The Function of Plakophilin 1 in Desmosome Assembly and Actin

Filament Organization
Mechthild Hatzfeld, Christof Haffner, Katrin Schulze, and Ute Vinzens
Molecular Biology Group of the Medical Faculty, University of Halle, 06097 Halle/Saale, Germany

Abstract. Plakophilin 1, a member of the armadillo type with the formation of filopodia and long cellular
multigene family, is a protein with dual localization in protrusions, where plakophilin 1 colocalized with actin
the nucleus and in desmosomes. To elucidate its role in filaments. This phenotype was strictly dependent on a
desmosome assembly and regulation, we have analyzed conserved motif in the center of the armadillo repeat
its localization and binding partners in vivo. When over- domain. Our results demonstrate that plakophilin 1
expressed in HaCaT keratinocytes, plakophilin 1 local- contains two functionally distinct domains: the head do-
ized to the nucleus and to desmosomes, and dramati- main, which could play a role in organizing the desmo-
cally enhanced the recruitment of desmosomal proteins somal plaque in suprabasal cells, and the armadillo re-
to the plasma membrane. This effect was mediated by peat domain, which might be involved in regulating the
plakophilin 1’s head domain, which interacted with des- dynamics of the actin cytoskeleton.
moglein 1, desmoplakin, and keratins in the yeast two-
hybrid system. Overexpression of the armadillo repeat Key words: keratinocytes • desmoglein • armadillo •
domain induced a striking dominant negative pheno- cell adhesion • cell motility

Introduction
Desmosomes are adhering junctions that anchor interme- The intracellular domains of the desmosomal cadherins
diate filaments to sites of cell–cell contact. Biochemically, associate with a number of plaque proteins that establish
they are distinct, but are related to the adherens junctions the link to the intermediate filament system (Troyanovsky
that anchor actin filaments. They contain two types of et al., 1993, 1994a, 1996; Mathur et al., 1994; Chitaev et al.,
transmembrane proteins of the cadherin superfamily, the 1996; Kowalczyk et al., 1996; Witcher et al., 1996). Plako-
desmogleins (Dsgs)1 and desmocollins (Dscs). There are at globin and desmoplakin are essential components of the
least three different desmogleins, and three different des- plaque. Plakoglobin associates with both types of desmo-
mocollin genes (Dsg1-3 and Dsc1-3) that are differen- somal cadherins and binds to several Dsg and Dsc iso-
tially expressed (Koch and Franke, 1994; Schmidt et al., forms (Mathur et al., 1994; Troyanovsky et al., 1994a,b,
1994). Whereas Dsg2 and Dsc2 are ubiquitously expressed 1996; Chitaev et al., 1996; Wahl et al., 1996; Witcher et al.,
in all cells that possess desmosomes, the expression of 1996). The binding of plakoglobin to E-cadherin is a pre-
Dsgs1 and 3 and Dscs1 and 3 is restricted to stratified epi- requisite for desmosome formation, and its COOH termi-
thelia (Schmidt et al., 1994; Garrod et al., 1996). Expres- nus is involved in regulating desmosome size (Ruiz et al.,
sion patterns of isoforms of the desmosomal cadherins 1996; Lewis et al., 1997; Palka and Green, 1997). More-
overlap, and individual desmosomes can contain more over, a signaling role in the Wnt pathway, which is similar
than one isoform (North et al., 1996). Clustering of desmo- to that of ␤-catenin and the Drosophila homologue arma-
somal cadherins and desmosome formation depends on dillo, has been reported (Karnovsky and Klymkowsky,
both Dsgs and Dscs (Chitaev and Troyanovsky, 1997). 1995; Rubenstein et al., 1997). Direct interactions between
desmosomal cadherins and desmoplakin have been re-
ported only in vitro (Smith and Fuchs, 1998), and it ap-
pears that plakoglobin is necessary to link these proteins
Address correspondence to M. Hatzfeld, Molecular Biology Group of the in vivo (Kowalczyk et al., 1996, 1997). Desmoplakin binds
Medical Faculty, University of Halle, Magdeburger Strasse 18, 06097 to intermediate filaments through its COOH-terminal do-
Halle/Saale, Germany. Tel.: 49-345-557-4422. Fax: 49-345-557-4421. E-mail: main and connects desmosomes to the cytoskeleton (Stap-
mechthild.hatzfeld@medizin.uni-halle.de
1
Abbreviations used in this paper: arm, armadillo; DP-NTP, des-
penbeck et al., 1993, 1994; Kouklis et al., 1994; Bornslae-
moplakin NH2-terminal polypeptide; Dsc, desmocollin; Dsg, desmoglein; ger et al., 1996; Meng et al., 1997). Thus, intermediate
ONPG, o-nitrophenyl-␤-D-galactopyranoside. filaments seem to be linked to the plasma membrane

 The Rockefeller University Press, 0021-9525/2000/04/209/14 $5.00

The Journal of Cell Biology, Volume 149, Number 1, April 3, 2000 209–222
http://www.jcb.org 209
through a linear sequence of interactions between ker- scriptase–PCR, with expand reverse transcriptase and expand high fidelity
atins, desmoplakin, plakoglobin, and the cytoplasmic tail polymerase (Roche Diagnostics). Suitable restriction sites for cloning
were included in the primer sequences. PCR products were either directly
of cadherins. cloned into the expression vectors or first ligated into the PCRII vector
Additional components of the desmosomal plaque are using the TOPO TA cloning kit (Invitrogen BV). All PCR products were
plakophilins 1, 2, and 3, and p0071 (Hatzfeld et al., 1994; sequenced completely.
Heid et al., 1994; Hatzfeld and Nachtsheim, 1996; Mertens Prokaryotic expression was performed in the pRSET A vector, which
includes an NH2-terminal His tag (Invitrogen Corp.). Expression in eu-
et al., 1996; Bonne et al., 1999; Schmidt et al., 1999). These karyotic cells was performed with the following vectors: pCMV5 (with
proteins contain a central domain that consists of a series NH2-terminal T7 tag; Andersson et al., 1989), pCMVscript (without tag;
of 45 amino acid repeats (arm repeats), and are members Stratagene), pOPRSV (without tag; Stratagene), and pEGFP (with NH2-
of the p120ctn family of armadillo (arm) -related proteins terminal GFP tag; CLONTECH Laboratories). Vectors for the expression
(Reynolds et al., 1994; Hatzfeld and Nachtsheim, 1996; of GAL4 fusion proteins in yeast were the pBD and pAD vectors (Strat-
agene) and the pAS2-1 and pGAD424 vectors (CLONTECH Laborato-
Daniel and Reynolds, 1997). Plakophilin 1 is a major com- ries). Fig. 1 a gives an overview of the plakophilin 1 constructs used in this
ponent of desmosomes from stratified and complex epi- study. The intracellular domains of human Dsg1, Dsg2, and Dsg3, and
thelia, and it is predominantly expressed in the suprabasal Dsc1a, 1b, Dsc2a, 2b, and Dsc3a, 3b were amplified by reverse tran-
layers (Kapprell et al., 1988). It binds to keratins in vitro scriptase–PCR from HaCaT cell RNA and cloned into the pGAD424 vec-
tor. Dsg1 domains IA (amino acids 568–643), CS (amino acids 644–764),
(Kapprell et al., 1988; Hatzfeld et al., 1994; Smith and and the Dsg-specific domain (Dsg; amino acids 765–1,049) were amplified
Fuchs, 1998), but the significance of this interaction in vivo by PCR and cloned into the pGAD424 vector. Human keratins 8, 18, 6,
has not yet been established. More recently, plakophilin 1 and 17 and their individual domains in pAD and pBD have been de-
has been described as a widespread nuclear protein that is scribed elsewhere (Schnabel et al., 1998). The complete coding sequence
also expressed in nondesmosome-bearing cells, where it of human ␤-actin was amplified by PCR from HeLa cell RNA and cloned
into pGAD424.
accumulated in the nucleoplasm (Schmidt et al., 1997;
Klymkowsky, 1999). The function of plakophilin 1 in the
nucleus remains unknown so far. An essential role in des- In Vitro Mutagenesis
mosome organization and stability has been suggested re- In vitro mutagenesis was performed to delete the 5–amino acid motif
cently on the basis of a genetic skin disease. Patients lack- ENCMC (amino acids 452–456) in the plakophilin 1 arm repeat domain.
The reaction was performed with the QuickChange site-directed mu-
ing plakophilin 1 suffer from a skin fragility syndrome. tagenesis kit (Stratagene). The deletion was verified by sequence analysis.
Desmosomes in their skin are small and poorly formed
with widening of keratinocyte intercellular spaces and per-
Two-Hybrid Assays
turbed desmosome/keratin filament interactions (McGrath
et al., 1997). Desmoplakin was found predominantly cyto- Plasmids were transformed into the yeast strain YRG2 (Stratagene) by
electroporation. Double transformants were grown on plates lacking leu-
plasmic in these patients, suggesting a role for plakophilin cine and tryptophane. Expression of the His reporter gene was analyzed
1 in organizing suprabasal desmosomes. These findings on plates lacking histidine in addition to leucine and tryptophane. lacZ re-
point to an essential role of plakophilin 1 in establishing porter gene expression was analyzed in the colony lift filter assay and
stable cell contacts, desmosomal plaque size, and organi- quantitated using the ONPG (o-nitrophenyl-␤-D-galactopyranoside) sub-
zation. In a recent paper, a direct interaction between the strate as described in the yeast protocols handbook (CLONTECH Labo-
ratories).
desmoplakin NH2 terminus and plakophilin 1 and a role of
plakophilin 1 in recruiting desmoplakin to the membrane
was described (Kowalczyk et al., 1999). It was proposed Recombinant Protein Purification and
that this interaction may be important for clustering of Antibody Production
desmosomal components through lateral interactions. The plakophilin 1 head and arm repeat domains in pRSET were ex-
To learn more about the function of plakophilin 1, we pressed in BL21 DE3 bacteria and purified under denaturing conditions
on Ni-NTA resin (Qiagen). The purified protein fragments were used for
have analyzed its function in desmosome assembly in immunization of rabbits.
more detail. Wild-type plakophilin 1 recruited endogenous
desmosomal proteins into the plaque when overexpressed Cell Lines, Wound Healing Assay, and Transfections
in keratinocytes. This function was mediated by its head
HeLa and HaCaT (Boukamp et al., 1988) cells were routinely cultured in
domain, and we show that this domain interacts with Dsg1, DME supplemented with 10% FBS. Normal human epidermal kerati-
desmoplakin, and keratins. In contrast, the arm repeat do- nocytes cells and keratinocyte medium were obtained from Promocell.
main had a dominant negative phenotype: it promoted for- Wound healing assays were performed with HaCaT cells grown to conflu-
mation of filopodia and long cellular protrusions, where it ency, and a wound was inserted by scraping. Cells were analyzed 24 h after
colocalized with actin filaments. Deletion of a conserved wounding. For transient transfection experiments, cells were plated 12–16 h
before transfection. Cells were transfected either by the calcium phos-
motif in the center of the arm domain abolished the ability phate precipitation method (5prime→3prime, Inc.) or using the DOSPER
of plakophilin 1 to modulate cellular morphology and to liposomal transfection reagent (Roche Biochemicals). Cells were fixed
associate with actin. Our data suggest that plakophilin 1 is and processed for immunofluorescence analysis after 20–44 h.
involved in regulation of desmosome assembly as well as
dynamics of the actin cytoskeleton. Antibodies and Immunofluorescence Microscopy
Cells grown on coverslips were rinsed in PBS and fixed in methanol at
⫺20⬚C for 10 min, followed either by acetone treatment for 1 min or by
Materials and Methods treatment in 0.2% Triton X-100 in PBS for 20 min. Alternatively, cells
were fixed in 3.7% formaldehyde in PBS freshly prepared from paraform-
RNA Isolation and Plasmid Constructs aldehyde and permeabilized in 0.2% Triton in PBS. Cells were washed in
PBS and incubated with 1% BSA in PBS before antibody application.
RNA was prepared according to the LiCl/urea extraction method (Auf- Plakophilin 1 and its fragments were detected by the polyclonal rabbit
fray and Rougeon, 1980), and cDNAs were synthesized by reverse tran- sera against the head and repeat domains. Alternatively, plakophilin 1

