science pharmaceutical expo biotech jobs
Originally published online as doi:10.1189/jlb.0207080 on June 6, 2007

Published online before print June 6, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures
Right arrow All Versions of this Article:
jlb.0207080v1
82/3/542    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Su, W.-H.
Right arrow Articles by Jen, C. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Su, W.-H.
Right arrow Articles by Jen, C. J.
(Journal of Leukocyte Biology. 2007;82:542-550.)
© 2007 by Society for Leukocyte Biology

Polymorphonuclear leukocyte transverse migration induces rapid alterations in endothelial focal contacts

Wen-Hong Su, Hsiun-ing Chen and Chauying J. Jen1

Department of Physiology, College of Medicine, National Cheng Kung University, Tainan, Taiwan, Republic of China

1 Correspondence: Department of Physiology, College of Medicine, National Cheng Kung University, Tainan, 701 Taiwan, ROC. E-mail: jen{at}mail.ncku.edu.tw


arrow
ABSTRACT
 
Transmigrated polymorphonuclear leukocytes (PMNs) usually undergo subendothelial transverse migration before penetrating into inner tissue layers. Whether or how endothelial cells (ECs) respond to the PMN migrating underneath them is unknown. A tissue flow chamber was used to establish a fMLP gradient and to observe PMN transverse migration along with its associated endothelial responses in culture (on a collagen gel) or in vascular tissues. Our results indicated that transversely migrating PMNs were in direct contact with the basal side of ECs. Contrasting to focal adhesion kinase (FAK) or proteins with phosphorylated tyrosine, paxillin disappeared rapidly (<1 min) from endothelial focal contacts after encountering the leukocyte’s leading edge and soon rejoined them after the PMN had left. In addition, FAK moved away or became dephosphorylated when PMNs remained at the same subendothelial location for longer than 10 min, leaving actin filaments apparently unaltered. Unlike PMN transendothelial migration, PMN transverse migration did not induce any detectable endothelial calcium signaling. Taken together, our findings indicated that PMN transverse migration interrupted endothelial-matrix interactions and induced rapid alterations in endothelial focal contact composition.

Key Words: endothelium • PMN • calcium • FAK • paxillin


arrow
INTRODUCTION
 
Endothelium is strategically located at the interface between circulating blood components and the vascular tissue. It is desirable to characterize the interactions between leukocytes and endothelial cells (ECs), as the interactions play a crucial role in controlling the initiation steps of inflammation—leukocyte transvascular migration. Although we know much about the intricate details of the molecular and cellular interactions, which mediate the early interactions between leukocytes and ECs, less is known about the actual transmigration mechanisms regarding how leukocytes squeeze through the endothelium [1 , 2 ]. Moreover, only sparse information is presently known about the behavior of post-transmigration leukocytes [3 ]. The perivascular basement membrane provides a distinct barrier to emigrating leukocytes, and once leukocytes reach this structure, their continued emigration is delayed until the basement membrane is penetrated [1 , 3 ]. It is surprising that after the completion of transendothelial migration, polymorphonuclear leukocytes (PMNs) are rather mobile in the subendothelial tissue [4 ]. That is, PMNs apparently undergo lateral migration underneath the endothelium with a speed approximately 10 µm/min [4 ]—a process referred as "transverse migration" in this report.

Although the endothelial responses to PMN transverse migration underneath are unknown, one may try to draw some analogy from cell adhesion and migration. The mechanistic information about the latter is well-appreciated [5 ]. Cell adhesion or migration usually involves the formation and dissolution of adhesion complexes. Out of the many cell-matrix adhesion complexes, focal contacts are dynamic structures, which organize around activated integrin clusters and serve as the attachment points of actin stress fibers to the cell membrane (for a review, see refs. [5 , 6 ]). Focal contacts contain numerous protein components, and some have phosphorylated tyrosine residues. As many focal contacts are connected to the cell’s interior via actin fibers, they play important roles in cell mobility and cell-matrix interactions. Using GFP-tagged paxillin as a marker, small focal contacts in cultured fibroblasts have been reported to translocate with a mean rate of 19 µm/h [7 ]. However, largely as a result of technical considerations, most of the available data about cell adhesion or migration was obtained from experiments using fibroblast-like cells cultured on rigid, two-dimensional matrices (for a review, see ref. [5 ]). Few of the derived model pathways have been validated thoroughly using tissues or other cell types cultured on soft substratum.

It is well known that mobile PMNs require intracellular calcium signaling to release their rear adhesion complexes (uropods) and to recycle integrins to their leading edges [8 ]. In addition, one of the most intriguing features of EC-PMN interactions is the PMN transmigration-associated endothelial calcium signaling [9 ], which is a requirement for PMN transmigration toward the chemoattractant underneath the EC monolayer. Although ECs pretreated with a permeable Ca2+-chelating agent still maintain their adhesive properties for PMNs, PMN transendothelial migration is abolished. We have shown that the calcium signaling is restricted to ECs immediately adjacent to a transmigrating PMN [10 ]. This endothelial calcium signaling happens during an early stage of PMN transmigration and is essential for the PMN cell body to pass through. In fact, PMN transmigration across the endothelial monolayer, but not rolling or adhesion, is accompanied by endothelial calcium signaling [10 ].

