2014 _ kim _ jbn

Delivered by Publishing Technology to: Sang Ho JoIP: 165.194.103.14 On: Mon, 10 Feb 2014 05:23:15

Copyright: American Scientific Publishers

Copyright 2014 American Scientic PublishersAll rights reservedPrinted in the United States of America

ArticleJournal of

Biomedical NanotechnologyVol. 10, 10301040, 2014

www.aspbs.com/jbn

Filopodial Morphology Correlates to the CaptureEfciency of Primary T-Cells on Nanohole Arrays

Dong-Joo Kim2 , Gil-Sung Kim2 , Jin-Kyeong Seol2, Jung-Hwan Hyung2, No-Won Park1,Mi-Ri Lee2, Myung Kyu Lee4, Rong Fan35, and Sang-Kwon Lee11Department of Physics, Chung-Ang University, Seoul 156-756, Republic of Korea2Basic Research Laboratory (BRL), Department of Semiconductor Science and Technology, Chonbuk National University,Jeonju, 561-756, Republic of Korea3Department of Biomedical Engineering, Yale University, New Haven, CT 06511, USA4Bionanotechnology Research Center, Korea Research Institute of Bioscience and Biotechnology (KRIBB),Daejeon, 305-806, Republic of Korea5Yale Comprehensive Cancer Center, New Haven, CT 06520, USA

Nanostructured surfaces emerge as a new class of material for capture and separation of cell populations includingprimary immune cells and disseminating rare tumor cells, but the underlying mechanism remains elusive. Although ithas been speculated that nanoscale topological structures on cell surface are involved in the cell capture process, thereare no studies that systematically analyze the relation between cell surface structures and the capture efciency. Herewe report on the rst mechanistic study by quantifying the morphological parameters of cell surface nanoprotrusions,including lopodia, lamellipodia, and microvilli in the early stage of cell capture (



Kim et al. Filopodial Morphology Correlates to the Capture Efciency of Primary T-Cells on Nanohole Arrays

nanometer-scale topography not only inuences diversecell behavior such as cell adhesion, motility, prolifera-tion and differentiation,1618 but represents a new routeto capture and separate rare cell populations.1924 Theseinteresting ndings have motivated us to develop a novelnano-platform for separating CD4 T-lymphocytes fromprimary mouse splenocytes via high afnity streptavidin(STR)-biotin conjugation with immuno-functionalized sil-icon nanowire arrays.19 Despite the remarkable improve-ment in the capture efciency using nanostructure-basedcell capture devices, the underlying mechanism that allowsnanostructured-surfaces to achieve signicantly enhancedcell separation efciency as compared to planar substrateremains elusive.Here we report on the study of how the morphologies

of lopodia or microspikes, which are needle-like actin-rich protrusions from the cell surface, are correlated tothe cell capture efciency using a periodic 2D nanoholearray model surface. Filopodia are involved in a wide vari-ety of functions including absorption, secretion, cellularadhesion, and mechnotransduction.25 Because cell captureis a rapid process and the cells captured by the immobi-lized antibodies can interact with the substrate surface andrest/spread quickly within 20 min, we suspect the earlystage cell-substrate interaction plays a critical role in thecell capture process. We prepared a set of nanohole arrays(NHAs) with the hole size ranging from 140 to 550 nm andobserved that the nanostructures indeed directs the forma-tion of lopodia (100300 nm in diameter)26 in the caseof CD4+ T-lymphocytes in contact with biofunctionalizedNHA substrates that further correlates to the efciencyof CD4+ T-cell separation. While cross-sectional imagingindicates that cell microvilli, which were previously sus-pected to be playing a crucial role in enhancing cell cap-ture efciency, were not observed at the interface betweencell and nanostructured surface. Although this may notexclude the role of microvilli in the process of initialcontact between immune T-cells and the nanostructures,our results suggest that the subsequent cell spreading andactin-rich lopodia formation is a more favorable mech-anism that leads to signicant enhancement of rare cellcapture efciency.

MATERIALS AND METHODSMaterialsColloidal polystyrene suspensions (200, 300, 430, and750 nm in diameter) were purchased from Thermo Scien-tic (Fremont, CA, USA). Cr etchant (CR-7) was suppliedby Cyantek Corporation (Fremont, CA, USA). N -methyl-2-pyrrolidone, (3-aminopropyl)-triethoxysilane (APTES),glutaraldehyde (GA), streptavidin (STR), and osmiumtetroxide were purchased for Sigma-Aldrich (St. Louis,MO, USA). C57BL/6/mice were supplied by Nara-Biotech (Seoul, Republic of Korea). Biotinylated anti-CD4mAb (clone GK 1.5), FITC (518 nm emission)-labeled

mAb-CD3, PE (575 nm emission)-labeled mAb-CD4(clone: RM4-4), and PerCP (690 nm emission)-labeledmAb CD19, and uorescence dye labeled-CD3, CD4,CD8, and CD19-mAbs were purchased from eBioscienceInc. (San Diego, CA, USA). All other chemicals were ofanalytical grade.

