945
Journal of Cell Science 107, 945-953 (1994)
Printed in Great Britain © The Company of Biologists Limited 1994
Identification of a β-type adaptin in plant clathrin-coated vesicles
Susanne E. H. Holstein, Martin Drucker and David G. Robinson*
Abteilung Cytologie des Pflanzenphysiologischen Instituts, Universität Göttingen, Untere Karspüle 2, D-37073 Göttingen, FRG
*Author for correspondence
SUMMARY
Plant clathrin-coated vesicles (CCV), suitably protected
against proteolysis, were isolated from zucchini hypocotyls,
and screened for the presence of adaptin-like polypeptides
using monoclonal antibodies prepared against α, β(β′) and
γ-adaptins of bovine brain. An immunoreactive polypeptide in plant CCV was only detected in the case of the β(β′)adaptin antibody. This polypeptide has a molecular mass
of 108 kDa in SDS-PAGE, and gives rise to a major
cleavage product of 70 kDa after proteolysis with trypsin.
Gel filtration of 0.75 M MgCl2-dissociated coat proteins
showed that the plant β(β′)-type adaptin eluted with other
polypeptides in a manner similar to the adaptor complexes
of brain CCV. Upon subsequent hydroxyapatite chro-
matography the immunoreactive polypeptide eluted in
fractions corresponding to Golgi (HA-I) rather than
plasma membrane (HA-II) brain adaptor complexes. In
addition, this polypeptide did not shift to a higher
molecular mass when subjected to urea-SDS-PAGE. Confirmation of the presence of a β-type adaptin in plants was
provided by dot and Southern blotting experiments using
genomic DNA from zucchini hypocotyls and a β-adaptin
cDNA clone from human fibroblasts.
INTRODUCTION
Of the various coat proteins of plant CCV only the triskelions have been identified (see Coleman et al., 1991, for a
review). These have the same morphology as their counterparts
from animal sources, but both the heavy and light chains of
plant clathrin are some 10-15 kDa larger (Lin et al., 1992;
Demmer et al., 1993). Although claims have been made for the
in vitro assembly of coat proteins from plant CCV into cages
(Coleman et al., 1987; Wiedenhoeft et al., 1988), the existence
of adaptor complexes in plant CCV has not been demonstrated.
In this paper we have used monoclonal antibodies prepared
against individual adaptor polypeptides (adaptins) from bovine
brain (Robinson, 1987; Ahle et al., 1988) to identify a putative
β(β′)-type adaptin amongst the coat polypeptides of plant
CCV.
Clathrin-coated vesicles (CCV) appear to be present in all
types of eukaryotic cells (see Brodsky, 1988; and Robinson
and Depta, 1988, for reviews). They originate at the plasma
membrane and trans-Golgi network in both animal and plant
cells where they participate in endocytosis (Goldstein et al.,
1985; Robinson and Hillmer, 1990) and lysosomal (Kornfeld
and Mellman, 1989) or vacuolar (Hinz et al., 1993) protein
transport, respectively.
Research on mammalian cells has established that CCV are
specific carriers of receptor-ligand complexes (see Pley and
Parham, 1993, for a recent review). The cytoplasmic tails of
these transmembrane receptors have domains that interact with
coat protein complexes termed adaptors by Pearse and
Bretscher (1981). These in turn possess binding sites for
clathrin triskelions (e.g. see Ahle and Ungewickell, 1989;
Lindner and Ungewickell, 1991), which constitute the
outermost shell of the coat. Two types of adaptors have been
isolated through hydroxyapatite chromatography and termed
HA-I (or AP-1) and HA-II (or AP-2) (e.g. see Pearse and
Robinson, 1984; Manfredi and Bazari, 1987). Immunofluorescence studies (Robinson, 1987; Ahle et al., 1988) subsequently demonstrated their location at the Golgi apparatus
(HA-I) and plasma membrane (HA-II). Since adaptors are also
known to promote the assembly of triskelions into clathrin
cages in vitro (e.g. see Zaremba and Keen, 1983; Keen, 1987),
it is apparent that these protein complexes play a key role in
the structure and function of CCV.
Key words: β-adaptin, clathrin-coated vesicle, pea cotyledon,
proteolysis, zucchini hypocotyl
MATERIALS AND METHODS
Materials
Hypocotyl segments from etiolated zucchini hypocotyls (Cucurbita
pepo L. var. ‘Cocozelle von Tripolis’) and developing pea cotyledons
(Pisum sativum L. var. ‘Kleine Rheinländerin’ and var. ‘Cador’) were
obtained as previously given (Demmer et al., 1993). Brains from
freshly slaughtered cattle were purchased at a local abattoir. Both
plant and animal tissues were frozen in liquid N2, stored at −80°C,
and quickly thawed before use.
Isolation of clathrin-coated vesicles (CCV)
A modified version of the procedure described by Demmer et al.
(1993) was employed. The composition of the homogenization
946
S. E. H. Holstein, M. Drucker and D. G. Robinson
medium was changed to: 0.1 M MES, pH 6.4, 1 mM EGTA, 3 mM
EDTA, 0.5 mM MgCl2, 1 mM o-phenanthroline, 2 µg ml−1 aprotinin,
1 µg l−1 E-64, 0.5 µg ml−1 leupeptin, 0.7 µM pepstatin, 2% (w:v) fatty
acid-free BSA. The crude CCV fraction (a 40,000-120,000 g pellet)
was further purified by centrifuging in Ficoll/sucrose according to
Campbell et al., (1983) and then by isopycnic centrifugation in a
sucrose density gradient using a vertical rotor (160,000 g; 2.5 hours;
Depta et al., 1991). CCV-enriched fractions (collecting at 40% to 45%
sucrose) were removed, pooled, and pelleted. When not immediately
analyzed, CCV-fractions were stored at −80°C.
