Microfluidics-based side view flow chamber reveals tether-to-sling transition in rolling neutrophils

Abstract

Neutrophils rolling at high shear stress (above 6 dyn/cm² ) form tethers in the rear and slings in the front. Here, we developed a novel photo-lithographically fabricated, silicone(PDMS)-based side-view flow chamber to dynamically visualize tether and sling formation. Fluorescently membrane-labeled mouse neutrophils rolled on P-selectin substrate at 10 dyn/cm² . Most rolling cells formed 5 tethers that were 2-30 μm long. Breaking of a single tether caused a reproducible forward microjump of the cell, showing that the tether was load-bearing. About 15% of all tether-breaking events resulted in slings. The tether-to-sling transition was fast (<100 ms) with no visible material extending above the rolling cell, suggesting a very low bending modulus of the tether. The sling downstream of the rolling cell aligned according to the streamlines before landing on the flow chamber. These new observations explain how slings form from tethers and provide insight into their biomechanical properties.

R

A

P

,

<

Neutrophil granulocytes are the rst responder immune cells in bacterial and fungal infections. Most of their immune response activity occurs outside of the blood vessels, requiring extravasation13. As a rst step of extravasation, streaming neutrophils start rolling along the blood vessel wall. Rolling is mediated by selectins and modulated by integrins4,5. Unlike nave lymphocytes, neutrophils are capable of rolling at high wall shear stress (WSS), 6. This ability of neutrophils is closely related with the (i) catch-bond behavior of selectins, (ii) deformability of neutrophils, which reduces the hydrodynamic drag and increases the area of contact with the wall, (iii) formation of tethers that bear the loads of force and torque7, and (iv) formation of slings that can serve as self-adhesive substrates8.

Tethers are long sub-micron diameter tubes pulled out from neutrophils rolling under hydrodynamic drag. Tethers form behind neutrophils, originating from microvilli, and bind to the vessel wall through P-selectin glycoprotein ligand -1 (PSGL-1) expressed on the tips of the microvilli binding to selectins8. From in vitro experiments, tether tension force is estimated to reach ~80 pN6, such that a single tether is expected to signicantly contribute to the forces resisting the hydrodynamic drag on the neutrophil (~470 pN at 10dyn/cm2). Mechanical properties of the tethers allow them to substantially stretch at largely unchanged tension force as the neutrophil rolls forward7. Eventually the bonds that anchor the tether to the substrate break under the pull of the rolling neutrophil, which we call tether breaking.

Tethers were rst identied with DIC microscopy in ow chamber experiments9,10 behind neutrophils rolling on functionalized glass substrates. Tethers of neutrophils with uorescently labeled membrane were imaged using TIRF microscopy with the quantitative dynamic foot printing (qDF) method6,11. In the high resolution qDF images, the tether anchoring points appeared as bright dots up to 16m behind the rolling cell, suggesting that tethers can be at least 16m long. The same technique was used to discover slings, up to 22 m long structures found on the substrate in front of rolling neutrophils. However, the TIRF microscopy imaging did not allow observing the tether-to-sling transition directly. Here, we tested the hypothesis that slings originate from tethers that swing around the rolling cell aer detaching from the selectin substrate.

For a neutrophil rolling on a horizontal substrate on a regular vertical (upright or inverted) microscope, capturing an entire tether or sling is difficult, because these structures have large extensions along the vertical (Z-) axis, whereas the imaging is always in a horizontal (XY-) plane. Confocal Z-axis optical sectioning has low temporal and spatial resolution12 and may also induce photo-bleaching and photo-toxicity. Therefore, it would be a poor match for the fast dynamic process of neutrophil rolling.

