Hindawi Publishing Corporation

Biochemistry Research International

Volume 2012, Article ID 672705, 14 pages
doi:10.1155/2012/672705

Research Article
Impaired Caveolae Function and
Upregulation of Alternative Endocytic Pathways Induced by
Experimental Modulation of Intersectin-1s Expression in
Mouse Lung Endothelium
Dan N. Predescu,1 Radu Neamu,2 Cristina Bardita,1 Minhua Wang,1 and Sanda A. Predescu1
1 Department
2 Division

of Pharmacology, Rush University, 1735 W Harrison, Chicago, IL 60612, USA

of Pulmonary, Allergy and Critical Care Medicine, Emory University School of Medicine, Atlanta, GA 30322, USA

Correspondence should be addressed to Sanda A. Predescu, sanda predescu@rush.edu

Received 15 August 2011; Accepted 16 November 2011
Academic Editor: Brian P. Head
Copyright 2012 Dan N. Predescu et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
Intersectin-1s (ITSN-1s), a protein containing five SH3 (A-E) domains, regulates via the SH3A the function of dynamin-2 (dyn2)
at the endocytic site. ITSN-1s expression was modulated in mouse lung endothelium by liposome delivery of either a plasmid
cDNA encoding myc-SH3A or a specific siRNA targeting ITSN-1 gene. The lung vasculature of SH3A-transduced and ITSN-1sdeficient mice was perfused with gold albumin (Au-BSA) to analyze by electron microscopy the morphological intermediates and
pathways involved in transendothelial transport or with dinitrophenylated (DNP)-BSA to quantify by ELISA its transport. Acute
modulation of ITSN-1s expression decreased the number of caveolae, impaired their transport, and opened the interendothelial
junctions, while upregulating compensatory nonconventional endocytic/transcytotic structures. Chronic inhibition of ITSN-1s
further increased the occurrence of nonconventional intermediates and partially restored the junctional integrity. These findings
indicate that ITSN-1s expression is required for caveolae function and ecient transendothelial transport. Moreover, our results
demonstrate that ECs are highly adapted to perform their transport function while maintaining lung homeostasis.

1. Introduction
ITSN-1s is a multimodular protein, evolutionary conserved
and widely expressed [1, 2]; it consists of two NH2 -termiAnal EH domains, a central coiled-coil domain, and five
consecutive COOH-terminal SH3 domains, SH3A-E, [3, 4].
Similarly to Dyn, ITSN-1s localizes to endocytic clathrincoated pits and caveolae at the plasma membrane and associates preferentially with the neck region of caveolae [5, 6].
The simultaneous presence of multiple SH3 and EH domains, best known for their function in endocytosis, as
well as the subcellular localization of ITSN-1s, led to the
early assumption that ITSN-1s may function as an adaptor/scaold of the general endocytic machinery [3, 5]. Subsequent studies have shown that ITSN-1s is capable of binding essential endocytic proteins, Epsin1/2, Eps15 [7], both

the neuronal and ubiquitously expressed Dyn isoforms,

[3, 5, 6, 8], stonin 2 [9, 10], and the signaling proteins RaIBPassociated Eps15-homology domain protein [11], mSos [12,
13], and 5-phosphatase SHIP2 [14]. ITSN-1s binds several
Dyn molecules simultaneously and clusters them at the
endocytic sites, creating a high concentration of Dyn required for collar formation around the necks of endocytic
vesicles [6, 8]. This is a crucial endocytic event since the
GTPase activity of Dyn is allosterically dependent on Dyn
protein concentration [15, 16] and GTP-dependent selfassembly of Dyn into rings around the necks of budding
vesicles plays a key role in membrane fission [17, 18]. Protein
overexpression, loss-of-function mutations, knockdown or
knockout approaches have provided compelling evidence
for the crucial role of ITSN-1s in scaolding the general
endocytic machinery, recruitment of Dyn, and regulation of

Biochemistry Research International

its function at the endocytic sites [6, 8, 16, 1921]. Evidence

for a possible role of ITSN-1s in regulation of Dyn function
during synaptic vesicles endocytosis has been reported in different species and experimental conditions [16, 2123]. The
ortholog of mammalian ITSN-1s in Caenorhabditis elegans
has a negative regulatory eect on a Dyn-controlled pathway
or on Dyn itself [22]. Drosophilas ITSN, Dyn-associated
protein (Dap160), scaolds endocytic machinery and promotes ecient endocytosis by maintaining a high concentration of Dyn and allowing allosteric activation of its GTPase
activity at synapses [15, 16, 23]. Moreover, in lamprey giant
reticulospinal synapse, acute perturbation of ITSN-Dyn interaction suggested that ITSN may control the amount of
Dyn released from the synaptic vesicle cluster to the periactive zone and that ITSN may regulate fission of clathrincoated intermediates [21]. ITSN-1s knockout mice established the requirement of ITSN-1s for ecient synaptic
vesicle endocytosis and tracking [20]. Moreover, the SH3A
domain of ITSN-1s has the unique ability to regulate Dyn2
oligomerization and its GTPase activity at the neck region
of caveolae [8]. Overexpression of full-length or truncated
ITSN-1s in cultured ECs caused severe impairment of caveolae endocytic function. Electron microscopy analysis of
these cells revealed a wide range of morphological changes
such as caveolae with pleomorphic necks, many of them surrounded by Dyn collars, caveolae clustering, and impaired
membrane fission [6, 8]. In particular, expression of the SH3
region, assumed to function in a dominant negative manner
by sequestering Dyn away from the caveolae endocytic sites
caused significant decrease in caveolae number [6]. Recent
studies revealed the association of ITSN-1 with arfaptin2, Src
homology 3 domain Growth factor receptor-bound 2-like
(endophilin) Interacting Protein 1 (SGIP1), and Fer/Cip4
homology domain only proteins 1 and 2 (FCHO) with membrane-deforming/bending activity essential for initiating
clathrin-coated pits maturation and coated vesicle formation
[11, 24, 25].
In the present study, we addressed the in vivo role of
ITSN-1s in the regulation of caveolae function in transendothelial transport by experimental modulation of ITSN-1s
expression in mouse lung endothelium. Our findings provide
evidence that ITSN-1s is required for ecient caveolae functioning during transendothelial transport. Moreover, modulation of ITSN-1s expression and the subsequent impairment of caveolae internalization and tracking upregulated
alternative endocytic/transcytotic pathways to compensate
for deficient caveolae function.

MA). SureBlue Reserve TMB Microwell Peroxidase Substrate

and horseradish-peroxidase- (HRP-) conjugated reporters
were from KPL (Gaithersburg MA). HyBlot CL films were
from Denville Scientific Inc., (Metuchen, NJ). All electron
microscopy reagents were from Electron Microscopy Science
(Forth Washington, PA). The tracers, DNP-BSA, and 8 nmBSA complexes were prepared in our laboratory as previously
described [26, 27]. Full-length human ITSN-1 cDNA (Gift
from Suzana de la Luna, Center for Genomic Regulation,
UPF, and Centro de Investigacion Biomedica en Red de
Enfermedades Raras (CIBERER-ISCIII), Barcelona, Spain),
was used as a template to generate C-terminal myc-his-SH3A
(residues 740806) as in [8]. SMARTpool reagents and
ITSN-1 siRNA duplex [GGACAUAGUUGUACUGAAAUU
(sense sequence) and 5 -UUUCAGUACAACUAUGU CCUU
(antisense sequence)] for knocking down ITSN-1s expression were from Dharmacon (Lafayette, CO).

