Dopamine Prevents Ultraviolet B–induced Development and Progression of Premalignant Cutaneous Lesions through its D₂ Receptors

Dopamine is a member of the catecholamine family (1). In addition to its role in the nervous system, DA also regulates critical physiologic and pathologic processes in the periphery by acting through its D₁ and D₂ class of receptors present on target cells (1, 2).

VEGF-mediated angiogenesis plays a critical role in several diseases, including malignant tumors (3–5). Although DA is reported to regulate VEGF-induced angiogenesis and thereby growth of well-established human malignant tumors of the stomach, breast, and colon in preclinical animal models (6–10), its role in premalignant lesions, including that of skin, is unknown. Therefore, it would be interesting to determine whether DA-stimulated angiogenesis plays any role in the pathogenesis and progression of ultraviolet B (UVB)-induced premalignant cutaneous lesions in a well-established murine model (11, 12).

This study is aimed to identify the role of DA as a regulator of UVB-induced skin tumorigenesis and investigate whether it plays a role in regulating VEGF-induced angiogenesis in the pathogenesis and progression of premalignant cutaneous squamous cell lesions.

Primary human dermal microvascular endothelial cells (HMVEC-D; catalog no. CC-2543, Lonza) were cultured in microvascular endothelial cell growth medium-2 (catalog no. CC-3202, Lonza). These primary endothelial cells were authenticated through short tandem repeat profiling, and Mycoplasma was assessed via PCR. For in vitro experiments, HMVEC-D cells were deprived of serum and growth factors for 12 hours and were then either treated with selective DA D₂ receptor agonist quinpirole (catalog no. Q102, Sigma) or pretreated with selective DA D₂ receptor antagonist eticlopride (catalog no. E101, Sigma) followed by quinpirole treatment. Recombinant human VEGF was purchased from R&D Systems (catalog no. 293-VE, R&D Systems, MN, USA) and used as a positive control to stimulate the VEGF pathway in the HMVEC-D cells. Sarcoma 180 (S180) cells were from ATCC, and human umbilical vein endothelial cells were from ATCC.

Six- to eight-week-old male Skh-1 mice (Charles River Laboratories) were used for the study. The Institutional Animal Care and Use Committee of the Ohio State University (Columbus, OH) approved all the animal procedures. The mouse premalignant skin lesions were induced through chronic UVB irradiation. The animals were dorsally exposed to UVB thrice a week with a consistent dose of 2,240 J/m² for 16 weeks, using Phillips FS40UVB lamps (American Ultraviolet Company) fitted with Kodacel filters (Eastman Kodak) at 290–320 nm. As previously reported, the UVB dose was determined using a UVX radiometer (UVB, Inc.; ref. 11). Thereafter, mice were randomly divided into three groups: vehicle treated control; treated with selective DA D₂ receptor agonist quinpirole at a dose of 10 mg/kg/day, i.p., for 14 consecutive days; pretreated with selective DA D₂ receptor antagonist eticlopride at 10 mg/kg/day, i.p., followed by quinpirole for 14 days (13). Skin tumors of >1 mm were enumerated and measured using a caliper in each mouse before the first treatment and every alternate day throughout the treatment period. After the study, mice were euthanized, and multiple dorsal skin lesions and tumors were harvested and fixed in 10% neutral buffered formalin, as reported previously (11, 12).

Deidentified human normal and premalignant (actinic keratosis) skin tissue sections were used in accordance with recognized ethical guidelines after gaining Institutional Review Board approval from Ohio State University (Columbus, OH). Both human and mouse skin sections were at first probed with anti-VEGF antibody (catalog no. ab46154, Abcam, 1:100) or anti-CD31 antibody (catalog no. ab28364, Abcam, 1:100) or anti-Ki67 antibody (catalog no. ab15580, Abcam, 1:100) overnight at 4°C and then with respective secondary antibodies (Vector Laboratories). Images were acquired using an Axio Scope upright light microscope (Carl Zeiss, Oberkochen; ref. 8). The VEGF staining intensity was assessed using the following grading system: grade 0, no expression; grade 1, moderate or focal expression; grade 2, strong and focal expression; and grade 3, strong and diffuse expression (14). Microvessel density which is enumerated by CD31 staining, a well-established surrogate marker of angiogenesis, was determined from 10 random microscopic fields per section using ImageJ software (NIH) and subjected to quantitative analysis by two investigators in a blinded manner, as described previously (15). Ki67-positive staining cells were manually counted from 10 random microscopic fields per section, and the percentage of Ki67-positive cells was calculated (16).

Terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) staining was performed on premalignant skin lesions from vehicle or quinpirole-treated mice using the ApopTag Peroxidase In Situ Apoptosis Detection Kit (catalog no. S7100, Millipore) according to the manufacturer's instructions. Positive staining cells were manually counted from 10 random microscopic fields per section, and the percentage of TUNEL-positive cells was calculated (17).

Skin tissue sections were incubated with two unconjugated primary antibodies, anti-CD31 (catalog no. sc-1506, Santa Cruz Biotechnology) and anti-DA D₂ receptor antibody (catalog no. sc-9113, Santa Cruz Biotechnology) or anti-CD31 (catalog no. sc-1506, Santa Cruz Biotechnology) and anti-phospho-VEGF receptor-2 (p-VEGFR-2;catalog no. ab5473, Abcam) or anti-CD31 (catalog no. sc-1506, Santa Cruz Biotechnology) and anti-nidogen 2 (NID₂;catalog no. ab14513, Abcam) or anti-CD31 (catalog no. sc-1506, Santa Cruz Biotechnology) and anti-Apelin (catalog no. 11497–1-AP, Life Technologies) at 4°C overnight. All the antibodies were used at a dilution of 1:100. The corresponding fluorochrome-conjugated secondary antibodies were applied for 1 hour at room temperature on the next day, and DAPI was used to counterstain the nuclei. Slides were mounted and visualized under a confocal scanning microscope (FBV-1000; Olympus Corporation; refs. 6, 13).

Hematoxylin and eosin (H&E)-stained tumor tissue sections from mice were graded in a blinded manner by board-certified pathologists, as described previously. The tumor area was determined by multiplying the length and width of each tumor. Tumor burden per mouse was then determined from the total tumor area per mouse, which was the sum of all tumor areas from each mouse (11, 12, 18, 19).

Three tumors per UVB-radiated mouse were further histologically examined for hyperplasia, papilloma (grades 1–3) and squamous cell carcinoma. Papillomas were exophytic tumors without evidence of stromal invasion. In contrast, a more endophytic appearance was seen in squamous cell carcinoma, with stromal invasion and loss of basement membrane continuity. A grade 1 papilloma was composed primarily of epithelium without a prominent papillary pattern, a grade 2 papilloma was a well-differentiated papillary mass, and a grade 3 papilloma was similar to a grade 2 papilloma, except that a few finger-like projections of atypical cells were present at the base of the mass. Hyperplasia and papillomas of all grades were considered benign (11).

HMVEC-D cells were deprived of serum and growth factors for 12 hours and then pretreated with 1 μmol/L of quinpirole (13) or PBS for 5 minutes, followed by treatment with 10 ng/mL of VEGF for 8 hours. PBS-treated cells were used as the untreated control. RNA was isolated using the RNeasy Mini Kit (QIAGEN) following the protocol of the manufacturer. Total RNA yield and the A_260/280 and A_260/230 ratios were determined using a NanoDrop spectrophotometer (Thermo Scientific). RNA quality was assessed using BioAnalyzer 2100 (Agilent). Reverse transcription was then performed using the RT² First Strand Kit (QIAGEN) using 250 ng RNA input per sample. RT-PCR was then performed using the ABI 7900 SDS Real-Time instrument (Life Technologies) in a 384-well format. C_t values of target genes were normalized to those of ACTB and GAPDH (reference genes). Fold changes in the expression levels of Eps15 homology domain-containing protein 3 (EHD3), Protocadherin-17 (PCDH17), PR domain zinc finger protein 1 (PRDM1), and Thrombomodulin (THBD) in the Quinpirole + VEGF, PBS + VEGF, or vehicle (PBS) groups were analyzed using the online web application of QIAGEN (https://www.qiagen.com/us/shop/genes-and-pathways/data-analysis-center- overview-page/), as described previously (20).

HMVEC-D cell lysates were harvested and used for the Western blot assay to assess dopamine D₂ receptors (catalog no. sc-9113, Santa Cruz Biotechnology, 1:200). The immunoblot signals were first developed using SuperSignal chemiluminescent substrates (Thermo Scientific, USA, catalog no. 34580) and visualized and captured using the C-DiGit Blot Scanner (LI-COR Biosciences, Model# 3600; ref. 13).

