In this study, we used two types of EGFR inhibitors, cetuximab and gefitinib, to induce EGFRIs-derived skin side effects on in vitro and ex vivo, respectively. In addition, we treated cetuximab or gefitinib with rhEGF at the same time, to determine whether EGF could improve the side effects of EGFRIs. To this end, it was necessary to confirm whether the EGFR signaling was activated when EGFR inhibitors (cetuximab and gefitinib) were treated simultaneously. So, after simultaneously treating cetuximab and rhEGF, pEGFR expression was simultaneously treated with gefitnib and rhEGF, and then AKT and PI3K phosphorylation related to EGFR signaling was examined. pEGFR expression was examined after simultaneous treating of cetuximab and rhEGF, and AKT/PI3K phosphorylation related to EGFR signaling was identified in gefitinib and rhEGF co-treated cells. As a result, as shown in Fig. 1, the expression of pEGFR was increased in the group simultaneously co-treated with cetuximab and rhEGF, compared with the cetuximab alone. In addition, EGFR expression was also increased by rhEGF in a concentration-dependent manner. Furthermore, we compared the binding affinity for human EGFR between cetuximab and rhEGF. As a result, rhEGF binds to hEGFR faster than does cetuximab. Our guess is that although cetuximab maintained binding with hEGFR for a longer time than rhEGF, EGFR phosphorylation occurred in cetuximab and rhEGF co-treated cells because of the faster binding rate of rhEGF with hEGFR than that of cetuximab. Gefitinib is one of the EGFR-TKI inhibitors, and the inhibition of EGFR phosphorylation and phosphorylation of downstream effectors AKT in cells treated with over 0.1 M concentration of gefitinib were observed [21]. Similarly, in this study, gefitinib treatment inhibited the phosphorylation of AKT and PI3K. However, in gefitinib and rhEGF co-treated cells, phosphorylated AKT and PI3K were also observed. We estimate that the lowest concentration of rhEGF (1 ng/mL) in this study exceeds the ability of TKI inhibition of 1 uM gefitinib, and further study is needed.
The skin barrier function is mainly provided by keratinocytes, and it is maintained by a tightly controlled balance between the proliferation and differentiation of keratinocytes [22]. EGFR signaling plays an important role in the final differentiation of keratinocytes by inducing the activation of Transglutaminase (TGase) via Phospholipase C-γ (PLCγ) and Protein kinase C (PKC) [23]. Also, EGF induces keratinocyte proliferation via Raf-MEK-ERK signaling pathway. In this study, in cetuximab- or gefitinib-treated tissue, low expressions of both ki-67 and filaggrin were observed. However, in rhEGF and cetuximab or gefitinib co-treated tissue, those expressions were high, and the expressions were similar to those of un-treated control tissue. These results mean that rhEGF treatment could normalize the proliferation and differentiation of keratinocytes via activating EGFR signaling pathway. Keratins are principal structural proteins in epidermis, and the primary function of keratin is structural and mechanical support. In addition, keratins also modulate the growth, adhesion, migration, and invasion of epithelial cells. So, the dysfunction or mutations of keratin proteins are associated with a remarkable variety of skin disorders, including skin blistering, and inflammatory disorders [22]. In normal skin, epidermis expresses K1 and K10 in suprabasal layers, and K5 and K14 in basal layer. Whereas, in inflammatory skin, such as atopic dermatitis and psoriasis, expression of inflammatory related-keratin, including K6, K16, and K17, are increased [24]. Terrinoni et al. reported that abnormal expression of K10 with respect to control skin was observed in the lesional skin of patient who were under epidermal erythrodermic hyperkeratosis [25]. In this study, K5 expression was inhibited by EGFRIs, and that decrement was induced by rhEGF cotreatment (Fig. 2 (c)). Whereas, the aggregation of K10 in EGFRIs-treated tissues was observed, but in rhEGF- and EGFRIs-treated tissues, aggregation of K10 was not observed (Fig. 2 (d)). These results mean that interruption of EGFR signaling by EGFRIs also affects keratin expression, and further studies of other inflammatory keratins, such as K6, K16, and K17, in the lesional skin of patient under EGFRIs-induced skin toxicity are needed.
EGFR is also implicated in regulating cell–cell contacts. Expression of claudin-3 was reduced in mice deficient for Egfr keratinocytes (EGFRdEP) epidermis, as was expression of claudin-1 in human lesional skin[26]. As we expected, we also confirmed the reduction of claudin-1, -3, and occludin expression in EGFRIs-treated tissues. Meanwhile, there were no dramatic changes of tight junction abnormal in rhEGF- and EGFRIs-treated tissues (Fig. 3). However, weak normalization of tight junction expression was observed in rhEGF and EGFRIs co-treated tissues. Therefore, these results mean that rhEGF may indirectly affect the formation of tight junctions during normalization of the proliferation and differentiation of keratinocytes.
Our previous studies reported that topical treatment of rhEGF relieved S. aureus-induced inflammation and AD-like skin lesions in Nc/Nga mice, and rhEGF treatment attenuated P. acnes-induced inflammatory responses, at least in part, through the modulation of TLR2 signaling in human epidermal keratinocytes [17, 19]. In this study, proinflammatory cytokines, including IL-1α, IL-8, and TNF-α expression, were increased by EGFRIs, and those were down-regulated by rhEGF (Fig. 4). EGFR signaling in keratinocytes affects the expression of chemokines induced by TNF-α through the ERK1/2 signaling pathway. Activated ERK1/2 has been shown to decrease chemokine mRNA stability and decrease chemokine mRNA expression, suggesting that EGFR-dependent ERK1/2 activity in keratinocytes also participates in homeostatic mechanisms that regulate dermatitis response [27].
The skin is not only a physical barrier protecting from infection, but is also an environmental niche hosting a plethora of commensal organisms. Combined with the physical epidermal barrier, antimicrobial peptides (AMPs) are major components of the active innate immune defense against invading microbes in the skin. Several reports revealed that the expression of some kinds of AMPs is under the control of EGFR. Helicobacter pylori virulence effector CagA or LPS function via EGFR signaling to either suppress or upregulate the expression of hBD3, respectively [28, 29]. Treatment of human epidermal keratinocytes with erlotinib reduced the expression of hBD3, RNase 7, and CAMP. Similarly, in egfr-deficient keratinocytes, it was observed that expression of murin β-defensin 14, the mouse homolog of hBD3, was reduced [30]. In this study, we confirmed that mRNA expression of AMPs, including hBD-2, -3, LL37, and RNase7, was inhibited by EGFRI treatment (Figs. 5 (a)–(d)). Moreover, mRNA expression of AMPs was increased in rhEGF and EGFRI co-treated cells (Figs. 5 (a)–(d)). However, only hBD-2 and − 3 protein expressions were affected by EGFRIs and rhEGF (Figs. 5 (e) and (f)). These results mean that rhEGF may be attributed to improve the innate immunity of the epidermis via inducing the expression of not cathelicidins, but defensins, especially hBD-2 and − 3, through EGFR signaling.
Clinical trial results confirmed that EGFRIs decreased the proliferation and differentiation of keratinocytes. Also, EGFRIs induced inflammatory reactions in skin. As expected, treatment of EGF ointment induced the re-epithelization and proliferation of keratinocytes, and reduced infiltrated inflammatory cytokines in dermis. These results are remarkably similar to the in vitro and ex vivo results in this study. Nevertheless, further studies are needed, because in this study, the expression of tight junction and AMPs in lesional skin was not confirmed.