Whether or how ECs sense and respond to PMN transverse migration underneath them is an intriguing question. To address this issue, we used a tissue flow chamber to establish a gradient of fMLP, which attracted PMN transmigration across an EC monolayer. Parameters of interest included the turnover of endothelial focal contact components and the possible endothelial calcium signaling. We focused on focal adhesion kinase (FAK) and paxillin; both are capable of recruiting intracellular signaling molecules and thus, play pivotal roles in mediating cell spreading and motility [11 ]. Furthermore, as the vast amount of current knowledge about cell-matrix interactions is derived from studies using cultured cells [5 ], this study examined how endothelial focal contacts were arranged in vascular tissue specimens and how they responded to the transversely migrating PMNs.


arrow
MATERIALS AND METHODS
 
Materials
Collagenase, EC growth supplement, fMLP, Ficoll-Hypaque, FITC, and heparin were purchased from Sigma Chemical Co. (St. Louis, MO, USA). FBS and medium M199 were from Gibco (Gaithersburg, MD, USA). Antibodies against laminin, paxillin, FAK, FAK with phosphorylated tyrosine397 (FAK-pY397), and proteins containing pY proteins were obtained commercially from DakoCytomation (Denmark; Z0097), BD Biosciences (San Jose, CA, USA; Clones 165 and 77), Biosource International (Camarillo, CA, USA; 44-624), and Cell Signaling Technology (Beverly, MA, USA; Clone 9411), respectively. Alexa Fluor-conjugated phalloidin (for labeling actin fibers), BAPTA-AM (for chelating intracellular calcium ions), Fura-2 AM (for detecting intracellular calcium ions), Hoechst 33342 (for labeling nuclei), and fluorescence-tagged, secondary antibodies (for labeling various focal adhesion components) were purchased from Molecular Probes (Eugene, OR, USA). Other reagents were purchased from Merck (Darmstadt, Germany).

Preparation of PMNs
PMNs were isolated from human peripheral blood by centrifugation on a discontinuous Ficoll-Hypaque gradient according to a method described previously [12 ]. Briefly, two layers of Ficoll-Hypaque with densities of 1.077 and 1.119 were formed carefully in a centrifuge tube, and the blood specimen (anticoagulated with 1/10 vol 106 mM sodium citrate) was layered onto the upper gradient. After centrifugation at 800 g for 50 min, the PMNs were separated from other blood cells as a distinctive band sandwiched between two Ficoll-Hypaque densities. Purified PMNs were resuspended in Krebs-Ringer HEPES (KRH) buffer (125 mM NaCl, 5 mM KCl, 1 mM KH2PO4, 1 mM MgSO4, 2 mM CaCl2, 25 mM HEPES, 6 mM glucose, 10% FBS, pH 7.4) and held on ice for no more than 3 h prior to use.

Studies about cultured ECs
Primary cultured ECs were isolated from human umbilical vein by collagenase (0.02%) digestion and were grown to confluence on a plastic dish in medium M199 containing 10% FBS, 10 U/ml heparin, and 25 µg/ml EC growth supplement. The endothelial monolayer was then trypsinized and resuspended in M199 and seeded on a 0.3% collagen gel-coated cover glass [10 ]. The first-passage cells reached confluence within 2 days and were subsequently used in 3 or 4 days. Before the experiment, the specimens were incubated with KRH buffer overnight and immersed in 1 µM fMLP and 5 µM Fura-2 AM (for measuring endothelial calcium signaling in certain experiments) for 1 h. The specimens were then incubated in Fura-2-free buffer for an additional 20 min to remove extracellular Fura-2 AM. The cover glass containing the fMLP-treated, Fura-2 fluorescence-labeled specimen was mounted on a modified flow chamber, which accommodated the thickness of collagen gel [10 ]. Then, the flow chamber was placed on a fluorescence microscope (Diaphot 300, Nikon, Japan, or Axioscope 2, Zeiss, Thornwood, NY, USA) or on a confocal microscope (TCS SP2, Leica, Heidelberg, Germany) and perfused continuously with fMLP-free buffer to establish a fMLP gradient across the endothelial monolayer. When a PMN suspension was perfused through the chamber, PMN transendothelial migration happened shortly after their contact with the ECs. The time of PMN transmigration across the EC monolayer along with the subsequent transverse migration steps were traced under phase-contrast optics.

Endothelial cytosolic calcium levels were derived from Fura-2 fluorescence intensity ratios between two excitation wavelengths [10 , 13 ]. Briefly, a video imaging system was set up on a microscope to automatically control the excitation filter wheel, which alternated between two filters of 340 nm and 380 nm. The fluorescence emission light at 510 nm was collected by a cooled charged-coupled device camera (CoolSnap fx, Roper Scientific, Duluth, GA, USA). The images were registered by the MetaFluor software (Universal Imaging, Detroit, MI, USA) at 10-s intervals with an exposure time of 0.25 s each to minimize photobleaching of the specimens. At the end of each experiment, the calcium concentration was calibrated by applying ionomycin (10 µM) in the presence of 5 mM EGTA, followed by 10 mM CaCl2. All signals were corrected by subtracting the autofluorescence, which was determined by exposing to 5 mM manganese to quench the cytosolic Fura-2 fluorescence. Finally, these fluorescent images were stored and processed off-line for ratio images of and converting to intracellular calcium ion concentration in individual ECs by a built-in formula. When necessary, we were able to monitor the kinetics of PMN transverse migration and endothelial calcium signaling by switching between phase-contrast and fluorescence optics. A correlation between each leukocyte-endothelial interaction step and the corresponding change of endothelial cytosolic calcium level, if any, could thus be established. All transmigration experiments were carried out at 37°C.

