Technical Advance: New in vitro method for assaying the migration of primary B cells using an endothelial monolayer as substrate

The use of activated endothelial cell monolayers as a substrate provides a novel method for analysis of primary B cell migration in vitro.

Abstract

Migration of B cells supports their development and recruitment into functional niches. Therefore, defining factors that control B cell migration will lead to a better understanding of adaptive immunity. In vitro cell migration assays with B cells have been limited by poor adhesion of cells to glass coated with adhesion molecules. We have developed a technique using monolayers of endothelial cells as the substrate for B cell migration and used this technique to establish a robust in vitro assay for B cell migration. We use TNF-α to up-regulate surface expression of the adhesion molecule VCAM-1 on endothelial cells. The ligand VLA-4 is expressed on B cells, allowing them to interact with the endothelial monolayer and migrate on its surface. We tested our new method by examining the role of L-plastin (LPL), an F-actin-bundling protein, in B cell migration. LPL-deficient (LPL^−/−) B cells displayed decreased speed and increased arrest coefficient compared with wild-type (WT) B cells, following chemokine stimulation. However, the confinement ratios for WT and LPL^−/− B cells were similar. Thus, we demonstrate how the use of endothelial monolayers as a substrate will support future interrogation of molecular pathways essential to B cell migration.

Introduction

The migration of B cells is important for their development, their ability to perform immunosurveillance, and their interactions with other immune cells [1, 2]. Appropriate homing of B cells to their proper niches requires their migration in response to chemoattractants [3–6] and the engagement of B cell adhesion molecules with other cells and extracellular matrix [5, 7–10]. A variety of methods have been used to identify and characterize the molecules that control cell migration and related processes essential to adaptive immunity. Leukocyte migration has been studied with multiphoton intravital imaging of exposed live tissue [11], live-cell microscopy of ex vivo tissue or in vitro cell culture [12, 13], in vitro Transwell migration assays [14], and in vitro adhesion assays [15].

In vitro assays of leukocyte migration often use substrate-coated glass coverslips, but the literature using this technique has largely focused on T cells, neutrophils, and macrophages [6, 16–23]. Unfortunately, primary B cells do not adhere well to substrate-coated coverglass, limiting our ability to perform and interpret similar in vitro assays. Our review of the published literature found only 2 other studies using this approach with primary B cells [24, 25].

Here, we describe a novel protocol for assessing B cell migration in vitro using an activated monolayer of endothelial cells as a substrate, which is modified from a protocol used to study transendothelial migration by lymphocytes [26]. We established conditions that supported effective engagement of murine B cells with human endothelial monolayers. With the use of this protocol, we demonstrated directed B cell migration in vitro. We then used the protocol to observe and quantify the requirement for LPL during B cell migration. LPL is a leukocyte-specific, actin-binding protein that connects a pair of actin filaments through its C-terminus, thus creating bundles of actin filaments. LPL may also participate in intracellular signaling via N-terminal serine and threonine phosphorylation sites [27, 28]. In prior work, B cells isolated from LPL^−/− mice displayed significantly reduced CXCL12- and CXCL13-mediated migration across a Transwell membrane [29]. With the use of our new B cell migration assay, we now show that LPL^−/− B cells exhibit decreased speed and increased arrest coefficient without changes in confinement ratio.

We propose that this B cell migration assay offers a new tool for delineating the function of actin-binding and signaling proteins involved in primary B cell trafficking and provides an easy method to characterize primary B cell movement in a more physiologic setting.

MATERIALS AND METHODS

Mice

LPL^−/− mice have been described [30, 31]. C57BL/6 WT and LPL^−/− mice were used at 8–14 wk old and cohoused in specific pathogen-free barrier animal facilities at WUSM. Animal experiments were approved by the WUSM Institutional Animal Care and Use Committee.

Antibodies

Fluorochrome-conjugated antibodies (clone ID) against the following antigens were purchased from BioLegend (San Diego, CA, USA): B220-APC-Cy7 (RA3-6B2), CD21/35-PE (7E9), CD23-APC (B3B4), and CD45.2-PE-Cy7 (104). hVCAM-1-hF_c (R&D Systems, Minneapolis, MN, USA) and mVCAM-1-hF_c (R&D Systems) were also used as a primary probe with goat anti-human IgG pAb-PE (eBioscience, San Diego, CA, USA). Mouse α-hVCAM-1 mAb (BBIG-V1; R&D Systems) was used with goat anti-mouse IgG pAb-PE (BD Biosciences, San Jose, CA, USA) to detect VCAM-1 on HDMVECs.

