Partial loss of Smad signaling during in vitroprogression of HPV16-immortalized human keratinocytes

Foreskin specimens, derived from elective routine circumcision of neonate boys, were collected in a non-identified fashion from a local hospital. The protocol for foreskin tissue collection and use (PHA #2008-10) was reviewed by the Palmetto Health (Columbia, SC) Institutional Review Board, which determined that this protocol does not constitute human subjects research because it uses non-identified discarded surgical tissue.

Normal HKc were isolated as described previously [26]. Briefly, neonatal foreskins were collected in transport medium: MCDB153-LB medium supplemented with 5% fetal bovine serum (FBS) and 10 μg/ml gentamicin. Connective tissue and fat were removed using a scalpel and the cleaned foreskin was incubated with the epidermis side facing up in 0.25% trypsin (Gibco BRL, Grand Island, NY) diluted 4:1 in complete medium (CM) for 16 to 24 h at 4°C (composition of CM is described below). Epidermis was detached from the underlying dermis using sterile forceps, chopped into small pieces and further disrupted by gentle up and down pipetting in CM. HKc were collected by centrifugation, resuspended into CM and plated in 100-mm tissue culture dishes. Cells were incubated in an atmosphere of 5% carbon dioxide and 95% air at 37°C. Media was changed 24 h post plating and every 48 h thereafter. CM consists of serum-free MCDB153-LB medium supplemented with hydrocortisone (0.2 μM), triiodothyronine (10 nM), transferrin (10 μg/ml), insulin (5 μg/ml), epidermal growth factor (EGF) (5 ng/ml), bovine pituitary extract (BPE) (35–50 μg/ml protein) and gentamicin (50 μg/ml).

Normal HKc were immortalized by transfection with a plasmid containing a dimer of the full-length HPV16 DNA sequence as described in detail previously [16, 26]. HPV16 immortalized cells lines (HKc/HPV16) were derived from four different foreskin donors and have been designated HKc/HPV16-d1, -d2, -d4 and -d5 [16, 26]. From each of the four HPV16 immortalized lines, growth factor independent cells (HKc/GFI) were selected in CM lacking EGF and BPE, referred to as growth factor depleted medium (GFDM). Furthermore, differentiation resistant cells (HKc/DR) were obtained from HKc/GFI that were selected in CM supplemented with 1.0 mM calcium chloride and 5% FBS [16]. All cell lines were routinely split 1:10 when confluent, medium was changed 24 h after passaging and every 48 h thereafter.

Cells were grown to about 70% to 90% confluency in 100-mm tissue culture dishes and washed two times with ice-cold phosphate-buffered saline (PBS). Cells were then placed on ice and lysed in 400 μl of RIPA buffer (50 mM Tris–HCl pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS) that was supplemented with protease inhibitors (complete protease inhibitor cocktail, Roche, Ltd.). Plates were scraped and lysates collected into Eppendorf tubes and mixed for 1 min. After 30 min of incubation on ice, the lysates were mixed again for 1 min and centrifuged at 14,000 × g for 20 min at 4°C. The supernatants were aliquoted and stored in a −80°C freezer until used. Samples used for the assessment of phospho-Smad2 were lysed in RIPA buffer that was supplemented with the protease inhibitor cocktail and with the phosphatase inhibitors sodium fluoride (50 mM) and sodium orthovanadate (1 mM).

Protein concentrations of whole cell lysates were determined using the BCA Protein Assay Kit (Pierce, Part No. 23227) using a microplate format and according to the manufacturer’s recommendations.