La Jolla Institute for Allergy and Immunology, La Jolla, CA, USA. Department of Physics and University of California San Diego, La Jolla, CA, USA. Department of Bioengineering, University of California San Diego, La Jolla, CA, USA. Correspondence and requests for materials should be addressed to K.L. (email: [email protected])

SCIENTIFIC REPORTS

1

www.nature.com/scientificreports/

Figure 1. Side view ow chamber. The ow chamber is assembled by vacuum clamping a silicone (PDMS) microuidic chip against a #1.5 cover glass. (A) Schematic of the ow chamber; large arrow indicates ow direction; rolling cells are symbolized by (not to scale) red dots. (B) Cross section of one test channel in Y-Z plane. For side and bottom view, cells rolling on the cover glass bottom and lateral silicone wall, respectively, are imaged. (C) Calibration of wall shear stress (WSS) as a function of perfusion pressure with and without cells. (D) Low magnication (20x) brighteld image of rolling leukocytes in the side view ow chamber. White arrows indicate the cells visible from side view and red arrow indicates a cell visible from bottom view perspective. Scale bar represents 45m.

Side view ow chambers were previously implemented using 45 tilted mirrors1315 or a horizontally mounted objective16. In another approach, rolling cells were imaged on a vertical wall of an ~0.6mm round capillary with agarose walls17, revealing WSS-dependent deformation of human neutrophils. These setups did not allow imaging with high-resolution (high numerical aperture), low working-distance objectives. Our new side view ow chamber can be interrogated with high numerical aperture immersion objectives for maximum spatial resolution.

Here we use a microuidic ow chamber to study primary mouse neutrophils rolling on a vertical (XZ-plane) wall, such that entire tethers and slings are observed at high resolution in a single XY-plane with high numerical aperture objectives. This setup enables the rst direct observation of tether-to-sling transition on well-dened substrate at physiologically relevant WSS values.

The microuidic device consisted of a polydimethylsiloxane (PDMS) chip with micro-channels engraved on its surface and a #1.5 cover glass sealing the micro-channels (Fig.1). The chips were cast using a silicon wafer with a lithographically fabricated relief of a negative photoresist (SU8-2000 by Microchem, Westborough, USA) on its surface as a master mold18. The surface of PDMS was functionalized with reactive amino groups using silane chemistry. The microuidic device had one inlet, one outlet, and 12 identical parallel perfusion channels with width w = 45 m, height h = 50 m, and length of 5 mm (Fig.1A,B). In experiments with washed neutrophils, the neutrophil suspension and buer were held in reservoirs connected to the device inlet and outlet, respectively, through PE10 (Becton Dickinson, New Jersey, USA) tubing. The perfusion was driven by a positive dierential pressure, P, between the inlet and outlet that was applied hydrostatically by placing the inlet reservoir above the outlet reservoir. The channels were rst primed with PBS, then functionalized via perfusion with 1 g/ml P-selectin-Fc (R&D Systems Inc., Minneapolis, USA) PBS solution for 15 min, and nally blocked with casein in PBS blocking solution (Thermo Fisher Scientic, Waltham, USA) for 1 h. P-selectin PSGL-1 interaction dependence of rolling was tested via adding P-selectin (clone RB40.34, 5g/ml for 15min) or PSGL-1 (clone 4RB12, 5g/ml for 15min) blocking monoclonal antibody to the cell suspension. The monoclonal blocking antibodies were puried from hybridoma supernatant at the biomolecular facility of the University of Virginia (Charlottesville, USA). The coefficient of proportionality between P and WSS in the perfusion channels was determined by perfusing the device with an aqueous suspension of 500nm uorescent beads (Polysciences, Warrington, USA) and calculating their maximal velocity vmax vs. P. The value of vmax was calculated by dividing the length of the streak lines of the beads near the axis of symmetry of the channel by the exposure time, and WSS at the mid-plane of the channel (25 m from the bottom), , was calculated as =4.3max/w, where is the viscosity and 4.3 is a factor derived by numerical simulation.