2. Materials and Methods

2.5. Cationic-Liposome-Mediated Delivery of Myc-SH3A and

ITSN-1s siRNA to Mouse Lungs. Cationic liposomes were
prepared using DDAB and cholesterol by rotary evaporation
of lipid solutions in chloroform, at 42 C, under vacuum as
in [29, 30]. The lipid film was rehydrated for 1 h at RT in 5%
dextrose followed by sonication and sterile filtration through
0.47 m and 0.22 m membranes. Then, the suspension was
extruded through a 50 nm polycarbonate filter mounted
on a miniextruder device (Avestin, Inc., ON, Canada), to
assure unilamellar vesicles formation. DNA-liposomes complexes were prepared using ratio 8 nmoles liposomes: 1 g

2.1. Materials. Cholesterol and dimethyl dioctadecyl ammonium bromide (DDAB) and bovine serum albumin (BSA)
were from Sigma (St. Louis, MO). MicroBCA protein assay
kit and enhanced chemiluminescent substrate (ECL) were
from Pierce (Rockfort, IL). Relevant antibodies (Abs) were
purchased as follows: rabbit anti-DNP polyclonal Ab from
Invitrogen (Carlsbad, CA), ITSN-1 mAb from BD Transduction Laboratories (Lexington, KY), -actin monoclonal Ab
from Sigma, and myc mAb from Cell Signaling (Danvers,

2.2. Animals. CD1 male mice, 68 weeks old, 2025 g weight,

from Jackson Laboratory (Bar Harbor, ME), kept under
standardized housing and feeding conditions, were used.
The experiments were done under anesthesia, using ketamine (60 mg/kg), acepromazine + xylasine (2.5 mg/kg) +
(2.5 mg/kg), and xylazine (2.5 mg/kg) in 0.1-0.2 mL PBS. All
experiments were approved and performed in accordance
with the guidelines of Rush University Institutional Animal
Care and Use Committee.
2.3. Protein Extraction. Mouse lung tissue was homogenized
in a buer containing 150 mM NaCl, 50 mM Tris, pH 8.0
and protease inhibitors; total lysates were prepared by adding
NP-40, to a final concentration of 1.0%, for 2 h, at 4 C. The
resulted lysates were clarified by centrifugation in a Beckman
tabletop ultracentrifuge with a TLA-55 rotor, for 45 min, at
4 C and 45,000 rpm. Protein concentration was determined
by the microBCA method with a BSA standard.
2.4. Western Blot and Densitometry. Equal amounts of total
protein (70 g/lane) were subjected to 420% SDS PAGE,
transferred to nitrocellulose membranes, and then subjected
to Western blot analyses as in [28], using Abs against ITSN-1s,
myc tag, as well as actin as loading control. The bands of immunoreactivity were visualized using an appropriate HRPconjugated reporter Ab and ECL detection. HyBlot CL
films were subjected to densitometric analysis using NIH
ImageJ1.37v software.

Biochemistry Research International

myc-SH3A DNA. For silencing ITSN-1 gene expression in
mouse lungs we used Dharmacon SMARTpool reagents,
as in [19]. RNase free conditions and reagents were used
throughout the silencing experiments. Preliminary experiments were performed to identify the siRNA sequence of the
SMARTpool, most ecient in silencing ITSN-1s gene. Briefly, the siRNA sequences for knocking down ITSN-1s protein
expression were first delivered in cultured mouse fibroblasts using Dharmafect Transfection Reagent according to
the manufacturers instructions. The target sequence 5 GAGAGAGCCAAGCCG GAAA-3 was found to be the most
ecient in downregulation of ITSN-1s expression in mouse
fibroblasts and thus, used to generate the siRNA duplex for in
vivo delivery. Also, the siCONTROL functional nontargeting
siRNA sequence 5 -UAGCGACUAAACACAUCAA-3 was
used as a control for potential secondary eects caused by
siRNA transfection of mouse fibroblasts.
For silencing ITSN-1s gene, siRNA/liposomes complexes
we prepared using 100 g ITSN-1s/mouse, 1 : 3 (vol/vol)
siRNA : liposomes ratio. Mice were injected intravenously
(tail vein) with cationic liposomes complexes, to achieve
either transduction of the specific myc-SH3A construct or
knockdown of ITSN-1s in lung microvessels. The lungs were
collected at dierent time points after the first delivery of
myc-SH3A plasmid or ITSN-1s siRNA/liposomes complexes
to monitor myc-SH3A and ITSN-1s protein expression
levels.
2.6. DNP-BSA and 8 nm Au-BSA Perfusion. DNP-BSA
(10 mg/mL) was perfused as in [26], to assess lung vascular
permeability to albumin in control mice, mice expressing the
SH3A, or ITSN-1s-deficient mice. The biochemical quantification of transendothelial transport was done as in [26].
Briefly, after tracer perfusion, the lungs were hand minced
and homogenized in PBS, using a Brinkmann polytron (1 min, 4 C). The homogenates were centrifuged
(45,000 rpm, 45 min) to obtain the final supernatants, expected to contain the lung interstitial fluid and thus, the
DNP-derivatized BSA transported to the lung interstitia. The
amounts of DNP-BSA were assessed by ELISA using HRPconjugated anti-DNP Ab followed by a colorimetric assay
with TMB microwell peroxidase substrate as in [26]. The
plates were read in an ELISA reader at 650 nm. For quantitative assessment of DNP-BSA transendothelial transport,
a standard curve using known concentration of DNP-BSA
was generated. The data were expressed in ng DNP-BSA normalized per mg total protein and 10 min. Au-BSA complexes
(A520 = 1.0) were perfused in situ for 10 min, the unbound
tracer was flushed out by 3 min perfusion using oxygenated
PBS as in [27].
2.7. RT-PCR. Mouse lung total RNA was extracted using
RNeasy Mini Kit with on-column DNase digestion by RNasefree DNase set (Qiagen, Valencia, CA). The cDNA was synthesized by using the High-Capacity Reverse Transcription
Kits (Applied Biosystems, ABI, Carlsbad, CA). The RT-PCR
reactions were amplified for 35 cycles with the appropriate
primer sets for the target genes, mouse ITSN-1s, ITSN-2s

3
and ITSN-2L, and a house-keeping gene cyclophilin by using
GoTag Green Master Mix (Promega, San Luis Obispo, CA).
Two to 10 L of PCR products were subjected to 1.2% agarose
gel electrophoresis and visualized using an Alpha Innotech
AlphaImager II system. Primer pairs (Sigma, San Louis, MO)
used in PCR: mouse ITSN-1s (forward: 5 -CAGGCTTGAAAAGTCCTCAAA-3 , reverse: 5 -GGTGATCATGCTGGAAGTCA-3 ); mouse ITSN-2s (forward: 5 -CTGGGCACAATGCAGCCTACG-3 , reverse: 5 -CAAAAAGTTGCAGGCCGTCCATG-3 ), mouse ITSN-2L (forward: 5 -CCAGGACGTTCTGTGTCTCACTATG-3 , reverse: 5 -CAGTAGTAGTCGGCGGGTGGT-3 ), and mouse Cyclophilin
(forward: 5 -GGCAAATGCTGGACCAAACAC-3 , reverse:
5 -TTCCTGGACCCAAAACGCTC-3 ).
2.8. Real-Time qPCR. reactions were performed using Power
SYBR Green PCR Master Mix (ABI) and analyzed in triplicate with a MyiQ Single Color Real-Time PCR Detection
System (Bio-Rad, Hercules, CA). The relative expression level
for each gene was calculated by a mathematical delta-delta Ct
method developed by Applied Biosystems. The amounts of
RNA were normalized to the house-keeping gene cyclophilin
for all time points and reported relative to control mouse
lungs (set at 1).
2.9. Electron Microscopy. 8 nm Au-BSA perfused mouse
lungs were fixed in situ by 15 min with 4% (wt/vol) formaldehyde and 2.5% glutaraldehyde in 1% tannic acid in 0.1 PIPES
buer pH 7.2, and then excised and fixed for 1 h at RT, in the
same fixative mixture. Fixed lungs were used to collect small
tissue blocks that were processed for electron microscopy by
standard procedures including 4% OsO4 alter fixation as in
[26, 27]. Selected specimens were dehydrated in increasing
concentrations of ethanol, followed by propylene oxide, and
embedded in Epon 812. Thin sections cut from blocks were
mounted on nickel grids, then stained with uranyl acetate
and lead citrate, and finally examined and micrographed in a
Joel 1220 transmission electron microscope.
2.10. Morphometric Analyses. Morphometric analyses of
normal and altered caveolar profiles as well as nonconventional endocytic/transcytotic structures were performed as
described previously [6]. Briefly, 1520 sections per grid
and 35 grids per block, from 8 Epon812 blocks, chosen at
random from control and ITSN-1s siRNA-treated mice, were
used. The morphological appearance was recorded on 150
micrographs (30,000 standard magnification) for each experimental condition. The measurement of the length plasma membrane on the electron micrograph used was performed as in [31]. The data were normalized per 100 m
length plasma membrane and are averages SD of three
independent experiments.
2.11. Statistical Analysis. Data were compared using one way
analysis of variance, Students t-test with a Bonferoni correction for multiple comparisons, and a two-way analysis
of variance where needed. Statistical significance was set at
P < 0.05.