Cell lysates were mixed with 5× loading buffer; (Tris–HCl (312.5 mM; pH 6.8), SDS (5%), glycerol (50%), bromophenol blue (0.05%) and beta-mercaptoethanol (β-ME) (25%)). All samples were then adjusted to equal volumes with 1× loading buffer prepared by diluting 5× loading buffer in RIPA buffer. Samples were denatured at 100°C for 5 min and cooled on ice. Samples were run using a 5% stacking gel and resolved on a 10% SDS-polyacrylamide gel. Electrophoresis was performed at 150 V for approximately 1 h using a Mini-PROTEAN II gel apparatus (Bio-Rad Laboratories, Inc) and standard running buffer (25 mM Tris base, 192 mM glycine and 0.1% SDS). The protein standards used for molecular weight assessment were the Precision Plus Protein Standards (Bio-Rad Laboratories). After electrophoresis, the stacking gel was removed, the separating gel was equilibrated for 10 min in refrigerated transfer buffer (25 mM Tris, 192 mM glycine and 20% methanol) and a polyvinylidene fluoride (PVDF) membrane (Bio-Rad Laboratories) was soaked in 100% methanol for 30 sec and equilibrated in refrigerated transfer buffer for 10 min. Proteins were transferred to the PVDF membrane at 80 mA during 16 h using transfer buffer at 4°C. After completion of protein transfer, PVDF membranes were briefly rinsed with 3 changes of PBS containing 0.05% Tween 20 and then blocked at room temperature for 1.5 h using a solution of 5% fat-free milk in the PBS-Tween buffer. For membranes utilized for phospho-Smad2 detection the blocking buffer was supplemented with sodium fluoride (50 mM). Membranes were then incubated overnight at 4°C with primary antibody that was diluted with fresh blocking solution, except for anti-phospho-Smad2 antibody that was added to a solution of 5% bovine serum albumin (BSA) in PBS-Tween. The primary antibodies used were a mouse monoclonal anti-Smad2 diluted 1:2,000 (L16D3; Cell Signaling Technology, Inc.), a rabbit polyclonal anti-Smad3 diluted 1:1,000 (LPC3; Zymed), a mouse monoclonal anti-Smad4 diluted 1:5,000 (B-8; Santa Cruz Biotechnology, Inc.), a rabbit polyclonal anti-Smad7 diluted 1:2,000 (ab5825; Abcam, plc.), a rabbit polyclonal anti-actin diluted 1:20,000 (A2066, Sigma-Aldrich Co. LLC.) and a rabbit monoclonal anti-phospho-Smad2 diluted 1:2,000 (138D4; Cell Signaling Technology, Inc.). The rabbit monoclonal anti-phospho-Smad2 antibody specifically detects endogenous phosphorylated Smad2, with phosphates at serines 465 and 467, which are the phosphorylation target sites of the TGF-β-activated receptor kinase TGFBR1 [27]. Blots probed with anti-actin antibodies were used to confirm equal protein loading. The membranes were then washed at room temperature 5 times with 0.05% Tween 20 in PBS for 5 min per wash and then were incubated at room temperature for approximately 3 h in either an anti-mouse (Vector Laboratories, Inc. - Cat. No. PI-2000) or an anti-rabbit secondary antibody (Vector Laboratories, Inc. - Cat. No. PI-1000). The secondary antibodies were conjugated with horseradish peroxidase (HRP) and were used at a 1:10,000 dilution in 5% fat-free milk/PBS-Tween blocking buffer. A second series of 5 washes at room temperature with 0.05% Tween 20 in PBS for 5 min per wash were followed by chemiluminescence detection using ECL Western Blotting Detection Kit (Amersham Biosciences), according to the manufacturer’s instructions. Subsequently, membranes were placed on X-OMAT AR films (Eastman Kodak) that were developed after exposure. Finally, quantitation of bands was performed by densitometry using ImageJ software [28].

A stock concentration of 10 μg/ml TGF-β1 was prepared by adding 200 μl of 4 mM HCl solution containing 1 mg/ml BSA to a vial containing 2 μg of lyophilized recombinant human TGF-β1 (R&D Systems, Inc. - Cat. No. 240-B). The TGF-β1 solution was aliquoted and stored at −20°C until use. The TGF-β1 stock solution was freshly diluted 10,000-fold in growth factor depleted medium (GFDM) supplemented only with EGF, but not with BPE (since BPE contains TGF-β) to obtain the 40 pM TGF-β1 concentration utilized in the experiments.

Cells investigated for phospho-Smad2 by Western blotting were grown on 60-mm tissue culture dishes to around 50-60% confluence and then incubated for 24 h in BPE-free CM. Next, cells were treated for 0, 0.5, 1, 2, 4 and 6 h with 40 pM TGF-β1 in BPE-free CM and cell lysates prepared as described above.

Glass coverslips (12 mm) were pre-coated using a 0.01% poly-L-lysine solution (Sigma-Aldrich Co. LLC, product number P4707), according to the manufacturer recommendations, and then placed into a 24-well tissue culture plate where they were soaked overnight in media. The next day normal HKc (10,000 cells), HKc/HPV16 (10,000 cells), or HKc/DR (5,000 cells) were plated on coverslips and allowed to grow until about 50% confluence. Cells were then incubated for 24 h in BPE-free CM and then treated for 0, 5, 15, 30, and 60 min with 40 pM TGF-β1 in BPE-free CM.