The Journal of Cell Biology, Volume 149, 2000 210

head and repeat domains, which were expressed in the pCMV5 vector,
were detected with a T7 mAb (Novagen, Inc.). For double labeling the
following antibodies were used: antidesmoplakin 1 and 2, DP1&2 2.15 or
DP1&2 mix (2.15 ⫹ 2.17 ⫹ 2.20); anti-Dsg1 ⫹ 2 DG3.10, anti-Dsc3 Dsc3
U114 were obtained from Progen. Antiplakoglobin and anti–Pan-cadherin
were from Sigma Chemical Co.; the keratin antibody RCK107 was from Dr.
F. Ramaekers (University Hospital, Rotterdam, The Netherlands).
Secondary antibodies were donkey anti–rabbit or anti–mouse coupled
to Cy2 or Cy3 (Jackson ImmunoResearch Laboratories, Inc., through Di-
anova) or Alexa 488 goat anti–mouse or goat anti–rabbit IgG (Molecular
Probes, Inc.). Actin filaments were visualized by incubation with FITC- or
TRITC-labeled phalloidin (Sigma Chemical Co.).
Microscopy was carried out with a Nikon Eclipse E600 microscope with
narrow band filters.

Laser Scanning Microscopy

Cells processed for immunofluorescence microscopy were analyzed using
a Zeiss LSM 510 laser scanning microscope equipped with a helium-neon
and an argon laser and a Plan-Apochromat 63⫻ objective. Excitation
wavelengths were 488 nm for Alexa 488 and 543 nm for Cy3. The used de-
tection filters were BP505-530 for Alexa 488 and LP560 for Cy3. Fluores-
cence was recorded using the multitracking procedure to get complete
separation of the fluorescence signals.

Western Blot Analysis

Total protein extracts were prepared by adding SDS sample buffer heated
to 100⬚C to the cell culture dishes. Yeast protein extracts were prepared
according to the SDS-urea method in the presence of the complete pro-
tease inhibitor cocktail tablets (Roche Diagnostics), as described in the
Yeast Protocols Handbook. Samples were separated on 8 or 10% acryl-
amide gels and transferred to nitrocellulose. Filters were blocked in 5%
nonfat dry milk in TBS with 0.05% Tween 20. Primary antibodies were
Figure 1. (a) Schematic representation of plakophilin 1 (PKP-1)
applied for 2 h at room temperature or overnight at 4⬚C. Filters were and its deletion mutant constructs used in this study. The PKP-1
washed and incubated with alkaline phosphatase–coupled secondary anti- head comprises amino acids 1–286; PKP-1 ⌬C1, amino acids 1–213;
bodies, and bound antibodies were visualized either with the CDP-Star PKP-1 ⌬C2, amino acids 1–168; PKP-1 ⌬N1, amino acids 70–286;
chemiluminescence reagent (Tropix) or with NBT/BCIP (Boehringer In- and PKP-1 ⌬N2, amino acids 147–286. The PKP-1 arm repeat do-
gelheim Bioproducts). In some experiments, the ECL detection system main contains amino acids 287–726, and the PKP-1 headless frag-
(Amersham Pharmacia Biotech) was used. ment contains amino acids 224–726. The ⌬ENCMC fragments
carry an additional internal deletion comprising amino acids 422–
426. (b) Specificity of plakophilin head and arm repeat domain
Results antibodies. Total protein extracts (lanes 1 and 5) as well as Tri-
ton-soluble (lanes 2 and 6), high salt soluble (lanes 3 and 7), and
Plakophilin 1 Constructs and Antibodies insoluble fractions (lanes 4 and 8) of HaCaT keratinocytes were
To address the function of plakophilin 1 in desmosome as- separated on 8% SDS gels, transferred to nitrocellulose, and
sembly and structure, we studied targeting of its domains probed with the plakophilin head (lanes 1⬘–4⬘) and repeat do-
main antibodies (lanes 5⬘–8⬘). Both antibodies reacted with a sin-
in epithelial cells and analyzed its direct binding partners
gle band of 80 kD in total cell extracts (lanes 1⬘ and 5⬘). The ma-
in the yeast two-hybrid system. Fig. 1 a summarizes the jority of the protein was in the insoluble fraction (lanes 4⬘ and 8⬘).
plakophilin 1 constructs tested in transfection assays and
in the yeast two-hybrid system. The GFP constructs of all
domains were analyzed in parallel with nontagged or T7-
tagged constructs to verify that the GFP tag did not inter- dual localization in desmosomes and in the nucleus
fere with intracellular sorting. (Schmidt et al., 1997), we analyzed intracellular targeting
Rabbit polyclonal antibodies against the plakophilin 1 of the protein after overexpression. Attempts to obtain
NH2-terminal domain and the arm repeat domain were clonal cell lines that strongly overexpress plakophilin 1, or
generated and tested for their specificity by Western blot- its head, or repeat domain, thus far, have been unsuccess-
ting on total cellular extracts. Fig. 1 b shows that both anti- ful. This may be due to the phenotype that is caused by
bodies reacted with a single band of 80 kD, demonstrating strong overexpression of plakophilin 1 or its fragments
that they did not cross-react with related proteins, such as (see below). Therefore, we have used transient transfec-
plakophilin 2 (96 kD) and 3 (86 kD), p120ctn (various iso- tion studies to analyze the function of plakophilin 1 and its
forms of 96–115 kD), or p0071 (130 kD). The majority of domains in a cellular context. Wild-type plakophilin 1,
the protein was detected in the insoluble protein fraction. which was overexpressed in HaCaT keratinocytes, local-
ized predominantly to the nucleus and to cell borders in
Wild-Type Plakophilin 1 and Its Head Domain confluent monolayers (Fig. 2 a), which is in agreement
Associate with Desmosomes and Enhance Recruitment with the intracellular localization of the endogenous pro-
of Desmosomal Proteins to the Plasma Membrane in tein (Schmidt et al., 1997). The balance between nuclear
HaCaT Cells localization and plasma membrane association appeared
Since plakophilin 1 has been described as a protein with similar in transfected and nontransfected cells. Double la-