At the proper time-points during PMN transverse migration underneath the endothelial monolayer, the flow chamber assembly was removed rapidly from the microscope stage, immersed into ice water, and perfused with ice-cold fixative/permeabilizing agent (0.4% paraformaldehyde with 0.5% Triton X-100 for 3 min, followed by 4% paraformaldehyde overnight). As this process took less than 10 s, the morphology of fixed specimens was essentially the same as shown in the last phase-contrast image. Moreover, the possible postfixation protein degradation as a result of PMN protease activation [14 ] was avoided. Finally, fixed specimens were removed from the flow chamber and were processed further for immunostaining of various protein components in focal contacts. Briefly, the specimens were labeled with primary antibodies at 25°C for 30 min, washed away unbound primary antibodies, and labeled with Alexa Fluor-conjugated secondary antibodies for another 30 min. The blocking reagent, 20% human serum, was present in the whole labeling process. The nuclei were labeled at the end with Hoechst 33342 to identify the location of PMN and ECs.

To monitor the relative amounts of paxillin and FAK-pY397 in each endothelial focal contact, fixed specimens containing transversely migrating PMNs and ECs were double-labeled with antibodies against these two proteins, e.g., green fluorescence (Alex Fluor 488) for paxillin and red fluorescence (Alexa Fluor 546) for FAK-pY397. First, fluorescence images of two different colors were acquired for the same specimen, which was fixed immediately after taking the final phase-contrast image. Second, the PMN migration path along with its momentary cell contour were traced from phase images taken every 10 s and mapped on the fluorescence images. Third, the areas of interest were divided into five zones, i.e., naïve zone (areas never encountering any transversely migrating PMN), leading zone (areas at the front end of PMNs), central zone (areas covering the central part of PMNs), recovering zone (areas just behind the PMNs), and previously exposed zone (Areas 1–3 cell distance behind the PMNs). Fourth, the green and red fluorescence intensities were recorded for individual focal contacts by using the MetaMorph software (Universal Imaging). Finally, the green:red fluorescence ratio values for individual focal contacts in each zone were grouped together and normalized by those in the naïve zone (the averaged ratio value from 20 naïve focal contacts was set as unity). Although there could be several transversely migrating PMNs located in the same image taken under low magnification, the fluorescence intensities in different high-magnification images were not identical. Therefore, the normalization procedure was carried out for each set of green and red fluorescence images under high magnification.

The FITC-labeled collagen was used in certain experiments to verify whether transversely migrating PMNs were indeed sandwiched between ECs and the collagen gel. The FITC-labeled collagen gel was prepared by mixing 1% collagen monomer (150 µl in 25 mM acetic acid) with 500 mM FITC (50 µl in double-distilled water) and neutralized by adding 1 N NaOH (10 µl) and PBS (290 µl) to form a gel with a final collagen concentration of 0.3%. The gel was washed repeatedly with PBS to remove unbound FITC, and it was finally immersed in M199. ECs grew equally well on unlabeled collagen gel or on FITC-labeled collagen gel.

Studies about vascular tissue segments
Freshly isolated vascular tissues (human umbilical cords, rabbit carotid arteries, rat aortas) were flushed with KRH buffer and cut into 1- to 2-cm vascular segments. The vascular segments were trimmed carefully to remove excess adventitial tissue, longitudinally opened, and mounted temporarily on a silicon sheet. The handling process was performed without tissue dehydration or endothelial damage. Finally, after being immersed in the fMLP solution for 1 h, the specimens were mounted on a tissue flow chamber, similar to that used previously [4 , 13 ]. Perfusion of the fMLP-free buffer through this tissue flow chamber also created a fMLP gradient, which allowed the transmigration of PMNs through the vascular endothelium [4 ]. At the end of transmigration experiments, specimens were fixed and processed further for observation of transversely migrating PMNs buried in the vascular tissue. Immunostaining of focal contact proteins was carried out, and the resulting fluorescence images were analyzed to verify whether certain protein components were missing in the neighborhood of transversely migrating PMNs.


arrow
RESULTS
 
Endothelial focal contacts on collagen gel and in various vascular tissues
Studies using cells cultured on protein-coated glass have generated vast amounts of information about focal contacts. However, largely, as a result of insufficient resolution under a light microscope, relatively little focal contact information is available currently about cells on soft substratum or in tissues. We have overcome this difficulty by simultaneously applying fixing and permeabilizing reagents to the specimens of interest. To visualize the distribution pattern of focal contacts, we applied fluorescence immunostaining methods to label several protein markers, such as FAK, FAK-pY397, paxillin, and pY proteins. The focal contact patterns in cultured ECs were different from those in a variety of vascular tissues, including rabbit aorta, rabbit carotid artery, rat aorta, and human umbilical vein (Fig. 1 ). Individual focal contacts of ECs grown on collagen gel were clearly distinguishable: typically, rod-shaped structures scattering across the entire cell-covered area (Fig. 1A and 1B) . Although vascular tissues were not nearly as flat as cultured cells, and thus, potions of their images were often off-focused, their focal contact proteins were located mainly at the endothelial periphery and formed a continuous line or belt along the cell contour in all vessel types examined (Fig. 1C 1D 1E 1F) . They also showed additional, web-like distribution or evenly dispersed, dot-like distribution in different vascular specimens.