Flow cytometry

Cells were blocked with FcR block (CD16/32, clone 93; eBioscience) and stained with the indicated antibodies in flow buffer (2% FCS, 1 mM EDTA, PBS). B Cells, tested for VCAM binding, were stained in PBS, with or without MnCl₂ (500 µM). Cells were acquired with either a BD LSRFortessa X-20 (BD Biosciences) flow cytometer or a modified FACScan (BD Biosciences) flow cytometer with DxP multicolor upgrades by Cytek Development (Fremont, CA, USA). Two additional lasers enabled acquisition of eight fluorophores by the modified FACScan. Cytometric data were analyzed with FlowJo software (v10.1; FlowJo, Ashland, OR, USA).

Cell culture

HDMVECs from neonatal skin (Thermo Fisher Scientific, Waltham, MA, USA) were cultured in Endothelial Basal Medium-2 (EBM-2; Lonza, Basel, Switzerland), supplemented with Endothelial Cell Growth Medium-2MV (EGM-2MV) SingleQuot growth factors (Lonza) at 37°C. Two days before imaging, 5 × 10⁴ HDMVECs were seeded onto the glass surface of imaging dishes (P35G-1.5-14-C; MatTek, Ashland, MA, USA) that had been coated with human fibronectin (10 µg/ml; Corning, Corning, NY, USA). The night before imaging, with the HDMVEC culture at or near 100% confluence, the culture media were supplemented with human TNF-α (20 ng/ml; Shenandoah Biotechnology, Warwick, PA, USA).

B Cell purification and activation

Splenic B cells were isolated (confirmed by flow cytometry ≥95% B220⁺) using a negative B Cell Isolation Kit (Miltenyi Biotec, San Diego, CA, USA), per the manufacturer’s protocol. When indicated, B cells were activated by overnight incubation in R10 medium [RPMI 1640 (Thermo Fisher Scientific), supplemented with FCS (10%; GE Healthcare HyClone; GE Healthcare Life Sciences, HyClone Laboratories, Logan, UT, USA), HEPES (10 mM; Thermo Fisher Scientific), penicillin/streptomycin (1×; Thermo Fisher Scientific), 2-ME (50 µM; Sigma-Aldrich, St. Louis, MO, USA), and L-glutamine (1×; GlutaMAX; Thermo Fisher Scientific)], supplemented with α-IgM Fab (10 µg/ml; ImmunoResearch Laboratories, West Grove, PA, USA) and mouse IL-4 (10 ng/ml; PeproTech, Rocky Hill, NJ, USA).

Migration assay

Before imaging, naive or activated B cells (10⁷ cells/ml) were rested (37°C, 5% CO₂) for 1 h in reduced-serum medium (R1; RPMI, 1% FCS, 10 mM HEPES). Concurrently, TNF-α-activated HDMVEC media were replaced with fresh endothelial culture media (without TNF-α) and incubated for 1 h. When indicated, the HDMVEC media were supplemented with the chemokine CXCL12 (100 ng/ml; R&D Systems) immediately before imaging (see Fig. 2). To induce directed B cell migration (see Fig. 3), chemokine (CXCL12 or CXCL13; final concentration in agarose 10 µg/ml) was diluted into prewarmed (37°C), sterile 0.5% soft agarose in PBS containing 1% BSA. A single 20 µl drop of agarose with chemokine was placed on the dry plastic ring of the MatTek dish without disturbing the monolayer, cooled for 10 min at room temperature, and then returned to the tissue-culture incubator. After 1 h of incubation, the media were changed, and the HDMVEC monolayer and solid chemoattractant agarose drop were both covered with fresh media. The B cells (1.5 × 10⁵ in 15 µl) were added to the HDMVEC monolayer immediately before imaging.

Figure 2. Optimal conditions for in vitro analysis of B cell migration included α-IgM/IL-4 activation of B cells, TNF-α activation of HDMVECs, and CXCL12 inclusion.

Results from 3 conditions are shown: 1) α-IgM/IL-4-activated B cells on TNF-α-activated HDMVECs without CXCL12, 2) α-IgM/IL-4-activated B cells on TNF-α-activated HDMVECs with CXCL12, and 3) naive B cells on unactivated HDMVECs. (A) Representative frames from movies of B cells migrating on endothelial monolayers. Cell treatments were as indicated. Original scale bars, 10 μm. (B) Flower plots of the first 45 min for the 10 longest tracks of WT B cells for each of the 3 indicated conditions. Axis scale in micrometers. (C) Track speed, arrest coefficient, and confinement ratio for tracked WT B cells. Each symbol represents 1 cell; lines indicate median values of all cells tracked. Cells were obtained from at least 2 independent fields for each condition shown. P values determined using one-way ANOVA.

