Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`Am J Physiol Lung Cell Mol Physiol 284 L187L196
`2003
`101152ajplung001522002
`
`Quantitative
`
`analysis of albumin uptake and transport
`in the rat microvessel endothelial monolayer
`
`THERESA A JOHN 1 STEPHEN M VOGEL 1 CHINNASWAMY
`ASRAR B MALIK 1 AND RICHARD D MINSHALL
`of 1Pharmacology and 2Anesthesiology College of Medicine
`Departments
`60612
`of Illinois
`Chicago Illinois
`
`12
`
`TIRUPPATHI
`
`1
`
`University
`
`Submitted
`
`15 May 2002 accepted in final
`
`form 14 August
`
`2002
`
`involving
`
`of
`
`by high
`
`nonspecific
`
`John Theresa A Stephen M Vogel Chinnaswamy
`Tiruppathi Asrar B Malik
`D Minshall
`and Richard
`Quantitative analysis of albumin uptake and transport in the
`endothelial monolayer Am J Physiol Lung
`rat microvessel
`Cell Mol Physiol 284 L187L196 2003 101152ajplung
`001522002We
`depen
`determined
`the
`concentration
`dence of albumin binding uptake and transport
`in conflu
`ent monolayers
`cultured
`lung microvascular
`of
`rat
`endothelial cells RLMVEC Transport of
`1 albumin in
`125
`occurred at a rate of 72 fmol
`RLMVEC monolayers
`ffnin ffi
`transport was inhibited
`106 cells
`Albumin
`by cell surface
`depletion of the 60kDa albumin binding
`gp60
`glycoprotein
`of caveolae using methyl
`and
`by disruption
`fficyclodextrin
`gp60 activation by means of gp60 cross linking
`By contrast
`I al
`using primary and secondary antibodies increased 125
`bumin uptake 23 fold At 37°C 125
`Ialbumin
`uptake had a
`time of 10 min and was competitively
`by
`half
`inhibited
`1 M Using a two site model we
`unlabeled albumin
`IC50 ffl
`mal capacity B of albumin
`the affinity KD and maxi
`by Scatchard
`analysis
`uptake to be 087
`M Koi
`M KD2 and 202
`and 933
`and 047 pmo1106 cells Bmaxl
`At 4°C we also observed two popula
`ffl 135 nM 1
`pmo1106 cells Bmax2
`tions of specific binding sites with high KDi
`ffl 16 M affinity
`On the basis of
`of the total and low KD2
`these data we propose a model
`in which the two binding
`and unclustered
`the
`clustered
`gp60
`affinities
`represent
`forms The model predicts that
`in caveo
`fluid phase albumin
`and trans
`lae accounts for
`the bulk of albumin
`internalized
`
`estimated
`
`if
`
`ported in the endothelial monolayer
`
`permeability
`capillary
`min binding glycoprotein
`
`vesicular
`
`transport
`
`caveolae albu
`
`gp60
`
`from morphological
`IS ACCUMULATING
`THERE
`EVIDENCE
`and functional
`studies that albumin is continuously
`exchanged between the microvascular
`and intersti
`by receptor mediated
`compartments
`tial
`transcytosis 8 11 13 1619 23 2729 Recent
`studies show that albumin molecules can associate
`with surface glycoproteins such as gp60 albondin in
`endothelial cells which can activate the transport of
`albumin 2 68 13 19 20 23 25 28 29 The
`
`albumin
`
`signaling pathways mediating transcytosis
`
`are not
`
`its
`
`lial
`
`that
`
`well understood The pathway
`the hetero
`trimeric G protein G and the activation
`of Src kinase
`albumin endocytosis and its
`signaling may activate
`from the luminal
`to the abluminal
`transport
`cell
`surface 1214 27 Interestingly
`albumin transcy
`by antibody induced
`cross link
`tosis was augmented
`ing of gp60 13 2729 These studies suggest that
`under physiological conditions albumin can bind to
`cell surface
`the albumin binding
`pro
`receptors
`teins ABPs and initiate
`transcytosis
`issues concerning the mode of
`However
`important
`and transport
`albumin uptake
`by a
`transcellular
`remain
`unresolved Fluid
`phase transport
`route
`albumin in vesicles or via the paracellular
`pathway
`not be expected
`to be a saturable process
`would
`whereas transport of albumin either bound to recep
`by receptors should be saturable
`tors or regulated
`nature of ABPs such as
`In addition the high affinity
`range 19 25 relative
`gp60 in the nanomolar
`to
`plasma and interstitial
`concentrations
`of albumin
`1 mM implies that 1 albumin uptake and trans
`processesand 2 bound
`cytosis should be constitutive
`albumin may not dissociate from its vesicular carri
`ers upon exocytosis These questions could not be
`previously addressed with any rigor because deter
`minations
`of albumin to the endothelial
`of affinity
`plasmalemma were
`hampered
`binding 2 6 7 19 This masked any saturable and
`the present study we