Figure 1
View larger version (63K):
[in this window]
[in a new window]

  		Figure 1. Fluorescent images showing endothelial focal contacts on collagen gel or in various vascular tissues. (A) Paxillin in HUVECs cultured on collagen gel; (B) FAK-pY397 in HUVECs cultured on collagen gel; (C) endothelial paxillin in rabbit aorta; (D) endothelial pY proteins in rabbit carotid artery; (E) endothelial pY proteins in rat aorta; (F) endothelial FAK-pY397 in human umbilical vein. Images were taken under wide-field optics. Note that regardless of which protein was used as an indicator for focal contacts, the focal contact patterns in vascular tissues (C–F) were different from those in cultured ECs (A and B).

PMN transverse migration underneath ECs
Confocal microscopy was applied to verify PMN transverse migration underneath cultured ECs (Fig. 2 ). Fibers in the collagen gel were clearly visible if collagen molecules had been prelabeled with FITC before gel formation. Transversely migrating PMNs with small and multilobed nuclei were sandwiched between collagen gel and ECs (with large nuclei and organized actin fibers). The collagen gel was indented where transversely migrating PMNs were located. Moreover, collagen fibers were undetectable between ECs and transversely migrating PMNs. When a PMN eventually penetrated into the collagen gel, it was surrounded completely by collagen fibers without having endothelial nuclei or actin filaments in its immediate proximity. As cultured cells often synthesize and secrete laminin to their lateral side, whether the transversely migrating PMN was sandwiched between ECs and the matrix materials, or it was inserted between laminin and collagen is another intriguing question. Experiments using specimens stained for laminin instead of collagen supported the former viewpoint; i.e., the transversely migrating PMNs were sandwiched between ECs and laminin (Supplemental Fig. S1). Similarly, transversely migrating PMNs in vascular tissues could be located by labeling cell nuclei and focal contacts in a piece of a longitudinally opened, vascular segment (Fig. 3 ).


Figure 2
View larger version (171K):
[in this window]
[in a new window]

  		Figure 2. Confocal images showing transversely migrating PMNs underneath an EC monolayer in culture. Optical sections were taken along the z direction from the apical surface of ECs (0 µm) toward the bottom of the collagen gel (A–H). Three PMNs with small and multilobed nuclei are marked; two are just underneath the endothelial monolayer (arrowheads), and one already penetrated into the collagen gel (arrow). Two vertical scans across a transversely migrating PMN are shown (I and J). (I) The vertical scan along the dotted line; (J) along the dashed line. F-actin is shown in red, collagen in green, and nuclei in blue. Although the PMN nucleus appeared to be located at the same level of the neighboring endothelial nucleus, the PMN was actually located underneath the endothelial actin network. The transversely migrating PMN was in direct contact with the basal side of the EC monolayer, as the collagen fibers were absent in regions between them.


Figure 3
View larger version (62K):
[in this window]
[in a new window]

  		Figure 3. Confocal images showing transversely migrating PMNs underneath the endothelium in a rabbit carotid artery. Optical sections were taken along the z direction from the apical surface of endothelium (0 µm) toward the smooth muscle layer of the vessel (A–H). pY proteins are shown in red and nuclei in blue. Three PMNs with small and multilobed nuclei are marked in the nuclear-stained image (I). The nuclei of two PMNs were underneath endothelial pY proteins (arrowheads), and one PMN nucleus was underneath the neighboring EC nucleus (arrow). Although the luminal surface of a vascular tissue was uneven at the microscopic level, smooth muscle cells with nuclei perpendicular to EC nuclei were absent in these images. As the optical sections only focused in the intimal layer, these PMNs were sandwiched between the endothelium and the basement membrane.

Partial reorganization of endothelial focal contacts during PMN transverse migration
In our hands, PMNs traverse underneath endothelium at a high speed (13.4+1.0 µm/min, n=37), comparable with their migrational speed on the apical side of ECs (15.4+0.6 µm/min, n=53). Presumably, a transversely migrating PMN would use the subendothelial path with lowest resistance. PMN would thus squeeze between endothelial focal adhesions or inevitably encounter some of them. As we seldom observed snake-shaped PMNs, they should be able to open the endothelial-matrix adhesions on their way. It is surprising that our overall impression was that endothelial focal contacts were grossly unchanged during PMN transverse migration underneath. However, when multiple-stained specimens were examined closely under high magnification, where a submerged PMN was located, the focal contact-associated paxillin stain largely diminished from FAK-associated spots (Fig. 4 ). Focal contacts in the newly encountered area (PMN leading zone and central zone) showed a reduced paxillin:FAK-pY397 ratio as compared with naïve regions (color changing from yellow to orange to red as the green fluorescence of paxillin stain weakened). Cumulative results from 11 PMN transverse migration events in six separate experiments are also shown in Figure 4 . The PMN transverse migration induced dynamic and reversible changes in focal contact components. We used the following approach to estimate the dissociation time of focal contact components. The exposure time of individual focal contacts to a transversely migrating PMN could be calculated from their relative locations to the PMN in the final snap-shot and the migrating speed of the PMN, parameters obtainable from the PMN tracings acquired under the phase-contrast optics. As the PMN transverse migration speed often reached 10–20 µm/min, paxillin disappeared and reappeared again (relative to FAK-pY397) in the time course of less than 1 min when encountering or recovering from a PMN exposure underneath. As a control, specimens double-labeled with FAK-pY397 and pY proteins did not show differential color changes (picture not shown).