SCIENTIFIC REPORTS

2

www.nature.com/scientificreports/

The experimental animal procedures of this study were performed in accordance with approved guidelines. These guidelines were approved by the Institutional Animal Care and Use Committee of La Jolla Institute for Allergy and Immunology, which is an AAALAC international accredited facility. For each experiment, bone marrow from a wild type female mouse was collected, resuspended in PBS and ltered through a 40 m nylon cell strainer (Biologix, Lenexa, USA). From that sample, neutrophils were enriched at room temperature via negative selection with a customized neutrophil enrichment kit (StemCell Technologies, Vancouver, Canada). FACS measurements indicated that >70% of cells were neutrophils (Ly-6G+), and the major contaminating cells were monocytes. The neutrophil isolate was resuspended at 2.5 million cells/ml in RPMI supplemented with 10% fetal calf serum (Gemini, West Sacramento, USA) and 0.1% penicillin-streptomycin (Gibco Thermo Fisher Scientic, Waltham, USA). To label the cell membrane, 4m Vybrant-DiO (Invitrogen, Carlsbad, USA) or 1 m Cell Mask Green (Invitrogen, Carlsbad, USA), or 5 ng/l (1:100) anti Ly-6G antibody conjugated with Alexa Fluor 488 (clone RB68C5, eBioscinece, San Diego, USA) was added to the cell suspension. Aer 10 min incubation, with washing (DiO) or without washing (Cell Mask Green, Ly-6G-AF488), the suspension was perfused through the ow channels. As soon as rolling cells were visible, the perfusion solution was switched to PBS to reduce the background signal by washing out the solution containing free dye.

Fluorescence imaging of rolling neutrophils was performed with a Leica SP5 confocal microscope (Leica Microsystems Inc., Bualo Grove, USA) using a 63 1.3 glycerol immersion objective. For bottom view images, cells rolling on the bottom surface were imaged, while for side view, cells rolling on the side walls were imaged (Fig.1B). A resonant scanner was used to record rolling neutrophils at 10 frames/second; a conventional scanner was used to image Z-stacks of photoxed cells. For photoxation, the epiuorescence light source was turned on for 5seconds, which resulted in intermittent or permanent cell arrest without visible damage to the cell. DIC imaging was done with Zeiss LSM 780 microscope (Carl Zeiss Microscopy, Thornwood, USA). Basic image processing and distance measurements were performed with Fiji image analysis freeware (website: ji.sc). Z-stack reconstruction and cell tracking was done with Imaris image analysis soware (Bitplane, Windsor, USA). For statistical analyses Students T test or correlation analysis was performed.

Primary mouse bone marrow neutrophils rolled at 5.21(SEM)m/s at WSS=610dyn/cm2 (n= 28) (Supplementary Movie 1). Neutrophil rolling was specic as it was not observed in non-coated channels, channels pre-incubated with P-selectin blocking or PSGL-1 blocking antibody in the cell suspension.

Bottom view images of rolling neutrophils showed tether anchor points behind the cell body as expected. Due to the limited optical section thickness, only short segments of tethers were visible (Fig.2A). Neutrophils rolling at 610dyn/cm2 had 410 tethers (median value 5, n= 12). The tethers were distributed across most of the apparent neutrophil width (Fig.2C) with the numbers of tethers declining towards the neutrophil periphery (Fig.2D).

Side view images taken immediately aer neutrophil arrest revealed the whole tether segments between the cell body and the side wall (Fig.2E). The side view images show many but not all tethers of each neutrophil, because the optical section thickness is about 2m, but tethers are distributed across ~7m of the cell width as shown by bottom view images (Fig.2C). Analysis of 193 tether side view images showed that at 610 dyn/cm2 WSS, the average apparent tether length, determined as the distance between the point where the tether appears to emerge from the cell body and the point where the tether reaches the wall of the ow channel was 9.80.4(SEM)m; the longest tether length was 30.1m and the shortest was 1.2m (Fig.2F). The tether length measured in side view was signicantly (p < 0.0001) longer than the tether length measured in bottom view (6.50.4(SEM)m, ranging from 1 to 24.8m; n= 58), suggesting systematic undersampling of long tethers in bottom view mode.

In some experiments, rolling neutrophils were photoxed to obtain a whole-cell side view Z-stack (Fig.2G and Supplementary Movie 2). The side view images showed that some tethers have grape-like structures along their length (Fig.2H). To show that tether formation is not an artefact caused by membrane labeling with membrane intercalating dyes (DiO and Cell Mask Green), non-labeled rolling neutrophils were imaged with DIC microscopy (Supplementary Fig. 1) or rolling neutrophils labeled with Ly-6G-AF488, a uorescently labeled antibody against a neutrophil surface marker, were imaged with confocal microscopy (Supplementary Fig. 2). Both of these methods showed tethers similar to the ones observed with the DiO and Cell Mask Green dyes. The side view images also conrmed the previously reported elongated shapes of rolling neutrophils under shear stress19,20.