Hatzfeld et al. Function of Plakophilin 1 211

 enriched for plakophilin 1 (Fig. 3, e and e⬘). These data
demonstrate that plakophilin 1 is able to recruit various
desmosomal plaque proteins to the plasma membrane,
and that this effect is mediated by its head domain.
In addition to its plasma membrane association, the
head domain showed very strong nuclear localization. Sur-
prisingly, some desmoplakin, Dsg, and Dsc were also de-
tected in the nucleus, suggesting that plakophilin 1 coim-
ported a fraction of these proteins into the nucleus.
To analyze the recruitment of desmosomal plaque pro-
teins to the plasma membrane in more detail, we used la-
ser scanning microscopy on HaCaT cells expressing the
head domain of plakophilin 1. Whereas plakophilin 1 and
E-cadherin staining overlapped only very little (Fig. 4 a), a
high degree of overlap was found between plakophilin 1
and desmoplakin staining (Fig. 4 b), demonstrating that
the major portion of overexpressed plakophilin 1 head do-
main does not localize to adherens junctions. These data
indicate that plakophilin 1–mediated recruitment of pro-
teins occurs primarily in desmosomes. To investigate the
effect of the recruitment on desmosome size and number,
we quantitated desmoplakin staining at cell borders by
scanning along plasma membrane stretches (Fig. 4, b⬘ and
b⬘⬘, arrows). The data are displayed as fluorescence inten-
sity profiles below the corresponding image. The number
and size of the peaks within these profiles were signifi-
cantly higher when recorded along cell borders of two
transfected cells (Fig. 4 b⬘), compared with the cell border
between transfected and nontransfected cells (Fig. 4 b⬘⬘).
Figure 2. Expression of full-length plakophilin 1 in HaCaT cells. Assuming that each peak represents a desmosome or a
Cells were fixed in methanol 30 h after transfection, and double group of desmosomes, these data indicate that plakophilin
labeled with the plakophilin 1 head domain antibody (a) and the 1–mediated recruitment of plaque proteins might result in
desmoplakin 2.15 antibody (a⬘). In confluent monolayers, plako- the generation and enlargement of desmosomes.
philin 1 accumulated in the nucleus and at the plasma membrane, To determine the region within the plakophilin 1 head
and desmoplakin was recruited to the plasma membrane of the
domain responsible for desmosome association, several
transfected cells. Labeling of endogenous desmoplakin in non-
transfected adjacent cells was comparatively weak. Bar, 20 ␮m. fragments were constructed (Fig. 1 a). Whereas all of them
were still able to associate with desmosomes in HaCaT
cells (Fig. 5), only the ⌬N1, ⌬N2, and ⌬C1 fragments were
capable of significantly enhancing the recruitment of en-
beling with desmoplakin antibodies revealed a strong in- dogenous desmoplakin (Fig. 5) and other desmosomal
crease of endogenous desmoplakin at the plasma membrane plaque proteins (data not shown) to the cell membrane.
in the transfected cells compared with nontransfected cells The ⌬C2 fragment did not recruit endogenous desmo-
(Fig. 2 a⬘). somal proteins, although it associated with desmosomes.
To identify the domains that target plakophilin 1 to A major portion of the ⌬N2 and ⌬C2 fragments remained
desmosomes and to the nucleus, the head and the arm re- cytoplasmic (Fig. 5). All fragments were still able to enter
peat domains of plakophilin 1 were expressed separately. the nucleus, but nuclear targeting was more efficient with
Whereas the arm repeat domain colocalized with the actin the ⌬N2 and ⌬C2 constructs. These experiments show that
cytoskeleton (see below), the head domain, like the full- at least one region mediating plasma membrane targeting
length protein, was detected in the nucleus and along the of plakophilin 1 as well as a signal directing nuclear local-
cell periphery (Fig. 3, a–e) and strongly enhanced recruit- ization is retained in all head deletion constructs.
ment of desmoplakin to the plasma membrane (Fig. 3 a⬘).
Costaining for other desmosomal proteins revealed that
recruitment of Dsg (Fig. 3 b⬘), Dsc (Fig. 3 c⬘) and, to a
Wild-Type Plakophilin 1 and Its Head Domain
lesser extent, plakoglobin (Fig. 3 d⬘) was also enhanced.
Accumulate in the Nucleus of HeLa Cells
The amount of recruited protein roughly correlated with When overexpressed in simple epithelial HeLa cells, pla-
the size of the membrane pool of plakophilin 1. In cells kophilin 1 accumulated in the nucleus, but was not re-
with a large membrane pool of plakophilin 1, recruited cruited to the plasma membrane (Fig. 6 a), suggesting that
proteins were detected continuously along the plasma the nuclear function of plakophilin 1 is conserved among
membrane (Fig. 3, a–d’). In other cells, the typical punc- all cells, whereas its function in stabilizing intercellular
tate pattern of individual desmosomes was retained (Fig. junctions is restricted to certain cell types. The lack of des-
4). Costaining for keratins showed colocalization of a mosome association of plakophilin 1 in HeLa cells may be
small pool of these proteins to plasma membrane patches due either to the lack of an appropriate binding partner

The Journal of Cell Biology, Volume 149, 2000 212

 Figure 3. Expression of the plakophilin 1 head
domain in HaCaT cells. Plasmid DNAs encod-
ing the plakophilin 1 head domain in pCMV5
were transfected into HaCaT cells. Cells were
fixed in methanol and extracted in Triton X-100
and double labeled with the plakophilin 1 head
domain antibody (a–e) and antibodies against
desmoplakin (a⬘), desmoglein (b⬘), desmocollin
(c⬘), plakoglobin (d⬘), and keratin (e⬘). In a and
b, single transfected cells are in the center; ar-
rows in c–e denote the plasma membranes be-
tween two transfected cells. The plakophilin 1
head domain was found in the nucleus and at
cell borders; it enhanced the recruitment of des-
moplakin (a⬘), desmoglein (b⬘), desmocollin (ar-
rows, c⬘) and, to a lesser extent, of plakoglobin
(arrows, d⬘) to the plasma membrane. Keratins
colocalized with plakophilin 1 at the borders of
transfected cells (arrows, e⬘). Bar, 20 ␮m.

such as cell type–specific Dsg and/or Dsc isoforms, or to (Fig. 6 b⬘). A similar distribution of desmoplakin was seen
different regulatory mechanisms that control modification in mitotic cells (Fig. 6 b⬘, arrowheads), where desmosomal
and/or assembly of desmosomal proteins in HeLa cells. In proteins have been internalized in vesicles. Nontrans-
addition to its nuclear localization, plakophilin 1 was fected, nonmitotic cells revealed the punctate staining pat-
found along actin filaments, as demonstrated by double la- tern along the plasma membrane, which is typical of des-
beling with phalloidin (Fig. 6, a and a⬘). mosomes (Fig. 6 b⬘, arrows). The extent of cytoplasmic
Transfection studies with the plakophilin 1 head domain staining of desmoplakin seemed to correlate with plako-
in HeLa cells showed almost exclusive nuclear localization philin 1 expression levels. The cytoplasmic staining could
of the fragment (Fig. 6 b). Decoration of actin filaments be due to internalization of desmosomes and/or enhanced
was not observed, suggesting that the binding site for di- synthesis and assembly of desmosomal proteins in the cy-
rect or indirect actin filament association is located in the toplasm (Demlehner et al., 1995). The ⌬C1 (Fig. 6 c), ⌬C2,
arm repeat domain (see below). Desmoplakin staining was ⌬N1, and ⌬N2 (not shown) constructs showed almost ex-
strong in the transfected cells, but it appeared in a punc- clusive nuclear localization with the same effect on des-
tate pattern in the cytoplasm rather than in membranes moplakin distribution as described above.

Hatzfeld et al. Function of Plakophilin 1 213

 Since wild-type plakophilin 1 decorated actin filaments
in transfected HeLa cells, we analyzed plakophilin 1 local-
ization more carefully in nontransfected cells to distin-
guish whether this was an artifact due to heavy overexpres-
sion that disturbed the intracellular sorting mechanisms, or
whether it was connected to a novel function of plakophi-
lin 1. In a wound healing experiment with HaCaT cells,
colocalization of actin filaments and plakophilin 1 was ob-
served at the tips of cellular protrusions (Fig. 6, d and d⬘),
suggesting a role for plakophilin 1 in regulating actin fila-
ment organization. Association with stress fibers was not
observed.

The Plakophilin 1 Head Domain Binds to

Desmoglein1, Desmoplakin, and Keratins in the
Yeast Two-hybrid Assay
Plakophilin 1 has been shown to bind to Dsg1, Dsc1a, and
desmoplakin in in vitro overlay assays (Smith and Fuchs,
1998), and to Dsg1 and desmoplakin in the two-hybrid sys-
tem (Kowalczyk et al., 1999). To localize the binding sites
of these proteins in plakophilin 1, the cytoplasmic domains
of Dsgs1-3 and Dscs1a,b-3a,b and the NH2 terminus of
desmoplakin were tested in the yeast two-hybrid system.
From all the desmosomal cadherins, only Dsg1 interacted
with the plakophilin 1 head domain (Fig. 7 a) and with all
head domain deletion constructs (Fig. 7, b and c), although
the ⌬N2 and ⌬C2 constructs appeared somewhat less effi-
cient in reporter gene activation, suggesting that the Dsg1
binding site was not completely retained in these con-
structs. Desmoplakin binding was retained in the ⌬C1 and
⌬C2 fragments (Fig. 7 b), but not in the ⌬N1 and ⌬N2
fragments (Fig. 7 c), demonstrating that desmoplakin
binds close to the NH2 terminus of plakophilin 1. These re-
sults suggest that desmoplakin and Dsg1 do not compete
for the same binding site in the plakophilin 1 head.
Since plakophilin 1 and plakoglobin (Troyanovsky et al.,
1993; Chitaev and Troyanovsky, 1997) both bind to des-
moplakin and Dsg1, we wanted to analyze if these two
proteins provide alternative links between the cadherins
and the cytoskeleton, or if plakophilin 1 stabilizes the Dsg-
plakoglobin-desmoplakin interaction through additional
protein interactions. Therefore, we determined the plako-
philin 1 binding site in the Dsg1 cytoplasmic domain. The
plakophilin 1 head domain interacted with the intact Dsg1
cytoplasmic domain, the Dsg ⫹ CS domain and, although

