Myosin Vb Is Required For Trafficking Of Cftr In Rab11a Specific Apical Recycling Endosomes In ...

33 Pages·2007·4.9 MB·English

JBC Papers in Press. Published on April 26, 2007 as Manuscript M608531200
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M608531200
MYOSIN VB IS REQUIRED FOR TRAFFICKING OF CFTR IN RAB11A-
SPECIFIC APICAL RECYCLING ENDOSOMES IN POLARIZED HUMAN
AIRWAY EPITHELIAL CELLS
Agnieszka Swiatecka-Urban1, Laleh Talebian1, Eiko Kanno2, Sophie Moreau-Marquis1, Bonita
Coutermarsh1, Karyn Hansen1, Katherine H. Karlson1, Roxanna Barnaby1, Richard E. Cheney3,
George M. Langford4, Mitsunori Fukuda2,5, and Bruce A. Stanton1
1
Dartmouth Medical School, Department of Physiology, Hanover, NH 03755
2
The Institute of Physical and Chemical Research (RIKEN), Fukuda Initiative Research Unit, 2-
1 Hirosawa, Wako, Saitama, 351-0198, Japan
3
University of North Carolina at Chapel Hill, Department of Cell and Molecular Physiology,
Chapel Hill, NC 27599
4
Dartmouth College, Department of Biological Sciences, Hanover, NH 03755
5
Graduate School of Life Sciences, Tohoku University, Department of Developmental Biology
and Neurosciences, Laboratory of Membrane Trafficking Mechanisms, Aobayama, Aoba-ku,
Sendai, Miyagi 980-8578, Japan
Running title: Myosin Vb facilitates CFTR recycling
Address correspondence to: Agnieszka Swiatecka-Urban, Department of Physiology, Dartmouth
D
Medical School, Hanover, NH 03755, Phone [phone], FAX [phone], E-Mail: o
w
[email protected] n
lo
a
d
e
d
CFTR-mediated Cl- secretion across fluid- fro
Vb attenuated the plasma membrane m
transporting epithelia is regulated, in part, by h
modulating the number of CFTR Cl- channels expression of CFTR by arresting CFTR ttp
in the plasma membrane by adjusting CFTR recycling. The dominant negative effect was ://w
w
dependent on the ability of the myosin Vb tail w
emnedcohcayntiosmsiss tahnatd rergeuclyactlei nCgF. THRo rweecvyecrli,n gt hine fragment to interact with Rab11a. Taken .jbc.o
airway epithelial cells remain unknown, at least together, these data indicate that myosin Vb is brg/
in part, because the recycling itineraries of required for CFTR recycling in Rab11a- y g
CFTR in these cells are incompletely specific apical recycling endosomes in polarized ues
understood. In a previous study, we human airway epithelial cells. t on
A
demonstrated that CFTR undergoes trafficking p
in Rab11a-specific apical recycling endosomes INTRODUCTION ril 3
, 2
in human airway epithelial cells. Myosin Vb is a The cystic fibrosis transmembrane conductance 01
9
plus-end directed, actin-based mechanoenzyme
regulator (CFTR) is an ATP binding cassette
that facilitates protein trafficking in Rab11a- (ABC) transporter and a cAMP-activated Cl-
specific recycling vesicles in several cell model channel that mediates transepithelial Cl- transport
systems. There are no published studies
in many diverse tissues including the airways (1-
examining the role of myosin Vb in airway
3), where CFTR plays a critical role in
epithelial cells. Thus, the goal of this study was mucociliary clearance by regulating the formation
to determine whether myosin Vb facilitates
and maintenance of the airway surface liquid
CFTR recycling in polarized human airway
(Reviewed in Refs. (4,5). Cystic fibrosis (CF), a
epithelial cells. Endogenous CFTR formed a lethal genetic disease, is caused by mutations in
complex with endogenous myosin Vb and the CFTR gene (1,2) and lung disease is the major
Rab11a. Silencing myosin Vb by RNA mediated
cause of morbidity and mortality in CF patients
interference decreased the expression of WT-
(6,7). Numerous studies reveal that CFTR
CFTR and ∆F508-CFTR in the apical mediated Cl- secretion across polarized epithelial
membrane, and decreased CFTR mediated Cl-
cells is regulated by modulating channel activity
secretion across polarized human airway
and by adjusting the total number of CFTR
epithelial cells. A recombinant tail domain channels in the plasma membrane (reviewed in
fragment of myosin Refs. (8,9). The latter is achieved by the removal
1
Copyright 2007 by The American Society for Biochemistry and Molecular Biology, Inc.
(i.e. endocytosis) and the insertion (i.e. recycling) human airway epithelial cells in a Rab11a-
of CFTR channels into the plasma membrane. The dependent fashion.
mechanisms that regulate CFTR recycling in
airway epithelial cells have not been elucidated, at
MATERIALS AND METHODS
least in part, because the recycling itineraries of
Cell Lines and Cell Culture—Calu-3 cells,
CFTR in these cells are practically unknown.