`specific albumin binding
`In
`of albu
`dependency
`determined the concentration
`on the surface of confluent
`min binding
`endothelial
`of albumin We specifi
`monolayers and transport
`lung microvascular endothe
`cally used cultured
`rat
`cells RLMVEC becausethey were shown by us
`cell surface adsorption of
`to have low nonspecific
`albumin We show that 1 binding
`and uptake
`of
`albumin have saturable and fluid phase components
`2 cellular uptake of albumin is submaximal at phys
`serum albumin
`concentration
`iological
`implying
`it can be further
`increased in response to appro
`signals and 3 gp60 an ABP can activate
`priate
`albumin transport
`
`Address
`
`for
`
`reprint
`
`requests
`
`and other
`
`Minshall University
`
`of Illinois
`
`R D
`correspondence
`at Chicago Dept
`of Pharmacology
`mc 868 835 S Wolcott Ave Chicago
`60612 Email
`rminshuicedu
`
`IL
`
`of this article were defrayed in part by the
`The costs of publication
`of page charges The article must
`be hereby
`payment
`therefore
`18 USC Section 1734
`marked advertisement
`in accordance with
`
`solely to indicate this fact
`
`httpwwwajplungorg
`
`1040060503 $500 Copyright
`
`2003 the American
`
`Physiological
`
`Society
`
`L187
`
`Abraxis EX2014
`Actavis LLC v Abraxis Bioscience LLC
`1PR201701101 1PR201701103 1PR201701104
`
`

`

`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`L188
`
`METHODS
`
`Endothelial Cell Cultures
`
`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`488 and 568nm excitation
`lines to detect DAPIAlexa
`350 band pass 385470 nm emission FITCAlexa 488 band
`pass 505550 nm emission and rhodamineAlexa 568 fluo
`rescence long pass 585 nm emission in optical sections
`m thick pinhole set to achieve 1 Airy unit
`
`laser
`
`1
`
`Cell Surface gp60 CrossLinking
`RLMVEC monolayers were washed twice with HEPES
`buffered DMEM at 4°C and incubated for 30 min at 4°C with
`antigp60 antibody 20 gml and then with secondary an
`20 gml for 30 min to crosslink
`tibody goat anti rabbit
`gp60 Controls were treated with the preimmune
`antibody
`20 gml at 4°C for 30 min Cells were rewarmed to 37°C to
`endocytosis 27 To deplete ABP gp60 27 anti
`gp60 antibody was preincubated with cells for 2 h at 37°C
`
`activate
`
`Binding
`
`of 125
`
`IAlbumin
`
`106
`
`Confluent
`
`cell cultures
`
`in six well plates 1012
`cells35 mm were washed twice in 10 mM HEPESbuffered
`DMEM pH 74 and then serum deprived for 2 h by incuba
`tion in DMEM Cells were precooled to 4°C for 30 min and
`1251 albumin in the presence or absence of unlabeled albumin
`in HBSS was added for 30 min at 4°C Unbound
`ligand was
`removed by three washes with icecold HBSS and cells were
`lysed with 1 ml of 50 mM Tris fHCI buffer pH 74 containing
`1 Triton X100 and 05 SDS 13 Cell associated 125 1
`albumin counts were measured in a gamma counter Packard
`Instruments Downers Grove IL
`
`Uptake of 125
`
`IAlbumin
`
`RLMVEC
`Serum deprived
`see
`confluent
`monolayers
`reagents at 37°C for
`above were pretreated with indicated
`Ialbumin
`of 125
`in the pres
`periods before addition
`required
`in HBSS Uptake at
`ence or absence of unlabeled
`albumin
`37°C was allowed to proceed for
`times indicated
`and then
`on ice and washing three times with
`terminated
`by chilling
`icecold HBSS To remove cell surface bound 125
`I albumin
`cells were washed three times with 1 ml of acid wash buffer
`05 M NaCI and 02 M acetic acid pH 25 26 Cells were
`then lysed and counted as described above
`
`lmmunostaining
`
`Transendothelial
`
`121 Albumin
`
`Transport
`
`salt
`
`and 50
`
`gml streptomy
`
`RLMVEC VEC Technologies Rensselaer NY were cul
`tured in high glucose DMEM GIBCO BRL Grand Island
`with 5 fetal bovine serum Hyclone
`NY supplemented
`Logan UT plus 50 Uml penicillin
`previously 4 The cultures were main
`tained in 5 00295 room air at 37°C
`cin as described
`Drugs and Reagents
`All drugs and reagents were obtained from Sigma Chem
`ical St Louis MO unless stated otherwise Hanks
`balanced
`NaHCO3 42 mM and
`solution HBSS containing
`HEPES 10 mM was adjusted
`to pH 74 Bovine
`serum
`fraction V 99 pure endotoxin free cold alcohol
`albumin
`dissolved in HBSS in concentra
`precipitated was freshly
`of 001100 mgml Methylfficyclodextrin
`was dis
`tions
`solved in HBSS Rabbit antibody
`against bovine endothelial
`and labeled with
