JNK Regulates Cell-Substrate and Cell-Cell Adhesion On Substrates of Varying Compliance

<>It is well accepted that substrate rigidity can guide the behavior of cells including spreading, migration, proliferation and differentiation in vitro (1, 2) and tissue formation and maintenance in vivo (3). Substrate rigidity also has been shown to regulate the cross talk between cell-cell and cell-substrate adhesion (4). Our group demonstrated JNK phorphorylates b-catenin and regulates adherent junction formation, a-catenin/b-catenin interactions and actin re-organization in epithelial cells (5, 6). Here we provide strong evidence supporting the role of JNK as a regulator of cell-cell vs. cell-substrate adhesion on substrates of varying compliance.      <>Our in vitro studies show that the substrate rigidity regulates JNK activity of human primary keratinocytes. The levels of p-JNK were significantly reduced in the keratinocytes within 24 hrs after seeded on soft substrates (p-JNK1, 12±7%; p-JNK2, 23±9% of plastic wells). In the cells seeded on soft substrate, E-cadherin and b-catenin, two components of adherent junction complex, co-localized at the cell-cell contact, however, the expressions of p-FAK and p-Paxillin, two of the focal adhesion markers, at cells protrusion were reduced. Keratinocytes with constitutive active JNK could not form adherent junction on the soft substrate, and the focal adhesion was still detectable. However, keratinocytes knocked down JNK1, JNK2, or JNK1/2 form adherent junction even on the rigid plastic surface. In vivo, the negative correlation (R=-0.89) between adherent junction formation and JNK activity has been demonstrated in human foreskin and wild type mouse skin tissue. Keratinocytes closer to basal layer express more p-JNK and p-c-Jun, however, the adherent junctions did not formed properly. The focal adhesion markers, p-FAK and p-paxillin, were prominently expressed in the basal layer keratinocytes of skin epidermis, which is consistent with b1 integrin subunit distribution in human skin. The JNK1^-/- mouse skin epidermis and siRNA JNK1 bioengineered skin contained fewer layers of keratinocytes (mouse skin 3 vs.7, skin equivalents 6 vs.12). And compared to wild type, the epidermis is much thinner (mouse skin 33±2%, p<0.05; skin equivalents 27±2%, p<0.05 of WT). However JNK2^-/- mouse skin and siRNA JNK2 skin equivalent contain similar layers compared to wild type, and the epidermis is thicker than WT (mouse skin:116±5%, p>0.05, skin equivalents:130±8%, p<0.05 of WT). In either JNK1 and JNK2 deficient mouse skin or bioengineered skin, adherent junction is properly formed even in the basal layer confirming our results in vitro.

In conclusion, keratinocytes on rigid substrate (in vitro) or close to dermis (in vivo) express high levels of p-JNK, which inhibits adherent junction formation. On the contrary, keratinocytes on soft substrates or at suprabasal cell layers showed low p-JNK level, lack of focal adhesions and enhanced adherent junction formation. In summary, JNK plays a role as a switch regulating the cross talk between adherent junction and focal adhesion both in vitro and in vivo models.

Acknowledgements:  This work was supported by grants from NIH (R01 EB000876) and NSF (BES-0354626) to S.T.A. We thank Dr. Roger J. Davis (UMass Medical School) for kindly providing the JNK1^-/- and JNK2^-/- mouse skin samples.

References:

1.            Buxboim A, et al. 2010 Nat Methods 7:695-697

2.            Joy A, et al. 2011 Langmuir 27:1891-1899

3.            Guo WH, et al. 2006 Biophys J 90:2213-2220

4.            Tsai J, et al.2009 Biophys J 96:L39-41

5.            Lee MH, et al. 2009 FASEB J 23:3874-3883

6.            Lee MH, et al. 2010 FASEB J 25:613-623

7.            Koria P, et al. 2006 J Invest Dermatol 126:1834-1841