Figure 3. B Cells undergo directed cell migration on HDMVEC monolayers.

WT B cells were activated overnight with α-IgM/IL-4 and added onto TNF-α-activated HDMVEC monolayers. An agarose drop containing CXCL12 was placed at a fixed position at the edge of the monolayer, corresponding to the upper-right corner of the field in this example (orange stars). (A) Representative frames taken at 5 min intervals from movies of B cells migrating on the surface of the endothelial monolayer, comparing migration of B cells from before and after placement of the CXCL12-containing agarose drop. Original scale bars, 30 μm. (B) Flower plots for 10 randomly selected tracks of WT B cells for the 2 indicated conditions. Axis scale in micrometers.

DIC images were acquired with a 10× air objective on an Olympus IX73 inverted microscope, operated with Micro-Manager software [32]. Cells were maintained at 37°C, with 5% CO₂, with an environmental chamber (Stage Top Incubator; Tokai Hit, Shizuoka-ken, Japan) throughout imaging. Single-plane images were acquired at 20 or 30 s intervals for at least 50 min.

Image analysis

Digital video images were processed with TrackMate (ImageJ software) [33]. Crawling B cells, identified by extension of lamellipodia and translocation of the cell body, were tracked by an individual viewing the movies. The resulting X–Y tracking data were used to calculate average track speed, arrest coefficient (fraction of track where instantaneous speed was <2 µm/min), and confinement ratio [(track displacement/track length) ⋅ (track duration)^1/2] as parameters reflecting cell migration [34].

Transwell migration assay

Splenocytes (10⁷ cells/ml) were rested (37°C, 5% CO₂) for 1 h in reduced serum medium (R2; RPMI, 2% FCS, 10 mM HEPES). Cells (∼2.5 × 10⁶) were then incubated in Migration Medium [RPMI, 0.5% BSA (Sigma-Aldrich), 10 mM HEPES] in the upper chamber of Transwell inserts precoated with hF_c-mVCAM (1 µg/ml; R&D Systems). Migration Medium (600 µl), with or without CXCL12 (100 ng/ml), was placed in the Transwell bottom chamber. Chambers were incubated at 37°C, 5% CO₂, for 3 h. Migrated cells were recovered from the bottom chamber, counted by hemocytometer, and analyzed by flow cytometry.

Statistics

Statistical comparisons were performed with Mann-Whitney or one-way ANOVA with Tukey comparisons, as indicated (Prism v5.0; GraphPad Software, La Jolla, CA, USA).

RESULTS AND DISCUSSION

Establishing an in vitro B cell migration assay

Our new B cell migration assay is adapted from a previously described assay to quantify leukocyte transendothelial migration using monolayers of cultured primary endothelial cells, HDMVECs [26]. As binding of the integrin VLA-4, a primary B cell adhesion molecule [35], to VCAM-1 has been previously shown to increase the efficiency of B cell migration [29], we first wanted to confirm that murine VLA-4 would engage with VCAM-1 expressed on the surface of human HDMVECs (Fig. 1A). The high-affinity conformation of VLA-4 can be induced by exposure to the divalent cation Mn²⁺ [36] and readily binds VCAM-1 [37]. Therefore, we probed naive murine B cells with mVCAM-1-hF_c or hVCAM-1-hF_c in the presence of Mn²⁺. Flow cytometry analysis showed equivalent binding of hVCAM-1 and mVCAM-1 to WT murine B cells (Fig. 1A). Furthermore, binding of hVCAM-1 to B cells from WT and LPL^−/− mice was equivalent (Fig. 1A), consistent with prior reports that LPL was dispensable for integrin activation and binding [22, 29, 30]. The binding of hVCAM-1 to B cells, activated with anti-IgM/IL-4, yielded similar results (data not shown). We also confirmed that overnight treatment of HDMVEC monolayers with TNF-α up-regulated the cell-surface presentation of VCAM-1 (Fig. 1B), as shown previously [38].

Figure 1. VLA-4 expressed by murine B cells binds hVCAM-1.