fluorescence) and an antidesmoplakin antibody (red fluores-

cence). A high degree of colocalization is visible along cell bor-
ders of transfected cells (arrows), demonstrating that plakophilin
1 is recruited primarily to desmosomes. To test whether the size
or the number of desmosomes is affected by this recruitment, we
recorded fluorescence intensities in the desmoplakin channel by
Figure 4. Laser scanning microscopy analysis of HaCaT cells ex- scanning along two defined plasma membrane stretches (arrows,
pressing the plakophilin 1 head domain. (a) Cells were stained b⬘ and b⬘⬘). The scan results are displayed as intensity profiles be-
with the plakophilin 1 head domain antibody (red fluorescence) low the corresponding image. In the profile recorded along a cell
and the anti–Pan-cadherin antibody (green fluorescence). Over- border between two transfected cells (b ⬘), the peak size and num-
lay of both fluorescence signals showed only little overlap along ber are increased when compared with the profile recorded along
the plasma membrane, demonstrating that the major portion of a cell border between a transfected and a nontransfected cell
plakophilin 1 does not localize to adherens junctions. (b) Cells (b⬘⬘). Note that photobleaching during the scan accounts for the
were stained with the plakophilin 1 head domain antibody (green slight differences in the two fluorescent pictures. Bars, 10 ␮m.

The Journal of Cell Biology, Volume 149, 2000 214

 Figure 6. Plakophilin 1 associates with the actin cytoskeleton in
HeLa (a–c) and HaCaT (d) cells. Plasmids encoding wild-type
plakophilin 1 (a), the head domain (b), and the ⌬C1-GFP fusion
construct (c) were transfected into HeLa cells, and the cells were
fixed in formaldehyde (a) or methanol (b and c) and processed
for immunofluorescence. (a) Wild-type plakophilin 1 was pre-
dominantly in the nucleus and decorated actin filaments as re-
vealed by double labeling with FITC-phalloidin (a⬘). (b) The
head domain was almost exclusively in the nucleus. In transfected
cells, desmoplakin revealed a punctate staining pattern in the cy-
toplasm. A similar distribution of desmoplakin was seen in mi-
totic cells (arrowheads). In contrast, desmoplakin showed the
Figure 5. Expression of plakophilin 1 head domain fragments in punctate staining pattern along the plasma membrane that is typ-
HaCaT cells. Plasmid DNAs encoding the GFP-tagged plakophi- ical of desmosomes in other nontransfected cells (arrows, b⬘). (c)
lin 1 head domain fragments were transfected into HaCaT cells, The ⌬C1 construct showed a similar distribution as the head do-
and their ability to recruit desmoplakin to the cell membrane was main. It was almost exclusively nuclear, whereas desmoplakin la-
analyzed by immunofluorescence. Whereas desmosome localiza- beling was cytoplasmic (c⬘). (d) A wound was inserted into a con-
tion was found with all fragments, nuclear staining was strong fluent monolayer of HaCaT cells, and cells were fixed in
only with ⌬N2 and ⌬C2. ⌬N1 and ⌬C1 showed reduced nuclear formaldehyde and processed for immunofluorescence after 24 h.
staining. Desmoplakin recruitment to the plasma membrane, as Endogenous plakophilin 1 (d) colocalized with actin (d⬘) at the
revealed by continuous labeling along the cell periphery, was en- tips of cellular protrusions of cells next to the wound. Bars, 10 ␮m.
hanced with ⌬N1, ⌬N2, and ⌬C1. In ⌬C2-overexpressing cells,
desmoplakin staining showed no considerable increase. Bar, 20 ␮m.

Hatzfeld et al. Function of Plakophilin 1 215

to a somewhat lesser extent, with the Dsg domain alone
(Fig. 7 d), indicating that the plakophilin 1 binding site dif-
fers from the plakoglobin binding site in the CS domain.
The requirement of the CS domain for strong binding sug-
gests that the plakophilin 1 binding site is close to the pla-
koglobin binding site, and that simultaneous binding might
be prevented because of steric hindrance.
Since plakophilin 1 has been shown to bind keratins in
vitro (Kapprell et al., 1988; Hatzfeld et al., 1994; Smith and
Fuchs, 1998), we also examined several keratin constructs
for their interaction with the plakophilin 1 head domain. As
shown in Fig. 7 e, the type I keratins K17 and K18 strongly
interacted with the plakophilin 1 head, whereas of the type
II keratins tested, only K8 showed a weak interaction. Inter-
action studies with the head fragments revealed binding of
K17 and K18 to the ⌬C1 and ⌬C2 constructs, but not to
⌬N1 and ⌬N2. The K8 binding site appeared to differ since
binding was retained in the ⌬N1, ⌬C1, and ⌬C2 constructs,
but was lost in the ⌬N2 fragment (not shown).
Interactions between the plakophilin 1 head fragments
and desmoplakin and Dsg1 were quantitated by measuring
LacZ reporter gene activation with the ONPG substrate.
As shown in Fig. 7 g, the desmoplakin–plakophilin 1 and
the Dsg1–plakophilin 1 interactions were much stronger
with the plakophilin 1 head in the pBD vector compared
with the pAS2-1 vector, although this vector allows high
protein expression levels as verified by Western blotting
with Gal4 and plakophilin 1–specific antibodies (Fig. 8 a).
The high protein expression could either interfere with
correct folding of plakophilin 1, or the desmoplakin and
Dsg1 binding sites are masked by inter- or intramolecular
interactions after expression in the pAS vector. Dsg1 in-
teracted most strongly with the ⌬N1 construct, which lacks
the desmoplakin binding site. Interaction with the ⌬C1 Figure 7 (continues on facing page).
construct was somewhat weaker. In contrast, the ⌬C2 and
⌬N2 constructs showed considerably reduced reporter
gene activation, suggesting that these constructs did not
contain the entire Dsg1 binding site. Desmoplakin inter- lating the actin cytoskeleton. Therefore, we expressed the
acted most strongly with the ⌬C2 construct, which lacks plakophilin 1 arm repeat and the headless domain in HeLa
part of the Dsg1 binding site. The interaction was lost with and HaCaT cells and verified expression by Western blot-
the ⌬N1 and ⌬N2 constructs, indicating that the binding ting (Fig. 8 b). Cells with high levels of these plakophilin 1
site is in the NH2-terminal region. fragments displayed a highly unusual morphology with
Since all previous experiments had shown interactions formation of filopodia, lamellipodia, or long protrusions
between cadherins and the arm repeat domains, but not the (Fig. 9, a–d), which interfered with normal monolayer for-
end domains of arm proteins, we also analyzed whether the mation. The transfected cells often sat on top of the other
headless plakophilin 1 or the repeat domain interacted with cells. Cells with lower expression levels still displayed a
any of the desmosomal cadherins. As shown in Fig. 7 f, none normal cell morphology (Fig. 10, a–e). However, double
of the desmosomal cadherins interacted with headless pla- labeling with desmoplakin antibodies showed that des-
kophilin 1. The same result was obtained with the arm do- moplakin had been internalized, and disintegration of
main construct. Expression of the headless fragment in junctions had already begun (Fig. 10, e and e⬘). The plako-
yeast cells was verified by Western blotting with anti-Gal4 philin 1 arm repeat domain colocalized with actin in lamel-
antibodies (Fig. 8 a, lane 7⬘). Here, the headless fragment lipodia (Fig. 10, a, c, and d) and sometimes stress fibers
gave the strongest signal, indicating that a lack of protein (Fig. 10, c and c⬘), suggesting a role in regulating actin po-
expression did not account for the lack of binding. lymerization and filopodia formation. This phenotype was
observed in HeLa and HaCaT cells.
The armadillo Repeat Domain of Plakophilin 1
Associates with Actin and Induces the Formation of The Phenotype Produced by the Plakophilin 1 Arm
Filopodia and Long Cellular Protrusions Domain and Its Capacity to Associate with Actin
The colocalization of wild-type plakophilin 1 with stress fi-
Filaments Critically Depend on a Conserved Motif
bers as well as actin-rich structures at the tips of filopodia A similar phenotype, the formation of long dendritelike
(Fig. 6) pointed to a possible role of plakophilin 1 in regu- cellular protrusions, had been observed in transfection

