sign in sign up
<<
Home , CDH3 (gene), Desmosome

View metadata, citation and similar papers at core.ac.uk brought to you by CORE C Michels et al. provided by Elsevier - Publisher Connector Adherens Junctions Regulate

Classical Regulate Formation

Journal of Investigative Dermatology (2009) 129, 2072–2075; doi:10.1038/jid.2009.17; published online 5 march 2009

TO THE EDITOR (Tunggal et al., 2005) and transferred for initial but not later stages of desmo- Tight intercellular adhesion is an im- to medium containing high (1.8 mM) some assembly. portant aspect of the epidermal barrier Ca2 þ concentration, to allow for cad- We next asked whether P- and requires the cooperative activity of herin-dependent adhesion. In contrast and E-cadherin cooperate in junction intercellular junctions, such as adhe- to p120ctn, b-catenin and a-catenin formation by lentivirally silencing rens junctions and desmosomes (Yin levels are stabilized by the P-cadherin in mouse E-cadko keratino- and Green, 2004). Desmosomal cad- expression of classical cadherins. cytes (E-cadko;P-cadkd#1). Of the five herins link to the Therefore, as expected, a-catenin and hairpin RNAs tested, only one effi- system, whereas classical cadherins b-catenin levels were reduced upon the ciently silenced P-cadherin (c14, Figure regulate the actin through loss of E-cadherin, whereas p120ctn S1a). Phase-contrast images revealed a their cytoplasmic binding partners, b- levels remained unchanged (Figure remarkable change in the appearance catenin, a-catenin, and p120ctn (Perez- 1a). More importantly, in the early of contacts. Although control Moreno and Fuchs, 2006). In vitro stages of contact formation, E-cadko and E-cadko cells flattened, aligned studies have indicated a specific role cells showed an increased expression their membrane, and formed intimate for E-cadherin in the establishment of of P-cadherin (Figure 1a), accompanied cell contacts after 48 hours in high not only adherens junctions but also by increased membrane recruitment Ca2 þ concentration, E-cadko/P-cadkd#1 tight junctions and desmosomes (Gum- after 2 but not 48 hours when compared cells were more rounded and failed biner et al., 1988), whereas antibody- with control (Figure 1b). No difference to form intimate cell contacts, suggest- blocking studies in keratinocytes indi- was observed in the kinetics of adhe- ing the absence of cell junctions cated a cooperative role for E-cadherin rens junction formation between con- (Figure 2a). and P-cadherin (Lewis et al., 1994). In trol and E-cadko, as judged by the A nearly complete loss of membrane vivo epidermal inactivation of a-cate- recruitment of b-catenin (Figure 1b), recruitment of a-, b-catenin, and nin in mice resulted in a loss of a-catenin, or p120ctn (Figure S1b). p120ctn was observed in the absence adherens junctions and a reduction in Despite similar of both P-cadherin and E-cadherin desmosomes, thus partially confirming kinetics, desmosome formation was (Figure 2a, Figure S1b), whereas knock- the in vitro results (Vasioukhin et al., delayed in the E-cadko keratinocytes, down of P-cadherin in control kerati- 2001). However, junctions appeared as has been observed by Lewis et al. nocytes or control hair pins in E-cadko normal upon in vivo E-cadherin dele- (1994). Two hours after the induction of keratinocytes had no apparent effect on tion in the thyroid gland (Cali et al., adhesion, 3, , catenin levels or recruitment (Figure 2a, 2007) or upon P-cadherin deletion , and were re- Figure S1b). This shows that E-cadherin (Radice et al., 1997), whereas only tight cruited to the sites of cell–cell contact, and P-cadherin cooperatively deter- junctions were disturbed after epider- whereas these retained a mine adherens junction formation in mal loss of E-cadherin (Tunggal et al., diffuse distribution in E-cadko keratinocytes. In addition, unlike con- 2005). These results raise the question keratinocytes (Figure 1c, Figure S1c). trol shRNAs, staining for desmoplakin 2 þ whether classical cadherins are dispen- However, 48 hours after Ca switch, (Figure 2a), desmoglein 1/2, plakoglo- sable for desmosome formation or membrane recruitment of desmoplakin bin, and plakophilin 3 (Figure S1c) whether the other classical cadherin in (Figure 1c), desmoglein 1/2, plakoglo- remained diffuse, with very little mem- the , P-cadherin, can func- bin, or plakophilin 3 (Figure S1c) was brane recruitment 48 hours after Ca2 þ ko tionally compensate for the loss of comparable between E-cad and con- switch. Ultrastructural analysis showed E-cadherin. Here, we assessed whether trol. Only the recruitment of desmogle- few cell contacts in the absence of E-cadherin regulates the kinetics of in 3 appeared lower in the absence of E-cadherin and P-cadherin, and these junction formation and asked whether E-cadherin (Figure 1c). The ultrastruc- lacked an electron-dense inner plaque P-cadherin and E-cadherin cooperate in tural appearance of desmosomes and had no resemblance to desmosomes facilitating desmosome formation. formed after 48 hours was unaffected (Figure 2c). Levels of desmosomal To examine whether the loss of by the absence of E-cadherin (Figure proteins—for example, plakoglobin or E-cadherin affected de novo junction 2c). In contrast, P-cadherin knockdown desmoglein 3—were comparable for kd formation, keratinocytes were isolated (P-cad ) did not affect adherens junc- control, E-cadko, and E-cadko/P-cadkd from control and epidermal-specific tion or desmosome formation (data not cells (Figure 1a and Figure S1d), sug- ko E-cadherin knockout (E-cad ) mice shown). Thus, E-cadherin is necessary gesting that the lack of desmosomes cannot be explained by reduced des- Abbreviations: E-cad ko, E-cadherin knockout; P-cad kd, P-cadherin knockdown mosomal protein levels.