3. Results
3.1. Experimental Manipulation of ITSN-1s Expression in
Mouse Lungs. To evaluate in vivo the role of ITSN-1s in regulating caveolae function during transendothelial transport
we modulated the expression of ITSN-1s in mouse lung endothelium by liposome delivery of a plasmid cDNA encoding
the SH3A domain, or a specific siRNA duplex targeting
ITSN-1 gene. Myc-SH3A DNA plasmid : liposomes complexes, prepared as described under Materials and Methods,
were delivered intravenously to mouse lungs as in [29, 30].
The lungs were collected at dierent time points after the
first delivery of myc-SH3A lipoplexes to monitor myc-SH3A
protein expression. Western blot of mouse lung lysates using
myc mAb indicated ecient SH3A expression, at 48 h alter
DNA treatment (Figure 1(a)). For ITSN-1s knockdown, the
silencing eect of ITSN-1s siRNA started at 48 h, it was
obvious at 72 h alter siRNA administration and lasted for an
additional 72 h, as evaluated by Western blot of mouse lung
lysates with ITSN-1 mAb (Figure 1(b)). ITSN-1s protein expression recovered at 168 h alter siRNA delivery. Actin was
used as a loading control. ITSN-1s knockdown did not aect
the expression of caveolin-1 (Cav1), the structural protein
of caveolae coat [32]. Densitometric analyses indicated a
greater than 70% reduction of ITSN-1s protein levels at day
3, compared to wild-type untreated mice (Figure 1(c), asterisk). Intravenous injection of empty liposomes did not aect
ITSN-1s protein levels (not shown). Ecient knockdown
of ITSN-1s at 72 h following siRNA delivery was further
confirmed by real-time quantitative (q) PCR (Figure 1(d)).
ITSN-1s mRNA levels in ITSN-1s siRNA-treated mouse lung
samples were more than 3-fold lower by reference to controls
(Figure 1(d), asterisk). Delivery of empty liposomes (vehicle)
did not aect the levels of mRNA in mouse lungs, compared
to untreated mice. To get more insights into the eects of
ITSN-1s deficiency on caveolae function we also generated a
mouse model with chronic inhibition of ITSN-1s, by repeated delivery of ITSN-1s siRNA/liposome complexes, every
72 hr, for 24 consecutive days (Figure 1(e)). Western blot
analyses applied on mouse lung lysates of ITSN-1s siRNAtreated indicated continuous knockdown of ITSN-1s for 21
days. For accurate quantitative data, we used again qPCR
analysis. The results indicated that the real knockdown of the
target ITSN-1 gene was about 3-fold by reference to cyclophilin, the gene used as standard (Figure 1(d), double asterisk). In addition, conventional RT PCR applied on lung
samples prepared from controls and ITSN-1s-deficient mice
(Figure 1(f)) demonstrated that the knockdown of ITSN-1s
is specific and ecient. The expression of ITSN-1s mRNA
was not detected at days 10, 15 and 24 alter ITSN-1s siRNA
delivery, while ITSN-2s and ITSN-2L mRNAs were not
detectably aected by comparison to untreated controls.
These murine models generated by liposome delivery of the
cDNA plasmid encoding the SH3A domain and the ITSN1s siRNA were used for a further in-detail study of ITSN-1s
endocytic function.
3.2. Acute Perturbation of ITSN-1s Expression Induces Pleomorphic Endocytic/Transcytotic Structures and Increases

Biochemistry Research International

Mouse Lung Vascular Permeability. Once ecient expression
of myc-SH3A was achieved, (48 h alter administration of
myc-SH3A plasmid/liposome complexes), DNP-BSA was
perfused through mouse lung microvasculature and its transendothelial transport was quantified by ELISA, as in [26].
The expression of the SH3A in mouse lung endothelium
caused greater than 36% increase in transendothelial transport of DNP-BSA, compared to controls (Figure 2(a)), suggesting disruption of the interendothelial junctions (IEJs).
The opening of IEJs was further confirmed by perfusion of
8 nm Au-BSA, at 48 h alter myc-SH3A administration. Electron microscopy analysis of the transport pathway followed
by 8 nm Au-BSA from the lumen to the perivascular space
in control specimens indicated nearly exclusive association
between the tracer and caveolae (Figure 2(b)). After 10 min
perfusion, we detected the tracer in the vessel lumen, (Figure
2(b), t), in vesicles open to the luminal side of ECs (Figure
2(b), v1), in vesicles within the endothelial cytoplasm, presumably in transit across the endothelium, (Figure 2(b), v2),
in vesicles discharging their load in the perivascular space
(Figure 2(b), v3), and in variable number in the perivascular
space (not shown). We also determined the location of the
tracer in relation to IEJs. We found no tracer in the junctional
profiles or associated with their abluminal exit; IEJs were not
permeated by the tracer particles that, generated however, filtration residues in their luminal introit, (Figure 2(b), arrowhead). On the contrary, the transendothelial transport of
tracer particles in mouse lung endothelium expressing the
SH3A revealed open IEJs permeated by Au-BSA complexes
(Figures 2(c), and 2(c1), arrows) and in some cases, heavy
association of gold particles with the abluminal exit of IEJs
(Figures 2(c), and 2(c2)). Interestingly, in SH3A-expressing
mouse specimens, caveolae (Figures 2(c), 2(c3), and 2(c4),
arrows) and clathrin-coated pits (Figures 2(c), 2(c7), arrow)
with long necks, and caveolae whose narrow necks were surrounded by staining-dense collars (Figures 2(c), 2(c5), 2(c6),
arrows) were frequently observed. A clathrin-coated pit with
elongated neck is shown in Figures 2(c), 2(c8) (boxed area,
inset). In addition to caveolae, a large number of elongated
elements and tubular membranous structures, many of them
labeled by gold particles (Figures 2(c), 2(c1), and 2(c9) (black
boxed structures), and insets c9.1, c9.2) comprising caveolae
and coated pits (Figures 2(c), 2(c9), and 2(c10), white boxed
structures) as well as large and uncharacteristic caveolae
clusters (c10, arrow and inset c10.1) occupied the intracellular space, in close proximity of the plasma membrane. The
diameter of the tubules ranged between 25 nm to 60 nm,
while their length varied between 150 nm to 600 nm (n = 70
tubules). As previously shown by us [33], the tubular elements may be part of an endosomal tubulovesicular network
that may communicate with cell surface via caveolae and
clathrin-coated vesicles. The network may be upregulated by
expression of the SH3A and acute disturbance of ITSN-1s/
Dyn2 interaction. The observation also reflects impaired
membrane fission reaction and thus, a stimulatory eect of
impaired Dyn2 function on growth of dynamic tubular
structures as reported for cells lacking Dyn1/Dyn2 [34].
Delivery of empty liposomes did not aect the morphology
of caveolae or ECs (not shown). Based on these findings we

Biochemistry Research International

50
37
ITSN-1s siRNA
25

ITSN-1s

20

Actin

168 h

144 h

120 h

(b)

1.2
1.2
Normalized relative mRNA ITSN-1s

% of control (arbitrary densitometric units)

(a)

96 h

Ctrl.

10

72 h

Cav-1
48 h

myc-SA3A

24 h

15

1
0.8
0.6
0.4

0.2
0
wt-ctrl.

veh.

1
0.8
0.6
0.4
0.2
0

72 h
10 d
24 d
ITSN-1s siRNA

wt-ctrl.

(c)

veh.

72 h
10 d
24 d
ITSN-1s siRNA
(d)

Ctrl. veh. 10 d 15 d 24 d
ITSN-1s
ITSN-1s siRNA
Ctrl.

2d

6d

12 d

18 d

24 d

ITSN-2s

ITSN-1s

ITSN-2L

Actin

Cyclophilin
(e)

(f)

Figure 1: Experimental modulation of ITSN-1 expression in mouse lungs. (a) Myc-SH3A domain of ITSN-1s is eciently expressed in
mouse lungs, at 48 h alter delivery of the plasmid cDNA/liposome complexes. Mouse lung lysates (70 g total protein/lane) of control and
myc-SH3A-transduced mice were subjected to 420% SDS PAGE, transferred to nitrocellulose membrane and immunoblotted with myc
Ab. (b) siRNA-mediated knockdown of ITSN-1s in mouse lungs. Lung lysates (70 g total protein/lane) of controls and mice injected with
liposomes/ITSN-1s siRNA complexes were resolved on 420% SDS PAGE and the electrophoretograms were transferred to nitrocellulose
membrane followed by Western blot with Abs against ITSN-1, actin (as a loading control) and cav-1. Note the ecient knockdown of ITSN1s protein for 96 h, starting at day 3 (3 d) alter siRNA delivery. (c) Densitometric analysis of representative HyBlot CL films demonstrates
that treatment with ITSN-1s siRNA/liposome complexes downregulates eciently ITSN-1s protein in mouse lungs. (d) Quantitative RT
PCR of ITSN-1s mRNA in ITSN-1s siRNA treated mouse lungs. Values were normalized to the house-keeping gene cyclophilin and reported
relative to control lungs (set at 1). (e) Chronic inhibition of ITSN-1s expression for 21 consecutive days by repeated delivery of ITSN-1s
siRNA/liposomes complexes. Lung lysates of control and ITSN-1s-siRNA-treated mice were analyzed by Western blot with Abs against ITSN1s and actin, as loading control, at several time points alter delivery, as indicated. (f) Conventional RT-PCR of ITSNs isoforms expressed
in mouse lungs. RT PCR reactions were amplified as described under Materials and Methods using appropriate primer sets for the target
genes, ITSN-1s, ITSN-2s and ITSN-2L, and a house-keeping gene cyclophilin. 5-L of PCR products were subjected to 1.2% agarose gel
electrophoresis and visualized using an Alpha Innotech AlphaImager II system. All data shown are representative for 3 dierent experiments,
(3 mice/experimental condition).