(A) Naive murine B cells were incubated with Mn²⁺ and probed with hVCAM-1 or mVCAM-1. Histograms show hVCAM-1 (solid green line) or mVCAM-1 (solid blue line) bound to murine WT B cells or hVCAM-1 bound to LPL^−/− B cells (gray-filled histogram). Negative control shows WT B cells incubated with Mn²⁺ and 2° antibody but without hVCAM-1 (dashed gray line). (B) Histogram of surface hVCAM-1 on HDMVECs with (solid green line) and without (solid blue line) TNF-α activation. Negative control includes HDMVECs incubated with 2° antibody and Mn²⁺ but no α-hVCAM-1 (dashed gray line). (C) Representative field of B cells (white arrows) coincubated with HDMVEC monolayers (yellow arrows are nuclei). Original scale bar, 100 µm. (D) Representative time-lapse images of WT B cells (white arrows) tracked as they crawl on an HDMVEC monolayer (yellow arrows are nuclei). These images are an expanded view of the area highlighted in C. Original scale bar, 20 µm.

To establish a system in which to visualize B cell migration, we added purified murine B cells onto HDMVEC monolayers and collected time-lapse DIC images at intervals of 20 or 30 s. The small, round B cells (Fig. 1C and D, white arrows) were easily distinguished from the underlying large, flat and nearly transparent HDMVECs (Fig. 1C and D, yellow arrows). Many B cells settled on the monolayer and appeared to engage with HDMVECs: they became less round, they flattened, and they extended lamellipodia (Fig. 1C and D, white arrows). B Cell migration was then quantified from these time-lapse images by manual tracking, exemplified in Fig. 1D.

Optimizing an in vitro B cell migration assay

We tested several conditions for quantification of B cell migration on HDMVEC monolayers. Purified B cells were initially used after activation by overnight incubation with α-IgM Fab and mouse IL-4. HDMVEC monolayers were incubated overnight with TNF-α to up-regulate VCAM-1 before imaging. When activated B cells were coincubated with activated HDMVEC monolayers, the B cells settled onto the monolayer, flattened, extended lamellipodia, and began to crawl (Fig. 2A and Supplemental Movie 1). Crawling cells displayed a “random walk” pattern of motion with rapid changes in direction, as opposed to noncrawling B cells that moved in 1 direction, “drifting” or tumbling over the HDMVEC monolayer.

We tested the effect of the chemokine CXCL12 on activated B cells with HDMVEC monolayers (Fig. 2A and Supplemental Movie 2). CXCL12 signals through the receptor CXCR4, expressed by all B cells [39], and it effectively induces B cell migration [29]. In the presence of chemokine, we observed that activated B cells became flattened or elongated, they extended lamellipodia, and they crawled on the HDMVEC monolayer. A fraction of the cells again appeared to tumble or drift across the monolayer without engaging or crawling. On occasion, clumps of cells formed in the presence of chemokine; these clumps were excluded from analysis.

As a negative control, we examined the movement of naive B cells incubated with HDMVEC monolayers that had not been treated with TNF-α (Fig. 2A and Supplemental Movie 3). HDMVEC monolayers do not express high levels of VCAM-1 in the absence of TNF-α treatment (Fig. 1B), so we anticipated that B cells would not engage and crawl upon nonactivated HDMVEC monolayers. Indeed, most of the naive B cells added to nonactivated HDMVEC monolayers underwent the drifting, unidirectional movement observed for some of the cells in the prior 2 conditions. We found fewer B cells that appeared to be engaged with the monolayer, which we defined as cells that flattened and extended lamellipodia.

For each field of view, we tracked all of the cells that appeared to be extending lamellipodia and migrating on the monolayer (Fig. 2B), up to 100 cells per each movie. For clarity, Fig. 2B displays only the 10 longest tracks in “flower plots” from 1 representative movie of each cohort. The flower plot tracks are truncated to a 45 min duration. Consistent with our qualitative observations of the movies, directed lymphocyte migration was enhanced by activation of B cells and by TNF-α activation of HDMVEC monolayers. As a negative control, we also tracked a small number of B cells in the absence of activation of B cells and HDMVECs that appeared to be drifting passively.

From the tracking data, we quantified B cell speed, arrest coefficient, and confinement ratio (Fig. 2C). Activated B cells, coincubated with activated HDMVEC monolayers in the absence of CXCL12, migrated with a median speed of 2.9 μm/min (95% CI 2.8–3.7), a value similar to those of B cells following detection of cognate antigen [25]. Addition of CXCL12 increased the cell speed to a median value of 4.1 μm/min (95% CI 3.9–4.8), closer to what has been reported for intravitally imaged plasma cells [24]. Naive B cells incubated with nonactivated HDMVEC monolayers displayed more drifting than crawling, and they had higher cell speeds, with a median of 5.0 μm/min (95% CI 4.7–6.0).