Dissociation, gel filtration and hydroxyapatite
chromatography of coat proteins
Coat proteins from bovine brain CCV were dissociated by resuspending CCV in 0.5 M Tris-HCl, pH 7.0, containing 0.05 M MES,
0.5 mM EGTA and 0.25 mM MgCl2, and allowing to stand, with
occasional agitation, for 30 minutes at room temperature. For plant
CCV the dissociation medium contained 0.75 M MgCl2 in the same
buffer and the treatment period was 1 hour at 4°C. After centrifugation at 350,000 g for 30 minutes (Beckman TL-100.3 rotor) to remove
stripped vesicles, the supernatants were applied directly to an FPLC
Superose 6-HR 16/50 (Pharmacia, Freiburg, FRG) gel filtration
column and eluted (flow rate 0.5 ml min−1) with 0.5 M Tris-HCl containing dissociation buffer, and then to a Pharmacia HR-5/2 column
packed with Bio-Gel HT (hydroxyapatite; Bio-Rad, München, FRG)
and eluted with a 1 to 500 mM Na2HPO4 gradient (30 ml; flow rate
0.2 ml min−1). Polypeptides in individual fractions were precipitated
by adding TCA (final concentration 10%) and allowing to stand at
4°C overnight.
Gel electrophoresis
Separation of coat polypeptides on 7.5% or 11% SDS-PAGE and their
identification by staining with silver or Coomassie Blue was carried
out by standard procedures as previously described (Depta et al.,
1987; Demmer et al., 1993). SDS-PAGE with 6 M urea in the separating gel was performed according to Ahle et al. (1988).
Dot and western blotting
Dot blotting was carried out as previously described (Zhang and
Robinson, 1990). For western blotting, polypeptides were transferred
electrophoretically to nitrocellulose filters and the immunoreactive
species visualized with a peroxidase-coupled second antibody using
an ECL-kit (Amersham-Buchler, Braunschweig, FRG). The primary
antibodies employed in this study, and the dilutions used, were as
follows: (1) polyclonal antibodies prepared against the 100-116 kDa
adaptins of bovine brain CCV (Mosley and Branton, 1988); diluted
1:1000. Designated here pAb-100. (2) Monoclonal antibodies
prepared against the 100-116 kDa adaptins of bovine brain CCV
(Ahle et al., 1988); diluted 1:50. Designated here mAb-100/1, mAb100/2, mAb-100/3. (3) A monoclonal antibody (IgG2a isotype)
prepared against the 100-116 kDa HA-II group of adaptins of bovine
brain CCV (Robinson, 1987); diluted 1:100. Designated here mAbB1/M6.
Purification of a β-adaptin cDNA clone
Escherichia coli (strain DH-5α) was transformed with the β-adaptin
HF-1 clone from human fibroblasts (Ponnambalam et al., 1990) using
the Okayama-Berg vector. Isolation of the clone and preparation of
the β-adaptin probe (PstI/XbaI fragment; 975 bp from the 5′ end) were
performed by standard procedures (Sambrook et al., 1989). The PstI
and XbaI fragments were purified by the glass-milking method
according to the manufacturer’s instructions (Amersham-Buchler,
Braunschweig, FRG).
Dot and Southern blotting
Genomic DNA, isolated from zucchini hypocotyls, pea cotyledons
and bovine brain according to standard procedures (Sambrook et al.,
1989), was immobilized on Hybond-N (Amersham-Buchler) and
denatured with 1.5 M NaCl in 0.5 M NaOH. The filters were then
neutralized with 0.5 M Tris-HCl, pH 7.2, containing 1.5 M NaCl and
1 mM EDTA and incubated in 5× SSC buffer. All of these steps were
carried out at room temperature for 10 minutes. After drying, the DNA
on the filters was fixed by exposure to UV light for 5 minutes. The
β-adaptin probe was labelled with [α-32P]dATP according to the
random priming method, and used at a concentration of 1×106 cpm
ml−1 of hybridization medium. Preparation of dot and Southern blots,
as well as the hybridization procedure were carried out as given by
Sambrook et al. (1989). Low stringency labelling was performed by
incubating with the probe in the presence of 30% formamide, washing
in 2× SSC at room temperature; for high stringency labelling the incubation was done at 42°C in the presence of 50% formamide, followed
by a wash in 2× SSC at room temperature. Blots were exposed for 48
hours on β-max film (Amersham-Büchler).
Determination of proteolytic activity in plant extracts
This was measured in two ways:
(1) Plant ‘cytosol’ was prepared by homogenizing small amounts
(10 g) of zucchini hypocotyls/pea cotyledons in CCV-homogenizing
buffer (0.1 M MES, pH 6.4, 1 mM EGTA, 0.5 mM MgCl2), filtering
through Miracloth, and then centrifuging for 30 minutes at 438,000 g
in a Beckman TL 100.3 rotor. The supernatant (cytosol) was then
divided into 2 equal parts and the components of the antiprotease
cocktail (EDTA, o-phenanthroline, aprotinin, E-64, leupeptin,
pepstatin and BSA; see above for concentrations) were added to one
of them. Cytosol (10 µg protein) ± antiprotease cocktail was then
added to 3 µg bovine brain CCV (resuspended in homogenizing
buffer) in an Eppendorf tube and incubated at 4°C for 2 hours. The
incubation was terminated by freezing in liquid nitrogen and storing
at −70°C. Individual samples were thawed and the protein precipitated was with absolute ethanol at −20°C overnight, prior to SDSPAGE and western blotting with mAb-100/1 (see above).