Nanohole Array (NHA) FabricationTo produce the NHA, the liquid state of colloidalpolystyrene nanoparticles (PS NPs, 200, 300, 430, and750 nm in diameter) monolayer was rst carefullydeposited on the quartz (QZ) substrate using a modi-ed self-assembly technique (Fig. 1(A)) we developedpreviously.22 To create space between the spin-coated PSNPs, their sizes were rst reduced by O2 plasma (O2/Ar=35/10 sccm, RF power of 100 W and bias power of 50 W)for 10 s; 200 nm PS, 23 s; 300 nm PS, 35 s; 430 nm

Figure 1. (A) Schematic diagram of nanohole array (NHA)fabrication using polystyrene (PS) nanoparticles on a pla-nar quartz substrate. ((B)(E)) Scanning electron microscope(SEM) images of NHAs (140, 200, 270, and 550 nm in diameter)prepared with four different feature sizes of PS nanoparticles(200, 300, 430, and 750 nm in diameter). Scale bar is 0.5 m.Insets show the enlarged images of each NHA. These guresalso represent the cross-section view images of each NHAshown in the lower part of each SEM image. (F) Schematic dia-gram showing sequent surface functionalization with APTES,GA, and STR on NHA substrates, (G) showing cell suspension(30 l), which is containing CD4, CD8, B, NK, and NKT-cells,loading process onto NHA substrates, and (H) revealing thewashing process to remove unbound T-cells from the NHAsubstrates using a 3D-rocker.

J. Biomed. Nanotechnol. 10, 10301040, 2014 1031



Filopodial Morphology Correlates to the Capture Efciency of Primary T-Cells on Nanohole Arrays Kim et al.

PS, and 60 s; 750 nm PS, respectively, to control thediameter and spacing for preparing the nanohole arraysperforated into continuous metal lm. To fabricate thenanohole pattern, the reactive ion etching (RIE, SoronaInc., Pyeongtaek, Republic of Korea) process was thenperformed for 40 s (CF4/Ar = 40/5 sccm, RF power of100 W, and bias power of 50 W) after 25-nm Cr metalonto the coated surface of PS NPs, and the PS NPs weresubsequently removed by unltrosonication in N -methyl-2-pyrrolidone. Finally, the Cr metal layer was removed viaa lift-off process using Cr etchant (Fig. 1(A)). The typi-cal diameters/widths of the resultant nanoholes were about140/60, 200/100, 270/160, and 550/200 nm with depthsof 260, 270, 320, an 330 nm, respectively as shown inFigures 1(B)(E). The period of hole-arrays was deter-mined by the initial period of hcp nanospheres arrays andthe diameter of holes could be easily tuned by the O2 etch-ing time. Moreover, our proposed method provides a ex-ible and versatile route to the fabrication of QNP arrayson a at QZ substrate, with potential application in optics,electronics, sensing, and as building blocks for more com-plex nanostructures.

NHA Surface FunctionalizationNHA substrates (7 mm 7 mm) were carefully cleanedwith H2O2:H2SO4 (1:1) for 10 min to remove all ofthe organic materials and impurities on the surface. Wethen washed the substrates using a three-step clean-ing process (acetone, isopropyl alcohol, and distilledwater) and dried them with air. The NHA surface wastreated with O2 plasma for 20 s to confer the hydroxylgroups on the NHA surface after piranha cleaning process(96%H2SO4:30%H2O2 = 11) for 10 min. Next, the sur-face was applied by a three-step surface functionalizationprocess using 1% (v/v) (3-aminopropyl)-triethoxysilane(APTES) in ethanol for 30 min at room temperature,12.5% (v/v) glutaraldehyde (GA) in distilled water for4 hrs on a 3D-rocker, and 50 g/mL streptavidin (STR) inphosphate buffered saline (PBS) overnight in an incubator(37 C, 5%CO2) as shown in Figure 1(F).