obtained from the American Type Culture
Recycling is a dynamic trafficking event
Collection (Manassas, VA) were seeded at 2 x 106
during which proteins are transported from
on Transwell permeable supports (4.67 cm2, 0.4
intracellular compartments and inserted into the
µm pore size; Corning Corporation, Corning, NY)
plasma membrane. Recycling requires formation
coated with Vitrogen plating medium (VPM)
of macromolecular complexes with numerous
containing Dulbecco’s modified Eagle’s medium
adaptors, including myosin motors (10,11). The
(DMEM; Invitrogen, Carlsbad, CA), human
non-conventional processive myosin motors – a
fibronectin (10 µg/ml; BD Biosciences, San Jose,
class of mechanoenzymes that convert the energy
CA), 1% purified collagen (PureCol™; Inamed,
from ATP hydrolysis into mechanical work – take
Fremont, CA), and bovine serum albumin (BSA;
multiple steps on actin filaments (12-15) and move
10 µg/ml; Invitrogen) and maintained as polarized
cargo proteins contained in specific transport
monolayers in air-liquid interface culture as
organelles. The plus-end directed movement
D
described previously (28,29). Serum was removed o
characteristic of class V myosins supports w
anterograde trafficking, including recycling from the medium 24 hours before experiments to nlo
a
(reviewed in Refs. (16,17). The plasma membrane augment cell polarization and cell cycle ded
recycling in non-polarized cells and the apical synchronization (30,31). CFBE41o- cells fro
m
plasma membrane recycling in polarized epithelial (∆F508/∆F508), stably transduced with either WT- h
CFTR or ∆F508-CFTR were a generous gift from ttp
cells occur primarily in Rab11a-positive Dr. J.P. Clancy (32). The stably transduced ://w
endosomes (18-20). A recent study demonstrates w
CFBE41o- cells were seeded at 1 x 106 on VPM w
that recycling of CFTR in human airway epithelial .jb
cells occurs in the Rab11a recycling compartment coated Transwell (4.67 cm2) or Snapwell (1.12 c.o
(21). cm2, 0.4 µm pore size; Corning Corporation) brg/
Myosin Vb interacts specifically with the GTP permeable supports and maintained as polarized y g
u
bsuogugneds (tiin.eg. mtheamt bmraynoes ibno uVnbd ) mfoarym boef Rreacbr1u1itae d(2 2to) mdeosncorilbayeedr s pirne viaoiru-slilqyu id(2 1in,3te2r)f.a cIen cualdtduritei oans, est on A
CFBE41o- cells were seeded on VPM coated p
Rusaibn1g 1dao-mspiencainfti cn ergeactyivceli nfrga gemnednotsso omf ems.y oSstiund Viebs pcmla2s tipcl atitses;u eC cournltiunrge pClaotrepso (r1a txio n1)0.6 cTeol lsi npcerre a9s.4e ril 3, 20
demonstrate that myosin Vb, together with Rab11a 1
9
export from the endoplasmic reticulum (ER) and
facilitates protein recycling in several polarized
expression of ∆F508-CFTR at the plasma
and non-polarized cell models (23-26).
membrane, cells were cultured at 27 ºC for 36
Furthermore, a recent study demonstrates that
hours (21,32). FBS and the selection antibiotic
myosin Vb facilitates trafficking of the
were removed from the media 24 hours before
α-amino-3-hydroxy-5-methyl-4-isoxazole
experiments. HEK293 cells stably expressing WT-
propionate (AMPA)-type glutamate receptor
CFTR were a generous gift of Dr. Neil Bradbury
subunit GluR1 in subpopulations of neurons by a
(33). HEK293 cells were cultured in VPM coated
Rab11 dependent mechanism (27). However,
tissue culture flasks (3 x 106 cells per 25 cm2 flask;
nothing is known about the role of myosin Vb in
Corning Corporation) or plastic tissue culture
CFTR trafficking in human airway epithelial cells.
plates (1 x 106 cells per 9.4 cm2 plate) and
Thus, the objective of the present study was to
maintained as described previously (29,34). FBS
test the hypothesis that endogenous myosin Vb
and the selection antibiotic were removed from the
facilitates CFTR recycling in Rab11a-specific
media 24 hours before experiments. COS-7 cells
apical recycling compartment in polarized human
(American Type Culture Collection) were cultured
airway epithelial cells. We report that myosin Vb
in tissue culture dishes (0.75x106 cells per 55 cm2
is required for the recycling of WT-CFTR and
dish) in DMEM supplemented with 100 µg/ml
∆F508-CFTR to the apical membrane of polarized
2
streptomycin, 100 U/ml penicillin, and 10% FBS used as a target for the small interfering RNA
in a 5% CO / 95% air incubator at 37 °C as (siRNA) directed against human myosin Vb. The
2
previously described (35). Calu-3 cells were double stranded siRNA against human myosin Vb
studied because they express high levels of (siMyoVb) was synthesized by Qiagen. The
endogenous CFTR; however, Calu-3 cells are very double stranded non-silencing siRNA control
difficult to transfect. CFBE41o- cells were used (Non-sil.siRNA) targeted the following sequence,
because they are isogenic except for CFTR, and 5’-AATTCTCCGAACGTGTCACGT-3’, which
they are relatively easy to transfect. However, showed no significant homology with any other
CFBE41o- cells express WT-CFTR and ∆F508- gene known analyzed using BLAST search. Non-
CFTR transgenes, thus, studies on these cells do sil.siRNA was purchased from Qiagen (catalog
not examine endogenous CFTR. Finally, HEK293 #1022076). Transfection of siMyoVb or Non-
and COS-7 cells were also used because they are sil.siRNA into HEK293 and CFBE41o- cells was
very easy to transfect and are a well accepted conducted using HiPerFect (Qiagen) according to
model to conduct biochemical studies including the manufacturer’s instructions. HEK293 or
co-immunoprecipitation studies. However, CFBE41o- cells (1x106) were plated on 9.4 cm2
because these cells are fibroblasts and are not tissue culture plates and incubated with the
polarized, their use was limited. optimized transfection mixture (5 nM of either
RNA Isolation and Reverse Transcription- siMyoVb or Non-sil.siRNA and 10 µl of
D
o
PCR—Reverse Transcription (RT)-PCR studies HiPerFect) at 37°C. After 24 hours, the w
n
were conducted to examine the endogenous transfection mixture was replaced with fresh cell lo
a
d
expression of myosin Vb in human airway culture medium and cells were cultured for ed
epithelial cells (Calu-3 and CFBE41o-) and in different lengths of time. Because silencing of fro
m
HEK293 cells, as previously described (29). Total myosin Vb expression was maximum between 48 h
ttp
RNA was isolated from each cell line using the and 72 hours, studies were conducted at ~ 60 hrs ://w
RNeasy Mini Kit (Qiagen, Valencia, CA) and after transfection with siMyoVb. The efficiency of w
w
DNase treated (DNA-Free, Ambion, Austin, TX) silencing myosin Vb expression was assessed by .jb
c
to remove contaminating DNA. Total RNA was Quantitative RT-PCR (Q-RT-PCR) and by .o
rg
quantified using a NanoDrop spectrophotometer Western blotting. The specificity of silencing b/
y
(NanoDrop Technologies, Rockland, DE) and the myosin Vb expression was determined by g
u
e
quality was assessed using an Agilent 2100 examining the expression of myosin Vc – another st o
Bioanalyzer (Agilent Technologies, Wilmington, member of class V myosin family expressed in n A
p
DReEt)r.o sTcwriop-ts teRpe vReTrs-eP CTRr awnsacsr ippetarfsoer m(eAdm ubsioinng, eexppitrheeslsiiaoln coefl lms y(o3s7in), VaIn d– ab ym yeoxsaimn iknninogw nt htoe ril 3, 2
0
Austin, TX) with random decamers and Taq DNA facilitate CFTR endocytosis (29) by Western 1
9
Polymerase (Invitrogen). Oligonucleotide primers blotting. A predesigned siRNA, targeting another
flanking nucleotides 3692-4396 (sense: 5’- non-conserved region of the human myosin Vb
ACCAAGCCACGCAGAATAACT-3’ and gene (AB032945) was purchased from Qiagen
antisense: 5’-GGACCGTGACCTGCCTGTTGA- (Catalog #SI00653527). Because siMyoVb (target
3’) were designed using Oligo 6.1 Primer Analysis sequence 4227-4248) silenced myosin Vb protein
Software (Plymouth, MN) and synthesized by expression more effectively than SI00653527,
IDTDNA (Integrated DNA Technologies, siMyoVb was used in subsequent experiments
Coralville, IA). These primers were used to (myosin Vb expression was 44.3+10.4% and
amplify a single 705 bp myosin Vb PCR product. 17.2+9.2% of control after transfecting HEK293
The product was subcloned into pCR4-TOPO cells with SI00653527 and siMyoVb, respectively,
(Invitrogen) and the sequence verified by ABI as assessed by Western blotting; n=3 in each
PRISM dye terminator cycle sequencing (Applied group).