Figure 4
View larger version (60K):
[in this window]
[in a new window]

  		Figure 4. Rapid disappearance of paxillin from endothelial focal contacts during PMN transverse migration. Phase-contrast images in the upper panels show the direction and traces of a transversely migrating PMN underneath the cultured EC monolayer; (A) a wide-field, double-stained fluorescence image showing endothelial focal contact distribution at the last moment. The specimen was fixed within 10 s after taking the last phase-contrast image and subsequently processed for staining of focal contact components and nuclei. FAK-pY397 is shown in red, paxillin in green, and nuclei in blue. The boundaries of this transversely migrating PMN are marked (30 s in dashed line and 40 s in solid line). (B) The summarized results, indicating that the paxillin disappearance was associated with PMN transverse migration. The values of the paxillin:FAK-pY397 fluorescence ratio in focal contacts were obtained and classified according to their location: naïve zone (areas never encountering any transversely migrating PMN), leading zone (areas at the front end of PMNs), central zone (areas covering the central part of PMNs), recovering zone (areas just behind the PMNs), and previously exposed zone (Areas 1–3 cell distance behind the PMNs). The total numbers of focal contacts analyzed in each zone are marked in parentheses. Results are presented as mean ± SEM. One-way ANOVA indicated significant difference among groups (P<0.0001). Tukey’s multiple comparison test indicated significant differences between two groups (P<0.001) except for naïve zone versus previously exposed zone (P>0.05) and recovering zone versus previously exposed zone (P>0.05). Note the fluorescence ratio values in PMN-affected zones were normalized against those in the corresponding naïve zone in each experiment. As the fluorescence color images were distorted partially by the nucleus stain, data from focal contacts, which overlapped with nuclei, were excluded.

It was intriguing to investigate whether FAK-pY397 would be altered when PMN stayed beneath it for longer periods of time. To achieve this goal, the PMN adhesion/transmigration experiment was performed at 37°C at the beginning. Then, the temperature of the flow chamber was reduced to 25°C to retard the subsequent PMN transverse migration. The transmigrated PMNs usually remained at almost the same location for 20–30 min. The specimen was fixed after taking the last phase-contrast image and subsequently processed for staining of focal contact components and nuclei. As a final step, actin filaments were labeled to examine whether stress fibers became disorganized. With prolonged presence of mobility-retarded PMN underneath, FAK-pY397 in endothelial focal contacts moved away or became dephosphorylated in 10–20 min (Fig. 5 ). The pattern of actin filaments appeared normal under these circumstances, indicating that other focal contact components might still be present. If the chamber temperature were raised to 37°C again, some PMNs would resume their transverse migration process after various lag periods. FAK-pY397 and paxillin stains reappeared eventually (figure not shown).


Figure 5
View larger version (81K):
[in this window]
[in a new window]

  		Figure 5. Fluorescence images showing delayed disappearance of FAK from endothelial focal contacts during PMN transverse migration. The PMN adhesion/transmigration experiment was performed regularly at 37°C but subsequently reduced to 25°C to retard the PMN transverse migration. The specimen was fixed after taking the last phase-contrast image and stained further for focal contact components and nuclei. Actin filaments were also labeled to examine whether stress fibers became disorganized. Phase-contrast images in the upper panels show the traces of a mobility-retarded PMN, and fluorescence images in A and B show the distribution of endothelial focal contacts and actin filaments. FAK-pY397 is shown in red, paxillin in green, and nuclei in blue (A). The boundaries of two PMNs located subendothelially are marked. Please note that yellow dots were present only in regions away from these two PMNs, indicating the absence of paxillin or FAK in the PMN-affected region. (B) The actin filaments are not grossly altered in the region without these two adhesion components. Actin filaments are in red and nuclei in blue.

Similar experiments were performed using vascular tissues instead. As a result of the poor resolution under phase-contrast optics and high autofluorescence in tissues, it was difficult to trace the dynamic movements of PMN and hard to carry out multiple stains. Nevertheless, when we examined the proximity of a transversely migrating PMN in rabbit carotid artery, paxillin was absent around the endothelial focal contacts, as indicated by the interrupted line of paxillin staining (Fig. 6C ). In contrast, the staining pattern of pY proteins remained relatively unaltered by PMN (Fig. 6A) . Eleven out of 13 observations at the PMN transverse-migrating site showed continuous contour of pY proteins along the cell boundary (as in Fig. 6A and 6B ). We only found discontinuity of pY protein labeling in two out of 13 cases. In contrast, all six observations near the transversely migrating PMNs showed discontinuity of paxillin staining (such as in Fig. 6C and 6D ).


Figure 6
View larger version (162K):
[in this window]
[in a new window]

  		Figure 6. Fluorescence images showing disappearance of paxillin from endothelial focal contacts during PMN transverse migration in rabbit carotid artery. (A and B) Wide-field fluorescence images from the same location showing the distribution of pY proteins and nuclei, respectively. (C and D) The distribution of paxillin and nuclei, respectively. Please note that areas surrounding PMN nuclei (arrows) show normal pY protein stain (A and B) and interrupted paxillin stain (C and D).