Cell PreparationThe CD4 T-lymphocytes to be separated were mouseCD4+ T-cells from whole mouse splenocytes, which con-tain CD4+ T, CD8+ T, natural killer (NK), natural killer T(NKT), and B-cells. These splenocytes were prepared fromthe spleens of C57BL/6/mice as described previously.19

A certain quantity of cells (105 cells/mL) was countedusing a conventional hemocytometer (Hausser Scientic,USA). Prior to loading the cell suspension in the culturemedium, the cell population with a nal volume of 30 lwas rst reacted with biotinylated anti-CD4 mAb and incu-bated at 4 C for 20 min (Fig. 1(G)). Following incubationfor 20 min with STR-conjugated NHA substrates at 4 C,unbound cells were removed by rinsing with phosphate

buffered saline (PBS), while separated CD4+ T cells couldbind to STR-NHA surfaces due to adhesion enabled bythe STR-biotin interaction. This process was repeated threetimes for 10 min on a 3-D rocker to completely removenon-specically unbound cells from the NHA substrates(Fig. 1(H)). Remnant of unbound cells were then collectedand transferred to the tube for enumerating the populationby standard ow cytometry (Becton Dickinson, NJ, USA).

Fluorescence Activated CellSorter (FACS) AnalysisTo enumerate the remnant of unbound cell population,FACS analysis was performed. First, the non-specicallyunbound T-lymphocytes after STR-NHA separation plat-form were collected by FACS falcon round-bottom tubes(eBioscience, USA). The collected T-lymphocytes werethen stained by FITC (518 nm emission)-labeled mAb-CD3, PE (575 nm emission)-labeled mAb-CD4, andPerCP (690 nm emission)-labeled mAb CD19 at 4 Cfor 20 min. After reaction, the stained T-lymphocyteswere washed for removing the unbounded antibody usingcentrifuge (1500 rpm, 5 min, twice) and then xed by4% paraformaldehyde in PBS for 15 min. Finally thexed cell suspension were transferred to the ow cytome-ter (FACS, Calibur, Beckton Dickinson, USA) and ana-lyzed using CellQuest Pro software (BD Bioscience,USA) for the separation efciency of the targeting CD4+

T-lymphocytes.

Surface-Bound T-Cells Preparation UsingScanning Electron Microscopy (SEM) AnalysisFor the FE-SEM analysis, a solution of the cells conju-gated with biotin-conjugated CD4 mAbs in RPMI-1640(500 mL, Invitrogen, NY, USA) was pipetted onto STR-functionalized NHAs (140, 200, 270, and 550 nm in diam-eter, 7 mm7 mm) and STR-planar glass substrates withcell populations 105 cells/mL. After incubation at 37 Cand 5% CO2 for 20 min, the separated and immobilizedCD4+ T-cells on STR-functionalized NHAs and planarglass substrates were rst xed with 4% GA in the refrig-erator for 2 h, followed by a post-xing in 1% osmiumtetroxide for 2 h. The captured T-cells on STR-conjugatedsubstrates were then dehydrated by successive immersionin 25%, 50%, 75%, 95% and 100% ethanol for 5 min at4 C and followed by nal dehydration with 100% ethanoltwice for 10 min at 4 C. Final dehydration needed toprocess twice. Dehydrated T-cells were then frozen for3 h at 80 C. Subsequently, the cells with substrateswere slowly dried for 24 h using a vacuum desiccator.Subsequently, the cells with substrates were slowly driedunder vacuum for 24 h. Once dried, the surface-boundT-cells were then sputter-coated with a layer of platinum(56 nm) to prepare conductive samples before perform-ing the FE-SEM measurement. Quantitative characteriza-tion of the cellular morphologies including lopodia and

1032 J. Biomed. Nanotechnol. 10, 10301040, 2014




lamellipodia formation (length, width, and number of theprotruded lopodia etc.) for the surface-bound T-cells onSTR-functionalized NHAs and planar glass was performedon the FE-SEM images.

RESULTS AND DISCUSSIONT-cell Capture Efciency on STR-NHAsFluorescence activated cell sorter (FACS) revealed thatwhole mouse splenocytes contained 42.7% CD4+T lymphocytes (CD3+/CD4+ and 57.3% non-CD4+ T-cells (Fig. 2(A)). Each type of cells (e.g., CD4,CD8-T, B, NK, and NKT cells) from the suspension ofwhole mouse splenocytes was quantied using FACSmeasurement with different surface markers (uores-cence dye labeled-CD3, CD4, CD8, and CD19-mAbs).After completion of the cell separation process via aSTR-NHA platform, the percentages of CD4+ T-cells in

Figure 2. (A) Flow cytometric (FACS) analysis results of CD4+ T-cells using cell suspensions, containing B, NK, NKT, CD4+