(2) A 450 µl sample of plant cytosol ± antiprotease cocktail
(prepared as above) was added to 50 µl azocasein (stock solution of
50 mg ml−1 in water) and incubated for 3 hours at 4°C or 20°C. The
incubations were terminated with perchloric acid and total proteolytic
activity was determined colorimetrically as previously described
(Demmer et al., 1993).
Limited proteolysis of CCV coat polypeptides
Zucchini and bovine brain CCV were resuspended in homogenizing
buffer (0.1 M MES, pH 6.4, 1 mM EGTA, 3 mM EDTA, 0.5 mM
MgCl2) containing trypsin (L-1-tosylamide-2-phenylethylchloromethyl-ketone-treated; Sigma, Deisenhofen, FRG) to protein at a ratio
of 1:10 (zucchini), 1:20 (brain) (w:w). Incubations were performed at
room temperature for the times indicated, and terminated by the
addition of 5 mM (final concentration) PMSF. Protein in the digest
was then precipitated by adding 10% TCA (final concentration), and
subjected to 7.5% SDS-PAGE.
Electron microscopy
Negative staining of CCV fractions was performed as previously
described (Depta et al., 1991), and grids were examined in a Philips
CM10 electron microscope operating at 80 kV.
RESULTS
Reducing proteolysis in plant homogenates
A prerequisite for the recognition of a particular polypeptide
in the coat of plant CCV is that the CCV have been isolated
under conditions where proteolytic degradation has been kept
to a minimum. We have therefore designed a proteolysis assay
that is specific for CCV coat proteins. This has consisted of
Plant adaptins
Fig. 1. Assaying proteolytic activity in plant extracts using bovine
brain CCV as substrate. Brain CCV were resuspended in
homogenizing buffer (lane 3), in soluble extracts prepared from
zucchini hypocotyls (lanes 1, 2), or pea cotyledons (var. Kleine
Rheinländerin, lanes 4, 5; var. Cador, lanes 6, 7), in the presence
(lanes 2, 5, 7) or absence (lanes 1, 4, 6) of an antiprotease cocktail
(see Materials and Methods), and incubated for 2 hours at 4°C. After
SDS-PAGE the samples were probed by immunoblotting with the
β(β′)-adaptin antibody mAb-100/1.
exposing bovine brain CCV to the cytosol (438,000 g supernatant) of plant extracts (pea, zucchini) in the absence and
presence of an antiprotease cocktail (see Materials and
Methods). Incubation was for 2 hours at 4°C, a period considered to be ‘normal’ for the preparation of a crude CCV pellet
from plant tissues. As a monitor of the extent of proteolysis we
have performed western blots with a monoclonal antibody
known to recognize β(β′)-adaptins (mAb-100/1; Ahle et al.,
1988).
Control incubations (brain CCV resuspended in homogenizing buffer) revealed the expected doublet at 116 (β′) and
105 (β) kDa (Fig. 1, lane 3). The former antigen is either lost
entirely, or considerably reduced in amount, when brain CCV
are incubated with soluble plant extracts in the absence or
presence of an antiprotease cocktail, respectively (Fig. 1, lanes
1, 2, 4-7). The 105 kDa antigen was particularly affected by
pea cytosol when prepared from the Cador seed variety (Fig.
1, lane 6). Even in the presence of an antiprotease cocktail,
only a trace of this antigen could be discerned (Fig. 1, lane 7).
Incubations with soluble extracts from zucchini hypocotyls
and Kleine Rheinländerin peas led to the presence of two
immunoreactive degradation products at 70 and 77 kDa in
western blots (Fig. 1, lanes 1, 4). When the antiprotease
cocktail was present in the incubation mixture these polypeptides were no longer detected (Fig. 1, lanes 2, 5).
The results obtained with the brain CCV proteolysis assay
have been confirmed by measurements of total proteolytic
activity of soluble extracts of the plants using azocasein as a
substrate. It is clear from Table 1 that zucchini extracts have
much less proteolytic activity than those from the two varieties
of pea cotyledon, and that at 4°C in the presence of an antiprotease cocktail proteolysis in zucchini soluble extracts is effectively zero. Although the antiprotease cocktail was capable of
reducing the proteolytic activity in extracts of both pea
varieties by about 40-50%, there remained 8 (Kleine Rheinländerin) to 20 (Cador) times as much proteolytic activity at
4°C as in zucchini extracts without the protective effect of the
antiprotease cocktail at this temperature. Even high concentrations of soybean trypsin inhibitor (200 µg ml−1) were ineffective in reducing proteolytic activity in pea extracts.
Screening plant CCV for adaptins
Negatively stained, CCV-enriched fractions isolated from
zucchini hypocotyls and pea cotyledons under conditions
intended to minimize proteolysis are depicted in Fig. 2. By
particle counting their purity has been determined to be, on the
average, around 80%. Despite having been exposed to a proteolytic environment during their isolation, CCV from pea
cotyledons are morphologically indistinguishable from
zucchini CCV. Such preparations were screened for the
presence of polypeptides that are immunologically related to
the 116-100 kDa adaptin family of brain CCV. Western blots
with a polyclonal serum (Mosley and Branton, 1988) reveal
several cross-reacting polypeptides in the plant lanes (Fig.
3A,B). These have similar, but not identical, electrophoretic
mobilities to the brain CCV antigens.