Biosystems; Foster City, CA). Experiments were performed to silence the
RNA Mediated Interference—A sequence expression of myosin Vb in CFBE41o- cells
corresponding to a non-conserved region of human cultured on Transwell and Snapwell permeable
myosin Vb cDNA (AB032945; (36)) 4227-4248, supports. CFBE41o- cells (1x106) stably
5’-AACGAGAATCTGGACCTTAAA-3’ was expressing either WT-CFTR or ∆F508-CFTR
3
were plated on 9.4 cm2 tissue culture plates and gene expression (fg/ng cDNA). The expression
incubated with the optimized transfection mixture level of human myosin Vb was reported as a
(10 nM of either siMyoVb or Non-sil.siRNA and percent of control.
20 µl of HiPerFect) at 37 °C. After 24 hours, cells Plasmids and Transient Transfection—cDNA
were trypsinized and plated on VPM coated encoding the entire tail domain of mouse myosin
Transwell or Snapwell permeable supports and Vb (MyoVb-T) and cDNA encoding a truncated
cultured for an additional 6 days to establish globular tail region of mouse myosin Vb tail
polarized monolayers. CFBE41o- cells stably domain (MyoVb-GT) were amplified from the
expressing ∆F508-CFTR were cultured at 27 °C mouse cDNA library by RT-PCR as described
during the last 36 hours to increase trafficking and previously (38) using the following pairs of
expression of ∆F508-CFTR in the plasma o l i g o n u c l e o t i d e s : 5 ’ -
membrane. The efficiency of silencing myosin Vb GGATCCAGCTCCCCAGACAGCTACAGC -3’
expression in polarized CFBE41o- cells was (MyoVb-T sense primer), 5’-
assessed by Western blotting. TCAGACTTCATTGAGAAACT-3’ (MyoVb-
Quantitative RT-PCR—Quantitative RT-PCR T/GT anti-sense primer), 5’-
(Q-RT-PCR) studies were conducted to measure CGGATCCGAAAAGCTGGAGAAGAATGA-3’
silencing of myosin Vb expression by siMyoVb. (MyoVb-GT sense primer). Addition of the FLAG
The sequence of human myosin Vb (AB032945) tag to the N-terminus of the myosin Vb constructs
D
o
was submitted to the Assays-by-Design services and construction of the expression vector, pEF was w
n
(Applied Biosystems) for the primers and probe performed as previously described (38-40). The lo
a
d
design. The probe target was set to a predicted resulting fusion fragments consisted of part of the ed
exon-exon splice junction. The probe was 6-FAM proximal/medial tail region and the entire globular fro
m
dye-labeled with an MGB (minor groove binding) tail region of myosin Vb tail domain (aa 1231- h
ttp
modification and non-fluorescent quencher on the 1818; FLAG-MyoVb-T) and a truncated globular ://w
3' end. The primers (18 µM each; sense: 5’- tail region of myosin Vb tail domain (aa 1384- w
w
GGAACCTGGTGACAGACTTGAAG-3’; 1818; MyoVb-GT). A plasmid containing the GFP .jb
c
antisense: 5’-CCGGATGCACATGTAGAGGAT- tagged tail domain of human myosin Vc (aa 902- .o
rg
3’), the probe (5 µM; antisense: FAM- 1742; MyoVc-Tail) was generated as previously b/
y
CTGTGCCCGACAGCAT) combined with described using the eukaryotic expression vector g
u
e
TaqMan Universal Master Mix (Applied pEGFP-C2 (37). Constructs were sequence st o
Biosystems) and cDNA were placed in triplicates verified by ABI PRISM dye terminator cycle n A
p
icny cal e9r 6(-AwBelIl fPoRrmISaMt sp7e7c0t0ro fSlueoqruoemnceetr iDc ettheecrtmioanl sweaqsu epnecrinfogr. mTeradn safeccctoiordni nogf cteol lsm wanituhf apclatusrmeird’ss ril 3, 2
0
System, Applied Biosystems). Q-RT-PCR instructions. HEK293 cells were transfected using 1
9
products were size-verified on a low melting point Effectene® (Qiagen) or FuGENE6 (Roche
agarose gel, subcloned into pCR4-TOPO and Diagnostics, Indianapolis, IN) and transfection of
sequence-verified by ABI PRISM dye terminator CFBE41o- cells was performed using FuGENE6
cycle sequencing. The standard curve prepared (29). The above cells were harvested 48 hours
from the myosin Vb plasmid DNA isolated from after transfection. Transfection of COS-7 cells was
HEK293 cells (not siRNA transfected; control) performed using LipoFECTAMINE Plus (Life
demonstrated R2 >0.99 and was linear over a 9-log Technologies Inc., Rockville, MD) and the cells
range. Equivalent amplification efficiencies of were harvested 72 hours after transfection (35).