T and CD8+ T-cells after binding to STR-conjugated NHA substrates ranging of 140 to 550 nm in diameter. (B) Average cellseparation efciencies (top bar-graph) for four different sizes of NHAs with the summary (bottom table). The cell separationefciency evaluated from FACS is dened by the percentage of cells captured (shown in right-part of (A)) to cells initially loaded(shown in left-part of (A)), for different cell separation STR-functionalized NHAs (140, 200, 270, and 550 nm in diameter). Forthe comparison, the FACS results of STR-functionalized planar glass wafer were also included. (C) Florescence images and (D)purity distribution of CD4+ T-cells for stained with 4,6-diamidino-2-phenylindole and phycoerythrin using three different NHAsamples (140, 200, and 270 nm in diameter). For the comparison, the results from control sample are also presented in (C) and(D). In (B) and (D), the error bars represent the standard error of mean from four repeats (n = 4).

the cell suspension with four different NHAs (of 140,200, 270, and 550 nm in diameter) decreased to 2.59,1.32, 3.10, and 3.99%, respectively as shown inFigure 2(A). The cell-capture efciency, which is denedas the percentage of the target CD4+ T-cells successfullycaptured to the total number of cells in the cell suspension(105 cells) initially loaded and analyzed by FACS, isplotted for all four different NHAs in Figure 2(B) (topbar-graph) and also summarized in the table of Figure 2(B)(bottom table). For comparison, the STR-functionalizedplanar glass substrates as control samples were also testedand the results are shown in Figure 2(B) (top-right ofbar graph). The result shows that all STR-functionalizedNHA substrates (92.7%) show excellent performance(a factor of 1.4) in cell capture efciency as comparedto the control samples (65.2%), which is consistentwith previous studies.1922 FACS results demonstrated thathigh yield separation of CD4 T-cells was achieved with





all of the STR-functionalized NHAs, and the results werecomparable to those from other nanostructure-based cellcapture experiments previously reported.1922 The hypoth-esis why signicantly elevated cell capture efciency canbe achieved on STR-nanostructured surfaces as comparedto the planar glass substrates is the following. First, it islikely due to the relatively higher contact area betweenthe cells and the solid surfaces compared to the planarglass substrates as we reported previously.22 Second, itwas speculated to be associated with the adhesive andtraction force of cells on the substrate, which is uniquein the condition of higher degree of dimensionality inSTR-NHA surface compared to the planar substrates. Ourprevious results showed that the surface-bound T-cells onSTR-nanostructures have the higher traction force (e.g.,adhesion force) due to the three-dimensional (3-D) surfaceaccessibility (e.g., nanohole or nanopillar etc.), whilethe cells on planar substrates exhibits a lower adhesionforce at the same stage, corresponding to 2-D behavior.27

Consequently, it results in elevated cell capture efciencyon the STR-nanostructures with enhanced cell adhesionforce occurring at the similar stages (Fig. 2(B)).

Purity of Captured T-Cells on STR-NHAsFurthermore, to investigate the purity of the capturedT-cells on NHAs and also to assess cell integrity, thecaptured CD4 T-cells were stained by phycoerythrin (PE)-conjugated anti-CD4 mAb and 4-6-diamidino-2 phenylin-dole (DAPI, DB Bioscience, USA). DAPI (blue-350 nm)is a nuclear dye that stains for cells with intact nuclei. Thestained cells were subsequently enumerated for the CD4+

T-cells (PE+/DAPI+) out of total cells (DAPI+) boundto the STR-functionalized NHA substrates by uorescencemicroscopy (FM) analysis. Figure 2(C) shows uorescenceimages of both PE and DAPI-stained CD4+ T-lymphocytesfor the STR-conjugated NHAs and planar glass (controlsample). As shown in Figures 2(C), (D), STR-conjugatedNHAs (140, 200, and 270 nm in diameter) were found tohave a high purity (80633%), which is comparable tothat of the control samples (planar glass, 77707%).

Cellular Morphology of Captured T-CellsTo quantify the morphological properties of the capturedCD4+ T-lymphocytes bound to STR-conjugated NHA sub-strates, FE-SEM analysis using a cell freezing techniquewere performed. Quantitative characterization of the cel-lular morphologies (number of extended lopodia andprotruded length and width of lopodia etc.) on bothSTR-functionalized NHA substrates and planar glass (con-trol sample) was performed via analyzing the FE-SEMimages. Figure 3(A) shows representative FE-SEM imagesof CD4+ T-lymphocytes bound on the surfaces of fourdifferent-sized NHAs. It was observed that the capturedCD4+ T-cells possessed different forms of lopodia (0.10.3 m)26 on the four NHAs with different nanohole sizes.