2072 Journal of Investigative Dermatology (2009), Volume 129 C Michels et al. Adherens Junctions Regulate Desmosomes

Hours high Ca2+ Control Eko+c16 Eko/Pkd#1 0248 Hours high Ca2+ 0 2 48 0 2 48 0 2 48 E-cadherin Control P-cadherin P-cadherin β-Catenin ko α-Catenin E

p120ctn

Actin Control Desmoglein 1/2 β-Catenin Plakoglobin Eko Actin

Hours high Ca2+ 0 2 48 0 2 48

Control

Eko

Desmoglein 3 Desmoplakin

Figure 1. Adherens junction and desmosome formation in primary keratinocytes lacking E-cadherin. (a) Western blot analysis of adherens junction and desmosomal proteins in control, E-cadko (Eko), and E-cadko/P-cadkd (Eko/Pkd#1) keratinocytes. (b) Immunofluorescence analysis of the kinetics of adherens junction formation in control and E-cadko (Eko) keratinocytes using antibodies against P-cadherin and b-catenin. The cells were cultured in high-Ca2 þ medium for the indicated time points. (c) Immunofluorescence analysis of the kinetics of desmosome formation in primary keratinocytes using antibodies against ko 2 þ 2 þ desmoplakin and desmoglein 3. For all experiments, control and E-cad keratinocytes were switched from medium with 0.05 mM Ca (low Ca ) to 1.8 mM Ca2 þ medium (high Ca2 þ ) and further incubated for the indicated time points to allow for cadherin-mediated adhesion and for the formation of junctions. Bars ¼ 15 mm.