We have also performed western blots with monoclonal antibodies prepared against the β(β′)-adaptins (mAb-100/1), the βadaptin (mAb B1/M6), the α-adaptin (mAb-100/2) and the γadaptin (mAb-100/3) of brain CCV (Robinson, 1987; Ahle et
al., 1988). Of these four antibodies, only mAb-100/1 and
-B1/M6 yielded positive reactions with plant CCV polypeptides (Fig. 3C-E,H). By using 7.5% gels, which separate
polypeptides around 100 kDa more effectively than 11-13%
gels, one can see that the immunoreactive species in zucchini
CCV has a molecular mass in SDS-PAGE of 108 kDa, which
is slightly larger than that of the 105 kDa β-antigen in brain
CCV. By contrast the immunoreactive polypeptide in pea has
a molecular mass of around 116 kDa, corresponding to the
larger of the two antigens in brain CCV (Fig. 3G). Of the two
β(β′)-adaptin monoclonals, mAb-B1/M6 appeared to be more
effective in recognizing cross-reacting polypeptides in plant
CCV (Fig. 3H). In order to achieve a similar intensity of
reaction on western blots we have found it necessary to run
gels with at least 15 times more protein in the plant than in the
brain lanes. Even then, the intensity of the immunoreactive
Table 1. Proteolytic activity in soluble extracts prepared from zucchini hypocotyls and pea cotyledons
Proteolytic activity† in
Extract*/additions
a. Control (homogenizing buffer)‡
b. Control + 200 µg ml−1
soybean trypsin inhibitor
c. Control + antiprotease cocktail§
947
Zucchini
(20°C)
Zucchini
(4°C)
Cador peas
(4°C)
Kleine Rheinländerin peas
(4°C)
62.3±16.2
Not determined
5.4±2.0
Not determined
164.7±43.2
176.4±34
92.9±19.8
83.2±16.2
19.8±12.6
Not measurable
103±27
37.4±18.0
*Supernatant of homogenate centrifuged at 438,000 g (30 min).
†Activity given as ∆E340 102 ml−1 180 min−1.
‡0.1 M MES, pH 6.4, 1 mM EGTA, 0.5 mM MgCl2.
§2 µg ml−1 aprotinin, 1 µg ml−1 E-64, 0.5 µg ml−1 leupeptin, 0.7 µM pepstatin, 3 mM EDTA, 1 mM o-phenanthroline, 2% BSA.
948
S. E. H. Holstein, M. Drucker and D. G. Robinson
Fig. 2. Negative staining of the plant CCV fractions used in this study. (a) Zucchini; (b) pea. The small, hollow, particles in the background of
the preparation are contaminating ferritin. Bar, 100 nm.
band in the pea cotyledon CCV lane in Fig. 3B,G,H was always
much weaker than for zucchini CCV despite the fact that
identical amounts of plant CCV protein were applied to the
gels. This is presumably a consequence of the residual proteolysis in pea extracts.
Evidence for a putative β-adaptin DNA sequence in
plants
We have isolated genomic DNA from zucchini hypocotyls and
developing pea cotyledons and have looked for a β-adaptin
sequence by performing dot blots using a radiolabelled βadaptin cDNA probe (PstI/XbaI fragment from the HF-1
clone). Genomic DNA from bovine brain was used as a
standard, since there is a 90% homology amongst various
mammals at the 5′ end of the β-adaptin clone (Ponnambalam
et al., 1990). The upper row in Fig. 4A shows the results of the
hybridization under low stringency and the lower row under
high stringency conditions. There is a clear labelling of the
plant DNA, although this is much weaker than for DNA from
bovine brain. Labelling is weaker under high stringency conditions but is nevertheless positive for the two plants.
Fig. 3. Probing plant CCV for adaptin-like
polypeptides. Immunoblotting with
polyclonal (pAb-100: B) and monoclonal
(mAb-100/1, C, G; mAb-100/2, D; mAb100/3, E; mAb-B1/M6, H) antisera prepared
against brain adaptin polypeptides. Lanes: 0,
low molecular mass kit with standards at 94,
67, 43, 30 kDa; 1, bovine brain CCV; 2, pea
cotyledon CCV; 3, zucchini hypocotyl CCV.
(A) Coomassie-stained 11% gel used for
blots B-E. (F) Silver-stained 7.5% gel used
for blots G, H. Blots B-H, developed with
ECL kit; B-E, for 2 minutes; blots G, H,
developed 10 minutes. Arrow indicates
position of clathrin heavy chain. In blots BG: lane 1 had 2 µg protein; lanes 2 and 3 each
had 15 µg protein. In blot H: lane 1 had 2 µg
protein; lanes 2 and 3 each had 25 µg protein.
Plant adaptins
949
Fig. 5. Tryptic digestion of coat polypeptides from bovine brain (A)
and zucchini (B) CCV, immunoblotting with the β(β′)-adaptin
antibody mAb-100/1. Lane 1, controls, no digestion; lane 2, 10
minutes proteolysis; lane 3, 30 minutes proteolysis; 2 µg protein per
lane in A; 25 µg protein per lane in B.
Fig. 4. (A) Dot blots showing hybridization of genomic DNA from
brain (lane 1, 5 µg), pea cotyledons (lane 2, 15 µg) and zucchini
(lane 3, 15 µg) with a radioactively labelled β-adaptin probe from
human fibroblasts (PstI/XbaI fragment of the HF-1 cDNA clone).
Upper row, low stringency; lower row, high stringency conditions.
(B) Restriction enzyme-digested genomic DNA from zucchini
hybridized with the β-adaptin probe (PstI/XbaI fragment of the HF-1
cDNA clone). Southern blot prepared under low stringency
conditions; kb standards were fragments of λDNA after digestion
with HindIII.
Fig. 4B presents the results of a Southern blotting analysis
of zucchini genomic DNA that had been digested by three
different restriction enzymes and then hybridized with the βadaptin probe under low stringency conditions. After EcoRI
digestion there is one very strong and one very weak band
visible; with HindIII, two bands; and with BglI, three bands.