Next, we asked what role tethers played in neutrophil rolling. If individual tethers bear significant loads, the neutrophil rolling should significantly accelerate immediately aer breaking of a tether. Frame-to-frame neutrophil displacement was analyzed between the last frame where an intact tether was seen (frame T) and the two following frames (frames T + 1 and T + 2). The displacement between T and T+ 1 was 1.7 0.1(SEM) times greater than between T+ 1 and T+ 2 (n= 22), indicating that a rolling neutrophil greatly accelerates, eectively making a jump, immediately aer a tether breaks. In some rolling neutrophils, multiple consecutive tether breaks could be observed (Fig.3A), and large neutrophil displacements (jumps) immediately followed. In one cell (cell 2 in Fig.3A,B), multiple tethers broke within ~0.5second, resulting in a very large jump. These ndings suggest that individual tethers make signicant contributions to the resistance force preventing the cell from moving with the ow.

This side-view ow chamber allows for the rst time to visualize the formation of slings. Specically, we tested the hypothesis that slings derived from tethers, which required detailed observation of the formation of tethers and slings. In side view (Fig.4A and Supplementary Movie 3) and bottom

SCIENTIFIC REPORTS

3

www.nature.com/scientificreports/

Figure 2. Tethers of rolling/just arrested neutrophils. (A) Bottom view confocal micrographs of neutrophils rolling from le to right. The focal plane is just above the cover glass to capture the tether anchoring points, which appear as short lines (some marked by white arrows). (B) To quantify tether position along the neutrophil width the distance between neutrophil Y centerline and the Y projection of the tether was measured (Dtether).

(C) Each gray bar represents the apparent width of an individual neutrophil from the bottom view imaging. Black diamonds indicate the tethers Dtether values. Whereas the positions of tethers are widely distributed, their number is maximal in the central 2m section along the Y-axis. (D) Histogram of Dtether values, (normalizedto neutrophil half-width) mean (SEM)=0.350.03, min=0.01, max=0.96, n=73. (E) Side view confocal images of neutrophils taken immediately aer their arrest with averaging over several frames to enhance signal-to-background ratio. Only tethers within the 2m depth of eld are visible. (F) Histogram of the apparent tether length, the distance between the tether anchor point and the point where the tether merges with the cell body,as measured in side view immediately before tether breaks; mean (SEM)=9.80.41m, min=1.2m, max=30.1m, n=193. (G) 3D reconstruction image of a recently arrested neutrophil. A neutrophil rolling on the side wall was photo-xed and a confocal Z-stack was taken. (H) In some cases grape-like structures were observed on the tethers (indicated by white arrows). Scale bars are 8m and open arrows indicate the ow direction.

view (Supplementary Fig. 1 and Supplementary Movie 4) recordings, slings were observed to attach to the substrate in front of the cell aer tether breaks, suggestive of a tether-to-sling transition taking place. However, considerable delay (3.2 0.8(SEM) sec) between the two events made the evidence for the tether-to-sling transition

SCIENTIFIC REPORTS

4

www.nature.com/scientificreports/

Figure 3. Tether break results in micro-jump. (A) The position of the center of mass of rolling neutrophils was tracked from video recordings as a function of time. The last frames where a certain tether is visible and the rst frames where the tether is not visible any more are highlighted by grey columns, which are interpreted as time points of tether break events. Tether breaks are found to concur with greater cell displacements over the xed time interval between the frames, indicating signicant short-term accelerations or micro-jumps.(B) Six consecutive frames from Cell 2 sequence (range indicated by red dashed line on panel A) are shown. White arrows indicate tether anchoring points on the last frame where the tether is visible. The vertical lines indicate the front of the cell at dierent time points. Two short tether breaks precede a long tether break, with each break resulting in a micro-jump. The neck of the sling in front of the cell rotates clockwise with the cell, approaching the substrate (side wall) in the process. Scale bar is 8m.