Biochemistry Research International

400

100

8 nm Au-BSA, 10 min

364.515.6 ng

200

265.613.2 ng

300
267.412.2 ng

ng BSA-DNP/mg TP/10 min

t
Lumen
v2

Control

v1

Empty liposomes

v3

v1

pcs

SH3A-liposomes

(a)

(b)

8 nm Au-BSA, 10 min

(c1)

(c2)

(c3)

(c8)

(c7)

(c6)

(c9)

(c5)

(c4)

(c9.1)

(c9.2)

(c10)

(c10.1)
(c1
c10.1
00.1
..1)

(c)

Figure 2: Increased endothelial permeability, impaired membrane fission, and increased occurrence of nonconventional endocytic/
transcytotic intermediates in the mouse lungs expressing the SH3A domain of ITSN-1s. (a) ELISA applied on mouse lung lysates of control,
empty liposomes, and myc-SH3A-transduced mice perfused with 10 mg/mL BSA-DNP, for 10 min. Results were expressed in ng BSA-DNP/
mg total protein/10 min. Bars SD. Results were obtained in 4-5 dierent experiments, with three mice per experimental condition. (b)
Representative EM image showing the transendothelial transport of 8 nm Au-BSA in control mouse lung. Tracer particles were detected in the
vessel lumen, (t), in vesicles open to the luminal side of ECs (v1), in vesicles within the endothelial cytoplasm, (v2), and in vesicles discharging
their load in the perivascular space (pvs; v3). Tracer particles did not permeate the IEJ and generated filtration residues (arrowhead), in the
luminal introit of the junction. Scale bar: 100 nm. (c) Representative electron micrographs showing transendothelial transport of 8 nm
Au-BSA in SH3A-transduced mouse lung specimens as well as nonconventional endocytic/transcytotic intermediates in ECs expressing the
SH3A domain of ITSN-1. Open IEJ permeated by 8 nm gold BSA complexes (c1) and heavy association of gold particles with the abluminal
exit of an IEJ (c2). Caveolae (c3, c4, arrows) and clathrin-coated pits (c7, arrow) with very long necks, and caveolae with their narrow necks
surrounded by staining-dense collars (c5, c6 arrowheads) were frequently observed. A clathrin-coated pit with elongated neck is shown (c8,
boxed area, inset). Elongated elements and tubular membranous structures, labeled by gold particles (c1, c9, black boxed structures, c9.1,
a9.2) and comprising caveolae and coated pits (c9, c10, white boxed structures) as well as large and uncharacteristic caveolae clusters (c10,
arrow and inset). Bars: 200 nm(c1), (c3), (c7), (c8), (c9.1); 100 nm(c2), (c4), (c6), (c9), (c8) inset, (c10); 50 nm(c5), (c9.2).

Biochemistry Research International

7
Lumen
EC

IEJ

400
300

100

Control

pcs
(b)

47215.6 ng

200

319.6 13.2 ng

328.3 12.2 ng

ng BSA-DNP/mg TP/10 min

500

Lumen

IEJ
m
EC1

Control
Empty liposomes
ITSN-1s siRNA
(a)

IEJ

EC2

IEJ

EC3
ITSN-1s siRNA
(c)

ITSN-1s siRNA, 72 h
8 nm Au-BSA, 10 min

Lumen

EC
IEJ

pvs
pvs
(d1)

(d)

Lumen
IEJ

pvs

Air space
(e)

Figure 3: Increased endothelial permeability and impaired interendothelial junctional integrity in mouse lung endothelium acutely depleted
of ITSN-1s. (a) ELISA applied on mouse lung lysates of control, empty liposomes, and ITSN-1s siRNA-treated mice, (72 h alter siRNA),
perfused with 10 mg/mL BSA-DNP, for 10 min. Results were obtained in 4-5 dierent experiments, (3 mice/experimental condition), and
expressed in ng BSA-DNP/mg total protein/10 min. Bars SD. (b, c) EM morphology of IEJs in control (b) and ITSN-1s siRNA (c) mouse
lung endothelium. Arrows in B point to the dense protein matrix of the tight junctions sealing the interendothelial space. Arrowheads in (c)
point to three open IEJs. EC: endothelial cell; m: mitochondria. Bar: 150 nm. (d, e) Representative electron micrographs showing open IEJs
labeled throughout their length by 8 nm Au-BSA particles. Arrows in (d) and magnified d1, point to three-four Au-BSA particles located
close to each other in the same plan, indicative of the wide opening of the IEJ. Gold particles are also associated with the abluminal exit of
IEJs. Note also the limited number of caveolae and dilation of the pericapillary space (pcs; asterisks). Bars: 200 nm (d); 100 nm (e).

concluded that in the presence of SH3A, similar to cultured

ECs [8], Dyn2/Dyn2 interactions are stabilized, Dyn2 function is impaired and caveolae cannot detach from the plasma
membrane to form free transport vesicles. Thus, impaired
caveolae function triggered the expansion of dynamic tubular structures to compensate for deficient transport via
caveolae.
Next, the functional eects of acute ITSN-1s deficiency
(72 h) on lung microvascular permeability were assessed
by 10 min perfusion of 10 mg/mL DNP-BSA through
mouse lung microvasculature and quantifying by ELISA the

amounts of DNP-BSA tracer transported to lung interstitia,

as above. We observed a significant increase in DNP-BSA
transport in the ITSN-1s-deficient mouse relative to controls
(i.e., 472 ng DNP-BSA/mg total protein/10 min in knockdown mouse versus 328.3 ng DNP-BSA/mg total protein/
10 min, in controls), consistent with endothelial barrier dysfunction, Figure 3(a). We further analyzed by electron microscopy the morphology of IEJs and if the tracer, 8 nm Au-BSA,
can escape through open IEJs. Detailed analysis of ultrathin
sections prepared from randomly chosen mouse lung specimens acutely depleted of ITSN-1s showed frequently open

Biochemistry Research International

Lumen

pvs

pvs

ITSN-1s siRNA, 72h

8 nm Au-BSA, 10 min
(a)

(a1)

Lumen

(c)

(b)
Lumen

(d)

(e)

(e.1)

(f)

Figure 4: Acute perturbation of ITSN-1s expression induces pleomorphic endocytic/transcytotic intermediates. (af) Representative EM
images of membranous rings (a, a1, e, e.1, f), tubular elements (b, c, d, arrows) and enlarged endosomes (d, arrowheads), loaded with 8 nm
Au-BSA and associated with caveolae-like morphology. Two tubular structures open to the lumen are shown in (c) and (e). The EM image
shown in (e) was selected from a mouse lung specimen not perfused with 8 nm Au-BSA. Note also the severe dilation of the perivascular
space (pvs) and the proteinaceous edema (a). Bars: 250 nm (a, b); 200 nm (c, d); 150 nm (f); 100 nm (a1, e).

IEJs. The electron micrograph in Figure 3(c) shows a tricellular junction where one side is obviously open. For the
IEJs between the EC2 and EC3, the sectioning of the tissue is
less favorable and one cannot conclude on junctional restrictiveness. By contrast, in control specimens (Figure 3(b)) the
interendothelial space is sealed by the tight junctions, which
are the most apical IEJ structures (Figure 3(b), arrows).
These observations strongly demonstrated that the increased
permeability for BSA-DNP in the lung vasculature of ITSN1s-deficient mouse is due, similar to the SH3A expressing
mice, to the frequent presence of open IEJs. In ITSN-1s depleted specimens, perfused with 8 nm Au-BSA, the pattern
of vesicular labeling was similar to controls, (Figure 2(b)). By
distinction to controls, we detected open IEJs heavily labeled
throughout their length by Au-BSA particles. Morphometric
analysis indicated that about 48% of IEJs counted were open
and penetrated by 8 nm Au-BSA complexes (n = 227). Under
control conditions, mainly in postcapillary venules, we found

that approximately 30% of junctions were open to 6 nm,

thus allowing water, ions, and small molecules to take the
paracellular pathway to the interstitia, data that are in agreement with [35]. Given the size of Au-BSA complexes (the
overall diameter of Au-BSA complexes (gold + BSA molecules) is about 20 nm), we can estimate that the junctions
were widely open: if, for example, in some areas, three or four
gold particles are found in the same horizontal plan, labeling
the junctional space, the opening of the junction could be
estimated at about 60 nm80 nm width. We also detected
gold particles associated with the abluminal exit of the junction (Figures 3(d), 3(d1), and 3(e)). Extensive dilation of the
perivascular space and rich proteinaceous edema is another
phenotypic characteristic of these mouse lungs (Figures 3(d),
and 4(a), asterisks). A new interesting observation was the
frequent detection of pleomorphic morphological structures/ intermediates thought to function in alternative endocytic pathways, such as membranous rings (Figures 4(a),

