblast; qRT-PCR, quantitative real-time reverse transcriptase–PCR; YFP, yellow fluorescent protein.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

6. Figure 5
Early-phase wound repair is accelerated by blockade of CXCR2 in human ex vivo healing. (a, c) K14 staining showed a significant increase in epidermal wound closure 3 days after treatment of human ex vivo wounds with the CXCR2 antagonist SB265610, indicated by E. Arrows show wound edges. Scale bars = 200 μm. Combined treatment with CXCL1 and CXCL2 delayed healing, indicated by D. Combining SB with CXCL2 (indicated by G), but not CXCL1 (indicated by F), rescued the delay in wound healing, but SB did not rescue healing with combinatory ligand treatment (indicated by H). (b) K6 staining showed increased neo-epidermal proliferation (scale bar = 50 μm) after (d) SB treatment. The letters in a and b relate to the treatments in c and d. Dotted lines indicate neo-epidermis. n = 3 donors per treatment. Data show mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

7. Figure 6
Blocking CXCR2 significantly improves Db wound healing in vivo. (a) Hematoxylin and eosin staining showed significantly reduced (b) wound width and (c) area after SB treatment. Arrows show wound edges. Scale bars = 500 μm. (d, e) Immune cell infiltration was dampened, quantified from (d) neutrophil and (e) Mφ staining. (f) Macrophage (Mac3) senescence (p16+ macrophages) was also reduced after SB treatment. DAPI indicates nuclei, Alexa Fluor 488 (Thermo Fisher Scientific, Paisley, UK) indicates Mac3; and Alexa Fluor 647 indicates p16. (g, h) Representative images. Scale bar = 50 μm. Wounds collected at day 7 after injury. Arrows show positively stained cells. Data show mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < Db, diabetic; Mφ, macrophage.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

8. Supplementary Figure S1
Aged skin and wounds have high levels of p16. (a) Immunohistochemistry or p16 staining in normal skin (D0) and wounds at day 7 after injury (D7) from young, diabetic (Db), and aged mice. (b) Representative images show p16+ cells (brown, black arrows). Scale bars = 50 μm. Data show mean ± standard error of the mean. ∗P < 0.05, ∗∗P < 0.01.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

9. Supplementary Figure S2
Original zymography gel from Figure 3. Left to right: lane 1, molecular weight ladder; lane 3, NDb M0 CM; lane 4, NDb M1 CM; lane 5, NDb M2 CM; lane 6, Db M0 CM; lane 7, Db M1 CM; and lane 8: Db M2 CM. CM, conditioned media; Db, diabetic; NDb, nondiabetic.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

10. Supplementary Figure S3
Ectopic expression of CXCL2 drives up-regulation of senescent markers in human fibroblasts. (a, b) CXCL2-transfected fibroblasts show increased p16 (arrows, red; nuclei, blue) at day 6 after transfection. Scale bars = 50 μm. (c, d) SA-βGAL staining (arrows) is also increased in CXCL2-transfected fibroblasts at days 2 and 6. Scale bars = 100 μm. Data show mean ± standard error of the mean. ∗∗∗P < CTCF, corrected total cell fluorescence; D, day; SA-βGAL, senescence-associated β-d-galactosidase.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

11. Supplementary Figure S4
Original zymography gel from Figure 4. (Left to right) lane 1, molecular weight ladder; lane 2, MMP2 standard; lane 4; control day 2 CM; lane 5, control day 6 CM; lane 6, CXCL2 day 2 CM; lane 7, CXCL2 day 6 CM. CM, conditioned media.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

12. Supplementary Figure S5
CXCR2 inhibition promotes HaCaT scratch migration after 24 hours of treatment. (a) Crystal violet-stained scratches and (b) bar graph. The CXCR2 ligands CXCL1 (indicated by A) and CXCL2 (indicated by B) impaired in vitro HaCaT wound closure, whereas treatment with the CXCR2 antagonist SB (indicated by E), accelerated scratch closure (representative of three independent experiments). Furthermore, treatment with CXCL1 (indicated by F) and CXCL2 (indicated by G) with SB independently did not delay scratch closure, whereas SB failed to improve HaCaT migration when both ligands were combined (indicated by H). Letters in a relate to treatments in b. Scale bar = 250 μm. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P <
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions

13. Supplementary Figure S6
Complete representative images for human ex vivo wounding experiments. K14 and K6 images of ex vivo wounds treated with a vehicle (indicated by A), CXCR2 ligands (CXCL1 and CXCL2), or the CXCR2 antagonist SB (alone, indicated by E). Wounds were also treated with combinations of the CXCR2 ligands and SB The panel below the images explains the treatment groups labelled A–H. Arrows show wound edges (K14). Dotted lines show neo-epidermis (K6). K14 scale bars = 200 μm. K6 scale bars = 50 μm. Images are representative of three human donors.
Journal of Investigative Dermatology  , e6DOI: ( /j.jid )
Copyright © 2019 The Authors Terms and Conditions