Structural and functional characteristics of lung macro- and microvascular endothelial cell phenotypes Judy King, a Tray Hamil, b Judy Creighton, b Songwei Wu, b Priya Bhat, b Freda McDonald, a and Troy Stevens b, * a Department of Pathology, Center for Lung Biology, The University of South Alabama College of Medicine, Mobile, AL 36617, USA b Department of Pharmacology, Center for Lung Biology, The University of South Alabama College of Medicine, Mobile, AL 36688, USA Received 11 September 2003 Abstract Lung macro- and microvascular endothelial cells exhibit unique functional attributes, including signal transduction and barrier properties. We therefore sought to identify structural and functional features of endothelial cells that discriminate their phenotypes in the fully differentiated lung. Rat lung macro- (PAEC) and microvascular (PMVEC) endothelial cells each exhibited expression of typical markers. Screening for reactivity with nine different lectins revealed that Glycine max and Griffonia (Bandeiraea) simplicifolia preferentially bound microvascular endothelia whereas Helix pomatia preferentially bound macrovascular endothelia. Apposition between the apical plasmalemma and endoplasmic reticulum was closer in PAECs (8 nm) than in PMVECs (87 nm), implicating this coupling distance in the larger store operated calcium entry responses observed in macrovascular cells. PMVECs exhibited a faster growth rate than did PAECs and, once a growth program was initiated by serum, PMVECs sustained growth in the absence of serum. Thus, PAECs and PMVECs differ in their structure and function, even under similar environmental conditions. D 2004 Elsevier Inc. All rights reserved. Keywords: Endoplasmic reticulum; Calcium; Store-operated calcium entry; Proliferation; Lectins Introduction Although endothelium lines blood vessels throughout the circulation, it exhibits highly specialized functions in dif- ferent vascular sites. In the systemic circulation, permeabil- ity edema is prominent at post-capillary venules (Thurston et al., 2000). White blood cell recruitment to sites of inflammation occurs at high endothelial venules (Cavender, 1990; Colditz, 1985) and, while the blood brain barrier consists of endothelium with tight cell – cell junctions (Gloor et al., 2001) that are highly restrictive, both renal glomerular (Stan et al., 1999) and liver sinusoidal endothelium (Gri- sham et al., 1975) possess fenestrations that are highly permeable. It is clear that these distinct endothelial cell characteristics are at least partly directed by environmental cues (Stevens et al., 2001). The embryological origin of endothelial cells may also contribute to their site-specific function (Stevens et al., 2001). Studies in the developing lung suggest two distinct processes form the circulation (deMello and Reid, 2000; deMello et al., 1997; Hall et al., 2000; Schachtner et al., 2000; Schwarz et al., 2000), including angiogenesis of large vessels and vasculogenesis of small vessels. deMello et al. (1997) used a casting technique to temporally illustrate vascular tube formation. The earliest formed vascular structures were observed at embryonic day 14 (E14) in the developing mouse lung. These structures progressively branched from large vessels at sharp angles, consistent with angiogenesis, but did not form a contigu- ous vessel. The parallel growth of blood lakes/islands that were filled with precursor cells of hematopoietic origin was observed by transmission electron microscopy until, at E15, a fusion between angiogenic sprouts and vasculo- genic blood islands could be resolved using the vascular casting technique. This issue has also been addressed by assessing the temporal expression pattern of endothelial 0026-2862/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.mvr.2003.11.006 * Corresponding author. Fax: +1-251-460-7452. E-mail address: [email protected] (T. Stevens). www.elsevier.com/locate/ymvre Microvascular Research 67 (2004) 139 – 151
13
Embed
Structural and functional characteristics of lung macro - Penn Medicine
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
www.elsevier.com/locate/ymvre
Microvascular Research 67 (2004) 139–151
Structural and functional characteristics of lung macro- and
aDepartment of Pathology, Center for Lung Biology, The University of South Alabama College of Medicine, Mobile, AL 36617, USAbDepartment of Pharmacology, Center for Lung Biology, The University of South Alabama College of Medicine, Mobile, AL 36688, USA
Received 11 September 2003
Abstract
Lung macro- and microvascular endothelial cells exhibit unique functional attributes, including signal transduction and barrier properties.
