Combined treatment with sodium butyrate and PD153035 enhances keratinocyte differentiation

Abstract

Epidermal growth factor (EGF) receptor (EGFR) signaling is a critical determinant of keratinocyte proliferation and differentiation in both normal and diseased skin. Here we explore the effects of combined treatment with the differentiation-promoting agent sodium butyrate (SB) and the EGFR inhibitor (EGFRI) PD153035 on terminal differentiation of normal human epidermal keratinocytes (NHEKs). Cells treated with SB showed increased expression of the levels of mRNA and protein of the differentiation markers filaggrin and transglutaminase 1. Co-treatment with EGF significantly blunted these effects of SB. Combined treatment with SB and PD153035 alleviated these inhibitory actions of EGF, resulting in improved effects of decreased cell growth and increased terminal differentiation, relative to the individual treatments. These results indicate that the combined use of a differentiation-promoting agent and an EGFR inhibitor may offer an additional approach to the management of hyperproliferative skin diseases.

Background

Epidermal growth factor (EGF) receptor (EGFR) is a critical regulator of epidermal homeostasis (1). In addition to its action to enhance keratinocyte migration and proliferation, it has been shown to impede all major processes of differentiation, namely cornified envelope, lipid biosynthesis, and tight junction formation (2, 3). Levels of this receptor and its ligands are elevated in the suprabasal layers of the epidermis in inflammatory hyperproliferative skin diseases (4). Inhibitors of EGFR signaling induce keratinocyte growth arrest and promote terminal differentiation (5). However, the use of EGFR inhibitors (EGFRIs) to treat skin disease is believed to be contraindicated, as EGFRIs used to treat certain cancers often result in papular pustular or acneiform eruptions that may limit treatment, and interestingly, worsen the skin lesions in patients who also have psoriasis (1). Sodium butyrate (SB) is known to promote keratinocyte terminal differentiation (6), and the similarities of these effects with trichostatin A (TSA) suggest that this is due to its action as a histone deacetylase inhibitor (HDACI) (7). However, our work indicates that EGFR activation can inhibit calcium- or chemical-induced differentiation (2, 8), potentially limiting the benefit of differentiating-promoting drugs, and providing a rationale for the combined treatment with a differentiation-promoting agent and an EGFRI. The use of SB or another HDACI is especially supported by recent studies showing that such compounds can ameliorate the inflammatory reactions provoked by EGFR inhibition (9, 10).

Questions addressed

The aims of this study were to determine whether EGF blocked the differentiation-promoting effects of SB, and whether combined treatment with SB and PD153035 was more effective than either single treatment for the endpoints of decreased cell growth and increased terminal differentiation.

Experimental design

For RNA qRT-PCR, immunoblotting, and percentage cornified envelope (CE) formation, NHEKs were grown as previously described (8) and treated with SB (Calbiochem, San Diego, CA; 2 mM, unless otherwise indicated), PD153035 (Calbiochem, San Diego, CA; 300 nM), or EGF (Invitrogen, Carlsbad CA; 10 ng/ml), with each endpoint measured as previously described (2, 8). Levels of poly ADP-ribose polymerase (PARP) and cleaved PARP were detected by immunoblot using an anti-PARP antibody (Epitomics, Burlingame, CA). For cell proliferation assays, growth was measured every 24 h by MTT assay (Sigma-Aldrich, St Louis, MO). Statistical analyses were performed using one-way (qRT-PCR, immunoblotting, CE assay) or two-way (MTT) ANOVA with Tukey’s (one-way ANOVA) or Bonferroni (two-way ANOVA) multiple comparisons test.

Results

In NHEKs, treatment with SB significantly increased the levels of the mRNA of the differentiation markers filaggrin (FLG) (Fig. 1a, left; 37-fold, 171-fold) and transglutaminase-1 (TGM1) (Fig. 1a, right; 6-fold, 3-fold) in each of two donors, respectively. Co-treatment with EGF significantly blunted these effects of SB by 77–97% (FLG) and 38–73% (TGM1) (Fig. 1a). Similarly for protein, SB significantly increased the levels of profilaggrin (ProFLG) (Fig. 1b, left; 7-fold, 31-fold) and TGM1 (Fig. 1b, right; 2-fold, 4-fold) in each of two donors, respectively. Co-treatment with EGF significantly blocked these effects of SB by 71–88% (FLG) and 34–98% (TGM1) (Fig. 1b). These results demonstrate the action of EGF to abrogate SB-mediated enhancement of keratinocyte differentiation.

Figure 1.

Sodium butyrate enhances the expression of markers of keratinocyte differentiation, while co-treatment with EGF abrogates this effect. (a) mRNA expression of FLG and TGM1 relative to the vehicle control (- EGF, -SB) after 48 h of treatment, indicated below the x-axis, for donor 1 (D1) or donor 2 (D2). Plotted is the mean ± SD, n=3. (b) Upper panels, immunoblot analysis of ProFLG and TGM1 protein levels, expressed as relative density units, after 120 h of treatment, indicated below the x-axis. Relative density ratios of proFLG or TGM1, normalized for loading with β-actin, were plotted relative to the vehicle control for each protein and each donor, set to 1. Plotted is the mean ± SD, n=3. The * indicates a significant difference (P<0.05) compared to the vehicle control; the † indicates a significant difference (P<0.05) compared to the SB-treated group. Lower panel, representative immunoblot.