`gp60 antigp60 antibody was prepared
`elsewhere 25 27 Control
`Cy3 as described
`isotype
`preimmune se
`matched antibody was isolated from rabbit
`rum Monoclonal anticaveolin1
`antibody was obtained from
`Transduction Laboratories Lexington KY Goat anti mouse
`IgG labeled with Alexa 568 or 350 bovine serum albumin
`and cholera toxin subunit BAlexa 594
`Alexa 488 conjugate
`Probes Eugene
`conjugate were purchased
`from Molecular
`OR
`
`Albumin
`
`Iodination
`
`ION Pharmaceuticals
`Labeling of albumin with Na1261
`T 3 The tracer albumin
`was performed
`using chloramine
`formed was purified to 04 free 1251 using Sephadex G25
`fic radioactivity was 03 10
`columns Sigma Chemical Speci
`by 10 trichloroacetic
`Ci g albumin
`determined
`acid pre
`cipitation
`
`Methyl ffCyclodextrin Treatment
`
`Confluent
`
`RLMVEC
`were
`02 nM10 mM in HBSS for 15 min
`methylfficyclodextrin
`washed twice with HBSS and incubated with fresh medium
`tracer albumin
`for the desired periods
`containing
`
`monolayers
`
`incubated
`
`with
`
`in
`
`of albumin gp60 and caveolin1
`Cellular
`localization
`the plasma membrane and plasmalemmal vesicles exposed to
`albumin was determined
`by immunocytochemical
`labeling
`confocal microscopy LSM 510 Zeiss
`and
`laser scanning
`RLMVEC were serum deprived
`for 2 h washed three times
`red free DMEM and
`with HEPESbuffered HBSS or phenol
`exposed to 01 mgml albumin
`gml
`in the presence of 50
`Alexa 488 conjugated albumin 20 gml cholera toxin sub
`andor 35
`unit BAlexa 594 conjugate
`gml Cy3antigp60
`antibody for up to 30 min 13 Cells were then washed three
`times with HBSS fixed with 4 paraformaldehyde
`in HBSS
`and blocked for 30 min in HBSS containing 5 goat serum
`01 Triton X100 and 001 NaN 3 Primary antibody
`label
`at 4°C with anticaveolin1
`ing was performed
`overnight
`antibody 1 gml Coverslips were then washed three times
`for 10 min in HBSS blocked for 30 min with 5 goat serum
`
`and incubated for 2 hat room temperature with fluorescent
`and goat anti mouse secondary an
`labeled goat anti rabbit
`In some cases 4 6diamidino2phenylindole
`dihy
`tibody
`drochloride DAPI
`gml was added to visualize
`the
`using 364
`Confocal microscopy was performed
`nucleus
`
`1
`
`125
`
`well
`
`I
`
`RLMVECs
`were grown
`on
`clear microporous
`polyester
`Transwell membranes 12 mm diameter
`1 cm2 growth area
`m pore size Corning Costar Cambridge MA The
`04
`inserts were filled with a total of 05 ml of incu
`membrane
`Ialbumin
`in the presence or absence of
`bate containing
`unlabeled albumin The lower well was filled with 15 ml of
`incubate of the same osmolarity as the inner well Thus fluid
`levels and osmotic pressure in the upper and lower wells
`and osmotic effects
`were equalized to minimize hydrostatic
`and test agents were added to the upper
`Labeled albumin
`1251albu
`solution and transendothelial
`in equimolar
`was measured at 37°C Aliquots
`of 50
`min permeability
`were sampled from the lower chamber every 1015 min and
`was measured Albumin
`gamma radioactivity
`permeability
`was calculated
`radiolabeled
`albumin
`flux across the cell
`for
`as described elsewhere 5 22 Trichloroacetic
`monolayer
`and SOS PAGE of the tracer al
`analysis
`acid precipitation
`bumin 27 in the abluminal
`chamber showed that
`the radio
`activity measured on the abluminal side of the cell monolayer
`and that
`albumin was not
`remained
`attached
`to albumin
`in the uptake or transport
`
`hydrolyzed
`
`process
`
`AJPLung Cell Mol Physic
`
`VOL 284
`
`JANUARY
`
`2003 wwwajplungorg
`
`

`

`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`L189
`
`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`filter
`
`indicating
`in albumin
`
`of
`
`layers grown on Transwell
`inserts received vehi
`fficyclodextrin 02 nM 10
`cle without drug or methyl
`mM for 15 min in the upper chamber
`and tracer
`albumin was added immediately thereafter Methyl
`ffi
`125
`blocked
`1 albumin
`transendothelial
`cyclodextrin
`M Fig 2 The
`10
`permeability with an IC50 of
`of the inhibitor 10 mM reduced
`highest concentration
`permeability of the albumin tracer by 80 from 77
`to 15 fmol ffnin
`the impor
`ff106 cells ffi
`tance of the cellular pathway
`in
`transport
`RLMVEC
`transport by
`albumin
`Regulation of transendothelial
`gp60 Inasmuch
`of gp60 can stimulate
`as activation
`albumin