We therefore sought to identify structural and functional features of endothelial cells that discriminate their phenotypes in the fully
differentiated lung. Rat lung macro- (PAEC) and microvascular (PMVEC) endothelial cells each exhibited expression of typical markers.
Screening for reactivity with nine different lectins revealed that Glycine max and Griffonia (Bandeiraea) simplicifolia preferentially bound
microvascular endothelia whereas Helix pomatia preferentially bound macrovascular endothelia. Apposition between the apical
plasmalemma and endoplasmic reticulum was closer in PAECs (8 nm) than in PMVECs (87 nm), implicating this coupling distance in
the larger store operated calcium entry responses observed in macrovascular cells. PMVECs exhibited a faster growth rate than did PAECs
and, once a growth program was initiated by serum, PMVECs sustained growth in the absence of serum. Thus, PAECs and PMVECs differ in
their structure and function, even under similar environmental conditions.
PA) and grown to confluence. Cytosolic Ca2+ was esti-
mated with the Ca2+-sensitive fluorophore fura 2/acetox-
ymethylester (Molecular Probes, Eugene, OR) according
to methods previously described. Calculations of free
Fig. 1. Endothelial cell phenotypes in the intact lung can be discriminated by le
(arrows) represents staining of endothelial cells with Helix pomatia. No staining
microscope). Staining with Glycine max and Griffonia simplicifolia is absent in the
confocal microscope).
[Ca2+]i are routinely made using modifications of the
formula described by Grynkiewicz et al. (1985).
Cell growth
Endothelial cells were seeded at 1 � 105 cells per well
in six well plates at n = 3. Cells were seeded in normal
media containing DMEM, 10% FBS (or as otherwise
noted), and 1 � pen/strep. Every 24 h for 6 days after
the seeding date, cells were photographed, resuspended
using trypsin, and counted using a Coulter counter.
Data analysis
Numerical data are reported as mean F SEM. One-
way ANOVA was used to evaluate differences between
experimental groups, with a Student Newman–Keuls
post hoc test as appropriate. Significance was considered
P < 0.05.
Results
Lectin binding to lung macro- and microvascular
endothelial cells
Lectin binding has previously been utilized as an effective
method of discriminating between macro- and microvascular
endothelial cells (Abdi et al., 1995; Del Vecchio et al., 1992;
Fischer et al., 2000; Gumkowski et al., 1987; Lotan et al.,
1994; Magee et al., 1994; Norgard-Sumnicht et al., 1995;
ctin binding. Green fluorescent stain in the lumen of the pulmonary artery
was observed with Helix pomatia in the peripheral lung (40�, fluorescent
pulmonary artery but is present in peripheral lung capillaries (arrows; 67�,
Fig. 2. Endothelial cell phenotypes in vitro can be discriminated by lectin binding. Increased staining intensity is demonstrated by a right shift in the
fluorescence intensity of sorted cells. Helix pomatia exhibits a more intense fluorescence in PAECs than in PMVECs. Glycine max and Griffonia simplicifolia
exhibit a more intense fluorescence in PMVECs when compared to PAECs. Inset pictures show control cells without lectin treatment (yellow), and cells treated
with a-GalNAc to block Helix pomatia, h-GalNAc to block Glycine max, and a-galactose to block Griffonia simplicifolia (blue). Sugars were added accordingto Sigma recommendations. In blocking studies, cells were incubated for 30 min with the blocking sugar before a 15-min incubation with the lectin.
J. King et al. / Microvascular Research 67 (2004) 139–151 143
Schnitzer et al., 1994). Nine different lectins were therefore
screened for binding to the rat pulmonary artery and micro-
vascular endothelial cell surface (Table 1). Of the nine lectins
Fig. 3. Lectin-induced agglutination discriminates endothelial cell phenotypes in vi
Griffonia simplicifolia selectively agglutinate PMVECs. Cells were trypsinized an
presence of lectins. Arrows indicate cell clumps. Pictures are representative of fiv
examined, six did not distinguish between macro- and
microvascular cell types, while three demonstrated a prefer-
ential binding pattern. In vivo, FITC-labeled Helix pomatia
tro. Helix pomatia selectively agglutinates PAECs whereas Glycine max and
d triturated in to single cell suspensions, then allowed to agglutinate in the
e separate experiments.