The arrest coefficient, defined as the fraction of time B cells were migrating at <2 μm/min [40], was high for activated B cells coincubated with activated HDMVEC monolayers in the absence of chemokine (median 0.49; 95% CI 0.42–0.52; Fig. 2C). Again, this value is similar to that of B cells after detection of cognate antigen [25]. Addition of chemokine decreased the value of the arrest coefficient (median 0.26; 95% CI 0.25–0.34). Naive B cells on nonactivated monolayers had an even lower arrest coefficient, with a median value of 0.24 (95% CI 0.19–0.31).

Finally, we calculated the confinement ratio, defined as the distance between the initial and the final positions of the cell, divided by the cumulative path-length distance traveled by the cell (Fig. 2C). The confinement ratios of activated B cells incubated with activated HDMVEC monolayers, with and without CXCL12, were similar, with median values of 1.3 (95% CI 1.2–1.6) and 1.1 (95% CI 1.0–1.3), respectively. The confinement ratio of the naive B cells incubated with nonactivated monolayers was higher (median 2.3, 95% CI 1.5–2.5), consistent with cell drifting, rather than active cell migration.

To determine if our new protocol would allow us to observe directed cell migration by activated B cells moving across the HDMVEC monolayer, we added activated B cells to HDMVECs that had been pretreated with TNF-α. We added a solid drop of agarose containing CXCL12 at a fixed position at the edge of the monolayer (Fig. 3). Different positions for the agarose were selected for each experiment to eliminate potential effects of extraneous fluid currents from the heated stage. We again collected time-lapse DIC images at intervals of 20 s. After the initial addition of fresh medium, before the chemokine gradient had time to establish, B cells engaged with the monolayer, extended lamellipodia, and appeared to crawl with a random walk pattern of motion (Fig. 3A, upper; Fig. 3B, left; and Supplemental Movie 4). However, after 30 min of incubation with the chemokine-laden agarose drop, a gradient was established, and the B cells began to move toward the source of chemoattractant (Fig. 3A, lower; Fig. 3B, right; and Supplemental Movie 5). Similar results were obtained for the chemokine CXCL13 (data not shown). Thus, our system also supports and allows for observation of directed B cell migration.

As activation of B cells and HDMVECs and inclusion of CXCL12 provided the greatest number of B cells migrating with the track speed and arrest coefficients closest to what has been reported for B cells imaged intravitally [24], we used these conditions for subsequent in vitro studies of B cell migration, as diagramed in Fig. 4.

Figure 4. Flow chart of B cell and HDMVEC preparation for live-cell imaging.

The optimal protocol for imaging B cell migration was determined to include overnight activation of B cells with α-IgM/IL-4, activation of HDMVECs with TNF-α, and inclusion of CXCL12 during imaging.

B Cells lacking LPL migrate more slowly

To test the new in vitro B cell migration assay further, we extended our previous investigations of the requirement for LPL in B cell migration (Fig. 5). As before, WT B cells engaged and crawled over activated HDMVEC monolayers (Fig. 5A and Supplemental Movie 6). B Cells from LPL^−/− mice settled onto monolayers, but they did not spread as much, and they moved more slowly (Fig. 5 and Supplemental Movie 7). Flower plot tracks revealed restricted movement (Fig. 5B), and cell tracking revealed decreased track speeds and increased arrest coefficients for LPL^−/− B cells compared with WT B cells (Fig. 5C). Confinement ratios were similar for B cells from WT and LPL^−/− mice (Fig. 5C), suggesting that LPL is not required for lymphocyte turning, consistent with the normal meandering indices obtained when analyzing LPL^−/− T cells in lymph nodes by 2-photon microscopy [30].

Figure 5. LPL^−/− B cells exhibit reduced migratory speed.

WT or LPL^−/− B cells were activated overnight with α-IgM/IL-4 and added onto TNF-α-activated HDMVEC monolayers, with CXCL12 stimulation. (A) Representative frames taken at 5 min intervals from movies of B cells migrating on the surface of the endothelial monolayer, comparing B cells from WT or LPL^−/− mice. Original scale bars, 10 μm. (B) Flower plots of the first 45 min of the 10 longest tracks. Axis values are micrometers. (C) Track speeds, arrest coefficients, and confinement ratios for attached B cells. Each symbol represents a tracked cell with the line showing the median value for all cells analyzed in at least 10 independent fields. P values determined using one-way ANOVA.