The formation of the at lamellipodia extending fromthe captured T-cells was observed only on the surface ofNHAs exceeding 270 nm, as shown in Figure 3(A).To further examine the interface and binding propertiesbetween the captured CD4 T-cells and NHA substrates,gallium ion (Ga+) milling of the captured CD4 T-cellswas performed using focused ion beam (FIB). Figure 3(B)shows the cross-sectional images of the captured CD4-Tcells on four different NHAs, indicating that the capturedCD4 T-cells formed tight binding and adhesion on STR-functionalized NHAs. However the insertion of microvilliinto nanoholes, a mechanism speculated to account forenhanced cell capture efciency, are surprisingly rare.Figure 3(C) shows the cell size, which was calculatedwithout considering the area of lopodia and lamellipo-dia from the cells (n = 117) captured on a 270 nm NHAsubstrate, displays a Gaussian distribution, implying thatnon-specically bound cells were also counted. Accordingto cell purity measurement (Figs. 2(C), (D)) of cell sus-pension, 80.6% of cells bound on the NHAs substratescould be CD4+ T-cells, indicating that approximately 80%of the bound cells (2.84 m in cell-diameter) in thesize distribution graph could be the right target cellsCD4+ T-lymphocytes. To ensure that the evaluation of thelopodia morphology including the width, length, featuresize and the number of the lopodial laments per cell inthe early stage of cell adhesion (in our case, only 20 minincubation) is statistically sound, we quantied more than55 cells, which were 80% of the total bound cells wecounted (n=70) (Figs. 3(D), (E) and 4(A)(D)). To jus-tify the signicance of our correlation results, p valueswere calculated with neighboring column data. As shownin Figs. 3(D), (E), the protruded lopodia width for 140-nm NHAs (80.2 nm in width) exhibits similar trend insize to that of the 200-nm NHAs (83.1 nm in width)resulting in statically insignicant difference (Fig. 3(D)and Table I). With further increasing the diameter of NHAsfrom 270 to 550 nm, lopodia protruding from the T-cellswere observed to increase in width (P < 00001, 126140 nm in width, Figure 3(E) and Table I). This cor-relation develops even at the very early stages of adhe-sion (20 min incubation). Hence, this linear correlation(except below




Figure 3. (A) Scanning electron microscope (SEM) images of CD4+ T-cells bound on four different sizes of NHA substrates (low-and high-magnication top, tilt, and enlarged tilt view images). (B) Cross-sectional SEM images of surface-bound CD4+ T-cells onfour different NHAs (140, 200, 270, and 550 nm in diameter) with enlarged images in selective area marked 1, 2, 3, and 4 in rstcolumn of (B). The samples were prepared by Ga+ ion milling with a focused ion beam (FIB). All of the CD4+ T-cells on the NHAsare highlighted in yellow for easy differentiation. (C) Cell size distribution (only cell-body excluding lopodia and lamellipodia)of captured cells (n = 117) randomly bound on streptavidin (STR)-functionalized NHA substrate (270 nm in diameter). Scale baris 1 m. (D) Filopodia width distribution of CD4+ cells bound on the four different STR-functionalized NHAs substrates after only20 min incubation at 4 C. P values of < 00001 () considered statically signicant. An insignicant statistical difference isrepresented as NS. (E) Selected lopodia width distribution in which only 80% of lopodia width taken from (F) of the CD4+T cells on NHAs substrates, indicating that the lopodia width was not changed between the 140 and 200 nm in diameter andlinearly increased with increasing of NHA diameters up to 550 nm.





Table I. The summary of evaluated lopodia morphology (number, length, and width of lopodia) of CD4+ T-cells bound on thedifferent nanohole arrays (NHA) using SEM analysis. For the comparison, the results of control samples (planar glass substrates)were also included.

Number of lopodia Filopodia length (nm) Filopodia width (nm)

Samples Average STDEV Average STDEV Average STDEV

Control 944 339 5475 2478 896 360NHA140 nm 639 194 6003 2304 802 129200 nm 667 201 6168 2590 831 392270 nm 511 160 6759 5713 1259 445550 nm 342 148 6734 2618 1412 521

Notes: STDEV: Standard deviation; Control: STR-functionalized planar glass samples without nanostructure; NHA: Nanohole arrays.

versus nanohole size shown in Figures 3(D), (E). Anotherpossible explanation on the elevation in lopodia widthwith increasing the NHA diameters (200550 nm) is theincreasing STR-conjugated surface width (100270 nm).The amine containing in the STR-conjugated NHAs pro-mote the actin extension to guide the development oflopodial laments in the initial stage of the incubation.33