from the bottom view chamber experiments less convincing and also made it difficult to establish which specic tether became the sling. Unlike the bottom view chamber, where a sling only becomes visible aer it attaches to the substrate, the side view enabled observing slings in frames immediately following those with tether breaks, within ~100 ms of their formation and long before they could be seen in the bottom view. During that time interval, slings could be seen stretched along the stream lines in front of the cell (Fig.4B). Computational uid dynamics (CFD) showing the Y velocity map around the rolling cell conrmed that initially, slings experience a velocity toward the wall (Fig.4C). As the cell continued rolling, bringing the point of attachment of the sling to the cell closer to the substrate, the stretched sling got closer to the substrate as well. The side view time series indicated that about 15% of all breaking tethers (28 of 193) formed slings. Slings were also observed on Ly-6G-AF488 labeled rolling neutrophils (Supplementary Fig. 2), supporting that sling formation is not an artefact caused by membrane labeling with DiO or Cell Mask Green.

We measured the angle between the tether and bottom surface (tether). Consistent with the length of the tethers and the diameter of the neutrophil, the average tether was 22.3 1.3(SEM) (Fig.4D), ranging from 12 to 38. The angle between the sling and a line parallel to the wall averaged 61(SEM) (sling) (Fig.4B,D). CFD showed that the streamlines in front of the rolling cell were tilted downwards. However, as the cell rotated and the sling approached the substrate, CFD predicted an upward angle (Fig.4C). This was experimentally conrmed by looking at short slings near the wall (Fig.4B, lower panel).

We compared the lengths of tethers immediately before and slings immediately aer the tether-to-sling transitions (Fig.4E). The average nal tether length was 14.6 1(SEM)m, signicantly shorter (p< 0.05) than the average initial sling length of 12.31(SEM)m. The sling-forming tethers were signicantly (p<0.0001) longer (14.6m) than the average (9.80.41(SEM)m) (Fig.4F), suggesting that slings preferentially form from longer tethers. On average, the slings were 81 7% of the length of the tethers they originated from.

Here, we introduced a new microuidic side-view ow chamber compatible with high resolution microscopy that enabled the rst direct observation of the tether-to-sling transition, a process of fundamental importance for leukocyte adhesion under high shear stress. In previous studies, tethers were hypothesized to form slings [6], but there was no direct evidence for that. Tether breaks could not be reliably matched to sling formation, because slings are invisible under TIRF microscopy until they are very close to the substrate.

Our data indicate that about 15% of all tethers form slings. This might be an underestimate, because the side view ow chamber records only a ~2m section in the XZ plane and thus might miss slings that may have moved

SCIENTIFIC REPORTS

5

www.nature.com/scientificreports/

Figure 4. Tether-to-sling transition. (A) Key frames of a tether-to-sling transition from a side view record. Transition happens within 100ms, but it takes 1.2sec until this sling reaches the substrate. During that period the sling would not be visible from the bottom view perspective. (B) Side view images allowed measuring the angle between the bottom surface and tethers (tether) or slings (sling). Two consecutive frames with a tether-to-sling transition are shown. (C) Computer uid dynamic simulation of ow conditions around the neutrophilat 10dyn/cm2 wall shear stress. The heat map indicates the direction and magnitude of the ow component perpendicular to the bottom surface (VY). At the right upper quadrant of the cell VY points down (indicated by white arrow) and at the right lower quadrant of the cell VY points up (indicated by red arrow). The four white lines originating on the right side of the cell indicate how the slings would align under the modeled stream conditions. (D) Last tether/rst sling angle pairs were measured on tether-to-sling transitions. The angle pairs are shown as bar-doublets, where the gray bar represents tether angle and the black bar represents sling angle. Mean (SEM) last tether angle=22.31.2, min=11.8, max= 38.7. Mean (SEM) rst sling angle=61, min=0, max=24.3. (E) Final lengths of individual tethers (grey bars, min=1.8m, max=25.6m) and initial lengths of slings that originated from those tethers (black bars, min=2.3m, max=31.5m) for dierent tether-to-sling transitions. (F) Average (SEM) lengths of nal lengths of all tethers, tethers before tether-to-sling transitions (Transition tethers), and initial lengths of slings. Scale bars indicate 8m, large arrows indicates ow direction.