Biochemistry Research International

4(a1), 4(e), 4(e1), 4(f)), many of them heavily loaded with
8 nm Au-BSA particles (Figures 4(a), and 4(a1)), tubules and
tubulovesicular elements open to the lumen (Figure 4(b),
boxed structures, 4(c), 4(e)) or with no noticeable communication with the extracellular environment (Figures 4(b),
4(c), 4(d), arrows), and enlarged endocytic structures, many
of them fused with typical caveolae, apparently discharging
their load (Figure 4(d), arrowheads). The rings ranged between 150 nm and 400 nm diameter (n = 65 rings) and displayed a narrow lumen, (20 nm 4 nm). Finally, significant
decrease in caveolae number was also observed in mouse
lung ECs deficient of ITSN-1 by reference to controls
(Table 1). The findings strongly suggested that inecient
Dyn2 recruitment to the endocytic site caused by ITSN-1s
deficiency, and the resultant impaired formation of free
caveolae able to function in the transendothelial transport,
caused not only disruption of the IEJs, but also prompted
the ECs of the mouse lung vasculature to upregulate and employ some alternative/compensatory pathways for endocytic
internalization and transendothelial transport (Table 1). Accumulation of tubular elements may also reflect the eect of
inecient Dyn2 recruitment to the endocytic sites, as previously reported for cells lacking Dyn1/Dyn2 [34]. Alltogether,
the findings demonstrate that normal ITSN-1s expression
is crucial for Dyn-mediated membrane fission reaction of
caveolae, for maintaining the caveolae number and for their
transport function in lung microvascular ECs. Moreover,
experimental manipulation of ITSN-1s expression induces
alternative endocytic/transcytotic pathways.
3.3. Chronic Inhibition of ITSN-1s Expression Partially Restores Caveolae Number and IEJ Integrity While Maintaining
the Activation of Alternative Endocytic/Transcytotic Pathways.
To get more insight into the role of ITSN-1s in regulation
of caveolae transport function and endothelial permeability
we next used the mouse model with chronic inhibition of
ITSN-1s expression for 21 consecutive days (Figure 1(c)).
The functional eects of chronic ITSN-1s deficiency on lung
microvascular permeability were assessed by 10 min perfusion of DNP-BSA through mouse lung microvasculature and
quantifying by ELISA the amounts of DNP-BSA tracer transported to lung interstitia, as described above. The findings
were somehow intriguing; transendothelial transport of
DNP-BSA as quantified by ELISA, via DNP Ab returns to
control levels (Figure 5(a)).
The electron microscopy morphological assessment of
interendothelial junctional integrity in mice with chronic
inhibition of ITSN-1s expression indicated that the 8 nm AuBSA particles cannot penetrate the IEJs. Instead, they formed
filtration residues in the luminal introit of the junction
(Figures 5(b) and 5(d), arrows). Interestingly however, the
perivascular spaces showed some dilation and mild edema,
suggesting that the IEJs are still leaky, but not permeable to
the size of our tracer.
Another interesting observation was the frequent detection of multivesicular bodies (mvb), comprising internal
small vesicles of about 50 nm diameter, some of them loaded
with 8 nm Au BSA particles (Figures 5(c) and 5(d)). The

9
alternative endocytic pathways and their morphological intermediates are still present and better represented (Figure
6). Membranous rings with diameters between 100 nm to
400 nm, (Figures 6(a) and 6(c)), free in the cytosol or open to
the lumen and tubules (Figures 6(b), 6(d), and 6(e)), some
of them branching or discharging tracer in the perivascular
space, (Figure 6(b), boxed structure, and inset), and enlarged endocytic structures (most probably secondary lysosomes), many of them fused with typical caveolae (Figure
6(a), arrow) were frequently detected. Often, the tubular
structures connected to the plasma membrane displayed at
their cytosolic termini caveolae (Figures 6, 6(b1)) or coated
vesicles (Figure 6(e), arrowhead). We have also found membrane rues (Figure 6(f), arrow, and insets (f.1), (f.2)) and
macropinocytic vesicles (Figure 6(f), mpv1, mpv2), involved
in tracer uptake, as shown by their gold labeling, and thus,
possibly compensating for deficient caveolae transport.
Morphometric analyses indicated that the number of
endocytic/transcytotic structures is slightly increased at day
24, compared to 72 h time point, with the membranous
ring/tubules labeled by Au-BSA showing a further increase to
14-fold, compared to controls (Table 1). Caveolae number is
partially restored, only 17.8% decrease, by reference to control conditions. Collectively, these observations emphasize
that physiological levels of ITSN-1s are critical for transendothelial transport via caveolae and paracellular permeability in vivo.

4. Discussion
We have shown that ITSN-1s is required for ecient transendothelial transport via caveolae and provided in vivo ultrastructural evidence to suggest that modulation of ITSN-1s
expression activates the paracellular pathway as well as a
previously described, but less understood endocytic tubulovesicular network and membranous rings that seem to
eciently compensate for deficient caveolae transport function. So far, limited evidence regarding nonconventional
endocytic mechanisms were partly obtained in cultured cells
mainly by studies of toxins and viruses [3640]. In this study,
using a murine model expressing the myc-SH3A domain of
ITSN-1s, capable of regulating Dyn2 activity at the endocytic
sites [8], and mice acutely (72 h) and chronically (21 d)
depleted of ITSN-1s via siRNA, we directly demonstrate that
by interfering with expression of ITSN-1s, a key partner of
Dyn in endocytic membrane fission, the caveolae function
was disrupted and alternative endocytic/transcytotic pathways and their morphological intermediates were activated.
Prior studies by others and us indicated that ITSN-1s
functions by recruiting Dyn and regulating its function at the
endocytic site [6, 8, 16, 2123, 41]. Studies in cultured ECs,
complemented by in vitro assays demonstrated that the SH3A
domain of ITSN-1s stimulates both basal and assemblystimulated GTPase activity of Dyn2, as well as Dyn2 ability to
form oligomeric structures [8]. In the presence of SH3A at
the endocytic site, Dyn2/Dyn2 interactions are stabilized, the
lifetime of assembled Dyn2 is prolonged, and despite continuous GTP hydrolysis, Dyn2 oligomers cannot disassemble.
Membrane fission reaction and the subsequent detachment

10

Biochemistry Research International

Table 1: Chronic inhibition of intersectin-1s expression causes the activation of alternative endocytic/transcytotic pathways and partially
restores caveolae number.
Endocytic/transcytotic structures
Caveolae open to the lumen
Caveolae apparently free in the cytosol
Total caveolae (luminal and free in the cytosol)
Abnormal endocytic structures (enlarged endosomes)
Caveolae clusters
Membranous rings
Tubules

Control
106.6 9.5
173.5 12.0
280.1 21.5
2.65 .34
3.95 .4
1.33 .04
.88 .05

ITSN-1s siRNA, 72 h
50.46 4.8
77.41 6.3
127.86 11.1
11.29 2.7
5.64 1.6
9.47 2.4
4.0 1.3

ITSN-1s siRNA, 24 d
87.15 5.9
143.0 9.5
230.14 15.4
14.93 3.8
11.3 2.5
12.8 2.8
18.84 3.4

Results are normalized per 100 m EC length.