Immediately after treatment, cells were rinsed 2 times with ice-cold PBS and fixed for 30 min on ice with 4% neutral paraformaldehyde in PBS. After fixation, cells were washed with PBS (3 × 5 min) and then permeabilized with a solution of Triton X-100 (0.1%) and glycine (0.05 M) in PBS (2 × 10 min). Cells were washed with PBS (2 × 2 min) and subsequently blocked for 45 min with normal goat serum (5%) and BSA (1%) diluted in PBS (blocking solution). A mouse monoclonal anti-Smad4 antibody (B-8; Santa Cruz Biotechnology, Inc.) was diluted 1:150 in 5-fold PBS-diluted blocking solution, added to the cells, and then incubated overnight at 4°C. The following day, cells were washed with PBS (3 × 5 min) and incubated at room temperature for 1 h in a secondary antibody solution, which was prepared by diluting an Alexa Fluor 488 (AF488) conjugated anti-mouse antibody 1:250 in 5-fold PBS diluted blocking solution. Cells were rinsed with PBS (2 × 5 min) and then subjected to a second round of staining for Smad3 using the same conditions. The primary antibody used was a rabbit polyclonal anti-Smad3 diluted 1:200 (LPC3; Zymed) and the secondary antibody was an anti-rabbit conjugated with cyanine 3 (Cy3) diluted 1:250. After incubation with the secondary antibody, cells were rinsed with PBS (2 × 5 min); DNA was stained with DAPI (Molecular Probes) for 15 min and rinsed again with PBS (3 × 5 min). Finally, coverslips were mounted on glass slides using a DABCO containing mounting media. The edges of the coverslips were sealed with nail-polish and allowed to air dry. Cells were imaged on a LSM Meta 510 confocal microscope (Carl Zeiss, Inc.).

Nuclear Smad3 and nuclear Smad4 were later quantified in the acquired confocal pictures using ImageJ software [28]. First, nuclear areas were determined using threshold gating in the DAPI channel. The resulting nuclear areas were then copied to both the Cy3 (Smad3) and AF-488 (Smad4) channels and fluorescence intensities, as well as corresponding areas, were quantified. Intensities for each nucleus were then corrected to its corresponding area. An average of 71 nuclei (range 56–89) was quantified per time point per cell line. Finally, absolute levels were converted to relative values within each time course, having as reference (100%) the maximum level in the time course.

All four HKc/HPV16 and their corresponding HKc/DR lines were plated in 6 well plates and transiently transfected using TransFast (Promega, Madison, WI) in triplicate wells per experimental condition with p6SBE-Luc or P6SME-Luc reporter constructs, along with pRL-SV40 Renilla luciferase (Promega). The p6SBE-Luc and p6SME-Luc constructs, which contain six intact (p6SBE) or mutated (p6SME) Smad-binding-elements (SBE) cloned into the pGL3 plasmid (Promega), were a gift of Dr. Scott Kern. The next day, cells were treated without or with 40 pM TGF-β1 (R&D Systems). Cells were harvested after 22 h of treatment and the lysate assayed for luciferase activity using the Dual Luciferase Assay System (Promega) according to manufacturer’s instructions. The Firefly luciferase values, measured in Relative Light Units (RLU), were normalized against Renilla luciferase activity to control for transfection efficiency.

We performed Western blot analysis in order to determine whether or not basal protein levels of Smad2, Smad3, Smad4 and Smad7 were comparable among foreskin keratinocytes established from different donors. Protein levels of Smad2, Smad3, Smad4, and Smad7 were comparable among the eight individuals studied (data not shown).

We have previously reported that HKc/HPV16 progressively become resistant to the antiproliferative effects of TGF-β1 during in vitro progression through HKc/GFI and HKc/DR stages [21, 24]. To determine if altered protein levels of Smad2, Smad3, Smad4 and Smad7 may be a factor leading to TGF-β resistance, we studied the steady state levels of the Smads by Western blot analysis of whole cell lysates. We assessed Smad levels in low and high passage HKc/HPV16, HKc/GFI, and HKc/DR, in four independently established HKc/HPV16 lines originating from different keratinocyte donors [16] in comparison with normal HKc. Representative results are shown in Figure 1. Smad2 and Smad3 protein levels were found not to change during progression of HKc/HPV16 (Figure 1A, 1B). We observed a consistent increase of Smad4 protein expression in HKc/HPV16, HKc/GFI, and HKc/DR compared to normal HKc (Figure 1C). Finally, we found similar levels of Smad7 protein in HKc/HPV16 and HKc/GFI compared to normal HKc, with levels of Smad7 protein decreasing slightly in HKc/DR (Figure 1D).