`cells via caveolae
`in endothelial
`endocytosis
`13 21 27 we determined whether
`gp60 activation
`To
`tracer albumin
`also increased the transport
`com
`prevent
`of the antigen antibody
`internalization
`plex gp60 on the apical cell surface was cross linked
`as described elsewhere 27 using an antigp60 anti
`body for 30 min in cold conditions 4°C Cells were
`rewarmed to 37°C with a solution containing
`1 albumin 25 nM plus unlabeled albumin 15 M
`re
`uptake of the tracer Control
`isotypematched antibody
`in lieu of antigp60
`ceived
`antibody Cross linking of gp60 significantly increased
`permeability 23 fold above the control
`125
`1 albumin
`15 nl ffnin ffi
`value of 38
`Excess unlabeled
`ftMff2
`albumin 1 mM reduced 125
`1 albumin
`permeability to
`06 nl mini ftMff2 Fig 3 in the cells subjected
`12
`to gp60 cross linking To address the dependence of
`of tracer albumin
`on the cell surface
`basal
`transport
`gp60 RLMVEC monolayers were incubated with anti
`for 2 h at 37°C to deplete cell surface
`gp60 antibody
`gp60 9 27 This procedure reduced basal transendo
`to 73
`06 nl
`ftmff2 80 inhibition Fig 3
`
`rapidly
`125
`
`to activate
`
`cultures
`
`125
`
`1 albumin
`
`permeability
`
`thelial
`min ffl
`
`i5
`
`E
`
`4C 75
`
`Time frnin
`
`90
`
`Fig 1 Time course of
`1251 Albumin
`transport
`
`1251a bumin
`
`in rat
`
`transendothelial
`permeability
`was determined
`lung microvascular
`cells RLMVEC
`in
`endothelial
`to confluence
`in Transwell
`grown
`serts Tracer albumin 68 nM 1251a bumin was applied to the apical
`together with 1 5 M or
`1 5 mM free albumin which was
`chamber
`in the basal chamber
`to equalize osmotic pressure and
`also present
`samples from the basal chamber were taken every 1015 min for 90
`min 4 wells per
`time Specific transport
`of 1251 albumin
`incubation
`by 15 mM a bumin was 80 of total
`amount
`125
`inh bited
`transport was cumulative
`1 Albumin
`increas
`clearance
`albumin
`for up to 90 min Data are from 1 experiment which is
`of 3 experiments performed
`
`125
`
`ing linearly
`
`representative
`
`Statistical Analysis
`
`fitting
`
`programs
`
`Data were analyzed by the nonlinear
`least squares curve
`LIGAND
`Elsevier Biosoft and Microcal
`Origin Microcal Software Northhampton MA Student s
`level of P
`ttest was used to compare results at significance
`comparisons were made by ANOVA All sta
`005 Multiple
`tests were made using GraphPad Prism lnstat soft
`ware San Diego CA
`
`tistical
`
`RESULTS
`
`Albumin
`
`Transport
`
`in Endothelial Monolayers
`
`Time course of transendothelial
`For studies
`
`port
`
`of
`
`transendothelial
`
`1251 albumin
`trans
`125 1a 1 b u min
`
`in
`
`filter
`
`if
`
`apicaltobasolateral
`the
`transport
`direction
`RLMVEC were grown to confluence on Transwell
`inserts 1cm2 surface area Experiments
`began when
`the liquid in the upper well was replaced by a warm
`125 1a 1 bu min plus unlabeled
`37°C solution containing
`albumin 15
`M or 01 mgml During
`the 90 min
`period aliquots of media 005 ml
`in the lower well
`were sampled and analyzed for gamma radioactivity
`tracer albumin
`see METHODS Appearance
`of
`in the
`basal chamber vs time described a straight
`line Fig
`1 as predicted
`the backflux
`of tracer
`into upper
`chamber was negligible From the slope of the fitted
`line in Fig 1 we calculated the 125 1a 1 bu min flux of 72
`Excess unlabeled albumin 15
`fmol ffnin ffi
`ffl 06 cells ffl
`flux Fig 1
`mM or 100 mgml blocked 125
`1 al bu min
`between labeled and unlabeled
`competition
`plasmalemmal sur
`at
`the apical endothelial
`
`implying
`albumin
`
`face
`transport The effect
`Cell pathway regulates albumin
`of methylfficyclodextrin
`the cholesterol binding agent
`caveolae 10 on transendothelial
`125
`that disrupts
`albumin permeability was assessed RLMVEC mono
`
`Fi thy 11
`
`ycinolex i
`
`r
`
`M
`
`Fig 2 Inhibitory
`on flux of
`effect of methylfficyclodextrin
`albumin RLMVEC on Transwell
`inserts were preincubated
`for 15
`min with 02 nM 10 mM methylfficyclodextrin
`or vehicle and then
`balanced salt solution HBSS contain
`with 1251 albumin
`in Hanks
`M free a bumin
`ing 15
`for 15 min A dose dependent
`inh bition
`flux from a control value of 77
`03 fmol ffnin
`1251 albumin
`was observed Values are means
`SE from 4 replicates
`each point
`in 1 experiment which is representative
`of 4 experiments