standard and target molecules were observed. Gel Production of Affinity-purified Antibodies to
analysis, SYBR green melting curve dissociation Myosin Vb—Two polyclonal antibodies against
analysis and sequencing revealed a single PCR myosin Vb were raised in rabbits: antibody 2506,
product. Raw data were analyzed; baseline and a g a i n s t t h e s e q u e n c e
threshold values set and gene expression QDSKKVQAEPPQTDIDLDPN (aa 1174-1193)
interpolated using the standard curve. The cDNA and antibody 2507, against the sequence
generated during reverse transcription and used as AGRNAEPNINARSSWPNSE (aa1278-1296) in
a template, was quantified by the NanoDrop the proximal/medial tail region of the tail domain
spectrophotometer and data were calculated as (PickCell Laboratories BV, Amsterdam, The
4
Netherlands). These myosin Vb sequences were (Pierce Chemical Co., Rockford, IL) at 4 ºC. The
absent from any other known genes, as analyzed pre-cleared lysates (1000 µl) were added to the
by a BLAST search. Rabbits were injected and protein G or protein A Sepharose beads (wet
boosted two times. The antiserum was affinity volume 120µl) antibody complexes. CFTR was
purified and the specificity of the anti-myosin Vb immunoprecipitated by incubation with the mouse
antibodies was determined by the ability of the M3A7 antibody-protein G sepharose complexes
antibodies to detect endogenous myosin Vb and and myosin Vb was immunoprecipitated by
the FLAG-tagged myosin Vb fragments. The incubation with the rabbit 2506 antibody-protein A
selectivity of the antibodies for myosin Vb was Sepharose complexes. After washing the protein G
determined by the ability of the anti-myosin Vb or protein A Sepharose complexes with the IP
antibodies to detect silencing of endogenous buffer, immunoprecipitated proteins were eluted
myosin Vb. by incubation at 100 °C for 3 minutes in sample
Other Antibodies—The antibodies used were: buffer (BioRad Laboratories) containing 80 mM
rabbit anti-myosin Va (41) and rabbit anti-myosin DTT. Immunoprecipitated proteins were separated
Vc antibody 199 ((37), rabbit anti-myosin VI tail by SDS-PAGE using 7.5% or 15% gels (BioRad
antibody (a gift from T. Hasson; USCD, San Laboratories) and analyzed by Western blotting
Diego, CA; (42)), mouse anti-human CFTR C- with an appropriate primary antibody and an anti-
terminus specific, clone 24-1 (R&D Systems; mouse or anti-rabbit HRP secondary antibody. The
D
o
Minneapolis, MN), mouse anti-CFTR, clone immunoreactive bands were visualized with w
n
M3A7 (Upstate Biotechnology, Lake Placid, NY), Western Lightning Chemiluminescence Reagent lo
a
d
mouse anti-GFP, clone JL-8, mouse anti-ezrin (BD Plus (PerkinElmer LAS, Inc., Boston, MA). I n ed
Biosciences), mouse anti-FLAG M2, HRP- addition, a tissue blot containing lysates from fro
m
conjugated, mouse anti-FLAG M2, and rabbit anti- human brain (SDS-PAGE 5-20%; Chemicon h
ttp
FLAG F7425 (Sigma-Aldrich, St. Louis, MO), International, Inc., Temecula, CA) was used for ://w
rabbit anti-Rab11a (Zymed, South San Francisco, Western blotting, as described above. w
w
CA), mouse anti-Rab8 (BD Transduction GST Pull-down Assay—cDNA encoding .jb
c
Laboratories, Palo Alto, CA), goat anti-mouse and mouse Rab11a was subcloned into the pGEX-4T-3 .o
rg
goat anti-rabbit HRP secondary antibodies vector (Amersham Biosciences, Piscataway, NJ) b/
y
(BioRad Laboratories, Hercules, CA). All (47). GST-Rab11a was expressed in Escherichia g
u
e
antibodies were used at the concentrations coli JM109 and purified by standard protocols. st o
recommended by the manufacturer or as indicated Glutathione-Sepharose beads (wet volume 10 ml; n A
p
in thIe mfigmurue nleogepndrse. c i p i t a t i o n a n d Athme perusrhifaimed BGiSoTsc-Rieanbc1e1sa) wcoeurep liendc uwbaitthed 1 f0o rm 1g h oaft ril 3, 2
0
Immunoblotting—CFTR and myosin Vb were 4 °C with COS-7 cell lysates (400 ml) containing 1
9
immunoprecipitated from cell lysates by methods either the FLAG-Myo Vb-T or FLAG-Myo Vb-
described previously (21,29). Briefly, cultured GT in a binding buffer (50 mM HEPES-KOH, pH
cells were solubilized in an immunoprecipitation 7.2, 150 mM NaCl, 1 mM MgCl , 1% Triton X-
2
(IP) buffer containing 150 mM NaCl, 50 mM Tris, 100, 0.5 mM GTPgS, 0.1 mM
pH 7.2, 0.1% IGEPAL (Sigma-Aldrich), 5 mM phenylmethylsulfonyl fluoride, 10 µM leupeptin,
MgCl , 5 mM EDTA, 1mM EGTA, 30 mM NaF, and 10 µM pepstatin A). After washing the
2
1 mM Na VO , Complete Protease Inhibitor Glutathione-Sepharose beads-GST-Rab11a
3 4
cocktail (Roche), and 40 µM guanosine 5’O-[3- complexes with the binding buffer, the bound
thio]triphosphate (GTP-g-S; EMD Biosciences, proteins were eluted by incubation with a sample
San Diego, CA), a non-hydrolysable analog of buffer and analyzed by 10% SDS-PAGE followed
GTP (43-46). After centrifugation at 14,000 x g by immunoblotting with HRP-conjugated anti-
for 15 minutes to pellet insoluble material, the FLAG M2 antibody. The immunoreactive bands
soluble lysates were incubated for 10 minutes at were visualized with an Enhanced
30 ºC with an additional 40 µM GTP-g-S. After Chemiluminescence Kit (Amersham Biosciences).