The absence of endothelial calcium signaling during PMN transverse migration underneath cultured ECs
As PMN transverse migration led to alterations in existing focal contacts at the endothelial basal side, endothelial intracellular calcium signaling (as an index for cell activation) was expected to occur during this process. Despite our rigorous search, this parameter was found to be unaltered when PMNs underwent transverse migration underneath cultured ECs (Fig. 7 ). Only one out of more than 20 PMN transverse migration events (affecting more than 70 ECs in total) was accompanied by a single endothelial calcium signaling event. Even in that particular case, the endothelial calcium signaling could be just coincidence, as ECs in culture occasionally showed sporadic calcium signaling without PMN exposure.


Figure 7
View larger version (44K):
[in this window]
[in a new window]

  		Figure 7. The absence of endothelial calcium signaling during rapid PMN transverse migration. Four different ECs had encountered a transversely migrating PMN (upper panels). The PMN transverse migration path is marked in the right-most image (430 s). As in a neighboring, naïve EC, which never encountered the PMN, the intracellular calcium levels remained unchanged in four PMN-contacted ECs. Histamine exposure induced immediate calcium signaling in all ECs examined, indicating their capability of responding to a conventional stimulus. C, Control.


arrow
DISCUSSION
 
PMN transverse migration underneath the cultured EC monolayer or underneath the vascular endothelium provides a unique means for investigating the focal contact dynamics, as it is rapid, localized, and physiological. Being initiated by PMN-induced, local release of the physical contact between the receptors located at the EC basal membrane and the extracellular matrix (ECM), this process was accompanied by a rapid loss (less than 1 min) of certain focal contact components, such as paxillin. However, ECs were disturbed minimally by the transversely migrating PMN, as indicated by the continuous presence of FAK-pY397 for at least 10 min, seemingly unaltered actin stress fibers for even longer time periods, and the absence of EC calcium signaling. In contrast, the process of PMN squeezing through interendothelial junctions (transendothelial migration) appears to be a much stronger stimulus. It causes not only local rearrangements of various junction proteins (e.g., PECAM-1 and vascular endothelial-cadherin) but also the cell-wide EC calcium signaling [4 , 9 , 10 , 15 ].

How would the relative movements of endothelial focal contact-associated paxillin and FAK affect PMN transverse migration? Paxillin and FAK are tyrosine-phosphorylated in early cell-matrix adhesions and located in close proximity as measured by microscopy-based fluorescence resonance energy transfer [16 ]. They are believed to play crucial roles in cell-matrix signaling pathways, as both contain a number of motifs, which serve as the docking sites for various signaling molecules, such as cytoskeleton proteins, tyrosine kinases, serine/threonine kinases, and GTPase-activating proteins. It is intriguing that although these two proteins are known to promote cell adhesion or migration in many cases, they inhibit cell migration under some conditions [17 ]. Contrasting the fact that FAK alone is capable of recruiting some proteins (such as p130cas) to activate the small GTP-binding protein Rac1, the FAK/paxillin complex may inactivate Rac1, directly or indirectly [18 ]. The conditions dictating positive or negative regulation and the mechanism of action in these two scenarios are unknown presently. Nevertheless, the dissociation of paxillin from FAK may lead to elevated Rac activity, which is needed for the interactions between cytoskeleton and membrane, such as the formation of local membrane ruffles. Thus, a transient paxillin disappearance from endothelial focal contacts could potentially alter local membrane properties in a way facilitating the rapid movements of a PMN underneath. Whether the prevention of paxillin relocation would hinder PMN transverse migration remains to be investigated.

It has been reported that substrata made of different proteins or rigidities can influence the growth of cells, possibly as a result of the various shapes and compositions of the focal contacts [19 , 20 ]. According to our observation, large focal contacts were located mostly at the periphery of ECs. Besides, our impression was that even relatively large focal contacts showed dynamic changes when encountering a transversely migrating PMN, indicating that they might be structurally alike. Whether these larger focal contacts were functionally different from small ones or not remains to be investigated. The current classes of adhesion complexes (mostly observed in cells cultured on solid substrata) include focal adhesions, fibrillar adhesion, focal complexes, and podosomes [5 ]. Focal adhesions are the large, oval-shaped structures composing numerous proteins, including paxillin, vinculin, integrin, and tyrosine-phosphorylated proteins, whereas focal complexes are small adhesions near the leading edge of a protrusion, which are induced by Rac activation. Thus, focal contacts in conventionally cultured cells could be different, not only in their shape and size but also in their composition. It is interesting that it is the small adhesions at the cell front that drive migration, whereas the larger, more organized adhesions tend to inhibit migration [21 ]. Although we only monitored paxillin and FAK in this study, other adhesion components could undergo similar alterations by themselves or as a group. Little information is available at this stage about which components are most important or most labile under these conditions. Additional studies are needed to ascertain the physiological significance of this kind of adhesion alterations.

How would the opening and closing of existing endothelial focal contacts in this study be compared with the better-understood cell spreading and migration? The cell-surface contact and recognition steps during cell spreading occur on a subsecond time scale, and full cell spreading takes tens of minutes to hours [22 ]. Once integrins recognized the ECM components on the surface, numerous proteins are to be recruited and organized to form new focal contacts. Results from the current study showed clearly that the closing of existing focal contacts only caused minor disturbances, and it happened much faster than the spreading-associated focal contact formation process. As for cell migration, the uropod retraction step is essentially a focal contact dissolution process, which requires the activation of a calcium-dependent, phosphatase calcineurin [8 ]. In contrast, PMN transverse migration only induced minimal and transient alterations of endothelial focal contacts without stimulating EC calcium signaling. Taken together, the above-mentioned discrepancy could be regarded as two different modes of stimulation; i.e., although spreading and migration in general seemed to be "active", cellular processes, which involve complete reconstruction of focal contacts, PMN transverse migration only "passively" induced minor focal contact changes in locally affected ECs.