`performed
`
`125
`
`ffl
`
`of
`ffl 06
`
`for
`
`cellsffl
`
`AJPLung Cell Mol Physic
`
`VOL 284
`
`JANUARY
`
`2003 wwwajplungorg
`
`

`

`L190
`
`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`at
`
`calization
`
`ffl 202 pmo1106 cells From these data we deter
`Bmax
`mined the total albumin
`unlabeled
`uptake labeled
`forms as a function of albumin concentration Fig 5 C
`The cellular
`uptake of albumin was dependent
`on al
`bumin concentration
`region of concentra
`the steepest
`and 1Off3 M albu
`tion dependency was between 1Off5
`in Fig 5C predicts an EC50 of
`min The smooth curve
`M and saturation
`3 mM albu
`100
`of transport
`min
`Con focal fluorescence imaging of albumin uptake We
`the uptake of albumin
`in caveolae by colo
`visualized
`forms of albumin
`of
`fluorescent
`cholera
`toxin subunit B and antigp60 antibody Cholera toxin
`subunit B was used as a specific marker
`for caveolae
`gang lioside GM1
`because it
`labels the caveolaespecfic
`14 gp60 was previously
`shown to be associated with
`caveolae 13 After 30 min of
`endothelial
`incubation
`with these fluorescent probes at 37°C cells were acid
`washed to remove residual
`cell surface probe fixed
`and stained with the nuclear marker DAPI 1 gml or
`plus Alexa 350 labeled sec
`anticaveolin1
`antibody
`ondary
`antibody
`High resolution
`images
`confocal
`10 m optical
`in Fig 6A show internal
`thickness
`to
`and Cy3gp60 antibody
`ized Alexa 488 albumin
`gether with caveolin1
`near the apical
`immunostaining
`staining of gp60 and albumin
`cell surface Coincident
`three probes were apparent
`and colocalization
`of all
`in
`the overlay In other experiments fluorescent albumin
`was coincubated with fluorescent cholera toxin subunit
`B Figure 68 shows the confocal
`images of internalized
`Alexa 594 cholera toxin subunit B and Alexa 488 albu
`min together with DAPI The fluorescence of albumin
`and cholera toxin was coincident
`in the merged image
`albumin s association with caveolae
`indicating
`To address the effects of methyl fficyclodextrin
`on
`endothe
`albumin
`uptake and caveolin1
`distribution
`cells were pretreated with 10 mM methyl fficyclo
`incubated with HBSS containing Alexa
`and
`dextrin
`
`lial
`
`2e
`
`75
`
`90
`
`125
`
`1511
`Time m
`I a lbu min
`Fig 4 Time course of endothelial
`uptake Confluent
`RLMVEC were incubated for 590 min 6 wells per incubation
`time
`M albumin
`with 228 nM 1251a bumin
`and 15
`to quantify
`total
`or with 15 mM unlabeled
`a bumin
`nondis
`uptake F Total uptake of 1251 albumin
`to quantify
`125 1a bumin
`placeable
`in
`at 15 min and again at 75 min
`creased with time peaking initially
`10 min Data are representative
`of 4 separate experi
`half
`time ffl
`ments
`
`uptake
`
`777
`
`Fig 3 60kDa albumin binding
`protein gp60dependent
`1251albu
`in RLMVEC RLMVEC
`min transport
`monolayers were washed
`DMEM at 4°C and incubated for 30 min
`twice with HEPESbuffered
`with antigp60 antibody 10 gml at 4°C and with secondary anti
`10 gml for 30 min to induce cross linking
`body goat anti rabbit
`10 gml and
`Controls were treated with
`preimmune
`antibody
`secondary antibody Cells were rewarmed
`to 37°C to activate uptake
`of 125I a bumin
`and 15 M unlabeled albumin Crosslinking of gp60
`23 fold from a control value of
`increased 1251 albumin
`permeability
`cell surface gp60 RLMVEC
`38
`To deplete
`15 nl ffnin ffi
`ftrnff2
`DMEM and
`monolayers were washed
`twice with HEPESbuffered
`incubated with antigp60 antibody 10 gml for 2 h at 37 °C Deple
`by 85 1251 Albumin
`Ia lbu m in
`tion of gp60 reduced 125
`transport
`flux was also blocked by 95 when cells were coincubated with 15
`SE n ffl 4
`mM unlabeled albumin 100 mgml Values are means
`by ANOVA
`
`P 001 vs control
`
`Albumin Uptake in Endothelial Cell Monolayers
`
`in cell
`
`initial
`
`thelial
`
`125
`
`uptake RLMVEC
`Time course of
`tracer albumin
`monolayers were incubated
`for various periods with
`M unlabeled albu
`1 albumin in the presence of 15
`125
`min 125
`the cells was
`1 albumin
`accumulation within
`gamma radioactivity
`determined
`by counting
`lysates Lysates were prepared from acid washed cells
`to avoid contamination
`associated
`by tracer albumin
`with the cell exterior The time course of cellular
`125
`uptake was biphasic Fig 4 Specific tracer
`albumin
`uptake 1251 albumin uptake in the presence of 15 M