cooling to 4 °C, the soluble lysates were pre- Biochemical Determination of the Apical
cleared by incubation with protein G or protein A, Membrane CFTR—The biochemical determination
as appropriate, conjugated to Sepharose beads of apical membrane CFTR at steady state was
5
performed by domain selective cell surface Development of Anti-myosin Vb
biotinylation using EZ-Link™ Sulfo-NHS-LC- Antibodies—Three members of the myosin V
Biotin (Pierce), as described previously in detail family, encoded by three different genes are
(48,49). expressed in mammalian cells (37,55-57). Myosin
Endocytic and Recycling Assays—Endocytic Va is predominantly expressed in neuronal cells.
and recycling assays were performed in HEK293 In contrast, myosin Vb and Vc are enriched in
cells, essentially as described previously (49-53). epithelial cells. RT-PCR studies revealed that
For both assays, the plasma membrane proteins human airway epithelial cells (Calu-3 and
were first biotinylated at 4 °C using EZ-Link™ CFBE41o-) and HEK293 cells express a single
Sulfo-NHS-SS-Biotin (Pierce). For the endocytic myosin Vb product. The sequence of the PCR
assay, cells were warmed to 37 °C for 2.5, 5, 7.5, product was identical with the sequence of the
or 10 minutes after biotinylation and the disulfide human myosin Vb (Gene Bank accession number
bonds on Sulfo-NHS-SS-biotinylated proteins AB032945 for KIAA1119 protein (36)), except
remaining in the plasma membrane were reduced that the sequence in airway cells and HEK293
by L-glutathione (GSH; Sigma-Aldrich) at 4 °C. cells lacked exon 30 (Figure 1A & 1B). We
At this point in the protocol, biotinylated proteins developed two affinity purified, rabbit polyclonal
reside within the endosomal compartment. anti-myosin Vb antibodies, as described in
Subsequently, cells were lysed, biotinylated Materials and Methods. Antibodies 2506 and
D
o
proteins were isolated by streptavidin-agarose 2507, directed against unique amino acid w
n
beads, eluted into SDS-sample buffer, and sequences in the proximal/medial tail region of the lo
a
d
separated by 7.5% SDS-PAGE. For the recycling tail domain of myosin Vb (Figure 1A) recognized ed
assay, cells were warmed to 37 °C for 5 minutes a single endogenous product in Calu-3, CFBE41o- fro
m
after biotinylation to load endocytic vesicles with , and HEK293 cell lysates of the appropriate h
ttp
biotinylated proteins, including CFTR. molecular mass (~214-kDa; Figure 2A). In ://w
Subsequently, cells were cooled immediately to 4 addition, antibody 2507 recognized the FLAG- w
w
°C and the disulfide bonds on Sulfo-NHS-SS- tagged myosin Vb tail fragment containing the .jb
c
biotinylated proteins in the plasma membranes fragment of the proximal/medial tail region used .o
rg
were reduced by GSH at 4 °C. Subsequently, cells as an epitope for this antibody (FLAG-MyoVb-T; b/
y
were either lysed or warmed again to 37 °C for 2.5 Figure 1A, 1C, & 2B). The endogenous product g
u
e
or 5 minutes (to allow endocytosed, biotinylated recognized on Western blots as myosin Vb by st o
CFTR to recycle to the plasma membrane). Cells antibodies 2506 and 2507 did not result from n A
p
wboenred st hoenn cSouollfeod-N aHgaSi-nS tSo- b4i o°tCin,y alantde dth per odtiesiunlfsi dine cnreousrso-nraela cmtienmg bweri tho f mclyaosssi nV Vmay o–s itnhse –p rbiemcaaruislye ril 3, 2
0
the plasma membranes were reduced with GSH. antibodies 2506 and 2507 did not recognize 1
9
Recycling of endocytosed CFTR was calculated as myosin Va in human brain lysates (Figure 2A).
the difference between the amount of biotinylated Because myosin Vc is enriched in epithelial cells
CFTR after the first and second GSH treatment. and is expressed by Calu-3, CFBE41o-, and
Ussing Chamber Measurements—Ussing HEK293 cells (Figure 2A), studies were
chamber measurements were performed as conducted to examine the specificity of antibodies
previously described (54). 2506 and 2507 for myosin Vb using RNA
Data Analysis and Statistics—Statistical mediated interference. HEK293 cells were
analysis of the data was performed using transfected with double stranded, small interfering
GraphPad Prism version 4.0 for Macintosh RNA (siRNA) specific for a non-conserved region
(GraphPad Software Inc., San Diego, CA). The of the human myosin Vb sequence (siMyoVb;
means were compared by a two-tailed t-test. A P target sequence 4227-4248) or with the non-
value <0.05 was considered significant. Data are silencing double stranded RNA (Non-sil.siRNA)
expressed as mean ± S.E. control. siMyoVb decreased myosin Vb
expression as determined by Q-RT-PCR (Figure
RESULTS 3A). In contrast, Non-sil.siRNA had no effect on
Determination of Endogenous Expression of the endogenous expression of myosin Vb when
Myosin Vb in Human Airway Epithelial Cells and compared with the non-transfected cells (Mock).