out from the focal plane in the Y direction. This can be caused by random movement of rolling neutrophils in the Y direction, which is expected whenever the key anchor point is o-center21,22. When a tether breaks and another o-center tether becomes load-bearing, the cell will pivot about the Y-axis until the cell center is downstream from the new tether. Since the sling cannot provide any mechanical stability while detached (before reaching the wall), the sling will passively swing in the Y direction, forced by the Y movement of its neck where it is attached to the rolling cell2325.

The transition from a tether behind a cell to a sling in front of the cell is completed within less than one frame (100 ms). In the 28 tether-to-sling transitions, we could only see slings on the substrate or aligned along

SCIENTIFIC REPORTS

6

www.nature.com/scientificreports/

the ow streamlines in front of rolling cells and never above the rolling cell. This observation is consistent with the expected physical properties of slings, sub-micron diameter tubes with low resistance to bending7,2628 that

can only be stable (as needed to produce sufficiently sharp images under 100 ms exposure) under longitudinal tension, but not under transverse forces. Indeed when slings are in front of rolling cells, before landing on the substrate, they are closely aligned along the streamlines.

On most of the side view transition records, the sling can be reliably matched to a broken tether. Slings spend several hundred milliseconds hanging in the uid stream before landing. This explains the delay between tether break and sling formation observed with qDF. The hanging slings enclose an average angle of 6 1(SEM) with the bottom surface. This is consistent with the notion that the slings follow the ow prole in front of the cell and a previous intravital microscopy study with particle tracking velocimetry29. The oating sling does not reach the wall until the rolling neutrophil has rotated far enough so the origin (neck) of the sling is near the wall. This observation is consistent with slings being anchored to the cytoskeleton30,31, but the nature of this anchorage is currently unknown and awaits further research.

The side view chamber enables more accurate measurements of the tether lengths as compared with the bottom view chamber. However, the apparent lengths of the tethers in the side view ow chamber are still underestimates, because a segment of the tether close to its neck will be obstructed by the brighter cell, when the neck is not in the equatorial plane of the cell. If a tether originates from the cell body in a plane dierent from the equatorial plane, the part of the tether behind or in front of the cell equator is invisible and will not be measured. This is signicant as shown in (Fig.2), where many tethers originate in planes other than the equatorial plane. Thus, a complete understanding of rolling with tethers and slings is only possible through side AND bottom (or top) view imaging as reported here.

The side view images showed that some tethers have grape-like structures along their length, indicating that the grapes observed by electron microscopy of xed neutrophils with tethers6 are not xation artefacts. Similar grape-like structures were observed on tethers of rolling platelets32. The nature of these structures is unknown; one theory is that these are remnants of microvilli which are pulled from the cell with the membrane into the tether.

We observed that each break of a tether results in a major short-term acceleration of the cell rolling (a microjump). This indicates that, immediately before breaking, each tether is load-bearing, signicantly contributing to the forces opposing hydrodynamic drag on the cell, corroborating previous reports6,10.

In its current conguration, the side view ow chamber can show many rolling cells from perfusion with 250 l cell suspension containing 500,000 cells. The volume of cell suspension required can be further reduced in a device with a smaller number of perfusion channels. The PDMS walls can be readily coated with proteins thus making the side view ow chamber useful for studies of neutrophil arrest and interaction of other types of leukocytes33,34 or platelets35,36.

Here we studied neutrophils rolling on P-selectin with no integrin ligand. Tether and sling formation on more complex substrates has not been studied. Neutrophils rolling on E-selectin also form slings and tethers6. Little is known about tether and sling formation of other leukocytes, except that T-helper 1 (Th1) CD4 lymphocytes but not nave CD4 T lymphocytes rolling on P-selectin form slings and tethers6.