The Journal of Cell Biology, Volume 149, 2000 216

 Figure 7. Two-hybrid analysis of pro-
tein–protein interactions. (a) YRG2
yeast cells were double transformed with
the plakophilin 1 head in pAS2-1 and
desmoplakin-NH2 terminus (DP-NTP,
1), Dsg1 (2), Dsg2 (3), Dsg3 (4), Dsc1a
(5), Dsc1b (6), Dsc2a (7), Dsc2b (8),
Dsc3a (9), and Dsc3b (10) intracellular
domains. All cells grew on selection
plates lacking tryptophan and leucine
(⫺TL), indicating that they contain both
plasmids. Histidine reporter gene activa-
tion was analyzed on plates lacking histi-
dine (⫺TLH) and LacZ reporter gene
activation in a filter lift assay. DP-NTP
and Dsg1 activated both reporter genes.
(b) Double transformations of the ⌬C1
construct in pAS2-1 and Dsg1 (1), Dsc1a (2), Dsc1b (3), and DP-NTP (4), and of ⌬C2 with DP-NTP (5), Dsc1b (6), Dsc1a (7), and Dsg1
(8). DP-NTP and Dsg1 interacted with the ⌬C1 and ⌬C2 constructs, although LacZ reporter gene activation seemed weaker with Dsg1 ⫹
⌬C2. (c) Double transformations of the ⌬N1 construct in pAS2-1 with Dsg1 (1), Dsc1a (2), Dsc1b (3), and DP-NTP (4) and of ⌬N2 with
DP-NTP (5), Dsc1b, (6), Dsc1a (7), and Dsg1 (8). ⌬⌵1 and ⌬N2 reacted with Dsg1, whereas DP-NTP did not interact with the ⌬N1 and
⌬N2 constructs. Dsc1a and b did not interact with any of the head domain fragments. (d) Double transformants of the head domain with
Dsg1 deletion constructs containing the complete cytoplasmic domain (1), the IA domain (2), the CS domain (3), the IA and the CS do-
main (4), the Dsg domain (5), and the Dsg and CS domains (6). The head domain interacted strongly with the complete Dsg cytoplasmic
domain and with the Dsg⫹CS domain. The interaction with the Dsg domain alone was weaker. (e) Double transformations of the head
domain with K8 (1), K18 (2), K6 (3) and K17 (4) and of the arm domain with K8 (5), K18 (6), K6 (7) and K17 (8). The plakophilin 1
head domain interacted weakly with K8 and more strongly with K17 and K18. The arm repeats did not interact with any of the keratins
tested. (f) Double transformations of headless plakophilin 1 with the following intracellular domains: DP-NTP (1), Dsg1 (2), Dsg2 (3),
Dsg3 (4), Dsc1a (5), Dsc1b (6), Dsc2a (7), Dsc2b (8), Dsc3a (9), and Dsc3b (10). Although the His reporter gene was weakly activated
by some constructs, the LacZ reporter gene was not activated. (g) Interactions between the cytoplasmic domain of Dsg1 and DP-NTP
and the plakophilin 1 head domain fragments were quantitated using a ␤-galactosidase assay and the ONPG substrate. The bars repre-
sent three independent experiments each performed in triplet. None of the plakophilin 1 constructs activated the LacZ reporter gene on
its own. Dsg1 interacted with all constructs tested. However, the ⌬C2 and ⌬N2 constructs showed a strong decrease in reporter gene ac-
tivation suggesting that these constructs do not contain the entire Dsg1 binding site. DP-NTP interacted most strongly with the ⌬C2
construct and revealed no interaction with the ⌬N1 and ⌬N2 constructs.

studies with full-length p120ctn (Reynolds et al., 1996) and cell adhesion and signal transduction (Schmidt et al.,
␦-catenin (Lu et al., 1999). This suggested that the pheno- 1997). In the present study, we have determined the re-
type is conserved among p120ctn family members, and gions in plakophilin 1 responsible for binding of desmo-
might depend on the interaction with a common binding somal proteins and provide a functional analysis of the
partner that is involved in regulating actin filament organi- plakophilin 1 domains.
zation. To characterize this binding site in plakophilin 1,
we constructed a deletion mutant that lacks a central pen-
The Head Domain of Plakophilin 1 Mediates Binding to
tapeptide motif conserved among all p120ctn family mem-
Desmoplakin, Dsg1, and Keratins
bers. The motif (ENCM/VC) is specific for this family and
not detected in other arm related proteins. Transfection Plakophilin 1 is a desmosome-associated protein and has
studies with this mutant construct (plakophilin 1 arm been shown to bind Dsg1, Dsc1a, and desmoplakin in vitro
⌬ENCMC) showed that the mutant had lost its capacity to (Smith and Fuchs, 1998) and desmoplakin in vivo (Kowal-
induce changes in cell morphology and no longer associ- czyk et al., 1999). Using the yeast two-hybrid assay, we
ated with actin filaments in filopodia (Fig. 10, f and f⬘). In- have mapped the binding sites of desmosomal proteins
stead, it accumulated in the cytoplasm, sometimes in an and keratins within plakophilin 1. We show that it is the
aggregated form (Fig. 10 f). head domain that mediates the interactions between pla-
To analyze if the interaction between plakophilin 1 and kophilin 1 and Dsg1, desmoplakin, as well as keratins.
actin is direct, we used the two-hybrid system. These ex- Whereas desmoplakin binds close to the NH2 terminus be-
periments revealed no direct interaction between ␤-actin tween amino acids 1–70 of the head domain, the Dsg1
and the plakophilin 1 repeat domain (data not shown), binding site is located between amino acids 70 and 213
suggesting that the interaction either depends on an intact (Fig. 11). Our data suggest that these two sites do not act
microfilament or is mediated through an actin-associated independently. Dsg1 bound most strongly to the deletion
protein in vivo. construct lacking the desmoplakin binding site and vice
versa. This could be due to reduced accessibility of the
binding sites because of intramolecular interactions, which
Discussion are similar to those described for vinculin and ERM family
Plakophilin 1 was shown to localize to desmosomes and to members (Winkler et al., 1996; Tsukita et al., 1997), or in-
the nucleus, raising the possibility for a dual function in teractions with other proteins. A similar observation was

Hatzfeld et al. Function of Plakophilin 1 217

Figure 8. Expression of plakophilin 1 constructs in yeast (a) and
HeLa cells (b). (a) Yeast cell extracts were prepared as described
in Materials and Methods, and the cell extracts were stained with
Coomassie (lanes 1–7) or blotted with anti-GAL4 (lanes 1⬘–7⬘) or
plakophilin 1 head antibodies (lanes 2⬘⬘–6⬘⬘). (lane 1) YRG2
yeast cells without plasmid; (lane 2) plakophilin 1 head domain;
(lane 3) ⌬C2; (lane 4) ⌬C1; (lane 5) ⌬N1; (lane 6) ⌬N2; and (lane
7) headless plakophilin 1. Arrows denote the plakophilin 1 head
fragments reacting with the GAL4 antibody. (b) Total extracts
from HeLa cells transfected with the plakophilin 1 head (lanes
1–1⬘⬘) or the arm repeats (lanes 2–2⬘⬘) were prepared in SDS Figure 9. Expression of the arm repeat domain in HaCaT (a and
sample buffer and probed with the plakophilin 1 head and arm b), L6 (c), and HeLa (d) cells. Cells were transfected with the
repeat antibodies as indicated. Lanes 1 and 2 show the Ponceau arm repeat domain in pCMV5 (a, b, and d) or the headless con-
red–stained protein. struct in pEGFP (c). Transfected cells were visualized with the
T7 antibody (a and b) or the plakophilin 1 repeat antibody (d) or
by GFP fluorescence (c). Transfected cells showed changes in
cellular morphology, with the development of filopodia and long
made for plakoglobin, certain internal fragments of which cellular protrusions. Bars: (a and b) 40 ␮m; (c and d) 20 ␮m.
bound better to E-cadherin than the entire molecule (Chi-
taev et al., 1996). Alternatively, high expression of plako-
philin 1 could interfere with correct folding and thereby
prevent the interaction in an unspecified manner. In our two-hybrid assay, we could not confirm the inter-
The localization of the plakophilin 1 binding site within action between plakophilin 1 and Dsc1 reported by Smith
the Dsg1 cytoplasmic tail showed that it is distinct from and Fuchs (1998) using an overlay assay. Since most of the
the reported plakoglobin binding site (Mathur et al., 1994; two-hybrid vectors allow only low protein expression we
Chitaev et al., 1996). However, the close proximity of the were unable to detect expression of the Dsg, Dsc and kera-
two binding sites could prevent simultaneous binding. In tin fragments by Western blotting. Therefore, we cannot
contrast to plakophilin 1, other arm family members in- unequivocally rule out the possibility that we might have
cluding ␤-catenin, plakoglobin, and p120ctn associate with missed the plakophilin 1–Dsc1 interaction because of a lack
classical or desmosomal cadherins through their arm re- of protein expression or a lack of nuclear import of Dsc 1.
peat region (Hinck et al., 1994; Mathur et al., 1994; Aghib Alternatively, both proteins could associate in vitro after
and McCrea, 1995; Daniel and Reynolds, 1995; Sacco et al., their denaturation, but not under physiological conditions.
1995; Shibamoto et al., 1995; Aberle et al., 1996; Chitaev We also detected interactions between plakophilin 1
et al., 1996; Reynolds et al., 1996; Troyanovsky et al., 1996; and keratins. We found a weak binding of K8 and strong
Wahl et al., 1996; Witcher et al., 1996). Moreover, muta- binding of K17 and K18, suggesting a preference for type I
tional analysis has revealed that the arm repeat domains keratins that had also been proposed on the basis of in
of ␤-catenin and plakoglobin were sufficient to direct nu- vitro overlay assays (Kapprell et al., 1988). In contrast,
clear localization (Funayama et al., 1995; Karnovsky and Smith and Fuchs (1998) have reported that plakophilin 1
Klymkowsky, 1995). It is interesting that both characteris- binds preferentially to type II keratins. The controversial
tics, cell contact association as well as nuclear localization, data may be either due to the analysis of different keratins
are conserved between ␤-catenin, plakoglobin, and plako- (K5 and K14 versus K8, K18, K6, and K17) or the use of
philin 1, but the domains responsible for these functions different assay systems (in vitro overlay versus in vivo as-
are not in the conserved sequence region. says). Since plakophilin 1 is expressed in suprabasal cells