6
Western blot analysis with antibodies 2506 and data suggest that myosin Vb is coexpressed with
2507 confirmed silencing of myosin Vb CFTR in Rab11a-specific recycling endosomes in
expression (Figure 3B & 3C). siMyoVb did not polarized human airway epithelial cells.
decrease the expression of myosin Vc, as Endogenous Myosin Vb Facilitates the Apical
determined by Western blotting with the anti- Plasma Membrane Expression of WT-CFTR and
myosin Vc antibody (Figure 3B), indicating that ∆F508-CFTR in Polarized Human Airway
the anti-myosin Vb antibodies specifically Epithelial Cells—Expression of CFTR in the
recognized myosin Vb and did not cross-react with plasma membrane is determined, in part, by the
myosin Vc. Taken together the above data relative rates of CFTR endocytosis and recycling
demonstrate that the affinity purified rabbit (61-63). If myosin Vb facilitates CFTR recycling,
polyclonal anti-myosin Vb antibodies 2506 and it can be predicted that reduced expression of
2507 are specific and selective for myosin Vb. myosin Vb would attenuate CFTR recycling and
CFTR Co-imunoprecipitates with the thus, would decrease expression of CFTR in the
Endogenous Myosin Vb–Rab11a Complex in plasma membrane. To test this prediction,
Polarized Human Airway Epithelial expression of myosin Vb was silenced using RNA
Cells—Members of the myosin V family are mediated interference. CFBE41o- cells stably
recruited to distinct transport organelles by expressing WT-CFTR, were transfected with
organelle-specific Rab GTPases (58-60). Myosin siRNA specific for a non-conserved region of the
D
o
Vb is thought to be specifically recruited to human myosin Vb sequence (siMyoVb) or with w
n
Rab11a-specific recycling endosomes because the Non-sil.siRNA control as described in lo
a
d
myosin Vb interacts specifically with the GTP- Materials and Methods. As predicted, silencing ed
bound (i.e. membrane bound) form of Rab11a myosin Vb decreased expression of WT-CFTR fro
m
(22). We examined whether endogenous myosin specifically in the plasma membrane (Figure 5). h
ttp
Vb and Rab11a interact in polarized human airway Compelling evidence demonstrates that ://w
epithelial cells. Myosin Vb was protein trafficking in epithelial cells can be w
w
immunoprecipitated from polarized Calu-3 cells affected by the state of cell polarization (64,65). .jb
c
using a polyclonal anti-myosin Vb antibody Thus, additional studies were conducted to .o
rg
(2506). Western blot analysis of the examine the effect of myosin Vb silencing on the b/
y
immunoprecipitated complexes demonstrated that expression of WT-CFTR in the apical plasma g
u
e
endogenous myosin Vb interacts specifically with membrane in polarized CFBE41o- cells. st o
endogenous Rab11a (Figure 4A). CFBE41o- cells stably expressing WT-CFTR were n A
p
RabI1f1 ma-ysopseicni fVicb rfeaccyilcitlaintegs CenFdToRs otmraeffsi,c kCinFgT iRn cduayltsu raefdte or ntr asenmsfie-cpteiormn ewaibtlhe sgiMrowyothV bsu apsp odretssc rfiobre 7d ril 3, 2
0
should co-immunoprecipitate with the myosin in Materials and Methods. Under these conditions 1
9
Vb–Rab11a complex. To test this prediction, CFBE41o- cells formed polarized monolayers
CFTR was immunoprecipitated from Calu-3 cells (21,32). Silencing myosin Vb in polarized
using a monoclonal anti-CFTR antibody (M3A7). CFBE41o- cells decreased expression of WT-
Western blot analysis of the immunoprecipitated CFTR in the apical plasma membrane (Figure 6A-
complexes demonstrated that endogenous CFTR 6C).
interacts with myosin Vb and Rab11a in polarized Recent study reveals that, similar to WT-
Calu-3 cells (Figure 4B). These data confirm and CFTR, if ∆F508-CFTR is released from the ER in
extend our previous observation that CFTR human airway epithelial cells, it undergoes
coimmunoprecipitates with Rab11a in polarized trafficking to the plasma membrane in Rab11a-
human airway epithelial cells (21). Myosin Vc specific recycling endosomes (21). Thus, we
associates with the Rab8-specific, transferrin hypothesized that myosin Vb will also facilitate
accessible vesicular compartment in HeLa cells recycling of ∆F508-CFTR. CFBE41o- cells stably
but is excluded from the Rab11a-specific recycling expressing ∆F508-CFTR cells were cultured on
system (37). Rab8 did not coimmunoprecipitate semi-permeable growth supports for 7 days after
with myosin Vb and neither Rab8 nor myosin Vc transfection with siMyoVb. To increase the export
coimmunoprecipited with CFTR in human airway of ∆F508-CFTR from the ER and thus, the
epithelial cells (Figure 4B). Taken together, these expression of ∆F508-CFTR in the apical
7
membrane, cells were cultured at 27 ºC for 36 with Rab11a was examined by pull down
hours (21,32). As with WT-CFTR, silencing experiments. The affinity purified GST-Rab11a
myosin Vb resulted in decreased expression of (47) was immobilized on Glutathione Sepharose
∆F508-CFTR in the apical membrane (Figure 6D- beads and incubated with COS-7 cell lysates
6F). Taken together the above data are consistent containing either the FLAG-MyoVb-T or FLAG-
with the view that myosin Vb facilitates the MyoVb-GT. Western blot analysis of the protein
endocytic recycling and the apical membrane complexes eluted from the beads revealed that
expression of WT-CFTR and ∆F508-CFTR in only the FLAG-MyoVb-T formed a complex with
polarized human airway epithelial cells. GST-Rab11a (Figure 8A). These data confirm
Endogenous Myosin Vb Facilitates CFTR- previous results that multiple sites of contact are
mediated Cl- Secretion Across Polarized Human necessary for myosin Vb binding to Rab11a (22).
