**1. Introduction**

Head and neck cancer (HNC) is the ninth most commonly diagnosed cancer in the United States and is associated with significant morbidity, mortality, and economic loss [1,2]. Important racial disparities persist for HNC [3]. Historically, HNC was more common in Black Americans than Whites, however, the incidence rate in the latter has increased since the 1990s in parallel with human papillomavirus (HPV)—positive HNC, which is diagnosed more among White males [4]. Black Americans tend to have lower rates of HPV-associated HNC which generally present as oropharyngeal cancers and are associated with better response to treatment and improved five-year survival rates compared to HPV-negative HNC [5,6]. Overall, HNC in Black patients is associated with a much lower five-year survival rate, about 30% compared to 57% in White patients [5,6]. Disparate survival rates have

been investigated in recent years, yet consensus regarding the causative factors remains to be fully elucidated. Published reports document that Black HNC patients present with more advanced disease at initial diagnosis, experience greater delays in treatment initiation, and experience more gaps in health insurance coverage compared to Whites [7–9]. Still, our understanding of racial disparities in HNC remains incomplete. Research assessing potential biological mechanisms underlying HNC in racially diverse populations is lacking, as is information on the contribution of modifiable dietary factors to HNC incidence and progression. An inverse association between vitamin D status and cancer incidence, prognosis, and mortality has been reported in several cancer types, HNC included [10–14]. Moreover, it is well documented that Black Americans are at a higher risk for low vitamin D status, yet the link between HNC and vitamin D status has not been evaluated among Black patients [15,16].

Studies investigating vitamin D have identified a multitude of health benefits including improved calcium absorption, improved immune function, and an increase in bone density [17,18]. It has been speculated that vitamin D has potential anticarcinogenic effects through multiple biologic functions; however, clinical interventions and observational studies evaluating vitamin D have yielded mixed results across different cancers [10]. Nonetheless, a recent prospective cohort study targeting HNC in a population without race reported shows that higher vitamin D intake levels are associated with decreased risk of HNC recurrence [13]. Conversely, vitamin D deficiency has been linked to lymphatic metastasis among HNC patients [19].

The goal of this prospective pilot study was to assess vitamin D status in newly diagnosed Black and White HNC patients and to conduct in vitro studies to improve our understanding of mechanisms associated with the potential cancer inhibitory effects of vitamin D in the context of HNC. To our knowledge, there have been no studies evaluating the association between vitamin D levels and HNC in Black patients despite their increased risk of vitamin D deficiency and increased risk of HNC recurrence as well as reduced five-year survival rates compared to White patients. In the present analysis, we examine vitamin D levels and other characteristics in Black and White HNC patients treated for HNC in Miami, Florida. Further, to explore anti-cancer mechanisms of vitamin D microRNA (miR) and proteomic profiling was conducted on human HNC cancer cell lines following vitamin D treatment.

Our research findings support that Black HNC patients present at an earlier age and with reduced levels of circulating vitamin D levels compared to Whites even in sunny South Florida. In addition, none of our Black HNC patients had vitamin D levels greater than or equal to 30 ng/mL, the level associated with optimal regulation of parathyroid hormone (PTH), calcium absorption, and bone health. Parallel in vitro studies revealed that vitamin D treatment of HNC cell lines significantly altered miRs regulating genes involved in many cancer pathways (i.e., steroid biosynthesis, cell proliferation, angiogenesis, chemokine, stem cell pluripotency, MAPK, and WNT signaling). Similarly, proteomic profiling following vitamin D treatment revealed modulation of proteins with roles in HNC cancer progression, metastasis, chemoresistance, and cancer recurrence and further supporting a potential role vitamin D in targeted cancer prevention [20–23].

#### **2. Materials and Methods**

#### *2.1. Characteristics of the Head and Neck Cancer Patient Population*

Nineteen HNC patients (10 non-Hispanic Whites and 9 Blacks) were recruited from the Sylvester Cancer Center or Jackson Memorial Hospital in Miami, Florida. Patients 18 years of age or older were included if they had a new diagnosis of HNC cancer (including primary lesions of the oral cavity, excluding HPV-linked cancers of the oropharynx); primary lesions of stage I–IV were included. All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Institutional Review Board and Ethics Committee of the University of Miami (Project ID: 81110). Both male and female patients were enrolled in the study on a rolling basis. If the participant was

pregnant, suffered from other illnesses or used chronic medication known to impact vitamin D metabolism they were excluded from the study. Human subject inclusion codes were G1A, M1A, and C3A. Additional information collected included basic demographic and socioeconomic factors (i.e., age, sex, height, weight, residency, annual income, highest education level completed, risk factors, and occupational history).