`in 15 mM albumin reached an
`albumin minus that
`peak 60 fmo1106 cells within
`15 min ap
`cells at 3045 min
`proached a minimum 50 fmo110 6
`rose to a second peak 75 fmo110 6
`and finally
`cells at
`75 min The initial
`slope of the uptake curve was used
`to estimate the rate of tracer albumin
`uptake which
`was 6 fmol ffnin ffi
`this value was within
`ff106 cells ffl
`the same range as 125
`1 albumin transendothelial
`trans
`port of 72 fmol fthin ffl
`see above Endo
`ffl 06 cells ffl
`uptake was markedly
`1albumin
`inhibited
`at
`each time point by excess 15 mM unlabeled albumin
`Fig 4
`
`inhibition
`
`of
`
`tracer
`
`Fig 5A I C50 ffl 115 M Scatchard analysis Fig 58
`into its two com
`used to parse the sigmoidal
`function
`M high affinity
`ponents showed a 087
`component
`ffl 047 pmo1106
`with maximal binding capacity Bmax
`M low affinity
`cells and a 933
`component with
`
`Concentration
`maximal
`duced
`
`biphasic
`
`uptake Sub
`dependence of albumin
`albumin
`of unlabeled
`pro
`concentrations
`albumin
`uptake
`
`AJPLung Cell Mol Physiol
`
`VOL 284
`
`JANUARY
`
`2003 wwwajplungorg
`
`

`

`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`L191
`
`488 tagged albumin 50 gml plus unlabeled albumin
`15 M for 30 min Cells were acid washed fixed and
`ie anti
`stained for nuclei DAPI
`and for caveolin1
`caveolin1 monoclonal antibody at
`gml followed by
`goat antimouse Alexa
`568 labeled secondary anti
`body Methylfficyclodextrin
`abolished albumin
`inter
`by plasmalemmal vesicles Fig 7 In addi
`nalization
`
`1
`
`tion methylfficyclodextrin
`surface associated caveolin1
`
`tent with its effect
`
`in disrupting
`
`markedly
`immunostaining
`cell surface caveolae
`
`reduced the cell
`consis
`
`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`Albumin Binding to Endothelial
`
`Cell Surface
`
`125
`
`cell surface labeled
`
`Competition between endothelial
`and unlabeled albumin Agonist
`competition
`binding
`studies were made using RLMVEC in cold conditions
`4°C to abolish tracer endocytosis and internalization
`Cells were incubated with 125
`1 al bu min in the presence
`1 Off3 M for 30
`or absenceof unlabeled albumin 1 Off12
`min Unlabeled albumin reduced the specific binding of
`04 fmo1106 cells The
`1 albumin from 175 to 27
`fitted to the data were 45 nM
`IC50 and Hill
`coefficient
`and 028 respectively suggesting the existence of more
`than one albumin binding site on the RLMVEC surface
`Fig 8A Scatchard analysis of the binding data Fig
`binding sites with af
`two albumin
`8E3 demonstrated
`finity KD values of 135 nM and 16 M respectively
`Bmax values of 242 fmo1106 cells
`and corresponding
`and 24 pmo1106 cells The lower affinity
`site ac
`counted for 98 of the binding
`from cell surface binding
`of albumin
`Displacement
`of RLMVEC
`sites by antigp60 antibody Pretreatment
`for 1 h at 4°C with
`a 1200 dilution
`monolayers
`of
`antigp60 antibody decreased 125
`1 albumin
`binding by
`478 51 n ffl 3 relative to control antibody The
`and the specific antigp60
`albumin
`tracer
`antibody
`competed for the same membrane binding sites Thus
`binding sites detected on RLM
`high and lowaffinity
`VEC may represent different
`states of gp60
`
`DISCUSSION
`
`125
`
`exhibited
`
`RLMVEC monolayers
`a high degree of
`binding because 80 of
`1 albumin
`total
`specific
`binding was inhibited
`by excess unlabeled albumin
`binding site for albu
`a saturable lowaffinity
`Although
`min has been described in bovine pulmonary
`artery
`cells 23 and rat
`tissue microvessel
`endothelial
`fat
`cells 20 the finding
`of a second site of
`endothelial
`much higher affinity
`is a novel observation Competi
`tion studies of 125
`1 albumin
`binding under cold condi
`tions 4°C to prevent
`showed
`tracer
`internalization
`that albumin bound to the endothelial
`cell surface with
`M K02
`constants of 135 nM KDi
`and 16
`affinity
`On the basis of
`site represented 1 of the specific albu
`the corresponding
`Bmax values the
`high affinity
`105 sitescell Schnitzer
`min binding and averaged 5
`and Oh 21 and Tiruppathi
`et al 25 demonstrated
`that an antigp60 antibody displaced the specific bind
`125
`1 albumin
`from rat and bovine pulmonary
`ing of
`microvascular
`our
`endothelial
`cells consistent with
`
`g
`
`E
`
`o
`
`404
`
`20
`
`10
`
`14
`19
`41 n la baled Albumin Al
`
`10