Importantly, either E-cadherin or manner (Gottardi et al., 2001; Wong mechanisms could involve the regulation P-cadherin re-expression, but not GFP and Gumbiner, 2003). However, lenti- of cell surface stability of desmosomal as a control, was able to rescue b-catenin, viral expression of a construct similar to cadherins, vesicle transport of desmoso- desmoplakin, and plakophilin 3 recruit- the one used in these papers, the mal components, or cytoskeletal rearran- ment to cell–cell contacts (Figure 2d and E-cadherin cytoplasmic domain fused gements necessary to form stable e), indicating that classical cadherin to the IL-2R-receptor extracellular and adhesive structures (Yin and Green, levels, but not specificity, regulate adhe- transmembrane domains (IL-2R tail), 2004) (Perez-Moreno and Fuchs, 2006). rens and desmosome formation. This was did not rescue desmosome formation, In conclusion, together with pre- further confirmed using independently even though b-catenin was associated vious findings, our results show that (i) transduced and selected primary E-cadko with the membrane and was upregu- classical cadherins are important in the keratinocytes that showed a less-efficient lated to levels similar to those in cells initiation of intercellular junction for- downregulation of P-cadherin (E-cadko/ that were rescued by E-cadherin or mation and (ii) the regulation of desmo- P-cadkd#2). These cells still had some P-cadherin transduction (Figure 2d and some assembly by classical cadherins membrane recruitment of b-catenin and e and data not shown). Thus, unlike depends on levels but not on a specific desmoplakin, although the junctions growth and invasion, desmosome function of E-cadherin. These findings showed a more immature appearance formation appears to require adhesive are clinically relevant in light of recent (Figure 2a and b). Thus, as shown very interactions. observations that E-cadherin is a target recently in a similar study (Tinkle et al., An important question that remains is of and that 2008), classical cadherin levels deter- why, despite their intrinsic capacity to P-cadherin is mutated hypotrichosis mine the extent of adherens junction mediate Ca2 þ -dependent adhesion, des- associated with juvenile macular dys- and desmosome formation. mosomal cadherins are unable to estab- trophy (Sprecher et al., 2001). Given E-cadherin suppresses growth and lish productive adhesive structures in the that human mutations in several des- invasion in an adhesion-independent absence of classical cadherins. Possible mosomal proteins are also associated

www.jidonline.org 2073 C Michels et al. Adherens Junctions Regulate Desmosomes

48 hours high Ca2+ 48hours high Ca2+ P-cadherin β-Catenin Desmoplakin Phase contrast β-Catenin Desmoplakin Plakophilin 3

Control/ntc Control

ko kd Pkd E /P #1 GFP

ko kd Eko/ntc E /P #1 P-cad

ko kd Eko/Pkd#1 E /P #1 E-cad

ko kd Eko/Pkd#2 E /P #1 IL2R-tail

48hours high Ca2+ Control Eko/Pkd#1 2+ 48hours high Ca ntc GFP P-cad E-cad IL2R-tail ko Control E P-cad ntc Pkd ntc Pkd#2 Pkd#1 Control Eko Eko/Pkd#1 E-cad E-cad P-cad

Actin β-Catenin

Actin

Figure 2. Desmosome formation in primary keratinocytes depends on classical cadherin levels. (a) Immunolocalization of P-cadherin, b-catenin, and desmoplakin in control keratinocytes lentivirally transduced with a nonspecific shRNA (control/ntc) or shRNA against P-cadherin, P-cadkd (Pkd), and E-cadko keratinocytes lentivirally transduced with either a nonspecific shRNA (Eko/ntc) or shRNA against P-cadherin, E-cadko/P-cadkd#1 or #2 (Eko/Pkd). Right panel shows phase-contrast images of the different keratinocyte lines. Bar ¼ 15 mm for IF images and 50 mm for phase-contrast images. (b) Western blot analysis of classical cadherin expression in keratinocytes shown in panel a. (c) Ultrastructural analysis of representative desmosomes in control and E-cadko (Eko) and of characteristic cell–cell contact points found in E-cadko/P-cadkd#1 (Eko/Pkd#1) keratinocytes. Bars ¼ 100 nm. No desmosomal structures and only a few cell–cell contact points could be observed in the latter. (d) Rescue of desmosome formation by E-cadherin or P-cadherin but not by a membrane-tagged E-cadherin cytoplasmic domain. Immunofluorescence analysis of b-catenin, desmoplakin, and plakophilin 3 in control and E-cadko/P-cadkd (Eko/Pkd#1) keratinocytes lentivirally transduced with GFP as a negative control, E-cadherin, shRNA-resistant P-cadherin, or IL-2R-E-cadtail (IL-2R tail). Bar ¼ 15 mm. (e) Western blot of 2 þ 2 þ lentivirally transduced keratinocytes shown in panel d. For all experiments, keratinocytes were switched to 1.8 mM Ca medium (high Ca ) for 48 hours to allow for cadherin-mediated adhesion and the formation of intercellular junctions.