Intuitively, EC migration in response to shear stress may be regarded as another passive phenomenon, which happens under physiological conditions. Cultured ECs sense the shear stress applied on their apical surface and transmit the signal via cytoskeleton throughout the cell to the cell-matrix adhesions—promoting the formation of focal contacts in the flow direction and the disassembly of them at the rear [23 ]. Indeed, FAK and paxillin are involved in the shear-induced reorganization of focal contacts [23 , 24 ]. As the shear force is usually applied to the entire cell, it is not surprising that ECs in culture or endothelium in tissue respond to shear forces directly or indirectly with calcium signaling [25 , 26 ].

It would be interesting to compare the cell migration-associated forces and the adhesive forces between cell focal contacts and the substratum. Results from this study indicated that the adhesive force between endothelial focal contacts and matrix was apparently too small to retard any PMN transverse migration, thus providing a low-resistant pathway for a cell to migrate underneath cultured ECs. As our previous results showed that the speed of PMN transverse migration reached 11 µm/min in a piece of human umbilical vein tissue [4 ], a similar, low-resistance pathway was likely to be present in vascular tissues as well. A transversely migrating cell may take advantage of this property to find the proper spot for penetrating into the deeper tissue layers. Moreover, judging from the relatively short time scale needed in this event, the protease action does not seem to be involved in opening and reforming endothelial focal contacts. As most endothelial focal contacts, regardless of their sizes, showed similar, dynamic changes (Fig. 4 and Supplemental Fig. S2), a transversely migrating PMN might unlock adhesive sites such as opening a zipper tooth-by-tooth. It remains to be investigated whether transmigrating PMNs transiently form new adhesive sites with ECs above them to replace the original adhesive interactions between ECs and the matrix materials. Currently, relatively little is known about contact-induced cell-cell interactions in a three-dimensional tissue. How smooth muscle cells in the vascular medial layer respond to PMNs, which eventually broke through the basement membrane, deserves future research. As inflammation often takes hours or days to complete its course, numerous factors are expected to mediate the leukocyte-tissue interactions at later stages of it.

Finally, we believe that this is the first study reporting the organization of focal contacts in vascular tissues. The distribution patterns of endothelial focal contacts in situ were clearly distinct from those in cultured ECs. However, even with their well-organized arrangement, these focal contacts offered little resistance to the mobility of transversely migrating PMNs. This low-resistance subendothelial pathway appeared to be necessary for the rapid passage of PMN through the endothelial monolayer, which usually completes in less than 1 min [4 , 10 ]. Otherwise, if the transverse migration process is hindered by robust focal contacts, PMNs will be forced to penetrate into the medial layer of vascular tissue at the spot of transendothelial migration. As this basement membrane-piercing process probably takes longer time periods or may even involve the action of PMN-derived proteases, the interendothelial junctions will remain open during this period and lead to the breakdown of endothelial-barrier function.

In summary, our results supported the notion that a transversely migrating PMN was capable of interrupting endothelial-matrix interactions and inducing rapid alterations in endothelial focal contact composition.


arrow
ACKNOWLEDGEMENTS
 
This study was supported by grants from the National Science Council, Taiwan, ROC (NSC 94-2320-B-006-056 and NSC 95-2320-B-006-18).

Received February 1, 2007; revised April 19, 2007; accepted April 24, 2007.