Airway Epithelial Cells—Because silencing Furthermore, the data suggest that the FLAG-
endogenous myosin Vb decreased the expression MyoVb-T is sufficient for recruitment to Rab11a-
of CFTR in the apical membrane, we predicted specific recycling endosomes. Thus, if myosin Vb
that it would also inhibit the CFTR mediated Cl- facilitates trafficking of CFTR in Rab-11a-specific
secretion across polarized CFBE41o- cells. recycling endosomes, the FLAG-MyoVb-T should
CFBE41o- cells stably expressing WT-CFTR were have a dominant negative effect on CFTR
cultured on semi-permeable growth supports for 7 trafficking by displacing endogenous myosin Vb
D
o
days after transfection with siMyoVb. As from binding to the Rab11a-specific recycling w
n
predicted, siMyoVb inhibited the forskolin- endosomes. If the above predictions are correct the lo
a
d
sstuigmguelsatt etdh aIts ce (nFdioggueren o7u).s Tmaykoesni nto Vgebt heenrh tahnecsee sd tahtae FshLoAulGd- Malsyoo Vcob--iTm mbuunt onporte ctihpei taFteL AwGith-M CyFoTVRb.- GTTo ed fro
m
CFTR-mediated transepithelial Cl- secretion in test this hypothesis, HEK293 cells stably h
ttp
polarized human airway epithelial cells by a expressing WT-CFTR were transiently transfected ://w
mechanism that involves increasing the total with either the FLAG-MyoVb-T or FLAG- w
w
number of CFTR Cl- channels in the apical plasma MyoVb-GT. CFTR was immunoprecipitated with .jb
c
membrane. a mouse anti-CFTR antibody (M3A7). Western .o
rg
The Dominant Negative FLAG-MyoVb-T blot analysis of the immunoprecipitated complexes b/
y
Inhibits Expression of CFTR in the Plasma with a rabbit anti-FLAG antibody demonstrated g
u
e
Membrane—The direct interaction between the that, as predicted, only the FLAG-MyoVb-T co- st o
myosin Vb tail domain and Rab11a (22) that is immunoprecipitated with CFTR (Figure 8B). We n A
p
ptoro proescedy ctloi nogc cuer nudpoosno rmecersu itmsuegngt eosf tms yothsiant Vba hMyypooVthbe-sTiz edw othualdt thaerr edsotm CinFaTntR n ergeactyivceli nFgL AaGnd- ril 3, 2
0
recombinant, motorless myosin Vb fragment decrease CFTR expression in the plasma 1
9
capable of binding to Rab11a should be able to membrane. To test this hypothesis, HEK293 cells
displace endogenous myosin Vb from interacting stably expressing WT-CFTR were transiently
with the Rab11a-specific recycling endosomes. transfected with either the dominant negative
The myosin Vb fragment would be expected to FLAG-MyoVb-T or with the FLAG-MyoVb-GT
have a dominant negative effect on trafficking in as a control. As demonstrated in Figure 9A and
Rab11a-specific recycling endosomes. Two 9B, the FLAG-MyoVb-T decreased the plasma
Rab11a binding sites, located in the globular tail membrane expression of CFTR. The above data
region of myosin Vb tail domain are necessary for suggest that the FLAG-MyoVb-T decreased the
binding Rab11a ((22); Figure 1A). Thus, we plasma membrane expression of CFTR by
generated two recombinant FLAG-tagged myosin displacing endogenous myosin Vb from binding to
Vb fragments, one consisting of the entire tail Rab11a-specific recycling endosomes containing
domain and containing both Rab11a binding sites CFTR as cargo.
(FLAG-MyoVb-T) and a second fragment, Myosin Vc is excluded from the Rab11-
consisting of a truncated globular tail region and specific recycling compartment (37) and does not
containing only one Rab11a binding site (FLAG- coimmunoprecipitate with CFTR (Figure 4B).
MyoVb-GT) (Figure 1C). The ability of the Thus, it would be expected that a dominant
FLAG-tagged myosin Vb fragments to interact negative fragment of myosin Vc should not affect
8
the plasma membrane expression of CFTR. These minute time point. As illustrated in Figure 10C
studies were conducted to examine whether the and 10D, the FLAG-MyoVb-T did not increase
effect of the FLAG-MyoVb-T on the plasma CFTR endocytosis. Taken together, these data
membrane expression of CFTR was specific. indicate that the dominant negative FLAG-
HEK293 cells stably expressing WT-CFTR were MyoVb-T decreased the plasma membrane
transfected with either the GFP-tagged myosin Vc expression of CFTR by specifically inhibiting
tail fragment (GFP-MyoVc-T), previously shown recycling of CFTR.
to affect the trafficking of the transferrin receptor
(37) or with the GFP control. The GFP-MyoVc-T DISCUSSION
did not decrease the plasma membrane expression The major new observation in the present
of CFTR (Figure 9C & 9D). study is that myosin Vb regulates CFTR mediated
Myosin Vb Facilitates CFTR Recycling—The Cl- secretion across human airway epithelial cells
dominant negative FLAG-MyoVb-T could by facilitating the apical membrane recycling of
decrease the plasma membrane expression of WT-CFTR and ∆F508-CFTR. Our data provide
CFTR by either inhibiting CFTR recycling or by the first biochemical evidence that endogenous
stimulating CFTR endocytosis, or both. myosin Vb interacts with endogenous Rab11a in
Accordingly, studies were conducted to test the human airway epithelial cells.
hypothesis that the FLAG-MyoVb-T decreased
Previous studies have shown that myosin Vb D
o
expression of CFTR in the plasma membrane by w
and Rab11a facilitate recycling (18-20,22). n
inhibiting CFTR recycling. HEK293 cells stably lo
Lapierre et al first elucidated the role of myosin a
d
eGxFpPre scsoinntgr oWl, Tt-hCe FFTLRA wGe-Mre ytoraVnbs-fGecTte dc owntirtho l thoer Vb in the plasma membrane recycling of the ed fro
transferrin and polymeric IgA receptor (22). These m
dominant negative FLAG-MyoVb-T. CFTR h
investigators established the interaction between ttp
recycling was measured at 2.5 and 5 minutes as myosin Vb and the GTP-bound (i.e. membrane ://w
described in Materials and Methods. CFTR w
bound) form of Rab11a by yeast-two hybrid w
recycling was similar in cells transfected with the screening and demonstrated co-localization .jbc
FLAG-MyoVb-GT and the GFP control (Figure .o
between the GFP-tagged myosin Vb and rg
10A and 10D). These data confirm that this endogenous Rab11a and between the GFP-Rab11a by/
recombinant myosin Vb fragment that neither g
and endogenous myosin Vb in MDCK and HeLa u
e
interacts with Rab11a nor coimmunoprecipitates (22). Subsequent studies confirmed the role of st o
with CFTR serves as a good negative control in myosin Vb in the plasma membrane recycling in n A
p
othuer FstLuAdyG. -AMsy oilVlubs-trTa tdedec irne aFsiegdu rCeF 1T0RA raencdy c1li0nDg., MmyDoCsiKn aVnbd, HtoegLeat hceerl lsw (i6th6 ) Raanbd1 d1eam foancsiltirtaatteeds tthhaet ril 3, 2
0
The decrease in CFTR recycling is consistent with 1
recycling of the muscarinic receptor M in the 9
4
the decrease in the plasma membrane expression
neuronotypic PC12 cells (23) and the chemokine
of CFTR observed in cells transfected with the
receptor CXCR2 in leukemia 2H3 cells (24), and
dominant negative GFP-MyoVb-T and indicates
the canalicular formation of bile in hepatic WIF-
that myosin Vb facilitates CFTR recycling.