`
`5
`
`10
`
`15
`
`AlbAirriln Uptake pmolile cells
`
`ota I Albumin Added M
`
`kinetics
`
`of 125
`
`I albumin
`
`uptake
`
`by unlabeled
`albumin
`101f9
`by 34 nM
`followed
`1 h A
`at 37°C for
`increasing
`
`concentrations
`
`of unlabeled
`
`Scatchard
`
`max2
`
`analysis
`
`by KDi
`
`tracer
`
`Fig 5 Competition
`albumin
`Increasing
`10ff3 M were added to RLMVEC monolayers
`1251 albumin and monolayers were incubated
`I albumin was inhibited
`uptake of 125
`in the presence of
`to be 115 M
`IC50 was calculated
`of free albumin
`concentrations
`1 mM a bumin B
`was achieved with
`near maximal
`displacement
`was best
`analysis of 125
`fit by a
`I albumin
`displacement
`and 13maxl site 1 and KD2
`and
`2 site model
`represented
`site 2 where KD is affinity
`and Bm is maximal
`constant
`binding capacity C total
`uptake calculated from Scatchard
`a bumin
`of combined
`1251a bumin
`albumin
`plus unlabeled
`3 mM added
`shows
`saturation
`of albumin
`uptake with
`uptake
`albumin Data are representative
`of 4 separate experiments
`
`AJPLung Cell Mot Physic
`
`VOL 284
`
`JANUARY
`
`2003 wwwajplungorg
`
`

`

`L192
`
`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`A
`
`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`11
`
`I
`
`I
`
`10 pm
`
`Fig 6 Colocalization
`of a bumin with gp60 cholera
`toxin subunit B and caveolin1 A high magnification
`10 m optical
`thickness of RLMVEC
`image
`confocal
`uptake green bottom left
`shows Alexa 488 albumin
`Cy3gp60 ant body uptake red top left and caveolin1
`blue top right
`the apical sur
`immunostaining
`near
`of gp60 albumin and caveolin1
`face Overlay
`images
`shows
`of gp60 and
`bottom right
`coincident
`staining
`as yellow and as white where gp60 and albu
`albumin
`red green
`min are closely associated with caveolin1
`and blue B confocal
`image shows confluent RLMVEC
`staining with Alexa 594 cholera toxin sub
`monolayer
`unit B red top left Alexa 488a bumin green bottom
`blue top right
`left and DAPI
`for nuclear
`localization
`of a bumin
`and cholera
`Note
`colocalized
`fluorescence
`toxin in merged image yellow bottom right suggest
`ing a bumin
`by caveolae
`Images are represen
`uptake
`tative of 4 experiments
`
`AJPLung Cell Mol Physic
`
`VOL 284
`
`JANUARY
`
`2003 wwwajplungorg
`
`

`

`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`L193
`
`Control
`
`Cyclodextrin
`
`Albumin
`
`Caveolin1
`
`cell
`
`Fig 7 Effects of methylfficyclodextrin
`on albumin
`and
`cyclodextrin
`uptake
`surface
`caveolin1
`staining
`RLMVEC monolayers
`on glass cover
`slips were pretreated with vehicle or 2
`mM methylfficyclodextrin
`for 15 min
`HBSS containing
`50
`incubated
`with
`M
`and 15
`gml Alexa 488 albumin
`30 min fixed
`a bumin
`unlabeled
`for
`and stained with
`anticaveolin1
`anti
`gm1 and then with goat anti
`body 1
`mouse Alexa 568 and DAPI Uptake
`of
`albumin was blocked when
`fluorescent
`cells were first
`treated with methylffi
`cyclodextrin In addition methylfficy
`reduced
`clodextrin
`
`significantly
`
`cell
`Im
`surface caveolin1
`immunostaining
`of 4 experi
`ages are representative
`ments Scale bar 10 m
`
`inhibition
`
`analysis
`
`observations Thus the specific binding sites for albu
`min may include gp60
`Because the measured albumin
`uptake only repre
`sents that albumin
`by the cells
`tracer
`internalized
`which thus was present
`from acid
`in the lysate
`washed cells it
`that uptake data report
`the
`is likely
`endocytosis of the tracer At 10ff8 10ff3 M unlabeled
`albumin we observed progressive
`of tracer
`uptake Fig 5A From Scatchard
`albumin
`Fig 5B we determined the apparent albumin binding
`M low affinity
`constants of the transporter to be 087
`and 93 M ultra low affinity
`The lowaffinity
`uptake
`reported lowaffinity
`value is similar
`to other
`binding
`values 20 23 The low affinity
`binding may
`albumin
`albumin which was
`the tracer
`represent
`initially
`the cell surface such as
`bound to a saturable
`site at
`gp60 and was subsequently
`by endocyto
`internalized
`the ultralowaf
`component may
`sis By contrast
`finity
`re
`albumin
`tracer
`sequestered within
`a
`represent
`stricted membrane compartment with limited access
`caveolae which
`to the extracellular
`fluid Endothelial
`communicate with the extracellular
`via narrow
`fluid
`20 to 30nm necks 17 could qualify
`as such a