All of the DNA fragments recognized had a size lying in the
range 5-10 kb. The small number of bands in each case points
to a relatively high degree of specificity for the labelling. It
remains, however, unclear whether there is a singly copy of the
β-adaptin gene in zucchini, as suggested by the prominent band
in the EcoRI digest, or whether there might be two copies. In
the former case the β-adaptin gene could have two restriction
sites for BglI and one for HindIII, but the presence of the extra
bands in the HindIII and BglI lanes could also be taken as indicating the presence of a second β-adaptin gene.
Preliminary characterization of the plant adaptin
Since it is known that β(β′)-adaptins can be proteolytically
cleaved into a 60-70 kDa polypeptide (containing the NH2
terminus) and a 30-40 kDa polypeptide (Zaremba and Keen,
1985; Kirchhausen et al., 1989), we considered it appropriate
to prepare western blots of trypsinized plant CCV with mAbB1/M6. As shown in Fig. 5, limited proteolysis of zucchini
CCV coat polypeptides leads to the production of a 70 kDa
immunoreactive polypeptide at the cost of the immunoreactive
108 kDa polypeptide.
We have also performed SDS-PAGE of zucchini CCV in the
presence of 6 M urea, since, according to Ahle et al., (1988),
this leads to a change in the mobility of the 106 kDa (βadaptin) polypeptide of the brain HA II adaptor complex. Fig.
6 shows that whereas we have been able to reproduce this
effect for brain CCV, the immunoreactive plant polypeptide
did not shift in the presence of urea.
Purification of the plant adaptin
Dissociation of plant CCV coat polypeptides
It has previously been reported that the dissociation of coat
proteins with agents such as 2 M urea or 0.5 M Tris are much
less effective on plant than animal CCV (Robinson et al.,
1991a; Lin et al., 1992; Demmer et al., 1993). For a further
characterization of putative plant adaptins it was therefore
considered of great importance to attain higher yields of dissociated coat proteins. We have found 0.75 M MgCl2
(Woodward and Roth, 1978) to be most effective in this
950
S. E. H. Holstein, M. Drucker and D. G. Robinson
Fig. 6. Electrophoretic behaviour of β(β′) adaptins
in SDS-PAGE with (C, D) and without 6 M urea (A,
B); comparison of bovine brain (lane 2) with
zucchini (lane 1). (A,C) Silver-stained gels; (B,D)
immunoblots of the same with mAb-100/1. The 94
kDa standard is indicated in A and C (lane 3); and
the 116 kDa β′-adaptin of brain CCV.
regard. A 1 hour exposure to this agent leads to an almost
quantitative release of the immunoreactive β(β′)-adaptin-type
polypeptide from zucchini CCV (Fig. 7A,B).
Gel filtration of plant CCV coat polypeptides
Superose 6 gel filtration elution profiles for coat proteins dissociated from brain and zucchini CCV are given in Fig. 8A,B.
The form of the profile for the brain eluate is similar to that
previously published by Ahle and Ungewickell (1986), with
the major, first, protein peak representing clathrin triskelions.
Adaptor complexes, recognized by dot blotting with the βadaptin antibody mAb-100/1, elute as the minor, second peak.
The zucchini profile is distinctly different: firstly, there are two
peaks where brain triskelions elute; in SDS-PAGE it is the
second of these that contains the heavy and light chains of plant
clathrin (Demmer et al., 1993). Secondly, in the case of
zucchini coat polypeptides there is a very broad peak where
the brain adaptors elute. Nevertheless, the immunoreactive βtype adaptin polypeptide in zucchini CCV elutes only slightly
more slowly than the brain adaptor complexes.
We have also carried out gel filtration with dissociated coat
proteins from pea cotyledons (data not shown). The triskelion
peak elutes in the same manner as zucchini, but there is a very
large protein peak that elutes at the position of the adaptor
complexes of brain and zucchini. The major protein in this
peak is ferritin (Hoh and Robinson, 1993). Its presence has
hindered further attempts at adaptor purification.
Hydroxyapatite chromatography of adaptins
β(β′)-Adaptin-containing fractions from the Superose 6
column were pooled and applied to a hydroxyapatite column
and eluted with a phosphate gradient (Fig. 9A). Fractions
eluting between 135 mM and 470 mM Na2HPO4 were then
screened for α- and γ-adaptins using mAb-100/2 and -100/3,
respectively. With brain coat proteins a separation into two
major fractions was obtained corresponding to the HA-I and
HA-II adaptor complexes as previously described by Pearse
and Robinson (1984), Manfredi and Bazari (1987) and Ahle et
al. (1988). α-Adaptin, present only in the HA-II-adaptor
complex was detected in fractions 13-17 (263 mM to 400 mM
Na2HPO4), and the γ-adaptin of the HA-I complex was found
in fractions 7-10 (135 mM to 210 mM Na2HPO4). In the case
of zucchini the immunoreactive polypeptide was restricted to
fractions 7-10, i.e. they had hydroxyapatite binding character-
istics similar to the HA-I complex of brain CCV. Fig. 9B shows
the polypeptides present in the peak immunoreactive fractions
of the zucchini eluate in comparison with the corresponding
fractions from a brain eluate. It can be seen that in both lanes
prominent polypeptides are visible in the regions of 94-114
kDa, 50-60 kDa and below 20 kDa. Polypeptides with
molecular masses in these ranges are typical of the adaptor
complexes from bovine brain.