300

300.6 11.2 ng

100

296.6 12.8 ng

200
298.4 13.4 ng

ng BSA-DNP/mg TP/10 min

400

Lumen

Lumen

IEJ1
mvb

Control
Empty liposomes

pvs

ITSN -1s siRNA, 24 d

ITSN-1s siRNA, 21 d
(a)

(b)

(c)

Lumen
IEJ2

mvb

pvs

pvs

(d)

Figure 5: Chronic inhibition of ITSN-1s expression partially restores caveolae number and IEJ integrity. (a) ELISA applied on mouse lung
lysates of control, empty liposomes, and ITSN-1s siRNA-treated mice, 24 d alter delivery, perfused with 10 mg/mL BSA-DNP, for 10 min.
Results were obtained in 4-5 dierent experiments, (3 mice/experimental condition), and expressed in ng BSA-DNP/mg total protein/10 min.
Bars SD. (b, d) Filtration residues in the luminal introit of two IEJs (arrowheads). Mild dilation ( ) of the pvs suggests leakiness of the
IEJs. Multivesicular bodies (mvb), in close proximity of the plasma membrane, have some of their internal small vesicles, labeled by 8 nm
Au-BSA particles (c, d). Bars: 100 nm.

of caveolae from the plasma membrane leading to the

formation of free vesicular carriers were inhibited. Here, we
extended these studies and expressed the SH3A of ITSN-1s
in mouse lung endothelium, to address the in vivo regulatory
role of ITSN-1s on membrane fission and Dyn itself. Similar
to cultured cells, we see impaired caveolae release from the
plasma membrane, caveolae necks surrounded by stainingdense collars as well as elongated elements and tubular

membranous structures, associated with caveolae-like morphology. In addition, the deficient transcellular transport via
caveolae caused opening of the IEJs and activation of the
paracellular pathway, consistent with a functional crosstalk
between the trans- and paracellular transport pathways in
mouse lung endothelium. These observations support the in
vivo role of ITSN-1s, via its SH3A domain, in regulation of
Dyn2 function at the neck of caveolae and the importance

Biochemistry Research International

11

(a)

(b)

Lumen

Lumen

(b.1)
pvs
(d)

(c)

(e)

Lumen

pvs

pvs
(f.1)

(f)

mpv1

pvs
(f.2)

Lumen

mpv2

Figure 6: Chronic inhibition of ITSN-1s expression augments the activation of alternative endocytic/transcytotic pathways. (ae) tubulovesicular EM images of membranous rings with diameters between 100 nm to 400 nm, (a, c), free in the cytosol or open to the lumen
and tubules (b, d, e), some of them branching or discharging tracer in the perivascular space, (b, boxed structure, and inset), and enlarged
endocytic structures, many of them fused with typical caveolae (a). Note the tubular structures are connected to the plasma membrane
and displaying at their cytosolic termini caveolae (b1) or coated vesicles (e, arrowhead). Bars: 100 nm (b, c, d); 250 nm (a); 75 nm (e); (f)
membrane rues (f, arrow, insets f.1, f.2) and large macropinocytic vesicles (mpv1, mpv2) labeled by tracer particles. Bar: 150 nm.

of ITSN-1s/Dyn2 interaction in caveolae internalization and

tracking.
Acute siRNA-mediated knockdown of ITSN-1s, 72 h alter
liposomes/siRNA delivery caused 59% decrease in caveolae
number, disruption of IEJs, leading to 43% increase in DNPBSA transendothelial transport, compared to controls. Without ITSN-1s, Dyn2 a major interacting partner of ITSN-1s
and essential player in detachment of caveolae from the plasma membrane is not eciently recruited to the endocytic site
and thereby, caveolae detachment from the plasma membrane and formation of free vesicular carriers is disrupted
[19, 23]. Similar to Dyn, other endocytic proteins may be
mislocated and may interact less eciently aecting the dynamic nature of endocytic complexes [23]. Thus, regardless
of the molecular mechanisms involvedimpaired Dyn function, in the case of the SH3A expression, or inecient Dyn
recruitment to the endocytic site, in the case of ITSN-1s
knockdownthe membrane fission reaction is inhibited
and caveolae detachment from the plasma membrane does
not occur. ITSN-1s with its five SH3 domains is critical in
maintaining high Dyn concentration and thus, regulation of
its GTPase activity and oligomerization at the endocytic site.
Transendothelial transport (transcytosis) via caveolae is
the dominant mechanism of transendothelial transport; the
number of clathrin-coated vesicles in ECs is relatively small
and their contribution to the internalization process is minor
[6, 42, 43]. In ECs the true caveolae (vesicular carriers Cav1

positive) represent 95% of the whole vesicular population

and the non-Cav1 and nonclathrin vesicles identified in the
Cav1 null mouse are poorly represented and not yet well
defined as a Cav1 and clathrin-independent vesicular transport system [44, 45]. Our studies confirmed that indeed
caveolae are the main vesicular carriers of ECs functioning
in transendothelial transport. Similar to Cav1 null mice, that
lack morphologically identifiable caveolae, altered expression
of ITSN-1s, by decreasing caveolae number and impairing
their transport function, activated compensatory transport
pathways. We showed here that increased paracellular transport of tracer particles with diameter between 6 nm (DNPBSA) to 20 nm (Au-BSA) occurs in the ECs of mice with
acute modulation of ITSN-1s levels, while the high frequency
of pleomorphic tubulovesicular and ringlike structures labeled by Au-BSA, implied their direct participation in tracer
uptake and cellular tracking. ITSN-1s deficiency was able
not only to induce activation of classical interendothelial
pathway, but also to modulate the opening of the junctions.
Under conditions of ITSN-1s chronic inhibition, the paracellular pathway is still leaky but not permeable to 18 nm
diameter tracer. Apparently, ECs deficient of ITSN-1s limit
the activation of the paracellular pathway to small molecules, while the transport of macromolecules (Au-BSA and
DNP-BSA) reach similar values as the controls. This apparent return of junctional integrity may be part of a general protective mechanism against inflammation and lung

12
injury. Moreover, the increased presence of pleomorphic
tubulovesicular and ringlike structures labeled by Au-BSA
suggests their augmented participation in tracer uptake and
cellular tracking. To our knowledge, this is the first report
showing the in vivo modulation of alternative transport
pathways.
A large intracellular tubulovesicular network of endosomal origin that extended all the way from the nucleus to the
plasma membrane and involved in receptor recycling has
been previously reported [46]. Later on, a similar system
was characterized by Palades group and shown to comprise
G-protein-coupled receptors for thrombin and platelet activating factor [33, 40, 47, 48]. Structurally, it was shown that
this tubulovesicular network does not colocalize with the
early endosomes, Golgi and endoplasmic reticulum. Moreover, the network was shown to communicate with both
extracellular environment and nuclear compartment of ECs.
HRP uptake studies indicated that HRP labeled this tubular
network, in just minutes of treatment, consistent with the
idea it participates very early in the internalization process
[33]. In the mouse models used in this study, the alternative
intracellular transport routes are upregulated; deficient recruitment and impaired Dyn function at the endocytic sites
due to SH3A expression or ITSN-1s depletion may favor the
formation of tubular structures as shown in cells with conditional Dyn1/Dyn2 knockout [34]. In addition, an impaired
coordination between membrane invagination and vesicle
detachment may lead to formation of large invaginations.
Both ITSN-1s and 2 directly interact with the FCHo1/2 Bar
domain containing proteins [24, 49]. While no direct functional link has been established between ITSN-1s and F-BAR
domain containing proteins in caveolae budding, this interaction is likely to account for the altered caveolae morphology within the tubulovesicular structures and caveolae
clusters. Interestingly, chronic inhibition of ITSN-1s shows
no increase in the DNP-BSA transported to the interstitia.
Widening of IEJs is limited. The number of nonconventional
endocytic structures was greater, with the membranous ring/
tubules showing a 2-fold increase, compared to 72 h ITSN-1s
knockdown.
While our results can be explained by a model where
ITSN-1s modified expression and Dyn2 impaired function
may favor membrane tubulation, a direct action of ITSN-1s
on actin polymerization cannot be ruled out. We have shown
that the temporal and spatial actin polymerization and remodeling subadjacent to the plasma membrane plays an
important role in caveolae internalization [50]. ITSN-1s
binds via its SH3A the N-WASp, [51], an actin-nucleating
factor that acts at the plasma membrane upstream of Arp2/3
complex [52]. The physiological significance of WASp interaction with short ITSN-1, which does not possess the DHPH tandem as the long ITSN isoforms, is still not understood. Impaired caveolae internalization and tracking may
also have eects on cellular signaling. The upregulation of
endocytic membranous structures may provide distinct signaling platforms [53], capable of partially compensating for
caveolae signaling. Moreover, the caveolae number was partially restored, suggesting that the highly homologous ITSN2s isoform may partially compensate for ITSN-1s chronic

Biochemistry Research International

deficiency, a question that remains to be addressed in future
work.
Caveolae and Cav1are involved in important cellular
functions such as proliferation, apoptosis, cell dierentiation, cell migration, endocytosis, and transcytosis [44, 54,
55]. Caveolae are present in all cell types in the lung, including ECs, epithelial cells, smooth muscle cells, fibroblasts,
telocytes, and macrophages [55, 56]. Recent research implicated caveolae and Cav1 in pathogenesis of human lung
disease [45, 57, 58], neurodegenerative diseases [59], and
lung cancer [60, 61]. Not only Cav1, but also the ubiquitously
expressed ITSN-1s, a major protein of lung tissue, has been
related to lung inflammatory conditions [62], to endosomal
disorders associated with the earliest pathogenesis of Alzheimers disease and Parkinsons diseases [63] as well as cancer [64, 65]. Thus, a better understanding of the endocytic
function of ITSN-1s and how it regulates caveolae internalization and tracking may provide new insights for understanding of human disease and give new opportunities for
novel targets in therapeutic approaches.
In conclusion, our results provide new information on
the in vivo role of ITSN-1s as a scaold, recruiter and regulator of caveolae endocytic machinery and strengthen the
evidence for a functional endocytic tubulovesicular network
induced by ITSN-1s deficiency.