We next used indirect immunofluorescence microscopy to compare the nuclear accumulation of Smad3 and Smad4 in normal HKc, HKc/HPV16 and HKc/DR at various times (0 to 60 min) following TGF-β1 treatment. A representative example of the time course of the nuclear accumulation of Smad3 and Smad4 following TGF-β1 treatment is shown in Figure 2 for HKc/HPV16. Nuclear accumulation of Smad3 and Smad4 is evident as early as 5 min of TGF-β1 treatment, with marked nuclear accumulation by 15 min, and sustained nuclear localization up to 60 min (Figure 2). To quantify nuclear Smad3 and Smad4 accumulation over time in normal HKc and all four HPV16 immortalized lines, we used ImageJ software to analyze the immunofluorescence images. For comparison and normalization purposes, we set to 100% the maximum nuclear fluorescence signal obtained for Smad3 or Smad4 during the time course experiment. The intensities observed at the other time points were then expressed relative to the maximal intensity in each time course. A representative time course is shown in Figure 3: Smad3 began to accumulate into the nucleus as early as 5 min after the start of TGF-β1 treatment in normal HKc (Figure 3A, panel 1) and HKc/HPV16 (Figure 3A, panel 2). Maximal nuclear Smad3 accumulation was observed in normal HKc and HKc/HPV16 after 30 min of TGF-β1 treatment (Figure 3A, panels 1 and 2). In contrast, nuclear accumulation of Smad3 in HKc/DR was slightly delayed. Nuclear Smad3 levels remained unchanged in HKc/DR following 5 min of TGF-β1 treatment, although maximal nuclear accumulation was still observed at 30 min (Figure 3A, panel 3).

The time course of Smad4 nuclear accumulation was similar among normal HKc, HKc/HPV16, and HKc/DR, with maximal nuclear accumulation of Smad4 occurring following 30 min of TGF-β1 treatment (Figure 3B, panels 1–3).

We also investigated the kinetics of Smad2 phosphorylation after treatment with TGF-β1 in normal HKc, HKc/HPV16, and HKc/DR. Smad2 phosphorylation was assessed by Western blots of whole cell lysates from cells treated for various times (0 to 6 h) with TGF-β1. The maximum level of Smad2 phosphorylation in normal HKc and HKc/HPV16 was observed after 30 min of TGF-β1 treatment and began to decline by 60 min (Figure 4, panels A and B). In contrast, at 30 min HKc/DR had reached only 87% of maximal Smad2 phosphorylation and the peak of Smad2 phosphorylation did not occur until 1 h of TGF-β1 treatment (Figure 4, panel C). Densitometry analysis of multiple Western blots showed that these results were reproducible across six normal HKc strains examined, and four HKc/HPV16 lines with their corresponding HKc/DR lines. These results demonstrate that TGF-β1 signaling is somewhat delayed in HKc/DR compared to normal HKc and HKc/HPV16.

We compared the extent of Smad2 phosphorylation to total Smad2 protein in seven normal HKc strains derived from different donors, and four HKc/HPV16 lines with their corresponding HKc/DR following treatment with 40 pM TGF-β1 for 6 h. We observed comparable levels of Smad2 phosphorylation among the normal HKc strains, four of which are shown in Figure 5. Also, comparable levels of phospho-Smad2 between normal HKc and HKc/HPV16 were observed (Figure 5, panels A and B). In contrast, Smad2 phosphorylation was reduced in HKc/DR as compared to normal HKc and HKc/HPV16 (Figure 5, panels A and B). The levels of total Smad2 protein expressed after 6 h of TGF-β1 treatment were comparable among normal HKc, HKc/HPV16 and HKc/DR (Figure 5, panels A and B).

Finally, we explored the ability of TGF-β1 to induce the activity of a Firefly luciferase gene under the control of the 6SBE element (p6SBE-Luc). As a control, we used a plasmid structurally identical to the p6SBE-Luc, but in which all six SBEs had been mutated (p6SME-Luc). No induction of luciferase activity was detected in cells transfected with p6SME-Luc and treated with TGF-β1 (data not shown). Induction of luciferase activity was observed in all HKc/HPV16 and HKc/DR lines treated with TGF-β1 (Figure 6). However, while luciferase induction was 8- to 12-fold in HKc/HPV16, it was only 3.3- to 5.6-fold in HKc/DR.