In order to determine whether an EGFRI alleviated the inhibitory action of EGF on the SB-mediated promotion of differentiation, the individual and combined effects of SB, EGF and PD153035 were measured. In untreated NHEKs plated at low cell density, the cells doubled about every 24–36 h of culture. By 48 h, SB, PD153035, or SB plus PD153035 resulted in a significant decrease in cell number. Moreover, by 72 h, co-treatment with SB plus PD153035 resulted in a cell number significantly lower than with SB alone (Fig. 2a). Although treatment of NHEKs with 2.5 mM SB induces cell cycle withdrawal within 48 h (6), others have shown that concentrations above 3 mM significantly increase apoptosis in HaCaT cells (11). To determine whether SB caused a dose-dependent increase in apoptosis, NHEKs were treated with 0, 2, 4, or 6 mM SB for 120 h and levels of PARP and cleaved PARP were measured (Fig. 2b). No PARP cleavage was detected at any concentration of SB. However, treatment with the protein kinase inhibitor, staurosporine (STS, 1μM, 16h), clearly enhanced PARP cleavage, a marker of apoptosis (Fig. 2b). Thus, the growth inhibitory effects observed in Figure 2a most likely reflect cell cycle withdrawal and not apoptosis. Immunoblot analysis of ProFLG and TGM1 indicated that 2 mM SB was equi-effective or better than higher SB concentrations for induction of these markers of NHEK differentiation (Fig. 2c). In NHEKs from donor 1, SB (2 mM, 5 days) significantly increased the percentage CE formation, an endpoint of terminal differentiation, in comparison to untreated cells; co-treatment with EGF significantly abrogated this effect on differentiation (Fig. 2d, left, white bars). Of interest, this inhibitory action of EGF was released by the co-treatment with PD153035. Moreover, the percentage CE formation was greatest in the cells co-treated with SB and PD153035, either in the presence or absence of exogenously added EGF, indicating an enhanced effect of this co-treatment (Fig. 2d, left). In NHEKs from donor 2, SB failed to significantly increase the percentage CE formation (Fig. 2d, right); however, when cells were co-treated with SB and PD153035, a significant increase was observed in comparison to SB. Moreover, this effect of combined treatment was observed in the EGF-treated cells as well (Fig. 2d, right, gray bars). These results indicate that the combined therapy with a differentiation-promoting agent and an EGFR inhibitor may be effective even in individuals that exhibit low sensitivity to single agent treatment.

Figure 2.

Effects of sodium butyrate on keratinocyte proliferation and differentiation. (a) Time course of the effects of treatment with SB (2 mM) and/or the EGFRI PD153035 (300 nM) on the growth rates of NHEKs from donor 1 (D1) or donor 2 (D2). Plotted is the mean ± SD, n=4. * indicates significant difference (P< 0.01) compared to vehicle control; ** indicates significant difference (P<0.01) between single SB or PD153035 treatment and SB plus PD153035 combined treatment. (b) Immunoblot assessment of SB-induced apoptosis. NHEKs were treated with the indicated concentration of SB for 120 h or Staurosporine (STS, 1 μM) for 16 h and levels of PARP and cleaved PARP, a marker of apoptosis, were determined. (c) SB dose-response analysis for protein expression of ProFLG and TGM1 after 120 h of treatment. Relative density ratios of proFLG or TGM1, normalized for loading with β-actin, were plotted relative to the vehicle control for each protein and each donor, set to 1. Plotted is the mean ± SD, n=3. The * indicates a significant difference (P<0.05) compared to the vehicle control. (d) NHEK differentiation was determined as the percentage CE formation measured after 5 days of treatment, indicated below the x-axis, for donors 1 (D1) and 2 (D2). Plotted is the mean ± SD, n=4. The *, **, †, and †† indicate significant differences between the comparisons (P<0.05): * between the SB- and vehicle control-treated groups; † between the SB- and EGF-treated groups; ** between the SB- and SB plus PD153035-treated groups; †† between the SB plus EGF- and SB plus EGF plus PD153035-treated groups.

Conclusions

Our results show that SB-mediated promotion of keratinocyte differentiation is inhibited by EGF; yet combined treatment with SB and the EGFRI, PD153035, could overcome this inhibition. Moreover, this combined treatment was more effective at decreasing cell growth and increasing terminal differentiation. These results and reports showing that HDACIs can diminish the inflammatory effects of EGFRIs (9), suggest that combined treatment may benefit certain inflammatory hyperproliferative skin diseases. Studies of chemicals reported to inhibit both histone deactylases and EGFR signaling (12) are of similar interest.