The Journal of Cell Biology, Volume 149, 2000 218

 Figure 10. Expression of the
arm repeat domain in
HaCaT and HeLa cells. (a)
HaCaT cells were trans-
fected with plakophilin 1 arm
repeats and processed for im-
munofluorescence after 20 h.
Cells were stained with the
plakophilin 1 repeat anti-
body (a) and FITC-phalloi-
din (a⬘). (b) HaCaT cells
transfected with the GFP-
tagged headless construct.
Cells were fixed 20 h after
transfection and labeled with
FITC phalloidin (b⬘). (c)
HeLa cells were transfected
with the GFP-tagged head-
less construct. Cells were la-
beled with TRITC-phalloidin
(c⬘). (d) HeLa cells trans-
fected with the arm repeats
were double stained with the
arm repeat antibody (d) and
FITC-phalloidin (d⬘). (e)
HeLa cells transfected with
the arm repeats were double
stained with the arm repeat
antibody (e) and the des-
moplakin antibody (e⬘). (f)
HeLa cells transfected with
the rep⌬ENCMC construct
were double labeled with the
T7 tag antibody and FITC-
phalloidin (f⬘). Bar, 20 ␮m.

of stratified epithelia, its in vivo interaction partner is et al., 1995; Aberle et al., 1996; Chitaev et al., 1996; Rey-
probably one of the keratins specifically expressed in dif- nolds et al., 1996; Troyanovsky et al., 1996; Wahl et al.,
ferentiated keratinocytes such as K10. 1996; Witcher et al., 1996), we found that it is the head do-
main of plakophilin 1 that directs its localization to desmo-
somes as well as to the nucleus. This is consistent with the
Plakophilin 1 Enhances Recruitment of Desmosomal localization of the binding sites for desmosomal proteins
Proteins to the Plasma Membrane determined in the two-hybrid system. The nuclear local-
We have analyzed intracellular targeting of plakophilin 1 ization was observed in all cell types examined, indicating
after overexpression. We chose two different cell lines for that the nuclear function is conserved among different cell
our studies, HaCaT keratinocytes, which express endoge- types. In contrast, desmosome association was restricted
nous plakophilin 1 and consequently all its essential inter- to HaCaT cells, suggesting that binding to a cell type–spe-
action partners, and simple epithelial HeLa cells. These cific desmosomal protein might be essential for targeting,
cells possess desmosomes and express the ubiquitous des- or that regulatory mechanisms prevent the cell contact as-
mosomal proteins, but lack certain cell type–specific des- sociation of plakophilin 1 in simple epithelial HeLa cells.
mosomal proteins including Dsg1 and 3, Dsc1 and 3, and All the head domain fragments were still able to associate
plakophilin 1 (Schmidt et al., 1994). Moreover, desmo- with desmosomes and to enter the nucleus. With the ⌬C1
somes are less abundant and smaller in HeLa cells. and ⌬N1 fragments, desmosome association was preferred
In HaCaT cells, overexpressed plakophilin 1 was found over the nuclear localization. This may be due to better ac-
in the nucleus as well as plasma membrane associated, in cessibility of desmosomal binding sites in these constructs
agreement with the intracellular localization of the endog- (see above).
enous protein (Schmidt et al., 1997). Using deletion clones In HeLa cells, full-length plakophilin 1 also decorated
of plakophilin 1, we have determined which domains tar- actin filaments. The head domain alone localized only to
get plakophilin 1 to desmosomes (Table I). Whereas cell the nucleus, indicating that it is the arm repeat domain
contact association of other arm proteins including ␤-cate- that mediates the association with the actin cytoskeleton.
nin, plakoglobin, and p120ctn is mediated by their arm re- In HaCaT cells, we also found colocalization of the plako-
peat domain (Hinck et al., 1994; Mathur et al., 1994; Aghib philin 1 head domain with keratins along membrane
and McCrea, 1995; Daniel and Reynolds, 1995; Shibamoto patches. This could be due either to the direct interaction

Hatzfeld et al. Function of Plakophilin 1 219

 cleus. In this model, the induction of desmosomes would
depend on a putative signaling function of plakophilin 1.
Our experiments do not allow us to distinguish between
these two models, since all fragments that were able to in-
duce desmosome formation also revealed nuclear localiza-
tion and, therefore, might combine the signaling and struc-
tural functions. Further, both mechanisms might contribute
to the recruitment of endogenous desmosomal proteins.
Figure 11. Binding sites of plakophilin 1–associated proteins.
Plasma membrane association of desmoplakin and plako-
philin 1 after combined overexpression in COS cells (Kowal-
czyk et al., 1999) argues for a contribution of the recruit-
between plakophilin 1 and keratins, as shown in the two- ment mechanism.
hybrid system, or to recruitment via the keratin-binding Using laser scanning microscopy, we demonstrate that
protein desmoplakin. Nevertheless, these data, together plakophilin 1 preferentially associates with and recruits des-
with the results of the two-hybrid assay, suggest that pla- mosomal proteins, and that the recruitment of desmosomal
kophilin 1 interacts with keratins in vivo. Recruitment of components might result in the generation and the enlarge-
desmoplakin and keratins to the plasma membrane has ment of desmosomes. The possibility that expression of pla-
also been described in cells overexpressing a plakoglobin- kophilin 1 enhances desmosome formation in keratinocytes
synaptophysin chimera (Chitaev et al., 1996). is consistent with the observation that desmosomes of su-
In a recent report, Kowalczyk et al. (1999) showed that prabasal cells are larger than basal cell desmosomes, and
the desmoplakin NH2 terminus was recruited to the mem- that desmosomes are more numerous in suprabasal cells.
brane when overexpressed together with plakophilin 1 in Moreover, this finding explains why desmosomes were
COS cells. We extended these experiments and analyzed small and rare in a patient lacking plakophilin 1 (McGrath
the recruitment of various endogenous desmosomal pro- et al., 1997). Therefore, we propose that plakophilin 1 plays
teins in HaCaT cells overexpressing plakophilin 1. As an essential role in regulating desmosome organization and
judged by immunofluorescence recruitment of desmoplakin, size during keratinocyte differentiation.
Dsg and Dsc were strongly enhanced, and that of plako-
The Arm Repeat Domain of Plakophilin 1
globin was slightly enhanced, suggesting a major role for
Associates with Actin Filaments and Induces
plakophilin 1 in desmosome assembly. In cells with high
Formation of Filopodia
plakophilin 1 expression, we observed nuclear localization
of other desmosomal proteins including desmoplakin, In HeLa cells, overexpressed plakophilin 1 associated with
probably due to coimport mediated by plakophilin 1. This actin filament, suggesting that it might be involved in regu-
conclusion is supported by the fact that the ⌬N1 and ⌬N2 lating the actin cytoskeleton. This association was also ob-
constructs, which lack the desmoplakin binding site, never served in nontransfected cells, where plakophilin 1 colo-
coimported desmoplakin into the nucleus (Fig. 5). calized with actin in normal cells at the tips of plasma
There are two ways in which additional desmosomal membrane protrusions. A similar localization has been de-
proteins could be recruited to the plasma membrane. First, scribed for ␤-catenin, which interacts with the actin fila-
plakophilin 1 could bind to endogenous desmosomal pro- ment bundling protein fascin through its arm repeat do-
teins, target them to the plasma membrane, and thereby main (Tao et al., 1996). Actin filament association of
dramatically increase their stability. This is consistent with plakophilin 1 also appeared to be mediated by its arm re-
the finding that desmosomal proteins are usually synthe- peats. When overexpressed at high levels in HeLa and
sized in excess, and their cytoplasmic pool is rapidly de- HaCaT cells, the arm repeat domain induced the forma-
graded (Pasdar and Nelson, 1988, 1989). In this model, tion of long cellular protrusions, supporting a possible role
plakophilin 1 plays a structural role in desmosome assem- in the regulation of cell motility. This phenotype inter-
bly. Second, plakophilin 1 might be directly involved in fered with intercellular adhesion, and transfected cells
regulating the synthesis of desmosomal proteins in the nu- were separated from the monolayer. Since full-length pla-

Table I. Intracellular Localization of Plakophilin 1 and Its Fragments

IF- Actin-
Desmosome Nucleus Cytoplasm associated associated

PKP1 head ⫹⫹ ⫹⫹ (⫹)* (⫹)* ⫺

PKP1 head ⌬C1 ⫹⫹ ⫹ (⫹)* (⫹)* ⫺
PKP1 head ⌬C2 ⫹ ⫹⫹ ⫹ ⫹ ⫺
PKP1 head ⌬N1 ⫹⫹ ⫹ (⫹) (⫹)* ⫺
PKP1 head ⌬N2 ⫹⫹ ⫹⫹ ⫹ (⫹)* ⫺
PKP1 headless n.d.‡ (⫹)* ⫹⫹ ⫺ ⫹⫹
PKP1 arm repeat n.d.‡ (⫹)* ⫹⫹ ⫺ ⫹⫹
PKP1 arm repeat ⌬ENCMC ⫺ (⫹)* ⫹⫹ ⫺ ⫺
*This localization is only seen in a few cells but not in the majority of transfected cells.
‡
Cannot be determined due to the phenotype that is characterized by a loss of desmosomes.