J. King et al. / Microvascular Research 67 (2004) 139–151144
staining was observed selectively in macrovascular endothe-
lia. The green fluorescent stain can be seen lining the lumen
of the pulmonary artery (Fig. 1), but is absent in the
peripheral lung. FITC-labeled Glycine max and Griffonia
Green fluorescent stain can be seen lining the lumen of
capillaries in peripheral lung but is absent in the pulmonary
artery (Fig. 1).
Fluorescence-activated cell sorting (FACS) was used to
confirm that cells isolated and cultured in vitro retained their
in vivo phenotype. H. pomatia staining was prominent in
macrovascular cells, while G. max and G. simplicifolia
staining was prominent in microvascular cells (Fig. 2).
Controls (e.g., cells without FITC-labeled lectin) did not
fluoresce and specific sugars in competitive binding studies
prevented staining (inset, Fig. 2). To confirm specificity of
these lectins for their respective endothelial cell type,
agglutination studies were performed in which lectin-treated
endothelial cells were trypsinized and dispersed. As is seen
in Fig. 3, H. pomatia selectively agglutinated PAECs in the
presence of trypsin whereas G. max and G. simplicifolia
selectively agglutinated PMVECs in the presence of trypsin.
These findings suggest that rat lung endothelial cells possess
similar surface sugars that can be distinguished from lectin
binding, both in vivo and in vitro.
Lung endothelial cell morphology
Few studies have documented the morphological char-
acteristics of rat lung PAECs and PMVECs, particularly
under identical culture conditions. We therefore examined
ultrastructural characteristics of these cell types in vitro
and in situ. Both PAECs and PMVECs in culture exhibited
round to oval nuclei, few mitochondria, rough endoplasmic
reticulum, junctions between cells, and surface projections
(Fig. 4). Vesicles consistent with caveolae (50–80 nm)
were present in both cell types, and vesicles consistent
with clathrin-coated pits were observed in PAECs (100–
150 nm). Groups of filaments were observed along the
basal membrane of both cell types, however, they were
more prominent in the PAECs (data not shown). By
transmission electron microscopy PAECs measured 8.8–
38.3 Am in diameter and 2.5–7.1 Am in maximum height.
By transmission electron microscopy PMVECs measured
9.0–37.5 Am in diameter and 2.9–7.1 Am in maximum
height.
Similar characteristics were observed in situ (Fig. 4).
Native PAECs exhibited numerous vesicles and occasional
Weibel-Palade bodies along with some mitochondria and
Fig. 4. Endothelial cell morphology in vitro. Ultrastructural assessment of
PAECs [panel A] and PMVECs [panel B] in culture-demonstrated typical
appearance of mitochondria (M), rough endoplasmic reticulum (RER), and
nucleus (N). Caveolae- or clathrin-coated pits (C) were observed.
Organelles were similarly observed in perfusion fixed lung pulmonary
artery [panel C] and capillary [panel D]. F denotes filter; L denotes lumen.
Fig. 5. RER-plasmalemma coupling distinguishes endothelial cell phenotypes. (A) Typical cytosolic calcium response to activation of store operated calcium entry
using thapsigargin (1 AM) demonstrates a lower response in PMVECs than in PAECs. (B) Transmission electron micrograph reveals that RER can be observed
immediately adjacent to the apical cell membrane in PAECs. (C) In addition, RER can be observed immediately adjacent to a vesicle consistent in size with a
clathrin-coated pit (e.g.,c100 nm). RER were also observed nearby caveolae-like structures (50–80 nm) in PAECs. RER do not similarly approach the apical
plasma membrane (D) or vesicles (E) in PMVECs. (F) Summary data reveal the RER-apical membrane distance is approximately 100 nm in PAECs (n = 44
specimens) and 250 nm in PMVECs (n = 50 specimens). *Denotes significantly different from PAEC.
J. King et al. / Microvascular Research 67 (2004) 139–151 145
Fig. 7. PMVECs possess a greater proliferative index that do PAECs. (A)
Serum-stimulated (10%) cell growth was observed over a 6-day period.
After a 2-day lag phase both PAECs and PMVECs exhibited log phase
growth, although PMVECs grew faster than did PAECs. (B) Serum-
restriction (0.1%) inhibited the growth of both cell types. (C) However,
serum stimulation during the lag phase was sufficient to initiate PMVEC
growth, even when cells were serum-deprived during the log phase. Such
treatment inhibited the growth of PAECs. *Denotes different from PAECs.