Abbreviations
Ab:
Au-BSA:
caveolin-1:
DNP-BSA:
Dyn:
EC:
IEJs:
ITSN-1s:

Antibody
Gold-bovine serum albumin
Cav1
Dinitrophenylated-bovine serum albumin
Dynamin
Endothelial cell
Interendothelial junctions
Intersectin-1s.

Acknowledgments
This work was supported by National Institute of Health
Grants R01HL089462, R01HL089462-02S1 and American
Heart Association SDG0635175N (to S. A. Predescu).

References
[1] M. Guipponi, H. S. Scott, H. Chen, A. Schebesta, C. Rossier,
and S. E. Antonarakis, Two isoforms of a human intersectin
(ITSN) protein are produced by brain-specific alternative
splicing in a stop codon, Genomics, vol. 53, no. 3, pp. 369
376, 1998.
[2] C. Pucharcos, C. Casas, M. Nadal, X. Estivill, and S. de la
Luna, The human intersectin genes and their spliced variants
are dierentially expressed, Biochimica et Biophysica Acta, vol.
1521, no. 13, pp. 111, 2001.
[3] M. Okamoto, S. Schoch, and T. C. Sudhof, EHSH1/intersectin, a protein that contains EH and SH3 domains and binds to
dynamin and SNAP-25. A protein connection between exocytosis and endocytosis? The Journal of Biological Chemistry,
vol. 274, no. 26, pp. 1844618454, 1999.

Biochemistry Research International

[4] M. Yamabhai, N. G. Homan, N. L. Hardison et al., Intersectin, a novel adaptor protein with two Eps15 homology and
five Src homology 3 domains, The Journal of Biological Chemistry, vol. 273, no. 47, pp. 3140131407, 1998.
[5] N. K. Hussain, M. Yamabhai, A. R. Ramjaun et al., Splice variants of intersectin are components of the endocytic machinery in neurons and nonneuronal cells, The Journal of Biological Chemistry, vol. 274, no. 22, pp. 1567115677, 1999.
[6] S. A. Predescu, D. N. Predescu, B. K. Timblin, R. V. Stan, and
A. B. Malik, Intersectin regulates fission and internalization
of caveolae in endothelial cells, Molecular Biology of the Cell,
vol. 14, no. 12, pp. 49975010, 2003.
[7] A. S. Sengar, W. Wang, J. Bishay, S. Cohen, and S. E. Egan, The
EH and SH3 domain Ese proteins regulate endocytosis by
linking to dynamin and Eps15, EMBO Journal, vol. 18, no.
5, pp. 11591171, 1999.
[8] I. Knezevic, D. Predescu, C. Bardita et al., Regulation of
dynamin-2 assembly-disassembly and function through the
SH3A domain of intersectin-1s, Journal of Cellular and Molecular Medicine, vol. 15, no. 11, pp. 23642376, 2011.
[9] J. A. Martina, C. J. Bonangelino, R. C. Aguilar, and J. S. Bonifacino, Stonin 2: an adaptor-like protein that interacts with
components of the endocytic machinery, Journal of Cell Biology, vol. 153, no. 5, pp. 11111120, 2001.
[10] L. E. Kelly and A. M. Phillips, Molecular and genetic characterization of the interactions between the Drosophila stonedB protein and DAP-160 (intersectin), Biochemical Journal,
vol. 388, no. 1, pp. 196204, 2005.
[11] O. Dergai, O. Novokhatska, M. Dergai et al., Intersectin 1
forms complexes with SGIP1 and Reps1 in clathrin-coated
pits, Biochemical and Biophysical Research Communications,
vol. 402, no. 2, pp. 408413, 2010.
[12] X. K. Tong, N. K. Hussain, A. G. Adams, J. P. OBryan, and P.
S. McPherson, Intersectin can regulate the Ras/MAP kinase
pathway independent of its role in endocytosis, The Journal of
Biological Chemistry, vol. 275, no. 38, pp. 2989429899, 2000.
[13] X. K. Tong, N. K. Hussain, E. de Heuvel et al., The endocytic
protein intersectin is a major binding partner for the Ras exchange factor mSos1 in rat brain, EMBO Journal, vol. 19, no.
6, pp. 12631271, 2000.
[14] F. Nakatsu, R. M. Perera, L. Lucast et al., The inositol 5phosphatase SHIP2 regulates endocytic clathrin-coated pit
dynamics, Journal of Cell Biology, vol. 190, no. 3, pp. 307315,
2010.
[15] M. H. Stowell, B. Marks, P. Wigge, and H. T. McMahon,
Nucleotide-dependent conformational changes in dynamin:
evidence for a mechanochemical molecular spring, Nature
Cell Biology, vol. 1, no. 1, pp. 2732, 1999.
[16] B. Marie, S. T. Sweeney, K. E. Poskanzer, J. Roos, R. B. Kelly,
and G. W. Davis, Dap160/Intersectin scaolds the periactive
zone to achieve high-fidelity endocytosis and normal synaptic
growth, Neuron, vol. 43, no. 2, pp. 207219, 2004.
[17] P. Oh, D. P. McIntosh, and J. E. Schnitzer, Dynamin at the
neck of caveolae mediates their budding to form transport
vesicles by GTP-driven fission from the plasma membrane of
endothelium, Journal of Cell Biology, vol. 141, no. 1, pp. 101
114, 1998.
[18] J. E. Hinshaw and S. L. Schmid, Dynamin self-assembles into
rings suggesting a mechanism for coated vesicle budding,
Nature, vol. 374, no. 6518, pp. 190192, 1995.

13
[19] S. A. Predescu, D. N. Predescu, I. Knezevic, I. K. Klein, and
A. B. Malik, Intersectin-1s regulates the mitochondrial apoptotic pathway in endothelial cells, The Journal of Biological
Chemistry, vol. 282, no. 23, pp. 1716617178, 2007.
[20] Y. Yu, P. Y. Chu, D. N. Bowser et al., Mice deficient for the
chromosome 21 ortholog Itsn1 exhibit vesicle-tracking abnormalities, Human Molecular Genetics, vol. 17, no. 21, pp.
32813290, 2008.
[21] E. Evergren, H. Gad, K. Walther, A. Sundborger, N. Tomilin,
and O. Shupliakov, Intersectin is a negative regulator of dynamin recruitment to the synaptic endocytic zone in the central synapse, Journal of Neuroscience, vol. 27, no. 2, pp. 379
390, 2007.
[22] S. Rose, M. G. Malabarba, C. Krag et al., Caenorhabditis
elegans intersectin: a synaptic protein regulating neurotransmission, Molecular Biology of the Cell, vol. 18, no. 12, pp.
50915099, 2007.
[23] T. W. Koh, P. Verstreken, and H. J. Bellen, Dap160/intersectin
acts as a stabilizing scaold required for synaptic development
and vesicle endocytosis, Neuron, vol. 43, no. 2, pp. 193205,
2004.
[24] W. M. Henne, E. Boucrot, M. Meinecke et al., FCHo proteins
are nucleators of clathrin-mediated endocytosis, Science, vol.
328, no. 5983, pp. 12811284, 2010.
[25] E. Scappini, T. W. Koh, N. P. Martin, and J. P. OBryan, Intersectin enhances huntingtin aggregation and neurodegeneration through activation of c-Jun-NH2 -terminal kinase, Human Molecular Genetics, vol. 16, no. 15, pp. 18621871, 2007.
[26] S. A. Predescu, D. N. Predescu, and G. E. Palade, Plasmalemmal vesicles function as transcytotic carriers for small
proteins in the continuous endothelium, American Journal of
Physiology, vol. 272, no. 2, pp. H937H949, 1997.
[27] D. Predescu, R. Horvat, S. Predescu, and G. E. Palade, Transcytosis in the continuous endothelium of the myocardial
microvasculature is inhibited by N-ethylmaleimide, Proceedings of the National Academy of Sciences of the United States of
America, vol. 91, no. 8, pp. 30143018, 1994.
[28] S. A. Predescu, D. N. Predescu, and G. E. Palade, Endothelial
transcytotic machinery involves supramolecular protein-lipid
complexes, Molecular Biology of the Cell, vol. 12, no. 4, pp.
10191033, 2001.
[29] J. W. McLean, E. A. Fox, P. Baluk et al., Organ-specific endothelial cell uptake of cationic liposome-DNA complexes in
mice, American Journal of Physiology, vol. 273, no. 1, pp.
H387H404, 1997.
[30] K. Miyawaki-Shimizu, D. Predescu, J. Shimizu, M. Broman, S.
Predescu, and A. B. Malik, siRNA-induced caveolin-1 knockdown in mice increases lung vascular permeability via the
junctional pathway, American Journal of Physiology, vol. 290,
no. 2, pp. L405L413, 2006.
[31] D. Predescu, S. Predescu, J. Shimizu, K. Miyawaki-Shimizu,
and A. B. Malik, Constitutive eNOS-derived nitric oxide is
a determinant of endothelial junctional integrity, American
Journal of Physiology, vol. 289, no. 3, pp. L371L381, 2005.
[32] K. G. Rothberg, J. E. Heuser, W. C. Donzell, Y. S. Ying, J. R.
Glenney, and R. G. W. Anderson, Caveolin, a protein component of caveolae membrane coats, Cell, vol. 68, no. 4, pp.
673682, 1992.
[33] K. Ihida, D. Predescu, R. P. Czekay, and G. E. Palade,
Platelet activating factor receptor (PAF-R) is found in a large
endosomal compartment in human umbilical vein endothelial
cells, Journal of Cell Science, vol. 112, no. 3, pp. 285295, 1999.