The Journal of Cell Biology, Volume 149, 2000 220

kophilin 1 had no such effect, we conclude that desmo- tein interactions and their implications for cadherin function. J. Cell Bio-
chem. 61:514–523.
some association of the head domain is preferred. A simi- Aghib, D.F., and P.D. McCrea. 1995. The E-cadherin complex contains the src
lar phenotype has been described for p120ctn (Reynolds et substrate p120. Exp. Cell Res. 218:359–369.
al., 1996) and ␦-catenin (Lu et al., 1999) after expressing Andersson, S., D.L. Davis, H. Dahlback, H. Jornvall, and D.W. Russell. 1989.
Cloning, structure, and expression of the mitochondrial cytochrome P-450
the full-length protein. In the case of p120ctn, the arm do- sterol 26-hydroxylase, a bile acid biosynthetic enzyme. J. Biol. Chem. 264:
main was required for this effect (Reynolds et al., 1996), 8222–8229.
Auffray, C., and F. Rougeon. 1980. Purification of mouse immunoglobulin
suggesting that this function is conserved in the arm do- heavy-chain messenger RNAs from total myeloma tumor RNA. Eur. J. Bio-
main of p120ctn family members. We identified a 5–amino chem. 107:303–314.
acid motif (ENCMC) that is conserved among p120ctn fam- Bonne, S., J. van Hengel, F. Nollet, P. Kools, and F. van Roy. 1999. Plakophilin-3,
a novel armadillo-like protein present in nuclei and desmosomes of epithe-
ily members. Deletion of this motif in the plakophilin 1 lial cells. J. Cell Sci. 112:2265–2276.
arm repeat domain abolished the ability of the mutant to Bornslaeger, E.A., C.M. Corcoran, T.S. Stappenbeck, and K.J. Green. 1996.
associate with actin filaments and to induce the pheno- Breaking the connection: displacement of the desmosomal plaque protein
desmoplakin from cell–cell interfaces disrupts anchorage of intermediate fil-
type. This suggests that a protein–protein interaction me- ament bundles and alters intercellular junction assembly. J. Cell Biol. 134:
diated by this motif is responsible for this effect. Since we 985–1001.
Boukamp, P., R.T. Petrussevska, D. Breitkreutz, J. Hornung, A. Markham, and
were unable to detect a direct interaction between the pla- N.E. Fusenig. 1988. Normal keratinization in a spontaneously immortalized
kophilin 1 arm repeats and actin in the two-hybrid system, aneuploid human keratinocyte cell line. J. Cell Biol. 106:761–771.
the interaction either requires an intact microfilament, as Chitaev, N.A., and S.M. Troyanovsky. 1997. Direct Ca2⫹-dependent hetero-
philic interaction between desmosomal cadherins, desmoglein and desmo-
opposed to an actin monomer, or it is mediated by an ac- collin, contributes to cell–cell adhesion. J. Cell Biol. 138:193–201.
tin-binding protein. Chitaev, N.A., R.E. Leube, R.B. Troyanovsky, L.G. Eshkind, W.W. Franke,
The phenotype in patients lacking plakophilin 1 sug- and S.M. Troyanovsky. 1996. The binding of plakoglobin to desmosomal
cadherins: patterns of binding sites and topogenic potential. J. Cell Biol. 133:
gested an important role for plakophilin 1 in stabilizing in- 359–369.
tercellular adhesion, although the lack of hair follicles and Daniel, J.M., and A.B. Reynolds. 1995. The tyrosine kinase substrate p120cas
binds directly to E-cadherin but not to the adenomatous polyposis coli pro-
sweat glands suggests an additional role in certain differ- tein or alpha-catenin. Mol. Cell. Biol. 15:4819–4824.
entiation processes (McGrath et al., 1997). Our results Daniel, J.M., and A.B. Reynolds. 1997. Tyrosine phosphorylation and cadherin/
support the conclusion that plakophilin 1 has an important catenin function. Bioessays. 19:883–891.
Demlehner, M.P., S. Schafer, C. Grund, and W.W. Franke. 1995. Continual as-
structural function and explain the role of plakophilin 1 in sembly of half-desmosomal structures in the absence of cell contacts and
desmosome assembly at a molecular level. The localiza- their frustrated endocytosis: a coordinated Sisyphus cycle. J. Cell Biol. 131:
tion of desmosomal binding sites to the head domain cor- 745–760.
Funayama, N., F. Fagotto, P. McCrea, and B.M. Gumbiner. 1995. Embryonic
relates with the finding that this domain recruits endoge- axis induction by the armadillo repeat domain of beta-catenin: evidence for
nous desmosomal proteins to sites of cell contact, whereas intracellular signaling. J. Cell Biol. 128:959–968.
Garrod, D., M. Chidgey, and A. North. 1996. Desmosomes: differentiation, de-
the arm repeat domain reduced cell contacts and induced velopment, dynamics and disease. Curr. Opin. Cell Biol. 8:670–678.
the formation of motility-associated structures. In conflu- Hatzfeld, M., and C. Nachtsheim. 1996. Cloning and characterization of a new
ent keratinocytes, localization of plakophilin 1 to desmo- armadillo family member, p0071, associated with the junctional plaque: evi-
dence for a subfamily of closely related proteins. J. Cell Sci. 109:2767–2778.
somes is preferred over association with adherens junc- Hatzfeld, M., G.I. Kristjansson, U. Plessmann, and K. Weber. 1994. Band 6 pro-
tions and actin filaments. This is consistent with strong tein, a major constituent of desmosomes from stratified epithelia, is a novel
intercellular adhesion in these cells. However, in cells that member of the armadillo multigene family. J. Cell Sci. 107:2259–2270.
Heid, H.W., A. Schmidt, R. Zimbelmann, S. Schafer, S. Winter-Simanowski, S.
lack contact to adjacent cells, plakophilin 1 localizes to Stumpp, M. Keith, U. Figge, M. Schnolzer, and W.W. Franke. 1994. Cell
filopodia. Here, it may have a function in inducing junc- type-specific desmosomal plaque proteins of the plakoglobin family: plako-
philin 1 (band 6 protein). Differentiation. 58:113–131.
tion formation as soon as the tip of the cell contacts an op- Hinck, L., I.S. Nathke, J. Papkoff, and W.J. Nelson. 1994. Dynamics of cad-
posing cell. This idea is consistent with the finding that for- herin/catenin complex formation: novel protein interactions and pathways of
mation of actin-associated cell contacts precedes desmosome complex assembly. J. Cell Biol. 125:1327–1340.
Kapprell, H.P., K. Owaribe, and W.W. Franke. 1988. Identification of a basic
formation and is a prerequisite for desmosome formation protein of Mr 75,000 as an accessory desmosomal plaque protein in stratified
(Lewis et al., 1994, 1997). Plakophilin 1 could play a role in and complex epithelia. J. Cell Biol. 106:1679–1691.
recruiting desmosomal proteins from the cytoplasm to the Karnovsky, A., and M.W. Klymkowsky. 1995. Anterior axis duplication in Xe-
nopus induced by the over-expression of the cadherin-binding protein plako-
plasma membrane at sites of newly formed cell contacts. globin. Proc. Natl. Acad. Sci. USA. 92:4522–4526.
Klymkowsky, M.W. 1999. Plakophilin, armadillo repeats, and nuclear localiza-
We are grateful to K. Green for providing the desmoplakin construct for tion. Microsc. Res. Tech. 45:43–54.
the two-hybrid studies and to F. Ramaekers for keratin antibodies. We Koch, P.J., and W.W. Franke. 1994. Desmosomal cadherins: another growing
thank K. Green for many helpful suggestions and discussions and E. multigene family of adhesion molecules. Curr. Opin. Cell Biol. 6:682–687.
Bornslaeger, K. Green (both from Northwestern University, Chicago, IL), Kouklis, P.D., E. Hutton, and E. Fuchs. 1994. Making a connection: direct bind-
ing between keratin intermediate filaments and desmosomal proteins. J. Cell
and M. Osborn (MPI-Biophysical Chemistry, Göttingen, Germany) for Biol. 127:1049–1060.
critical reading of the manuscript. We would also like to thank C. Horn for Kowalczyk, A.P., J.E. Borgwardt, and K.J. Green. 1996. Analysis of desmo-
technical help, C. Nachtsheim for her contribution in the initial phase of somal cadherin-adhesive function and stoichiometry of desmosomal cad-
this study. We are also grateful to M. Iwig and the Zeiss company for help herin-plakoglobin complexes. J. Invest. Dermatol. 107:293–300.
Kowalczyk, A.P., E.A. Bornslaeger, J.E. Borgwardt, H.L. Palka, A.S. Dhaliwal,
with the confocal microscopy. C.M. Corcoran, M.F. Denning, and K.J. Green. 1997. The amino-terminal
This work was supported by grants from the Deutsche Forschungsge- domain of desmoplakin binds to plakoglobin and clusters desmosomal cad-
meinschaft (Ha 1791/3-1 and 3-2, Ha 1791/5-1) and the BMBF. herin–plakoglobin complexes. J. Cell Biol. 139:773–784.
Kowalczyk, A.P., M. Hatzfeld, E.A. Bornslaeger, D.S. Kopp, J.E. Borgwardt,
Submitted: 18 August 1999 C.M. Corcoran, A. Settler, and K.J. Green. 1999. The head domain of plako-
Revised: 16 February 2000 philin-1 binds to desmoplakin and enhances its recruitment to desmosomes.
Accepted: 23 February 2000 Implications for cutaneous disease. J. Biol. Chem. 274:18145–18148.
Lewis, J.E., P.J. Jensen, and M.J. Wheelock. 1994. Cadherin function is re-
quired for human keratinocytes to assemble desmosomes and stratify in re-
References sponse to calcium. J. Invest. Dermatol. 102:870–877.
Lewis, J.E., J.K. Wahl III, K.M. Sass, P.J. Jensen, K.R. Johnson, and M.J.
Aberle, H., H. Schwartz, and R. Kemler. 1996. Cadherin-catenin complex: pro- Wheelock. 1997. Cross-talk between adherens junctions and desmosomes