1,2,5 with hair disorders (reviewed in Lai- Christian Michels , SUPPLEMENTARY MATERIAL 1,5 Cheong et al., 2007), desmosomal Thomas Buchta , Supplementary Materials and methods dysfunction may explain why mutation Wilhelm Bloch3, Thomas Krieg1,2,4 1,2,4 Figure S1. Analysis of junction formation in the of P-cadherin leads to abnormal hair and Carien M. Niessen absence of classical cadherins. follicle ontogenesis. 1Department of Dermatology, University of Cologne, Cologne, Germany; REFERENCES 2Center for Molecular Medicine Cologne, Cali G, Zannini M, Rubini P, Tacchetti C, CONFLICT OF INTEREST University of Cologne, Cologne, Germany; D’Andrea B, Affuso A et al. (2007) Condi- The authors state no conflict of interest. 3German Sport University, Cologne, tional inactivation of the E-cadherin in Germany and 4Cologne Excellence thyroid follicular cells affects gland develop- ACKNOWLEDGMENTS Cluster on Cellular Stress Responses in ment but does not impair junction formation. We thank A. Schmitz and J. Scott for technical Aging-associated Diseases (CECAD), Endocrinology 148:2737–46 assistance and J. Scott and M. Niessen for critically University of Cologne, Cologne, Gottardi CJ, Wong E, Gumbiner BM (2001) E- reading the manuscript. We thank K Green Germany cadherin suppresses cellular transformation (Northwestern University) for providing the des- E-mail: [email protected] by inhibiting beta-catenin signaling in an 5 moplakin antibody. This work was supported by These authors contributed equally adhesion-independent manner. J Cell Biol DFG grants NI689/3-1 and SFB829. to this work 153:1049–60

2074 Journal of Investigative Dermatology (2009), Volume 129 EWM Verhoeven et al. Daily Stressors and Psoriasis

Gumbiner B, Stevenson B, Grimaldi A (1988) The Radice GL, Ferreira-Cornwell MC, Robinson SD, E-cadherin is essential for in vivo role of the molecule uvomor- Rayburn H, Chodosh LA, Takeichi M et al. epidermal barrier function by regulating ulin in the formation and maintenance of the (1997) Precocious mammary gland develop- tight junctions. EMBO J 24:1146– epithelial junctional complex. J Cell Biol ment in P-cadherin-deficient mice. J Cell Biol 56 107:1575–87 139:1025–32 Vasioukhin V, Bauer C, Degenstein L, Wise B, Lai-Cheong JE, Arita K, McGrath JA (2007) Sprecher E, Bergman R, Richard G, Lurie R, Shalev Fuchs E (2001) Hyperproliferation and de- Genetic diseases of junctions. J Invest Der- S, Petronius D et al. (2001) Hypotrichosis fects in epithelial polarity upon conditional matol 127:2713–25 with juvenile macular dystrophy is caused by ablation of alpha-catenin in . Cell Lewis JE, Jensen PJ, Wheelock MJ (1994) Cadherin a mutation in CDH3, encoding P-cadherin. 104:605–17 function is required for human keratinocytes Nat Genet 29:134–6 Wong AS, Gumbiner BM (2003) Adhesion-inde- to assemble desmosomes and stratify in Tinkle CL, Pasolli HA, Stokes N, Fuchs E (2008) New pendent mechanism for suppression of tumor response to calcium. J Invest Dermatol insights into cadherin function in epidermal cell invasion by E-cadherin. J Cell Biol 102:870–7 sheet formation and maintenance of tissue 161:1191–203 Perez-Moreno M, Fuchs E (2006) Catenins: keep- integrity. Proc Natl Acad Sci USA 105:15405–10 Yin T, Green KJ (2004) Regulation of desmosome ing cells from getting their signals crossed. Tunggal JA, Helfrich I, Schmitz A, Schwarz H, assembly and adhesion. Semin Cell Dev Biol Dev Cell 11:601–12 Gunzel D, Fromm M et al. (2005) 15:665–77