DISCUSSION
Research on plant CCV has lagged behind that of animal CCV
for a number of reasons. As previously noted (Robinson et al.,
1991a,b; Demmer et al., 1993) the recognition of individual
coat proteins in plant CCV has been hampered by inadequate
precautions against the release of vacuolar-localized proteases
upon homogenization. Limiting protection against proteolytic
degradation to the sole inclusion of PMSF (an inhibitor of
serine proteases) in the homogenizing medium, as often done
by plant cell biologists in the past, is of doubtful value now
that this substance has been shown to be incapable of reducing
proteolytic activity in homogenates prepared from maize roots
Fig. 7. Dissociation of coat proteins from zucchini hypocotyl CCV
using 0.75 M MgCl2. Lane 1, stripped vesicles; lane 2, released
proteins. (A) Immunoblot with mAb-100/1; (B) silver-stained 11%
Plant adaptins
Fig. 8. Gel filtration elution profiles for coat proteins dissociated
from bovine brain CCV (A) and zucchini CCV (B). Insets show
silver-stained SDS-PAGE of triskelion-containing fractions.
Fractions containing β(β′)-type adaptins are indicated by dot or
western blots with mAb-100/1 on the abscissa. CHC, clathrin heavy
chains; CLC, clathrin light chains.
(Abramycheva et al., 1993). Indeed, it has recently been shown
that PMSF can stimulate proteolysis of liver histone proteins
(Oka et al., 1992).
Here we confirm our previous observations (Demmer et al.,
1993) that zucchini hypocotyls have relatively low endogenous
levels of proteases and that proteolytic activity in homogenates
prepared from this tissue appear to be eliminated by the
951
inclusion of a suitable antiprotease cocktail. Nevertheless, as
our β(β′)-adaptin proteolysis assay has indicated (Fig. 1), some
factor(s) still remain active in ‘protected’ plant extracts that can
lead to the selective degradation of individual coat proteins
from brain CCV. It is therefore possible that a homologous
polypeptide in plant CCV would also suffer the same fate when
the plant cell is homogenized. Since our antiprotease cocktail
contains agents effective against serine (aprotinin, leupeptin),
thiol (E-64, leupeptin), acid (pepstatin), metallo-(EDTA, ophenanthroline), amino-(o-phenanthroline) proteases, as well
as high concentrations of BSA as an alternative substrate for
proteases, it is difficult to know what other measures might be
taken to eliminate this residual activity.
Our present data on proteolytic activity in pea cotyledon
extracts contradict those that we previously published
(Demmer et al., 1993). Regrettably, we cannot provide an
explanation for this; even proteolysis measurements performed
under conditions identical to those given by Demmer et al.
(1993) did not give comparable results. Because their principal
function is the synthesis and vacuolar deposition of storage
proteins (Müntz, 1989), developing cotyledons might logically
be expected to possess little proteolytic activity; at the most,
reflecting only those proteases involved in processing the proto mature forms of the storage polypeptides (e.g. see Scott et
al., 1992). The results presented in Table 1 show that this
assumption is not valid, and bring into question the validity of
previous claims as to the multiplicity of clathrin light chains
in this tissue (Lin et al., 1992), especially since these coat
polypeptides in brain tissue are well known to be very susceptible to proteolysis (e.g. see Kirchhausen and Harrison,
1984). Clearly, when working with pea cotyledons, the
inclusion of PMSF and 1 µg ml−1 soybean trypsin inhibitor or
leupeptin in the homogenizing medium, as Lin et al. (1992)
have done, is by no means an adequate protection against proteolysis. These difficulties are compounded by the fact that pea
cotyledon CCV fractions are also heavily contaminated with
ferritin (Hoh and Robinson, 1993). Some of this protein is
carried over into the supernatant when the coat proteins are dissociated and, with a molecular size of around 300 kDa, this
Fig. 9. Separation of HA-I
and HA-II adaptors by
hydroxyapatite
chromatography of gel
filtration-purified adaptor
complex fractions:
comparison of bovine brain
with zucchini.
(A) Immunoblotting of
hydroxyapatite fractions
with β(β′) adaptin (mAb100/1), α-adaptin (mAb100/2), γ-adaptin (mAb100/3) antibodies.
(B) SDS-PAGE (silverstained) of principal
immunoreactive fractions
(brain fraction 9, zucchini
fraction 8).
952
S. E. H. Holstein, M. Drucker and D. G. Robinson
protein elutes in the same fractions as do brain and zucchini
adaptor complexes. Such problems do not exist when working
with etiolated zucchini hypocotyls.
Despite the release of endogenous vacuolar plant proteases,
which is an inherent problem when working with plant tissue,
we have been able to accrue immunological and genetic
evidence for the presence of a β-type adaptin in plants. Two
monoclonals antibodies (mAb-100/1, Ahle et al., 1988; and
mAb-B1/M6, Robinson, 1987), which are directed against
epitopes on the NH2 terminus of bovine brain β(β′) adaptins
(Schröder and Ungewickell, 1991; M. Robinson, personal
communication), have been shown to recognize a putative
NH2-terminal 70 kDa fragment of a 108 kDa coat polypeptide
from zucchini CCV. This result, together with the successful
hybridization of an animal β-adaptin cDNA probe with plant
genomic DNA under conditions of high stringency, strongly
suggests the existence of a certain degree of homology between
the primary sequences of animal and plant adaptins. This is not
an altogether surprising inference, since Ponnambalam et al.
(1990) have shown that, at the cDNA level, there is a 90% conservation of structure in the β-adaptin at the NH2 terminus from
various mammalian species. Moreover, although adaptin
polypeptides have not yet been demonstrated in yeast, a gene
in this organism, analogous to the mammalian β-adaptin gene,
has been identified by Kirchhausen (1990). It has an open
reading frame encoding for an 80 kDa polypeptide and in the
NH2 terminus shows a 40% homology with the rat brain βadaptin. There is no coidentity in the COOH terminus and it
lacks the proline/glycine-rich hinge typical of mammalian
adaptins.