Hatzfeld et al. Function of Plakophilin 1 221

 depends on plakoglobin. J. Cell Biol. 136:919–934. in specific epithelial cells as desmosomal plaque components. Cell Tissue
Lu, Q., M. Paredes, M. Medina, J. Zhou, R. Cavallo, M. Peifer, L. Orecchio, Res. 290:481–499.
and K.S. Kosik. 1999. ␦-Catenin, an adhesive junction-associated protein Schmidt, A., L. Langbein, S. Pratzel, M. Rode, H.R. Rackwitz, and W.W.
which promotes cell scattering. J. Cell Biol. 144:519–532. Franke. 1999. Plakophilin 3—a novel cell-type-specific desmosomal plaque
Mathur, M., L. Goodwin, and P. Cowin. 1994. Interactions of the cytoplasmic protein. Differentiation. 64:291–306.
domain of the desmosomal cadherin Dsg1 with plakoglobin. J. Biol. Chem. Schnabel, J., K. Weber, and M. Hatzfeld. 1998. Protein-protein interactions be-
269:14075–14080. tween keratin polypeptides expressed in the yeast two-hybrid system. Bio-
McGrath, J.A., J.R. McMillan, C.S. Shemanko, S.K. Runswick, I.M. Leigh, E.B. chim. Biophys. Acta. 1403:158–168.
Lane, D.R. Garrod, and R.A. Eady. 1997. Mutations in the plakophilin 1 Shibamoto, S., M. Hayakawa, K. Takeuchi, T. Hori, K. Miyazawa, N. Kitamura,
gene result in ectodermal dysplasia/skin fragility syndrome. Nat. Genet. 17: K.R. Johnson, M.J. Wheelock, N. Matsuyoshi, M. Takeichi, et al. 1995. Asso-
240–244. ciation of p120, a tyrosine kinase substrate, with E- cadherin/catenin com-
Meng, J.J., E.A. Bornslaeger, K.J. Green, P.M. Steinert, and W. Ip. 1997. Two- plexes. J. Cell Biol. 128:949–957.
hybrid analysis reveals fundamental differences in direct interactions be- Smith, E.A., and E. Fuchs. 1998. Defining the interactions between intermedi-
tween desmoplakin and cell type-specific intermediate filaments. J. Biol. ate filaments and desmosomes. J. Cell Biol. 141:1229–1241.
Chem. 272:21495–21503. Stappenbeck, T.S., E.A. Bornslaeger, C.M. Corcoran, H.H. Luu, M.L. Virata,
Mertens, C., C. Kuhn, and W.W. Franke. 1996. Plakophilins 2a and 2b: constitu- and K.J. Green. 1993. Functional analysis of desmoplakin domains: specifi-
tive proteins of dual location in the karyoplasm and the desmosomal plaque. cation of the interaction with keratin versus vimentin intermediate filament
J. Cell Biol. 135:1009–1025. networks. J. Cell Biol. 123:691–705.
North, A.J., M.A. Chidgey, J.P. Clarke, W.G. Bardsley, and D.R. Garrod. 1996. Stappenbeck, T.S., J.A. Lamb, C.M. Corcoran, and K.J. Green. 1994. Phosphor-
Distinct desmocollin isoforms occur in the same desmosomes and show re- ylation of the desmoplakin COOH terminus negatively regulates its interac-
ciprocally graded distributions in bovine nasal epidermis. Proc. Natl. Acad. tion with keratin intermediate filament networks. J. Biol. Chem. 269:29351–
Sci. USA. 93:7701–7705. 29354.
Palka, H., and K. Green. 1997. Roles of plakoglobin end domains in desmo- Tao, Y.S., R.A. Edwards, B. Tubb, S. Wang, J. Bryan, and P.D. McCrea. 1996.
some assembly. J. Cell Sci. 110:2359–2371. ␤-Catenin associates with the actin-bundling protein fascin in a noncadherin
Pasdar, M., and W.J. Nelson. 1988. Kinetics of desmosome assembly in Madin- complex. J. Cell Biol. 134:1271–1281.
Darby canine kidney epithelial cells: temporal and spatial regulation of des- Troyanovsky, R.B., N.A. Chitaev, and S.M. Troyanovsky. 1996. Cadherin bind-
moplakin organization and stabilization upon cell–cell contact. I. Biochemi- ing sites of plakoglobin: localization, specificity and role in targeting to ad-
cal analysis. J. Cell Biol. 106:677–685. hering junctions. J. Cell Sci. 109:3069–3078.
Pasdar, M., and W.J. Nelson. 1989. Regulation of desmosome assembly in epi- Troyanovsky, S.M., L.G. Eshkind, R.B. Troyanovsky, R.E. Leube, and W.W.
thelial cells: kinetics of synthesis, transport, and stabilization of desmoglein Franke. 1993. Contributions of cytoplasmic domains of desmosomal cad-
I, a major protein of the membrane core domain. J. Cell Biol. 109:163–177. herins to desmosome assembly and intermediate filament anchorage. Cell.
Reynolds, A.B., J. Daniel, P.D. McCrea, M.J. Wheelock, J. Wu, and Z. Zhang. 72:561–574.
1994. Identification of a new catenin: the tyrosine kinase substrate p120cas Troyanovsky, S.M., R.B. Troyanovsky, L.G. Eshkind, V.A. Krutovskikh, R.E.
associates with E-cadherin complexes. Mol. Cell. Biol. 14:8333–8342. Leube, and W.W. Franke. 1994a. Identification of the plakoglobin-binding
Reynolds, A.B., J.M. Daniel, Y.Y. Mo, J. Wu, and Z. Zhang. 1996. The novel domain in desmoglein and its role in plaque assembly and intermediate fila-
catenin p120cas binds classical cadherins and induces an unusual morpholog- ment anchorage. J. Cell Biol. 127:151–160.
ical phenotype in NIH3T3 fibroblasts. Exp. Cell Res. 225:328–337. Troyanovsky, S.M., R.B. Troyanovsky, L.G. Eshkind, R.E. Leube, and W.W.
Rubenstein, A., J. Merriam, and M.W. Klymkowsky. 1997. Localizing the adhe- Franke. 1994b. Identification of amino acid sequence motifs in desmocollin,
sive and signaling functions of plakoglobin. Dev. Genet. 20:91–102. a desmosomal glycoprotein, that are required for plakoglobin binding and
Ruiz, P., V. Brinkmann, B. Ledermann, M. Behrend, C. Grund, C. Thalham- plaque formation. Proc. Natl. Acad. Sci. USA. 91:10790–10794.
mer, F. Vogel, C. Birchmeier, U. Gunthert, W.W. Franke, and W. Birch- Tsukita, S., S. Yonemura, and S. Tsukita. 1997. ERM proteins: head-to-tail reg-
meier. 1996. Targeted mutation of plakoglobin in mice reveals essential ulation of actin-plasma membrane interaction. Trends Biochem. Sci. 22:53–58.
functions of desmosomes in the embryonic heart. J. Cell Biol. 135:215–225. Wahl, J.K., P.A. Sacco, T.M. McGranahan-Sadler, L.M. Sauppe, M.J. Whee-
Sacco, P.A., T.M. McGranahan, M.J. Wheelock, and K.R. Johnson. 1995. Iden- lock, and K.R. Johnson. 1996. Plakoglobin domains that define its associa-
tification of plakoglobin domains required for association with N-cadherin tion with the desmosomal cadherins and the classical cadherins: identifica-
and alpha-catenin. J. Biol. Chem. 270:20201–20206. tion of unique and shared domains. J. Cell Sci. 109:1143–1154.
Schmidt, A., H.W. Heid, S. Schafer, U.A. Nuber, R. Zimbelmann, and W.W. Winkler, J., H. Lunsdorf, and B.M. Jockusch. 1996. The ultrastructure of
Franke. 1994. Desmosomes and cytoskeletal architecture in epithelial differ- chicken gizzard vinculin as visualized by high-resolution electron micros-
entiation: cell type-specific plaque components and intermediate filament copy. J. Struct. Biol. 116:270–277.
anchorage. Eur. J. Cell Biol. 65:229–245. Witcher, L.L., R. Collins, S. Puttagunta, S.E. Mechanic, M. Munson, B. Gum-
Schmidt, A., L. Langbein, M. Rode, S. Pratzel, R. Zimbelmann, and W.W. biner, and P. Cowin. 1996. Desmosomal cadherin binding domains of plako-
Franke. 1997. Plakophilins 1a and 1b: widespread nuclear proteins recruited globin. J. Biol. Chem. 271:10904–10909.

The Journal of Cell Biology, Volume 149, 2000 222