#### *2.2. Dietary Intake Information and Risk Factor Characterization*

Block Dietary Data Systems screeners were used to collect data on the patients' dietary intake including fruit and vegetable consumption as well as calcium and vitamin D intake. In addition to this, physical activity and sun exposure data were collected. Smoking status and alcohol consumption data were acquired by utilizing Tobacco Product Assessment Consortium (TobPRAC) tools.

### *2.3. Blood Measures*

Blood levels of both the circulating metabolite (25-hydroxyvitamin D) and active metabolite (1,25-dihydroxyvitamin D) of vitamin D were measured from serum. Plasma was utilized to determine calcium and PTH levels. Levels of 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D were determined using a two-step process involving rapid extraction and purification and (radioimmunoassay) RIA with specific antibodies (DiaSorin, Stillwater, MN, USA). PTH levels were determined utilizing the Immulite Immunometric assay (Diagnostic Products Corp., Los Angeles, CA, USA). Quest Diagnostics (San Juan Capistrano, CA, USA and Miramar, FL, USA) analyzed the samples utilizing well validated methods as previously reported [24]. Patients were categorized on the basis of four serum concentrations informed by the literature. 25-hydroxyvitamin D: <10.0 ng/mL, 10.0–19.9 ng/mL, 20.0–29.9 ng/mL, and >30.0 ng/mL. The first cut point is the classic definition of vitamin D deficiency and the second the level at which blood PTH homeostasis is achieved, the third a level proposed to classify an insufficient state and the fourth a level postulated for optimal regulation of PTH, calcium absorption and bone density [25]. The other two blood-based markers were assessed on a continuous scale. Students T-test were used to determine if there were statistically significant differences in blood or nutrient levels between Black and White patients. Chi-Square test was utilized to determine differences between racial groups by level and a *p*-value of 0.05 was considered statistically significant.

#### *2.4. Head and Neck Cancer Cell Lines*

SCC-25 (CR-1628™) and CAL-27 (CRL-2095™) cell lines were both obtained from American Type Culture Collection (ATCC). SCC-25 cells were isolated from a 70 year old White male who had squamous cell carcinoma of the tongue at T2N1 [26]. Likewise, CAL-27 cells were isolated from squamous cell carcinoma of the tongue in a White male who was 56 years old [27]. The two cell lines were plated at 3.5E06 and 1E06, respectively, and allowed to adhere for 30 h in T-75 flasks with DMEM:F12 complete medium, 10% FBS. Cells were then treated with either 2 μM of vitamin D in the form of cholecalciferol (vitamin D3, Sigma, St Louis, MO, USA) or vehicle (dilute EtOH) and harvested 24 h post-treatment utilizing a cell scraper. Cells were placed in D-PBS (without calcium or magnesium) and centrifuged at low speed (500× *g*) to pellet. The pellets were snap frozen in liquid nitrogen and stored at −80 ◦C until they were processed for RNA isolation, proteomic profiling, or evaluation of individual vitamin D linked proteins. The concentration of Vitamin D3 utilized for the HNSCC cell line studies were based on our preliminary viability investigations as shown (Figure S1) which show the LD50 in the range or 0.5 to 2uM as well as consideration of the published literature, both in preclinical and clinical studies [28–33]. As discussed above, circulating levels of vitamin D in humans have been targeted at >30 ng/mL (75 nmol/L or 0.075uM) [33] for optimal health. However, precise dose extrapolation from cell lines to physiological relevant human concentrations of vitamin D is complicated by the fact that levels may not accurately reflect tissue or intracellular concentrations [34]. Furthermore, human intervention studies or supplementation in free-living populations frequently involve daily vitamin D intake or in some intervention studies multiple times a day, whereas the in vitro

studies herein are based on a single dose of vitamin D. Considering these factors a concentration of 2 μM was chosen for all cell-based assays.

#### *2.5. RNA Isolation and microRNA Assay*

RNA was isolated utilizing standard phenol-chloroform extraction procedures as previously described [35]. RNA quality was determined by Nanodrop using the 8000 Spectrophotometer (Thermo Scientific, Wilmington, NC, USA) and RNA integrity and presence of the small RNA fraction was determined using the Bioanalyzer 2100 capillary electrophoresis system (Agilent, Santa Clara, CA, USA). Sixty nanograms of total RNA was reverse transcribed using the human Megaplex Primer Pools A and B and the TaqMan miR reverse transcription kit (Applied Biosystems, Foster City, CA, USA) [35]. Each sample was pre-amplified for 12 cycles using human pool A and B Taqman® Megaplex™ PreAmp Primers and PreAmp Master Mix (Applied Biosystems) and the preamplification reactions diluted, combined with TaqMan ® Gene-Expression Master Mix (Applied Biosystems) divided into eight aliquots and each aliquot was added to one of the eight sample ports of the TaqMan® Array A or B (v2.0), respectively. The TaqMan® Array Human miR Card Set v2.0 (Thermo Fisher Scientific, Waltham, MA, USA) enables detection of 667 human miRs, 3 miR endogenous reference controls, and 1 miR assay not related to human as a negative control. Table S1 includes relevant miR platform and sequence information. The real-time PCR reactions were run according to the manufacturer's instructions. RealTime Statminer Software (Integromics, Philadelphia, PA, USA) was used to analyze the data. The global geometric mean of all expressed miR assays was used to normalize the data. Significantly altered miRs were determined based on the most stringent criteria with a *p*-value cutoff of 1E-05 was used to determine statistically significant miRs.