The HF-1 clone used here has been shown by Ponnambalam
et al. (1990) to encode for the β-adaptin of the HA-II adaptor
complex. Our observation that this probe hybridizes strongly
with only one fragment in an EcoRI digest of zucchini genomic
DNA suggests that plants, like yeasts, might have only a single
β-adaptin gene. This contrasts with the situation in mammals
where two β-adaptin genes exist (Kirchhausen et al., 1989).
Correct CCV-mediated targeting in mammals requires two
different adaptor complexes but not necessarily two different
β-adaptins. In order to determine whether there are also two
different adaptor complexes in plants we must first attempt to
identify other adaptin polypeptides. The failure of other monoclonals (mAb-100/2, mAb-100/3) to cross-react with coat
polypeptides in plant CCVs is not unexpected, since mAb100/2 is directed against an epitope on the COOH terminus of
the α-adaptin and mAb-100/3 is directed against an epitope on
the hinge region of the γ-adaptin, and both domains are known
to be quite variable within the mammals (Schröder and
Ungewickell, 1991; Robinson, 1990). This means that we can
only speculate about the existence of α- and γ-type adaptins in
plants. cDNAs for αA-, αC-, β- and γ-adaptins (Ponnambalam
et al., 1990; Robinson, 1989, 1990) are, however, available and
future work in our laboratory will include attempts at eludicating homologous nucleotide sequences for these adaptins in
plants.
We gratefully acknowledge the gifts of antibodies from Drs D.
Branton, M. Robinson and E. Ungewickell. We especially thank the
latter colleague for useful discussions in the early stages of this investigation. Dr P. Parham is also thanked for kindly providing the HF-1
β-adaptin clone. Bernd Raufeisen prepared the drawings and Sybille
Hourticolon did the photographic work. We thank Heike Freundt for
her help in the preparation of the manuscript. This work is supported
by the Deutsche Forschungsgemeinschaft.
REFERENCES
Abramycheva, N. Y., Nesterenko, M. V. and Babakov, A. V. (1993). Factors
affecting the binding between fusicoccin and plasma membranes from maize
roots. Planta 189, 301-305.
Ahle, S. and Ungewickell, E. (1986). Purification and properties of a new
clathrin assembly protein. EMBO J. 12, 3143-3149.
Ahle, S., Mann, A., Eichelsbacher, U. and Ungewickell, E. (1988). Structural
relationships between clathrin assembly proteins from the Golgi and the
plasma membrane. EMBO J. 7, 919-929.
Ahle, S. and Ungewickell, E. (1989). Identification of a clathrin binding
subunit in the HA2 adaptor protein complex. J. Biol. Chem. 264, 2008920093.
Brodsky, F. M. (1988). Living with clathrin: its role in intracellular membrane
traffic. Science 242, 1396-1402.
Campbell, C. H., Fine, R. E., Squicciarini, J. and Rome, L. H. (1983).
Coated vesicles from rat liver and calf brain contain cryptic mannose 6phosphate receptors. J. Biol. Chem. 258, 2628-2633.
Coleman, J., Evans, D., Hawes, C., Horsley, D. and Cole, L. (1987).
Structure and molecular organization of higher plant coated vesicles. J. Cell
Sci. 88, 35-45.
Coleman, J. O. D., Evans, D. E., Horsley, D. and Hawes, C. R. (1991). The
molecular structure of plant clathrin and coated vesicles. In Endocytosis,
Exocytosis and Vesicle Traffic in Plants (ed. C. R. Hawes, J. O. D. Coleman
and D. E. Evans), pp. 41-63. Cambridge University Press.
Depta, H., Robinson, D. G., Holstein, S. E. H., Lützelschwab, M. and
Michalke, W. (1991). Membrane markers in highly purified clathrin coated
vesicles from Cucurbita hypocotyls. Planta 183, 434-442.
Demmer, A., Holstein, S. E. H., Hinz, G., Schauermann, G. and Robinson,
D. G. (1993). Improved coated vesicle isolation allows better
characterization of clathrin polypeptides. J. Exp. Bot. 44, 23-33.
Goldstein, J. L., Brown, M. S., Anderson, R. G. W., Russell, D. W. and
Schneider, W. J. (1985). Receptor-mediated endocytosis: concepts
emerging from the LDL system. Annu. Rev. Cell Biol. 1, 1-39.
Hinz, G., Hoh, B. and Robinson, D. G. (1993). Strategies in the isolation of
storage protein receptors. J. Exp. Bot. (suppl.) 44, 283-291.
Hoh, B. and Robinson, D. G. (1993). The prominent 28 kDa polypeptide in
clathrin coated vesicle fractions from developing pea cotyledons is
contaminating ferritin. Cell Biol. Int. 17, 551-557.
Keen, J. H. (1987). Clathrin assembly proteins: affinity purification and a
model for coat assembly. J. Cell Biol. 105, 1989-1998.
Kirchhausen, T. and Harrison, S. C. (1984). Structural domains of heavy
clathrin chains. J. Cell Biol. 99, 1725-1734.
Kirchhausen, T., Nathanson, K. L., Matsui, W., Vaisberg, A., Chow, E. P.,
Burne, C., Keen, J. H. and Davis, A. E. (1989). Structural and functional
division into two domains of the large (100- to 115 kDa) chains of the
clathrin-associated protein complex AP-2. Proc. Nat. Acad. Sci. USA 86,
2612-2616.
Kirchhausen, T. (1990). Identification of a putative yeast homolog of the
mammalian β chains of the clathrin-associated protein complexes. Mol. Cell.
Biol. 10, 6089-6090.
Kornfeld, S. and Mellman, I. (1989). The biogenesis of lysosomes. Annu. Rev.
Cell Biol. 5, 483-525.