Experimental data are expressed as mean ± SEM values. Student t test and ANOVA were performed to analyze data from experiments involving independent groups (21). Repeated measures ANOVA was performed to analyze data from experiments wherein data were obtained on days 0–14, as described previously (22). Differences for which P < 0.05 were considered statistically significant (23). Bonferroni correction was performed for multiple comparisons (24).

Among angiogenic factors, VEGF is a critical cytokine promoting angiogenesis (3–5). Therefore, we assessed VEGF and CD31 expression in human actinic keratosis lesions and murine cutaneous premalignant lesions induced through UVB irradiation. IHC revealed strong and diffuse VEGF expression (grade 3 staining intensity) and significant upregulation of CD31 (*, P < 0.05) in actinic keratosis lesions compared to normal skin tissues, which showed no expression of VEGF, that is, grade 0 staining (Fig. 1A–E). Similarly, strong and diffuse VEGF expression (grade 3 staining intensity) and significant upregulation of CD31 (*P < 0.05) were detected in UVB-induced premalignant skin lesions in Skh-1 mice in comparison with normal skin tissues, which showed no expression of VEGF, that is, grade 0 staining (Fig. 2A–E).

Because DA regulates VEGF-induced angiogenesis through its cognate receptors (1), we analyzed the expression of DA receptors in endothelial cells in actinic keratosis in humans and UVB-induced skin lesions in mice. Immunofluorescence results revealed that DA D₂ receptors are expressed in endothelial cells in both these lesions (Fig. 3).

The number of cutaneous tumors and the tumor burden significantly decreased in UVB-irradiated Skh-1 mice upon treatment with a selective DA D₂ receptor agonist quinpirole (10 mg/kg once daily i.p. for 14 days) in comparison with vehicle-treated controls (*P < 0.05). On the contrary, pretreatment of these animals with selective DA D₂ receptor antagonist eticlopride (10 mg/kg once daily i.p. for 14 days) abrogated the effects of quinpirole (⁺, P< 0.05), thereby further supporting that these actions were through DA D₂ receptors (Fig. 4A and B). We did not observe significant differences in the body weight between the control and treated animals.

Furthermore, semiquantitative histopathologic analysis of lesions in the vehicle-treated control mice indicated hyperplasia, grade 1–3 papilloma and squamous cell carcinoma in 13.3%, 26.5%, 23.3%, 20%, and 16.6% animals, respectively. On the contrary, we observed hyperplasia, grade 1 2 papilloma in 6.6%, 3.3% and 3.3%, respectively, in quinpirole-treated animals. Importantly, no grade 3 papilloma or cutaneous squamous cell carcinoma was seen in quinpirole-treated animals. These results demonstrate that treatment with a selective DA D₂ receptor agonist can considerably inhibit development and progression of UVB-induced premalignant squamous cell skin lesions.

Because DA D₂ receptors are expressed in the dermal endothelial cells (Fig. 3), we, therefore, examined the effect of the stimulation of these receptors on VEGF-induced angiogenesis in vivo. Skh-1 mice dorsally exposed to UVB thrice a week with consistent dosing of 2,240 J/m² for 16 weeks were treated intraperitoneally (i.p.) on week 17 with a selective DA D₂ receptor agonist quinpirole at a dose of 10 mg/kg once daily for 14 consecutive days. After treatment, that is, on day 14, our results indicate significant inhibition of CD31 (microvessel density) in the skin lesions upon the activation or stimulation of DA D₂ receptors in endothelial cells with a selective DA D₂ receptor agonist quinpirole in comparison with vehicle-treated controls (Fig. 4C and D; *, P < 0.05). However, quinpirole did not affect VEGF expression (Supplementary Fig. S1). Furthermore, quinpirole significantly inhibited epidermal proliferation in UVB-radiated animals compared with vehicle-treated controls (*, P < 0.05; Supplementary Fig. S2). We did not observe any significant differences in apoptosis following quinpirole treatment (Supplementary Fig. S3).