#### *2.6. MiR Gene Targets and Enrichment Analysis*

Validated gene targets from miRs significantly altered by vitamin D treatment were determined using miRTarBase (The Chinese University of Hongkong, Shenzhen, China) [36]. The miRTarBase software and database enables users to search for validated gene targets based on the miR ID. This database was chosen as it is relatively stable and miRTarBase has been shown to be broader and more comprehensive when compared to other miR target validation databases [37]. Database for Annotation, Visualization, and Integrated Discovery (DAVID, v6.8, Frederick National Laboratory for Cancer Research, Frederick, MD, USA) was utilized to analyze the validated gene targets regulated by up or down-regulated miRs. Additionally, MetaCore software (Clarviate Analytics, Philadelphia, PA, USA) was used to further explore functional analysis of common gene targets for CAL-27 and SCC-25. *p-*values are calculated for the terms in each ontology after enrichment; the terms are then tested as separate hypotheses. The resultant *q*-value illustrates the corrected-values accounting for the total terms in the ontology including the rank of each term. This provides an estimate of the Benjamini False Discovery Rate (FDR).

#### *2.7. Proteomic Profiling*

Protein resuspension was achieved in 2-D cell lysis buffer (30 mM Tris-HCl, pH 8.8, containing 7 M urea, 2 M thiourea, and 4% CHAPS). This mixture was then sonicated at 4 ◦C followed by shaking for 30 min at room temperature, centrifugation at 4 ◦C (14,000 rpm) for 30 min and supernatant collection. Bio-Rads protein assay was used to measure protein concentration. The protein lysate was further processed for global proteome profiling including CyDye labeling, running of SDS gels, gel imaging, resolving protein spots, spot digestion, and MALDI-TOF MS and TOF/TOF tandem MS/MS methods were conducted in collaboration with Applied Biomics (Hayward, CA, USA).

### *2.8. Lysate Collection and Western Blot Analysis of HNC Cancer Cells Following Vitamin D*

CAL-27 (7E5) and SCC-25 (7E5) cells were seeded in T-25 flasks (Corning, Thermofisher Scientific, Waltham, MA, USA) and adhered overnight prior to treatment with 2 μM vitamin D (Sigma Aldrich, Saint Louis, MO, USA) or vehicle (dilute ethanol) dissolved in phenol red free complete RPMI medium (Thermo Fisher Scientific, Scientific, Waltham, MA, USA). Cell lysates were harvested at 24 and 48 h post-treatment using RPPA lysis buffer (1% Triton X-100, 50 mM HEPES, pH 7.4, 150 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA, 100 mM NaF, 10 mM sodium pyrophosphate, 1 mM sodium orthovanadate, and 10% glycerol) with complete EDTA-free protease and PhosSTOP phosphatase inhibitors (Sigma Aldrich, Saint Louis, MO, USA). Protein was quantified using the DC protein assay (Bio-Rad, Hercules, CA, USA). Approximately 15 μg of protein was loaded in precast 4–20% Criterion TGX gels (Bio-Rad, Hercules, CA, USA), ran for 1 h, transferred to a PVDF membrane with the Trans-Blot® Turbo™ system (Bio-Rad, Hercules, CA, USA) for 30 min, blocked for 1 h at room temperature, incubated overnight with primary antibodies and incubated with the secondary antibody for 1 h. Images were captured via the ChemiDoc Molecular Imager and band quantification with ImageLab analysis software (both Bio-Rad, Hercules, CA, USA). Expression values were determined by chemiluminescent immunodetection and normalized to appropriate loading controls. Immunoblotting was performed using commercially available antibodies from Abcam (Cambridge, MA USA ): DAB2 (#ab33441; 1:500), Cell Signaling Technology (Danvers, MA USA ): GAPDH (#2118; 1:25,000), LsBio (Seattle, WA USA ): LRP2 (#LS-c667890; 1:250), Santa Cruz Biotechnology (Dallas, TX USA ): Vitamin D receptor (#sc-13133; 1:100), GAPDH (#sc-32233; 1:40,000), and Thermo Fisher Scientific, Scientific, Waltham, MA, USA): Vitamin D binding protein (#PA5-29082; 1:500).