(D) Phase contrast images illustrate that PAECs and PMVECs grow at
different rates. Each cell type was seeded in 6-well plates at 105 cells/well,
and grown in the presence of 10% serum. PMVECs reached confluence on
day 4, whereas PAECs reached confluence on day 6.
Fig. 6. Cell junctions differ between PAECs (A) and PMVECs (B) in vitro.
PMVECs occasionally exhibited cytoplasmic processes (CP) not found in
PAECs. RER was found nearby cell borders, although always closer to the
membrane in PAECs than in PMVECs. ‘‘A’’ denotes apical cell side. ‘‘F’’
denotes filter. ‘‘N’’ denotes nucleus.
J. King et al. / Microvascular Research 67 (2004) 139–151146
RER. The RERwas as close as 58 nm from the apical surface.
Projections extend from the apical surface of the native
pulmonary artery endothelial cells, with more numerous
projections at the junctions between cells. Many of the
cellular projections are thin. In situ PAECs measured 11.0–
29.7 Am in diameter and 2.1–7.2 Am in maximum height.
Maximal cell height was in the area of the nucleus, where the
cell extended into the lumen. Nuclei were primarily oval
shaped. The peripheral parts of the cell were thinner, mea-
suring as little as 62.5 nm in height when perfused fixed
vessels were examined. In situ capillaries of the lung paren-
chyma contained two to three endothelial cells in a vascular
cross-section. The cell’s periphery was very thin, measuring
only 15.8 nm in some areas; organelles were absent in this
thin periphery. Numerous vesicles were present throughout
the cells, with the exception of the thinnest regions. Rare
profiles of RER were present, located primarily in the thicker
portions of the cell. RER was seen as close as 89 nm from the
apical surface. A few scattered mitochondria were present.
Apical (luminal) surfaces of the endothelial cells were often
flattened, although scattered projections were observed. The
capillary endothelial cells measured 10.3–26.9 Am in diam-
eter and 2.6–7.2 Am in maximum height. Because of the
Fig. 7 (continued).
J. King et al. / Microvascular Research 67 (2004) 139–151 147
small caliber of the capillaries, the nucleus caused a distinct
bulging into the lumen.
RER-membrane coupling: relevance to calcium signaling
We have previously observed that the thapsigargin-in-
duced store operated calcium entry response is lower in
PMVECs than it is in PAECs (Kelly et al., 1998; Moore
et al., 1998b; Stevens et al., 1997, 1999). Store-operated
calcium entry pathways are activated by depletion of cal-
cium in the endoplasmic reticulum (Putney, 1986). At
present, the signal(s) linking calcium store depletion to
activation of store-operated calcium entry is unclear, al-
though three separate models have been developed to
J. King et al. / Microvascular Research 67 (2004) 139–151148
address putative coupling mechanism(s) (for review, see
(Parekh and Penner, 1997; Putney, 2001; Putney and
Ribeiro, 2000)). Both conformational coupling and secretion
coupling models implicate a ‘‘physical’’ relationship be-
tween the endoplasmic reticulum and plasmalemma in
activation of store-operated calcium entry channels. We
therefore examined the morphological distribution of endo-
plasmic reticulum in PAECs and PMVECs, to evaluate
whether physical coupling between the endoplasmic reticu-
lum and plasma membrane could provide a plausible expla-
nation for the decreased store operated calcium entry
response in PMVECs. RER was observed close to the
apical, lateral (junctions between cells), and basal cell
membranes in both cell types, although the pattern of
distribution differed significantly in PAECs and PMVECs.
Measurements taken from transmission electron micro-
graphs revealed that the RER was as close as 8 nm to the
apical cell membrane in PAECs and as close as 87 nm to the
apical cell membrane in PMVECs (Fig. 5). Summary data of
membrane associated organelles indicated that, on average,
RER is nearly 2.5-fold closer to the plasmalemma in PAECs
than in PMVECs. Together, these findings suggest that the
proximity of RER to the plasmalemma may contribute to the
differential calcium signaling responses seen in these cell
types.
PMVECs form a more restrictive macromolecular barrier
than do PAECs (Chetham et al., 1999; Kelly et al., 1998).