In summary, we directly show on neutrophils rolling on P-selectin that some tethers form slings that are aligned along streamlines in front of the cell before landing on the substrate; tethers originate from almost anywhere across the leukocyte width; tethers are load-bearing and thus directly relevant for cell rolling. We conclude that the new side and bottom view ow chamber is a versatile tool for studying cell rolling, adhesion and migration under ow.

1. Kolaczkowska, E. & Kubes, P. Neutrophil recruitment and function in health and inammation. Nature Reviews. Immunology 13, 159175 (2013).

2. Beyrau, M., Bodkin, J. V. & Nourshargh, S. Neutrophil heterogeneity in health and disease: a revitalized avenue in inammation and immunity. Open Biology 2, 120134 (2012).

3. Mocsai, A. Diverse novel functions of neutrophils in immunity, inammation, and beyond. The Journal of Experimental Medicine 210, 12831299 (2013).

4. McEver, R. P. Selectins: initiators of leucocyte adhesion and signalling at the vascular wall. Cardiovascular Research 107, 331339 (2015).

5. Schmidt, S., Moser, M. & Sperandio, M. The molecular basis of leukocyte recruitment and its deciencies. Molecular Immunology 55, 4958 (2013).

6. Sundd, P. et al. Slings enable neutrophil rolling at high shear. Nature 488, 399403 (2012).7. Pospieszalska, M. K., Lasiecka, I. & Ley, K. Cell protrusions and tethers: a unied approach. Biophysical Journal 100, 16971707 (2011).

8. Sundd, P., Pospieszalska, M. K. & Ley, K. Neutrophil rolling at high shear: attening, catch bond behavior, tethers and slings. Molecular Immunology 55, 5969 (2013).

9. Schmidtke, D. W. & Diamond, S. L. Direct Observation of Membrane Tethers Formed during Neutrophil Attachment to Platelets or P-Selectin under Physiological Flow. The Journal of Cell Biology 149, 719730 (2000).

10. Ramachandran, V., Williams, M., Yago, T., Schmidtke, D. W. & McEver, R. P. Dynamic alterations of membrane tethers stabilize leukocyte rolling on P-selectin. Proceedings of the National Academy of Sciences of the United States of America 101, 1351913524 (2004).

11. Sundd, P. et al. Quantitative dynamic footprinting microscopy reveals mechanisms of neutrophil rolling. Nature Methods 7, 821824 (2010).

12. Rullo, J. et al. Actin polymerization stabilizes alpha4beta1 integrin anchors that mediate monocyte adhesion. J Cell Biol 197, 115129 (2012).

13. Cao, J., Usami, S. & Dong, C. Development of a side-view chamber for studying cell-surface adhesion under ow conditions. Annals of Biomedical Engineering 25, 573580 (1997).

14. Dong, C. & Lei, X. X. Biomechanics of cell rolling: shear ow, cell-surface adhesion, and cell deformability. Journal of Biomechanics 33, 3543 (2000).

SCIENTIFIC REPORTS

7

www.nature.com/scientificreports/

15. Gonnermann, C. et al. Quantitating membrane bleb stiffness using AFM force spectroscopy and an optical sideview setup. Integrative Biology: Quantitative Biosciences from Nano to Macro 7, 356363 (2015).

16. Chaudhuri, O., Parekh, S. H., Lam, W. A. & Fletcher, D. A. Combined atomic force microscopy and side-view optical imaging for mechanical studies of cells. Nat Meth 6, 383387 (2009).

17. Yuan, J., Melder, R. J., Jain, R. K. & Munn, L. L. Lateral view ow system for studies of cell adhesion and deformation under ow conditions. BioTechniques 30, 388394 (2001).

18. Tkachenko, E., Gutierrez, E., Ginsberg, M. H. & Groisman, A. An easy to assemble microuidic perfusion device with a magnetic clamp. Lab on a Chip 9, 10851095 (2009).

19. Damiano, E. R., Westheider, J., Tozeren, A. & Ley, K. Variation in the velocity, deformation, and adhesion energy density of leukocytes rolling within venules. Circulation Research 79, 11221130 (1996).

20. Firrell, J. C. & Lipowsky, H. H. Leukocyte margination and deformation in mesenteric venules of rat. American Journal of Physiology -Heart and Circulatory Physiology 256, H1667H1674 (1989).