Effect of Daily Stressors on Psoriasis: A Prospective Study

Journal of Investigative Dermatology (2009) 129, 2075–2077; doi:10.1038/jid.2008.460; published online 5 February 2009

TO THE EDITOR Participants for this study were re- the least, (lowest EPCL score) was Many patients (37–88%) with psoriasis cruited from the Departments of Derma- determined. Prospective relationships believe that there is a causal relationship tology at the University Medical Centre between daily stressors and changes in between stressors and their skin disease St Radboud and the Canisius Wilhelmina disease outcome 4 weeks later were (Gupta et al., 1989; Nevitt and Hutch- Hospital, Nijmegen, the Netherlands. subsequently examined by calculating inson, 1996; Yosipovitch et al., 2000; This study was approved by relevant Pearson’s correlation coefficients be- O’Leary et al., 2004; Zachariae et al., ethics committees and conducted ac- tween the EPCL daily stressor scores 2004; Fortune et al., 1998, 2005), and cording to the Helsinki principles. Writ- and the change in disease severity (PASI) this relationship has received increasing ten inform consent was obtained from all and itch. For this purpose, residual gain attention over the years (Pacan et al., participants. For the purpose of this scores were used to measure the change 2003; Reich et al., 2003; Gaston et al., study, the participants were followed in PASI and itch. These scores take into 1991). Although some earlier studies up for over 6 months with monthly account the individual baseline levels have found that over 60% of the patients measures of disease severity (Psoriasis and the control for regression to the with psoriasis retrospectively report to Area and Severity Index (PASI); Freder- mean effects (Kerlinger, 1975). have experienced stressful life events in iksson and Pettersson, 1978), itch impact Mean levels of disease severity, itch, the month before the exacerbation of of skin diseases on daily life (ISDL; anddailystressorsinthemonthinwhich their skin disease (Gupta et al., 1988; Evers et al., 2008), and daily stressors patients experienced the highest and Pacan et al., 2003; Reich et al., 2003), (everyday problem check list (EPCL); lowest levels of daily stressors are pre- evidence for a prospective relationship Dekkers et al., 2001; Vingerhoets et al., sented in Table 1, showing significant between stressors and disease outcome 1989). The study sample consisted of 62 differences between the moments of is, thus far, lacking. The aim of this study patients with a mean age of 52.3 years highest and lowest reported daily stres- is therefore to investigate the relationship (SD 13.2 years, range 21.9–79.7) at the sors. When patients reported the highest between the experience of daily stressors startofthestudy.Ofalltheparticipants, level of daily stressors, they also experi- and the change in disease outcome 72.6% were male and 27.4% were enced significantly more itch and a more (disease severity and itch) 4 weeks later female.Furthermore,4.8,66.1,and severe disease than when they reported in patients with psoriasis. We hypothe- 29.1% of them had a primary, second- the lowest level of daily stressors. In sized that only at moments of relatively ary, and tertiary education level, respec- addition, Pearson’s correlation coeffi- high levels of daily stressors there would tively. For each participant, the month in cients between daily stressors and be a relationship between stressors and which the participant reported the most changes in disease severity (PASI) and an increase in itch and disease severity 4 daily stressors (highest EPCL score), as itch 4 weeks later are presented in weeks later. well as the month in which they reported Table 2. Only at the moment that patients reported the highest amount of daily stressors there was a positive, significant Abbreviations: PASI, psoriasis area and severity index correlation between daily stressors and

www.jidonline.org 2075