arrow
REFERENCES
 
    1
  1. Muller, W. A. (2003) Leukocyte-endothelial-cell interactions in leukocyte transmigration and the inflammatory response Trends Immunol. 24,327-334[Medline]
    2. 2
  2. Van Buul, J. D., Hordijk, P. L. (2004) Signaling in leukocyte transendothelial migration Arterioscler. Thromb. Vasc. Biol. 24,824-833[Abstract/Free Full Text]
    3. 3
  3. Dangerfield, J., Larbi, K. Y., Huang, M. T., Dewar, A., Nourshargh, S. (2002) PECAM-1 (CD31) homophilic interaction up-regulates {alpha}6ß1 on transmigrated neutrophils in vivo and plays a functional role in the ability of {alpha}6 integrins to mediate leukocyte migration through the perivascular basement membrane J. Exp. Med. 196,1201-1211[Abstract/Free Full Text]
    4. 4
  4. Su, W. H., Chen, H. I., Jen, C. J. (2002) Differential movements of VE-cadherin and PECAM-1 during transmigration of polymorphonuclear leukocytes through human umbilical vein endothelium Blood 100,3597-3603[Abstract/Free Full Text]
    5. 5
  5. Wiesner, S., Legate, K. R., Fassler, R. (2005) Integrin-actin interactions Cell. Mol. Life Sci. 62,1081-1099[CrossRef][Medline]
    6. 6
  6. Zamir, E., Geiger, B. (2001) Molecular complexity and dynamics of cell-matrix adhesions J. Cell Sci. 114,3583-3590[Medline]
    7. 7
  7. Zamir, E., Katz, M., Posen, Y., Erez, N., Yamada, K. M., Katz, B. Z., Lin, S., Lin, D. C., Bershadsky, A., Kam, B., Geiger, Z. (2000) Dynamics and segregation of cell-matrix adhesions in cultured fibroblasts Nat. Cell Biol. 2,191-196[CrossRef][Medline]
    8. 8
  8. Lawson, M. A., Maxfield, F. R. (1995) Ca2+- and calcineurin-dependent recycling of an integrin to the front of migrating neutrophils Nature 377,75-79[CrossRef][Medline]
    9. 9
  9. Huang, A. J., Manning, J. E., Bandak, T. M., Ratau, M. C., Hanser, K. R., Silverstein, S. C. (1993) Endothelial cell cytosolic free calcium regulates neutrophil migration across monolayers of endothelial cells J. Cell Biol. 120,1371-1380[Abstract/Free Full Text]
    10. 10
  10. Su, W. H., Chen, H. I., Huang, J. P., Jen, C. J. (2000) Endothelial [Ca2+]i signaling during transmigration of polymorphonuclear leukocytes Blood 96,3816-3822[Abstract/Free Full Text]
    11. 11
  11. Schaller, M. D. (2001) Paxillin: a focal adhesion-associated adaptor protein Oncogene 20,6459-6472[CrossRef][Medline]
    12. 12
  12. English, D., Anderson, B. R. (1974) Single step separation of red blood cells, granulocytes and mononuclear leukocytes on discontinuous density gradients of Ficoll-Hypaque J. Immunol. Methods 5,249-252[CrossRef][Medline]
    13. 13
  13. Huang, T. Y., Chu, T. F., Chen, H. I., Jen, C. J. (2000) Heterogeneity of [Ca2+]i signaling in intact rat aortic endothelium FASEB J. 14,797-804[Abstract/Free Full Text]
    14. 14
  14. Moll, T., Dejana, E., Vestweber, D. (1998) In vitro degradation of endothelial catenins by a neutrophil protease J. Cell Biol. 140,403-407[Abstract/Free Full Text]
    15. 15
  15. Shaw, S. K., Bamba, P. S., Perkins, B. N., Luscinskas, F. W. (2001) Real-time imaging of vascular endothelial-cadherin during leukocyte transmigration across endothelium J. Immunol. 167,2323-2330[Abstract/Free Full Text]
    16. 16
  16. Ballestrem, C., Erez, N., Kirchner, J., Kam, Z., Bershadsky, A., Geiger, B. (2006) Molecular mapping of tyrosine-phosphorylated proteins in focal adhesions using fluorescence resonance energy transfer J. Cell Sci. 119,866-875[Abstract/Free Full Text]
    17. 17
  17. Yano, H., Mazaki, Y., Kurokawa, K., Hanks, S. K., Matsuda, M., Sabe, H. (2004) Roles played by a subset of integrin signaling molecules in cadherin-based cell-cell adhesion J. Cell Biol. 166,283-295[Abstract/Free Full Text]
    18. 18
  18. Schaller, M. D. (2004) FAK and paxillin: regulators of N-cadherin adhesion and inhibitors of cell migration? J. Cell Biol. 166,157-159[Abstract/Free Full Text]
    19. 19
  19. Katz, B. Z., Zamir, E., Bershadsky, A., Kam, Z., Yamada, K. M., Geiger, B. (2000) Physical state of the extracellular matrix regulates the structure and molecular composition of cell-matrix adhesions Mol. Biol. Cell 11,1047-1060[Abstract/Free Full Text]
    20. 20
  20. Georges, P. C., Miller, W. J., Meaney, D. F., Sawyer, E., Janmey, P. A. (2006) Matrices with compliance comparable to that of brain tissue select neuronal over glial growth in mixed cortical cultures Biophys. J. 90,3012-3018[CrossRef][Medline]
    21. 21
  21. Beningo, K. A., Dembo, M., Kaverina, I., Small, J. V., Wang, Y. L. (2001) Nascent focal adhesions are responsible for the generation of strong propulsive forces in migrating fibroblasts J. Cell Biol. 153,881-888[Abstract/Free Full Text]
    22. 22
  22. Cohen, M., Joester, D., Geiger, B., Addadi, L. (2004) Spatial and temporal sequence of events in cell adhesion: from molecular recognition to focal adhesion assembly ChemBioChem 5,1393-1399[CrossRef][Medline]
    23. 23
  23. Li, S., Butler, P., Wang, Y., Hu, Y., Han, D. C., Usami, S., Guan, J. L., Chien, S. (2002) The role of the dynamics of focal adhesion kinase in the mechanotaxis of endothelial cells Proc. Natl. Acad. Sci. USA 99,3546-3551[Abstract/Free Full Text]
    24. 24
  24. Zaidel-Bar, R., Kam, Z., Geiger, B. (2005) Polarized downregulation of the paxillin-p130CAS-Rac1 pathway induced by shear flow J. Cell Sci. 118,3997-4007[Abstract/Free Full Text]
    25. 25
  25. James, N. L., Harrison, D. G., Nerem, R. M. (1995) Effects of shear on endothelial cell calcium in the presence and absence of ATP FASEB J. 9,968-973[Abstract]
    26. 26
  26. Jen, C. J., Jhiang, S. J., Chen, H. I. (2000) Effects of flow on vascular endothelial intracellular calcium signaling of rat aortas ex vivo J. Appl. Physiol. 89,1657-1662[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures
Right arrow All Versions of this Article:
jlb.0207080v1
82/3/542    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Su, W.-H.
Right arrow Articles by Jen, C. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Su, W.-H.
Right arrow Articles by Jen, C. J.