21. Pospieszalska, M. K., Zarbock, A., Pickard, J. E. & Ley, K. Event Tracking Model of Adhesion Identies Load-bearing Bonds in Rolling Leukocytes. Microcirculation 16, 115130 (2009).

22. Pospieszalska, M. K. & Ley, K. Dynamics of Microvillus Extension and Tether Formation in Rolling Leukocytes. Cellular and Molecular Bioengineering 2, 207217 (2009).

23. Xu, G. & Shao, J.-Y. Human neutrophil surface protrusion under a point load: location independence and viscoelasticity. American Journal of Physiology - Cell Physiology 295, C1434C1444 (2008).

24. Yu, Y. & Shao, J. Y. Simultaneous tether extraction contributes to neutrophil rolling stabilization: a model study. Biophysical Journal 92, 418429 (2007).

25. Shao, J. Y., Ting-Beall, H. P. & Hochmuth, R. M. Static and dynamic lengths of neutrophil microvilli. Proceedings of the National Academy of Sciences of the United States of America 95, 67976802 (1998).

26. Waugh, R. E. Membrane Tethers. In Leukocyte Adhesion (ed. Ley, K.) 324 (Academic Press, Amsterdam, 2009).27. Waugh, R. E. & Bausermann, R. G. Physical measurements of bilayer-skeletal separation forces. Ann. Biomed. Eng. 23, 308321 (1995).

28. Waugh, R. E. & Hochmuth, R. M. Mechanical equilibrium of thick, hollow, liquid membrane cylinders. Biophys. J. 52, 391400 (1987).

29. Pickard, J. E. & Ley, K. Micro-PTV measurement of the uid shear stress acting on adherent leukocytes in vivo. Biophysical Journal 96, 42494259 (2009).

30. Liu, B. & Shao, J. Y. Tangential tether extraction and spontaneous tether retraction of human neutrophils. Biophysical Journal 103, 22572264 (2012).

31. Datar, A., Bornschlogl, T., Bassereau, P., Prost, J. & Pullarkat, P. A. Dynamics of membrane tethers reveal novel aspects of cytoskeleton-membrane interactions in axons. Biophysical Journal 108, 489497 (2015).

32. Dopheide, S. M., Maxwell, M. J. & Jackson, S. P. Shear-dependent tether formation during platelet translocation on von Willebrand factor. Blood 99, 159167 (2002).

33. Coghill, P. A., Kesselhuth, E. K., Shimp, E. A., Khismatullin, D. B. & Schmidtke, D. W. Eects of microuidic channel geometry on leukocyte rolling assays. Biomedical Microdevices 15, 183193 (2013).

34. Angiari, S. Selectin-mediated leukocyte trafficking during the development of autoimmune disease. Autoimmunity Reviews 14, 984995 (2015).

35. Colace, T. V., Tormoen, G. W., McCarty, O. J. & Diamond, S. L. Microuidics and coagulation biology. Annual Review of Biomedical Engineering 15, 283303 (2013).

36. Combariza, M. E., Xinghuo, Y., Nesbitt, W. S., Mitchell, A. & Tovar-Lopez, F. J. Nonlinear Dynamic Modelling of Platelet Aggregation via Microuidic Devices. Biomedical Engineering, IEEE Transactions on 62, 17181727 (2015).

We are very thankful to Sara McArdle for her help with proofreading. This project was supported by NIH P01 HL078784.

K.L., A.G., A.M., E.G. and Z.M. designed the study. E.G. prepared the microuidic devices. A.M. performed the experiments. A.M. and Z.M. set up the microscopy and did the data analysis. K.L. and A.M. wrote the manuscript. A.G., E.G. and Z.M. edited the manuscript.

Supplementary information accompanies this paper at http://www.nature.com/srep

Competing nancial interests: The authors declare no competing nancial interests.

How to cite this article: Marki, A. et al. Microuidics-based side view ow chamber reveals tether-to-sling transition in rolling neutrophils. Sci. Rep. 6, 28870; doi: 10.1038/srep28870 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

SCIENTIFIC REPORTS

8