To investigate whether the actions of DA D₂ receptors are specifically mediated through inhibition of VEGF-mediated angiogenesis in vivo, we first examined the effects of DA D₂ receptor agonist quinpirole treatment on VEGF receptor-2 (VEGFR-2) phosphorylation, as VEGF binds VEGFR-2 receptors in vascular endothelial cells to induce angiogenesis (3, 6). We then investigated the effects of quinpirole treatment on the expression of the VEGF-dependent vascular genes NID₂ and Apelin, which are well-established surrogate markers of VEGF signaling in vivo (25). The results of our immunofluorescence colocalization experiment demonstrated significant inhibition of VEGFR-2 phosphorylation (Fig. 5A) and downregulation of NID₂ (Fig. 5B) and Apelin (Fig. 5C) in the endothelial cells of dermal precancerous tissues from quinpirole-treated mice compared with untreated controls. These results confirm that DA D₂ receptor agonist quinpirole acts specifically through inhibition of VEGF signaling in vivo.

To further determine whether the mechanisms underlying DA D₂ receptor-mediated suppression of angiogenesis (Fig. 4C and D) and the subsequent reduction in the number of cutaneous tumors and the tumor burden in UVB-irradiated Skh-1 mice (Fig. 4A and B) are occurring selectively through the inhibition of VEGF functions, dermal microvascular endothelial cells (HMVEC-D) were pretreated with either 1 μmol/L of a selective DA D₂ receptor agonist quinpirole or PBS for 5 minutes, followed by treatment with 10 ng/mL of VEGF for 8 hours in vitro. RT-PCR analysis revealed significant inhibition of VEGF-dependent proangiogenic genes, namely, EHD3, PRDM1, PCDH17, and THBD in the quinpirole + VEGF HMVEC-D group compared with the PBS + VEGF- HMVEC-D controls (Fig. 6A; *, P < 0.005; ⁺, P < 0.005; ref. 25). As our Western blot data demonstrated the presence of DA D₂ receptors in the HMVEC-D (Fig. 6B), these results confirmed that the inhibition of VEGF-dependent proangiogenic genes in these cells was through the activation of DA D₂ receptors.

Together, these results indicate that treatment with selective DA D₂ receptor agonist significantly reduces UVB-induced number of tumors and the tumor burden, including the development of malignant squamous cell cancer through the suppression of VEGF-induced angiogenesis.

This study is the first, to our knowledge, to report that DA acts as an angiogenic switch to inhibit the development and progression of UVB-induced skin lesions. Activation of DA D₂ receptors downregulated VEGF-dependent proangiogenic genes both in vivo and in vitro (25). These actions of DA through its D₂ receptors are novel and have not been previously reported (Figs. 5 and 6A).

Nonmelanoma skin cancer, which includes squamous cell carcinoma (SCC) of the skin, is commonly seen in the world and ultraviolet radiation, particularly its UVB component, increases cancer risk (26–28). The currently available treatment often does not give a complete cure, as the recurrence rate is high, causing a significant burden to the economy (29, 30). These limitations underscore the development of an alternative treatment approach. In this regard, DA D₂ receptor agonist-mediated inhibition of premalignant squamous cell skin lesion progression through suppression of VEGF-induced angiogenesis could be a suitable approach to prevent SCC of the skin.

Furthermore, and importantly, unlike the currently used anti-VEGF agents, DAD₂ receptor agonists do not cause hypertension, a significant side effect of currently used anti-VEGF agents due to their compensatory effects on the cardiovascular system (31). They are inexpensive and have also been used to treat other disorders for several years, with well-established safety profiles (31–33). Therefore, DA D₂ receptor agonists can be used in clinical trials to prevent and treat UVB-induced squamous cell premalignant cutaneous lesions, especially at sites where surgery can cause severe disfiguration.

The authors declare no competing interests exist.

K. Lu: Data curation, formal analysis, validation, investigation, visualization, methodology, writing–review and editing. M. Bhat: Data curation, investigation, visualization, methodology. S. Peters: Data curation, validation, investigation, visualization, methodology, writing–review and editing, histology. R. Mitra: Data curation, validation, investigation, visualization, methodology, writing–review and editing, histology. X. Mo: Data curation, validation, investigation, visualization, methodology, writing–review and editing, statistical analysis. T.M. Oberyszyn: Methodology, writing–review and editing. P.S. Dasgupta: Formal analysis, methodology, writing–review and editing. S. Basu: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

This research was supported by DOD, USA Grant (W81XWH-12–1-0443; NIH, USA grants R01CA169158 and R01HL131405; to S. Basu).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.