14
[34] S. M. Ferguson, A. Raimondi, S. Paradise et al., Coordinated
actions of actin and BAR proteins upstream of dynamin at
endocytic clathrin-coated pits, Developmental Cell, vol. 17,
no. 6, pp. 811822, 2009.
[35] M. Simionescu, A. Gafencu, and F. Antohe, Transcytosis of
plasma macromolecules in endothelial cells: a cell biological
survey, Microscopy Research and Technique, vol. 57, no. 5, pp.
269288, 2002.
[36] W. Romer, L. Berland, V. Chambon et al., Shiga toxin induces
tubular membrane invaginations for its uptake into cells,
Nature, vol. 450, no. 7170, pp. 670675, 2007.
[37] M. Marsh and A. Helenius, Virus entry: open sesame, Cell,
vol. 124, no. 4, pp. 729740, 2006.
[38] K. Sandvig and B. van Deurs, Membrane trac exploited by
protein toxins, Annual Review of Cell and Developmental Biology, vol. 18, pp. 124, 2002.
[39] M. Kirkham, A. Fujita, R. Chadda et al., Ultrastructural identification of uncoated caveolin-independent early endocytic
vehicles, Journal of Cell Biology, vol. 168, no. 3, pp. 465476,
2005.
[40] D. Predescu, K. Ihida, S. Predescu, and G. E. Palade, The
vascular distribution of the platelet-activating factor receptor,
European Journal of Cell Biology, vol. 69, no. 1, pp. 8698, 1996.
[41] D. R. Glodowski, C. C. Chen, H. Schaefer, B. D. Grant, and
C. Rongo, RAB-10 regulates glutamate receptor recycling
in a cholesterol-dependent endocytosis pathway, Molecular
Biology of the Cell, vol. 18, no. 11, pp. 43874396, 2007.
[42] G. E. Palade, M. Simionescu, and N. Simionescu, Structural
aspects of the permeability of the microvascular endothelium,
Acta physiologica Scandinavica. Supplementum, vol. 463, pp.
1132, 1979.
[43] N. Simionescu and M. Simionescu, Cellular interactions of
lipoproteins with the vascular endothelium: endocytosis and
transcytosis, Targeted Diagnosis and Therapy, vol. 5, pp. 45
95, 1991.
[44] S. A. Predescu, D. N. Predescu, and A. B. Malik, Molecular
determinants of endothelial transcytosis and their role in
endothelial permeability, American Journal of Physiology, vol.
293, no. 4, pp. L823L842, 2007.
[45] M. Drab, P. Verkade, M. Elger et al., Loss of caveolae, vascular dysfunction, and pulmonary defects in caveolin-1 genedisrupted mice, Science, vol. 293, no. 5539, pp. 24492452,
2001.
[46] C. R. Hopkins, Intracellular routing of transferrin and transferrin receptors in epidermoid carcinoma A431 cells, Cell, vol.
35, no. 1, pp. 321330, 1983.
[47] R. Horvat and G. E. Palade, Thrombomodulin and thrombin
localization on the vascular endothelium; their internalization
and transcytosis by plasmalemmal vesicles, European Journal
of Cell Biology, vol. 61, no. 2, pp. 299313, 1993.
[48] R. Horvat and G. E. Palade, The functional thrombin receptor is associated with the plasmalemma and a large endosomal
network in cultured human umbilical vein endothelial cells,
Journal of Cell Science, vol. 108, no. 3, pp. 11551164, 1995.
[49] S. Ahmed, W. Bu, R. T. Chuen Lee, S. Maurer-Stroh, and W.
Ing Goh, F-BAR domain proteins: families and function,
Communitative & Integrative Biology, vol. 3, no. 2, pp. 116
121, 2010.
[50] I. K. Klein, D. N. Predescu, T. Sharma, I. Knezevic, A. B. Malik,
and S. Predescu, Intersectin-2L regulates caveola endocytosis
secondary to Cdc42-mediated actin polymerization, The
Journal of Biological Chemistry, vol. 284, no. 38, pp. 25953
25961, 2009.

Biochemistry Research International

[51] N. K. Hussain, S. Jenna, M. Glogauer et al., Endocytic protein
intersectin-l regulates actin assembly via Cdc42 and N-WASP,
Nature Cell Biology, vol. 3, no. 10, pp. 927932, 2001.
[52] T. Takenawa and S. Suetsugu, The WASP-WAVE protein network: connecting the membrane to the cytoskeleton, Nature
Reviews Molecular Cell Biology, vol. 8, no. 1, pp. 3748, 2007.
[53] M. T. Howes, S. Mayor, and R. G. Parton, Molecules, mechanisms, and cellular roles of clathrin-independent endocytosis,
Current Opinion in Cell Biology, vol. 22, no. 4, pp. 519527,
2010.
[54] J. H. Chidlow Jr. and W. C. Sessa, Caveolae, caveolins, and
cavins: complex control of cellular signalling and inflammation, Cardiovascular Research, vol. 86, no. 2, pp. 219225,
2010.
[55] Y. Jin, S.-J. Lee, R. D. Minshall, and A. M.K. Choi, Caveolin1: a critical regulator of lung injury, American Journal of Physiology, vol. 300, no. 2, pp. L151L160, 2011.
[56] L. M. Popescu and M. S. Faussone-Pellegrini, TELOCYTES
a case of serendipity: the winding way from Interstitial Cells
of Cajal (ICC), via Interstitial Cajal-Like Cells (ICLC) to
TELOCYTES, Journal of Cellular and Molecular Medicine, vol.
14, no. 4, pp. 729740, 2010.
[57] X. M. Wang, Y. Zhang, P. K. Hong et al., Caveolin-1: a critical
regulator of lung fibrosis in idiopathic pulmonary fibrosis,
Journal of Experimental Medicine, vol. 203, no. 13, pp. 2895
2906, 2006.
[58] R. Mathew, Cell-specific dual role of caveolin-1 in pulmonary
hypertension, Pulmonary Medicine, vol. 2011, Article ID
573432, 12 pages, 2011.
[59] J. Y. Park, K. S. Kim, S. B. Lee et al., On the mechanism of
internalization of -synuclein into microglia: roles of ganglioside GM1 and lipid raft, Journal of Neurochemistry, vol. 110,
no. 1, pp. 400411, 2009.
[60] M. Li, H. Chen, L. Diao, Y. Zhang, C. Xia, and F. Yang, Caveolin-1 and VEGF-C promote lymph node metastasis in the
absence of intratumoral lymphangiogenesis in non-small cell
lung cancer, Tumori, vol. 96, no. 5, pp. 734743, 2010.
[61] H. L. Chen, L. F. Fan, J. Gao, J. P. Ouyang, and Y. X. Zhang,
Dierential expression and function of the caveolin-1 gene in
non-small cell lung carcinoma, Oncology Reports, vol. 25, no.
2, pp. 359366, 2011.
[62] S. Singla, D. Predescu, C. Bardita et al., Pro-inflammatory
endothelial cell dysfunction is associated with intersectin-1s
down-regulation, Respiratory Research, vol. 12, article 46,
2011.
[63] D. J. Keating, C. Chen, and M. A. Pritchard, Alzheimers disease and endocytic dysfunction: clues from the Down syndrome-related proteins, DSCR1 and ITSN1, Ageing Research
Reviews, vol. 5, no. 4, pp. 388401, 2006.
[64] J. B. Wang, W. J. Wu, and R. A. Cerione, Cdc42 and Ras
cooperate to mediate cellular transformation by intersectin-L,
The Journal of Biological Chemistry, vol. 280, no. 24, pp.
2288322891, 2005.
[65] D. N. Predescu, M. Wang, M. Patel, C. Bardita, and S. Predescu, Intersectin-1s deficiencyan early molecular pathogenetic event of lung cancer, in Proceedings of the American
Thoracic Society International Meeting, 2011, abstract.