Lin, H.-B., Harley, S. M., Butler, J. M. and Beevers, L. (1992). Multiplicity
of light chain-like polypeptides from developing pea (Pisum sativum L.)
cotyledons. J. Cell Sci. 103, 1127-1137.
Lindner, R. and Ungewickell, E. (1991). Light-chain-independent binding of
adaptors, AP180, and auxilin to clathrin. Biochemistry 30, 9097-9101.
Manfredi, J. J. and Bazari, W. L. (1987). Purification and characterization of
two distinct complexes of assembly polypeptides from calf brain coated
vesicles that differ in their polypeptide composition and kinase activities. J.
Biol. Chem. 262, 12182-12188.
Mosley, S. T. and Branton, D. (1988). An antibody against 100- to 116-kDa
polypeptides in coated vesicles inhibits triskelion binding. Exp. Cell Res.
174, 511-520.
Müntz, K. (1989). Intracellular protein sorting and the formation of protein
reserves in storage tissue cells of plant seeds. Biochem. Physiol. Pflanzen
185, 315-335.
Plant adaptins
Oka, T., Thakur, M. K., Miyamoto, K.-I., Sassa, T. H, Suzuki, I. and
Natori, Y. (1992). Phenylmethylsulfonyl fluoride stimulates proteolysis of
nuclear proteins from chick liver. Biochem. Biophys. Res. Commun. 189,
179-183.
Pearse, B. M. F. and Bretscher, M. S. (1981). Membrane recycling by coated
vesicles. Annu. Rev. Biochem. 50, 85-101.
Pearse, B. M. F. and Robinson, M. S. (1984). Purification and properties of
100 kD proteins from coated vesicles and their reconstitution with clathrin.
EMBO J. 3, 1951-1957.
Pearse, B. M. and Crowther, R. A. (1987). Structure and assembly of coated
vesicles. Annu. Rev. Biophys. Biophys. Chem. 16, 19-68.
Pley, U. and Parham, P. (1993). Clathrin: its role in receptor-mediated
vesicular transport and specialized functions in neurons. Crit. Rev. Biochem.
Mol. Biol. 28, 431-464.
Ponnambalam, S., Robinson, M. S., Jackson, A., Peiperl, L. and Parham,
P. (1989). Conservation and diversity in families of coated vesicle adaptins.
J. Biol. Chem. 265, 4814-4820.
Robinson, D. G. and Depta, H. (1988). Coated vesicles. Annu. Rev. Plant
Physiol. Plant Mol. Biol. 39, 53-99.
Robinson, D. G. and Hillmer, S. (1990). Endocytosis in plants. Physiol. Plant.
79, 96-104.
Robinson, D. G., Balusek, K., Demmer, A. and Holstein, S. E. H. (1991a).
Some current problems in plant coated vesicle research. Giorn. Bot. Ital. 125,
55-61.
Robinson, D. G., Balusek, K., Depta, H., Hoh, B. and Holstein, S. E. H.
(1991b). Isolation and characterization of plant coated vesicles. In
Endocytosis, Exocytosis and Vesicle Traffic in Plants (ed. C. R. Hawes, J. O.
D. Coleman and D. E. Evans), pp. 65-79. Cambridge University Press.
Robinson, M. S. (1987). 100 kD coated vesicle proteins: molecular
heterogeneity and intracellular distribution studied with monoclonal
antibodies. J. Cell Biol. 204, 887-895.
Robinson, M. S. (1989). Cloning of cDNAs encoding two related 100 kD
coated vesicle proteins (α-adaptins). J. Cell Biol. 108, 833-842.
Robinson, M. S. (1990). Cloning and expression of α-adaptin, a component of
953
clathrin-coated vesicles associated with the Golgi apparatus. J. Cell Biol.
111, 2319-1326.
Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Coning. a
Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Schröder, S. and Ungewickell, E. (1991). Subunit interaction and function of
clathrin-coated vesicle adaptors from the Golgi and the plasma membrane. J.
Biol. Chem. 266, 7910-7918.
Scott, M. P., Jung, R., Müntz, K. and Nielsen, N. C. (1992). A protease
responsible for the post-translational deavage of a conserved Asn-Gly
linkage in glycinin, the major seed storage protein of soybean. Proc. Nat.
Acad. Sci. USA 89, 658-666.
Wiedenhoeft, R. E., Schmidt, G. W. and Palevitz, B. A. (1988). Dissociation
and reassembly of soybean clathrin. Plant Physiol. 86, 412-416.
Woodward, M. P. and Roth, T. F. (1978). Coated vesicles: characterization,
selective dissociation and reassembly. Proc. Nat. Acad. Sci. USA 75, 43944398.
Zaremba, S. and Keen, J. H. (1983). Assembly polypeptides from coated
vesicles mediate reassembly of unique clathrin coats. J. Cell Biol. 97, 13391347.
Zaremba, S. and Keen, J. H. (1985). Limited proteolytic digestion of coated
vesicle assembly polypeptides abolishes reassembly activity. J. Cell
Biochem. 28, 47-58.
Zhang, Y.-H. and Robinson, D. G. (1990). Cell wall synthesis in
Chlamydomonas reinhardtii: an immunological study on the wild type and
well-less mutants CW2 and CW15. Planta 180, 229-236.
(Received 15 July 1993 - Accepted, in revised form,
23 December 1993)
Note added in proof
Camidge and Pearse (1994; J. Cell Sci. 107, 709-718) have
954
S. E. H. Holstein, M. Drucker and D. G. Robinson
provided evidence for the presence of a single β-adaptin in
Drosophila. A corresponding cDNA clone when expressed in
mammalian cells shows that this β-adaptin can colocalize with
α-adaptin at the plasma membrane and γ-adaptin at the Golgi
apparatus.