Our results are also similar to those in which small interfering RNA-mediated depletion of LPL reduced T cell velocity but not persistence when cells were plated on VCAM-1 [22]. Thus, analysis of B cells from WT and LPL^−/− mice migrating on HDMVECs in vitro supports the direct comparison of the function of actin-binding proteins in B and T cells and shows that the phenotype of LPL^−/− currently appears similar in B and T cells.

We demonstrated previously that B cells from WT and LPL^−/− mice express similar levels of CXCR4, the receptor for CXCL12, and that they exhibit similar levels of attachment to plate-bound VCAM-1 and CXCL12 [29]. Furthermore, B cells from WT and LPL^−/− mice also undergo similar activation with α-IgM and IL-4 treatment [29]. Thus, the reduced speed of LPL^−/− B cells observed using our new assay is not a result of reduced expression of the chemokine receptor or a failure of B cell activation.

LPL^−/− B cells display decreased migration toward CXCL12

To validate these novel results from our new in vitro B cell migration assay, we modified our methodology to reassess the Transwell migration of different splenic B cell populations from WT and LPL^−/− mice. Splenic cells can be divided into 3 populations based on expression of the markers CD23 and CD21 (Fig. 6A). Cells expressing high levels of CD23 and low CD21 are FO B cells, cells expressing low levels of CD23 and high levels of CD21 are MZ B cells, and cells expressing low levels of CD23 and CD21 are immature or NF splenic cells that have not yet completed maturation (Fig. 6A). By resting splenocytes at 37°C with 5% CO₂ for at least 1 h before exposure to chemokine, we greatly enhanced the responsiveness of B cells to CXCL12. After resting, an average of 49% of input WT B cells transmigrated toward the chemokine (Fig. 6B); previously, only 12% of input WT B migrated over the course of an experiment [29].

Figure 6. Reduced CXCL12-mediated Transwell migration of LPL^−/− FO B cells.

(A) Representative flow cytometric gating of splenic B cells showing NF, FO, and MZ B cells. (B) Transwell migration of B cells, expressed as a percentage of input cells, for total B cells, FO B cells, or MZ B cells, obtained from WT (closed circles) or LPL^−/− (open circles) mice. Each symbol represents the average of triplicates within each experiment, and the horizontal lines indicate the median value of 5 independent experiments. **P < 0.01, using Mann-Whitney.

In this modified Transwell migration assay, fewer LPL^−/− (32.5 ± 8.1%) compared with WT (49.0 ± 7.3%) splenic B cells migrated toward CXCL12 (Fig. 6B). This decrease in migrating cell number correlates with the decrease in cell speed that we observed when LPL^−/− B cells were crawling on HDMVECs (Fig. 5A), independently confirming the results from the new in vitro B cell migration assay. Interestingly, we found that LPL^−/− led to diminished migration of FO B cells, the predominant splenic B cell population, but had no effect on the migration of MZ B cells (Fig. 6B).

Summary and conclusions

In summary, we have developed a new method for analyzing the 2-dimensional crawling of primary B cells in vitro. This method provides information regarding cell direction, speed, and shape, which is not provided by traditional Transwell migration assays. With the use of a monolayer of endothelial cells as substrate, our assay should support the investigation of proteins important for integrin activation and adhesion, as well as proteins, such as LPL that primarily support locomotion. We also suggest that endothelial cells are a more physiologically relevant substrate than the artificial coated surfaces used in other assays.

AUTHORSHIP

S.C.M. conceived of the study. E.R.J., M.D.O., and S.C.M. designed experiments. P.J.S-H., T.P.S., and E.R.J. acquired and analyzed the data. P.J.S-H. and E.R.J. interpreted the data. P.J.S-H was the primary author of the manuscript, with revisions from M.D.O., J.A.C., and S.C.M. Funding was acquired by J.A.C. and S.C.M.

Glossary

APC

allophycocyanin

CI

confidence interval

DIC

differential interference contrast

FO

follicular

HDMVEC

human dermal microvascular endothelial cell

hF_c

human IgG Fc

hVCAM-1

human VCAM-1

LPL

L-plastin

LPL^−/−

L-plastin deficient

mVCAM-1

murine VCAM-1

MZ

marginal zone

NF

newly forming

WT

wild-type

WUSM

Washington University School of Medicine

DISCLOSURES

The authors declare no conflicts of interest.