Subsequently, the lopodia of the captured T-cells startto deliver more actin-rich lament along the shafts ofthe initial lopodia for covering different widths of STR-NHAs.33 Therefore, these results suggest that the lopo-dia of CD4+ T-cells interact with the NHA substrate viaan initial high-afnity STR-biotin conjugation19 followedby guided extension of lopodia on the nanostructuresof the NHAs (Fig. 3(E)). Furthermore, we also quanti-ed other morphological parameters such as the protru-sion length of lopodia and the average number of thelopodial laments per cell as summarized in Table Iand Figures 4(A)(D). In addition, the cell-capture ef-ciency (n = 4) is plotted for the four different NHAsin Figure 4(E). In Figures 4(A), (B), no statically sig-nicant difference (NS, not-signicant) was observed inthe plot showing the lopodial protrusion length of thecaptured CD4 T-cells versus nanohole dimensions, indi-cating that the length of the protruded lopodia is rela-tively independent of the dimensions of the NHAs in theearly stage of cell capture (




Figure 4. (A) and (B) Protruded length of lopodia, (C) and (D) the average number of extended lopodia per cell, and (E) cellseparation yield (efciency) of mouse primary CD4+ T-lymphocytes for four different sizes of STR-conjugated NHA substrates(140, 200, 270, and 550 nm). The protruded lopodia length and number of lopodia per cell shown in (B) and (D) were takenfrom 80% of CD4+ T-lymphocytes of all bound cells shown in (A) and (C). To ensure that the lopodia evaluation are staticallysignicant, at least 5262 cells, which are 80% of the bound cells (n = 7080) on NHAs substrates, were used for the evaluationof lopodia morphology as indicated in (C). In (E), the error bars represent the standard error of mean from four repeats (n = 4).P values of < 00001 () considered statically signicant. An insignicant statistical difference is represented as NS.

the STR-NHA substrates. This observation, which webelieve is novel, can be explained by contact guidanceand cell-adhesion force. When the T cells interact withSTR-conjugated NHA surfaces, they develop lopodia tosense the surface, and then preferentially adhere to theregion with features of similar dimensions. Afterwards the

lopodia of the captured T-cells start to extend rapidly onthe surface even within a short incubation time (20 min).

Filopodia Contact BehaviorThe length of lopodia including lamellipodia (Fig. 4(A))protruding from primary mouse T-lymphocytes on the





(A)

(B) (C)

(D)

Figure 5. (A) Fluorescence images and cell size histograms of live primary CD4+ T-lymphocytes bound on four different NHAsubstrates (140, 200, 270, and 550 nm in diameters) using confocal microscopy. (B)(C) Cell size distribution (histogram) of theCD4+ T-lymphocytes (n= 250300) on four different NHAs. The results shown in (C) were taken from 80% of CD4+ T-lymphocytesof all bound cells shown in (B). The actin and nuclei were stained by rhodamine-conjugated phalloidin (Life technologies Co.,USA) and 4,6-diamidino-2-phenylindole dihydrochloride (DAPI, BD Bioscience, USA), respectively. P values were calculated withneighboring column data (P < 00282, P < 00061, and NS, not signicant). (D) The summary of evaluated lopodia morphol-ogy (number, length, and width of lopodia) of CD4+ T-cells bound on the different nanohole arrays (NHA) using uorescencemeasurements. For the comparison, the results of control samples (planar glass substrates) were also included.





Figure 6. Fluorescence images of live single CD4+

T-lymphocyte bound on NHA substrate (d = 270 nm) usingconfocal microscopy. (A) Scanned uorescence images takenfrom the top to bottom of the bound CD4 T-lymphocyte,showing the cross-sectional view of the cell. ((B)(D)) and(F) show the enlarged images of selected areas marked 1,2, 3, and 4, respectively, representing clear actin-stainedlopodia shown in (F). (E) Schematic view of the cells onNHA substrate for confocal scanning measurement. The actinand nuclei were stained by rhodamine-conjugated phalloidinand 4,6-diamidino-2-phenylindole dihydrochloride (DAPI),respectively. DAPI is a uorescent dye for labeling DNA.

NHA surface was 0.60 to 0.67 m determined by FE-SEM (Figs. 4(A), (B)), which is relatively longer thanthe intrinsic nanoscale structures (e.g., microvilli) on thesurface of human T cells observed on planar cover-slip glass (0.25 to 0.78 m).35 The cell lopodiaand microvilli are interrelated because the lopodia ormicrospikes start to protrude when the cells contact thesubstrate surface via microvilli and started to spread eventhough their sizes were less than 1 m.25 However,in our study we observed that more specically lopodia-substrate interaction is correlated to cell capture efciencyand the microvilli at the interface of cell-substrate werenot extensively observed. To further conrm the develop-ment of the structural proteins of the actin-rich lopodiafrom surface-bound CD4 T-cells on the NHA substrates,FM analysis of live single CD4+ T-lymphocyte boundon NHA substrate (d = 270 nm) was also performed(Figs. 6(A)(F)). Scanned FM images taken from the topto bottom of the bound CD4 T-lymphocyte clearly exhibitactin-rich lopodial protrusions in the early stage of thecell-substrate interaction (




4. J. W. Murphy, M. R. Hidore, and S. C. Wong, Direct interactionsof human-lymphocytes with the yeast-like organism, cryptococcus-neoformans. J. Clin. Invest. 91, 1553 (1993).