Since macromolecular flux occurs at least partly through
intercellular junctions, we examined sites of cell–cell con-
tact in PMVECs and PAECs (Fig. 6). Electron dense
structures were observed at cell–cell borders. In both cell
types, RER could be resolved near cell–cell borders nearby
electron dense structures that contribute to cell–cell adhe-
sion, although the RER was closer to cell–cell borders in
PAECs than in PMVECs. RER was seen as close as 13 nm
from the cell membrane between PAECs and as close as 77
nm from the cell membrane between PMVECs. RER was
observed as close as 14 nm from the basal cell membrane in
PAECs and 76 nm from the basal cell membrane in
PMVECs. Store-operated calcium entry channels have not
presently been resolved within cell junctions. However,
these findings suggest that activation of store-operated
calcium entry may provide a calcium source nearby sites
of cell adhesion.
Endothelial cell growth
To further characterize lung endothelial cells, we evalu-
ated the growth rates of PAECs and PMVECs. Trypsinized
cells were triturated to single cell suspensions and re-seeded
in the presence of 10% serum. Both PAECs and PMVECs
exhibited a characteristic 2-day lag phase followed by log
phase growth (Fig. 7). PMVECs grew faster than PAECs.
Growth of both cell types was inhibited when cells were
incubated with 0.1% serum for 5 days. However, when cells
were incubated with 10% serum during lag phase growth
and then switched to 0.1% serum, PMVECs grew almost
normally whereas PAEC growth was significantly inhibited.
Together, these findings suggest that PMVECs possess a
unique pro-proliferative phenotype that is not present in
PAECs.
Discussion
Our present studies were founded on the hypothesis
that PAEC and PMVEC phenotypes are distinct, in part
due to their epigenetic origin. If this hypothesis is true,
then the cells should retain distinct functions in vitro
when their environments are similar. We approached this
hypothesis using structure–function analyses, evaluating
morphological characteristics of the cells along with
functional endpoints.
Lectins are protein agglutinins isolated from various
plant and animal sources that have proven useful for
distinguishing between cell phenotypes(Abdi et al., 1995;
Del Vecchio et al., 1992; Fischer et al., 2000; Gumkowski et
al., 1987; Lotan et al., 1994; Magee et al., 1994; Norgard-
Sumnicht et al., 1995; Schnitzer et al., 1994; Symon and
Wardlaw, 1996). Lectins interact with cell surface carbohy-
drates and therefore cell-specific lectin binding provides
important information regarding glycocalyx characteristics.
As in earlier studies, G. simplicifolia in particular selectively
interacted with PMVECs in vivo and in vitro. This lectin
exhibits affinity for a-galactose, indicating the PMVEC
glycocalyx is enriched with an a-galactose containing
carbohydrate. In contrast to prior studies, which primarily
observed H. pomatia binding to alveolar cells, we found H.
pomatia interacted with PAECs with preference over
PMVECs in vivo and in vitro. H. pomatia exhibits affinity
for a- and h-N-acetyl-galactosamine, indicating the rat
PAEC glycocalyx is enriched with an a- and h-N-acetyl-galactosamine carbohydrate. Since the glycocalyx contrib-
utes to cell–cell recognition, evidence for a differential
‘‘structure’’ of the PAEC and PMVEC glycocalyx suggests
these cells function distinctly in response to inflammatory
stimuli. Indeed, selectins bind homing receptors on the
endothelial cell glycocalyx (Symon and Wardlaw, 1996),
and bacteria interact with the endothelial cell surface
through adhesions that bind the glycocalyx (Hoppe et al.,
1997). The contribution of such distinct surface carbohy-
drate structures to site-specific inflammatory responses will
be important to resolve.
PAECs and PMVECs possessed significant morpholog-
ical distinctions, both in the intact circulation and in
culture. Association between the apical plasmalemma and
endoplasmic reticulum is closer in PAECs than in
PMVECs, implicating this membrane to organelle coupling
in calcium-mediated signal transduction. The principal
mode of calcium entry in endothelial cells is through so-
called store operated calcium entry pathways (Moore et al.,
1998b; Nilius and Droogmans, 2001), where calcium store
J. King et al. / Microvascular Research 67 (2004) 139–151 149
depletion in the endoplasmic reticulum activates calcium
entry across the plasmalemma. Neither the mechanism of
membrane channel activation nor the molecular identity of
membrane channels is well understood. However, certain