B9 cells (26). Furthermore, a recent study by Lise
Myosin Vb does not regulate CFTR
et al demonstrated that myosin Vb mediates
Endocytosis—As noted above, the decreased
trafficking of the glutamate receptor subunit
expression of CFTR in the plasma membrane
GluR1 by a Rab11a dependent mechanism in
caused by the FLAG-MyoVb-T could also result
neuronal subpopulations (27). However, there is
from an increase in CFTR endocytosis. HEK293
no biochemical evidence that endogenous myosin
cells stably expressing WT-CFTR were
Vb and Rab11a interact (22). Furthermore, the role
transfected with the GFP control, the FLAG-
of myosin Vb in CFTR trafficking has not been
MyoVb-GT control or the dominant negative
reported in respiratory epithelial cells. We
FLAG-MyoVb-T and CFTR endocytosis was
previously demonstrated that Rab11a facilitates
measured as described in Materials and Methods.
CFTR recycling in polarized human airway
CFTR endocytosis was linear between 0 and 5
epithelial cells (21). In the present study we
minutes in cells transfected with the GFP control
provide evidence that endogenous myosin Vb
(Figure 10B). Thus, data are reported at the 5-
9
coimmunoprecipitates with endogenous Rab11a myosin Vc is recruited to Rab8-specific vesicular
and facilitates CFTR recycling in human airway compartment and is excluded from the Rab11a-
epithelial cells. An inability to establish specific compartment (37). Rab8 has been
biochemically the association between myosin Vb implicated in regulating transport of proteins from
and Rab11a was attributed in the past to a weak or the trans-Golgi to the basolateral membrane
indirect nature of the myosin Vb–Rab11a (74,75). CFTR did not coimmunoprecipitate with
interaction or both (22). Higher expression levels either myosin Vc or Rab8 in human airway
of endogenous myosin Vb and Rab11a in human epithelial cells. The tail fragment of myosin Vc
airway epithelial cells (Calu-3) compared to (GFP-MyoVc-T), which perturbed transferrin
MDCK or HEK293 (Swiatecka-Urban et al, trafficking in HeLa cells (37), had no effect on
unpublished observation) may have contributed to CFTR trafficking in HEK293 cells. Thus, myosin
our success in demonstrating the Vc may mediate trafficking of a subpopulation of
coimmunoprecipitation between these proteins. endosomes from which CFTR is excluded in
Furthermore, we show that the interaction between HEK293 cells.
myosin Vb and Rab11a is dependent on the In summary, our data provide direct evidence
Rab11a binding sites in the myosin Vb tail domain that in polarized human airway epithelial cells,
(Figure 8A). The complementary data from this myosin Vb regulates CFTR dependent Cl-
and previous work, discussed above, are consistent secretion by facilitating the recycling of WT-
D
o
with the conclusion that myosin Vb together with CFTR and ∆F508-CFTR to the apical plasma w
n
Rab11a facilitates recycling in several cell types. membrane. We anticipate that elucidating the lo
a
d
We extend these observations and demonstrate mechanisms that regulate CFTR recycling will ed
that myosin Vb together with Rab11a facilitates help to identify unique therapeutic targets to fro
m
the recycling of CFTR in human airway epithelial modulate CFTR mediated Cl- secretion across h
ttp
cells. human airway epithelial cells. ://w
It remains unknown how myosin Vb interacts w
w
with Rab11a and how CFTR interacts with the ACKNOWLEDGMENTS .jbc
eRpaibth1e1laia-ml cyeolslsin. TVhbe sec oimntpelreaxct ioinn sh mumaya nb ea ieriwthaeyr This study was supported by NIH grant P20- b.org/
direct, as demonstrated by the Rab11a binding to RR018787 from the National Center for Research y gu
Resources (ASU), a Shwachman Award es
cargo proteins (67,68) or indirect and may be SWIATE03QO from the Cystic Fibrosis t on
facilitated by other interacting proteins (24,25,66). A
Foundation (ASU), NIH grant RO1-DC03299 p
Additional studies are needed to characterize these (REC), NIH grant RO1-DK45881 (BAS), NIH ril 3
interactions. , 2
grant RO1-DK34533 (BAS), NIH grant P20- 0
1
Our studies suggest that loss of myosin Vb 9
RR018787 from the National Center for Research
and/or Rab11a would be expected to attenuate Resources (BAS), and a Research Development
CFTR recycling and to decrease CFTR expression Program grant from the Cystic Fibrosis
in the apical plasma membrane in airway epithelial
Foundation (BAS). We would like to thank Dr.
cells. This effect could compromise the
John Wakefield from Tranzyme, Inc.
maintenance of the airway surface liquid, a (Birmingham, AL) who generated the CFBE41o-
situation observed in patients with cystic fibrosis
cells stably expressing the WT-CFTR or ∆F508-
(4). So far, neither mutations nor polymorphisms
CFTR, and Dr. J.P. Clancy from the University of
in the myosin Vb or Rab11a gene have been Alabama at Birmingham, AL for providing the
reported in humans or animals. The effects of stable CFBE41o- cells. We would also like to
altered Rab11a expression – observed during
thank Dr. Neil Bradbury from the Rosalind
treatment with chemotherapeutic agents (69), Franklin University of Medicine and Science in
infections (70), tumorogenesis (71,72), or hypoxia Chicago, IL for providing the stable HEK293 cell
(73) – on the function or plasma membrane
lines and Dr. Tama Hasson from the University of
expression of CFTR are currently unknown.
California at San Diego, CA for providing the
Our data do not support the role of myosin Vc anti-myosin VI antibody.
in CFTR recycling. Previous data indicate that
10