5. D. M. Phillips and T. Xin, Hiv-1 infection of the trophoblast cell-line bewoa study of virus uptake. AIDS Res. Hum. Retrov. 8, 1683(1992).

6. M. C. Mingari, F. Schiavetti, M. Ponte, C. Vitale, E. Maggi,S. Romagnani, J. Demarest, G. Pantaleo, A. S. Fauci, and L. Moretta,Human CD8(+) T lymphocyte subsets that express HLA classI-specic inhibitory receptors represent oligoclonally or monoclon-ally expanded cell populations. Proc. Natl. Acad. Sci. USA 93, 12433(1996).

7. R. Pearcepratt and D. M. Phillips, Studies of adhesion of lym-phocytic cellsimplications for sexual transmission of human-immunodeciency-virus. Biol. Reprod. 48, 431 (1993).

8. K. G. E. Scott, L. C. H. Yu, and A. G. Buret, Role of CD8(+)and CD4(+) T lymphocytes in jejunal mucosal injury during murinegiardiasis. Infect. Immun. 72, 3536 (2004).

9. G. W. Wolkersdorfer, T. Lohmann, C. Marx, S. Schroder, R. Pfeiffer,H. D. Stahl, W. A. Scherbaum, G. P. Chrousos, and S. R. Bornstein,Lymphocytes stimulate dehydroepiandrosterone production throughdirect cellular contact with adrenal zona reticularis cells: A novelmechanism of immune-endocrine interaction. J. Clin. Endocr. Metab.84, 4220 (1999).

10. H. Cantor, E. Simpson, V. L. Sato, C. Garrison, and L. A.Herzenberg, Characterization of subpopulations of T-lymphocytes. 1.separation and functional studies of peripheral T-cells binding dif-ferent amounts of uorescent Anti-Thy 1.2 (Theta) antibody usinga uorescence-activated cell sorter (FACS). Cell Immunol. 15, 180(1975).

11. K. Wang, B. Cometti, and D. Pappas, Isolation and counting of mul-tiple cell types using an afnity separation. Anal. Chim. Acta 601, 1(2007).

12. D. Huh, H. H. Wei, O. D. Kripfgans, J. B. Fowlkes, J. B. Grotberg,and S. Takayama (eds.), Gravity-driven microhydrodynamics-basedcell sorter (microHYCS) for rapid, inexpensive, and efcient cellseparation and size-proling. Proceedings of the 2nd IEEE-EMBSSpecial Topics Conference on Microtechnologies in Medicine andBiology, Madison, WI, USA, May (2002).

13. Z. G. Wu, A. Q. Liu, and K. Hjort, Microuidic con-tinuous particle/cell separation via electroosmotic-ow-tunedhydrodynamic spreading. J. Micromech. Microeng. 17, 1992(2007).

14. S. M. Prattis, D. H. Gebhart, G. Dickson, D. J. Watt, and J. N.Kornegay, Magnetic afnity cell sorting (Macs) separation andow cytometric characterization of neural cell-adhesion molecule-positive, cultured myogenic cells from normal and dystrophic dogs.Exp. Cell Res. 208, 453 (1993).

15. B. Schmitz, A. Radbruch, T. Kummel, C. Wickenhauser, H. Korb,M. L. Hansmann, J. Thiele, and R. Fischer, Magnetic activated cellsorting (MACS)a new immunomagnetic method for megakary-ocytic cell isolationcomparison of different separation techniques.Eur. J. Haematol. 52, 267 (1994).

16. S. J. Qi, C. Q. Yi, S. L. Ji, C. C. Fong, and M. S. Yang, Cell adhesionand spreading behavior on vertically aligned silicon nanowire array.ACS Appl. Mater. Inter. 1, 30 (2009).

17. K. W. Kwon, S. S. Choi, S. H. Lee, B. Kim, S. N. Lee, M. C.Park, P. Kim, S. Y. Hwang, and K. Y. Suh, Label-free, microuidicseparation and enrichment of human breast cancer cells by adhesiondifference. Lab. Chip 7, 1461 (2007).