`compartment Our evidence for this is as follows
`at
`4°C when internalization
`by endocytosis
`is completely
`component
`the ultralowaf
`was not detected Fig 8 2 at 37°C methylfficyclo
`dextrin which was shown to eliminate
`caveolae Fig
`7 markedly
`all components of 125
`1 al bu min
`inhibited
`uptake Fig 2 and 3 Alexa 488 albumin entered the
`
`inhibited
`
`finity
`
`binding
`
`because the fluorophore was
`caveolar compartment
`vesicles with Cy3
`shown to colocalize in endocytic
`labeled antigp60 antibody Fig 6A and Alexa 594
`cholera toxin subunit B Fig 6B the specific caveolae
`marker 15
`We used methylfficyclodextrin
`to ad
`as a reagent
`between albumin
`dress the relationship
`uptake and
`abolished
`albumin
`transport Methyl fficyclodextrin
`and transport Figs 2 and 7 The albumin
`uptake
`the transfer of albumin
`protocol monitored
`transport
`through the cell monolayer and into the ablumi
`tracer
`chamber
`this assay the backflux
`nal
`of
`tracer
`In
`compartment was mini
`albumin
`from the abluminal
`of the volume in the abluminal
`mal becauseof dilution
`albu
`thus allowing
`a measure of unidirectional
`well
`min flux Moreover we imposed no transendothelial
`thus render
`or oncotic pressure gradient
`hydrostatic
`ing negligible the convective movement of tracer We
`from the methylfficyclodextrin
`of
`sensitivity
`albumin
`that
`transendothelial
`transport
`flux of tracer albumin
`sional paracellular
`was minimal Fig 2
`The earliest
`step in the uptake and transport
`albumin
`cells is the docking
`endothelial
`through
`ABPs 2 6 7 13
`albumin
`on the caveolaeassociated
`1921 25 28 29 Inasmuch as the transendothelial
`flux of tracer albumin was abolished by excess unla
`beled albumin 15 mM the albumin
`internalized
`by
`the mono
`cells and transported
`through
`is displaceable Fig 1 In the presence of a much
`
`the diffu
`in RLMVEC
`
`of
`
`of
`
`infer
`
`endothelial
`
`layer
`
`AJPLung Cell Mol Physic
`
`VOL 284
`
`JANUARY
`
`2003 wwwajplungorg
`
`

`

`L194
`
`ALBUMIN ACTIVATED
`
`FLUID PHASE TRANSPORT
`
`to be 72 fmol albumin ffrrin ffl
`ffl 06 endothelial
`the tracer albumin up
`Importantly we observed that
`take rate of 6 fmol
`from
`ffl 06 cells ffl
`ffrrin ffl
`calculated
`the uptake curve in Fig 4 was
`slope of
`rate
`comparable to the transport
`Analysis of the time course of tracer albumin uptake
`showed peaks at 15 and 75 min Fig 4 A possible
`this complex uptake process is that
`of
`interpretation
`this represents the reuptake of albumin tracer that has
`involving up
`been exocytosed Recycling of the tracer
`could account
`and reuptake
`take extrusion
`for the
`dual uptake peaks
`Antigp60 antibody exposure of cells used to deplete
`the cell surface gp60 27 reduced albumin
`uptake and
`barrier at 37°C by 85
`Fig 3 This finding
`is consistent with a role of cell
`surface gp60 in mediating albumin transport via trans
`cytosis By contrast gp60 activation induced by a much
`briefer exposure to the antigp60 antibody plus a sec
`gp60 27 markedly
`to crosslink
`ondary
`antibody
`125 1a lb u mi n This method of
`increased the transport of
`induced in activating
`cross linking is analogous to that
`receptors which
`growth
`dimerize
`or glyco
`factor
`proteins which clus
`sylphosphatidylinositolanchored
`ter in rafts or caveolae on exposure to specific antibod
`
`Downloaded
`
`from
`
`ajplungphysiologyorg
`
`on
`
`December
`
`30
`
`2011
`
`ies 1 Because cross linking by the antigp60 antibody
`
`at 37°C but
`uptake and transport
`increased albumin
`blocked albumin binding at 4°C it
`and
`is an activating
`on the temperature
`a blocking antibody
`depending
`The data show that
`can compete with
`the antibody
`for binding to gp60 at 37°C as well as 4°C
`albumin
`at 37°C 98 of the internalized
`However
`albumin is
`of plasmalemma vesi
`in the fluid phase compartment
`than receptor bound because it
`cles rather
`required
`albumin
`more unlabeled
`to antagonize
`uptake Fig 5A than 1251
`125
`1 albumin
`competitively
`binding Fig 8A The data in Fig 58 show
`albumin
`has a
`the ultralowaf finity
`uptake of albumin
`that
`Bmax 202 pmolmillion
`much greater
`cells than the
`component 047 pmo1106 cells suggest
`low affinity
`represents 98 of
`fluid phase endocytosis
`ing that
`in RLMVEC Therefore
`the overall uptake of albumin
`
`significantly
`
`the initi