18. K. Kulangara and K. W. Leong, Substrate topography shapes cellfunction. Soft Matter. 5, 4072 (2009).

19. S. T. Kim, D. J. Kim, T. J. Kim, D. W. Seo, T. H. Kim, S. Y. Lee,K. Kim, K. M. Lee, and S. K. Lee, Novel streptavidin-functionalizedsilicon nanowire arrays for CD4(+) T lymphocyte separation. NanoLett. 10, 2877 (2010).

20. S. T. Wang, H. Wang, J. Jiao, K. J. Chen, G. E. Owens, K. I.Kamei, J. Sun, D. J. Sherman, C. P. Behrenbruch, H. Wu, and H. R.Tseng, Three-dimensional nanostructured substrates toward efcientcapture of circulating tumor cells. Angew. Chem. Int. Edit. 48, 8970(2009).

21. S. T. Wang, K. Liu, J. A. Liu, Z. T. F. Yu, X. W. Xu, L. B. Zhao,T. Lee, E. K. Lee, J. Reiss, Y. K. Lee, L. W. K. Chung, J. T. Huang,M. Rettig, D. Seligson, K. N. Duraiswamy, C. K. F. Shen, and H. R.Tseng, Highly efcient capture of circulating tumor cells by usingnanostructured silicon substrates with integrated chaotic micromix-ers. Angew. Chem. Int. Edit. 50, 3084 (2011).

22. D. J. Kim, J. K. Seol, Y. Wu, S. Ji, G. S. Kim, J. H. Hyung, S. Y. Lee,H. Lim, R. Fan, and S. K. Lee, A quartz nanopillar hemocytometerfor high-yield separation and counting of CD4(+) T lymphocytes.Nanoscale 4, 2500 (2012).

23. W. Q. Chen, S. N. Weng, F. Zhang, S. Allen, X. Li, L. W. Bao, R. H.W. Lam, J. A. Macoska, S. D. Merajver, and J. P. Fu, Nanoroughenedsurfaces for efcient capture of circulating tumor cells without usingcapture antibodies. ACS Nano 7, 566 (2013).

24. W. Q. Chen, Y. B. Sun, and J. P. Fu, Microfabricated nanotopologicalsurfaces for study of adhesion-dependent cell Mechanosensitivity.Small 9, 81 (2013).

25. P. Otteskog and K. G. Sundqvist, Characterization of the spreadingprocess in human lymphocytes-T. Exp. Cell Res. 149, 201 (1983).

26. P. K. Mattila and P. Lappalainen, Filopodia: Molecular architectureand cellular functions. Nat. Rev. Mol. Cell Bio. 9, 446 (2008).

27. D. J. Kim, J. K. Seol, G. H. Lee, G. S. Kim, and S. K. Lee, Celladhesion and migration on nanopatterned substrates and their effectson cell-capture yield. Nanotechnology 23, 395102 (2012).

28. C. J. Bettinger, R. Langer, and J. T. Borenstein, Engineering sub-strate topography at the micro- and nanoscale to control cell func-tion. Angew. Chem. Int. Edit. 48, 5406 (2009).

29. M. J. Dalby, Topographically induced direct cell mechanotransduc-tion. Med. Eng. Phys. 27, 730 (2005).

30. M. J. Dalby, N. Gadegaard, M. O. Riehle, C. D. W. Wilkinson,and A. S. G. Curtis, Investigating lopodia sensing using arrays ofdened nano-pits down to 35 nm diameter in size. Int. J. Biochem.Cell B 36, 2005 (2004).

31. A. Hart, N. Gadegaard, C. D. W. Wilkinson, R. O. C. Oreffo, andM. J. Dalby, Osteoprogenitor response to low-adhesion nanotopogra-phies originally fabricated by electron beam lithography. J. Mater.Sci.-Mater. M 18, 1211 (2007).

32. M. J. Dalby, M. O. Riehle, H. J. H. Johnstone, S. Affrossman,and A. S. G. Curtis, Polymer-demixed nanotopography: Controlof broblast spreading and proliferation. Tissue Eng. 8, 1099(2002).

33. U. Hersel, C. Dahmen, and H. Kessler, RGD modied polymers:Biomaterials for stimulated cell adhesion and beyond. Biomaterials24, 4385 (2003).

34. F. Haq, V. Anandan, C. Keith, and G. Zhang, Neurite devel-opment in PC12 cells cultured on nanopillars and nanoporeswith sizes comparable with lopodia. Int. J. Nanomed. 2, 107(2007).

35. S. Majstoravich, J. Y. Zhang, S. Nicholson-Dykstra, S. Linder,W. Friedrich, K. A. Siminovitch, and H. N. Higgs, Lymphocytemicrovilli are dynamic, actin-dependent structures that do not requireWiskott-Aldrich syndrome protein (WASp) for their morphology.Blood 104, 1396 (2004).


2014 _ kim _ jbn

Documents