Atomic Force Microscopy: HighResolution Dynamic Imaging ofCellular and Molecular Structurein Health and DiseaseDOUGLAS J. TAATJES,1* ANTHONY S. QUINN,1 JACOB H. RAND,2 AND BHANU P. JENA3

1Department of Pathology and Microscopy Imaging Center, College of Medicine, University of Vermont, Burlington, Vermont2Department of Pathology, Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York3Department of Physiology, Wayne State University School of Medicine, Detroit, Michigan

The atomic force microscope (AFM), invented in 1986, and a member of the scanning probe family of microscopes, offers theunprecedented ability to image biological samples unfixed and in a hydrated environment at high resolution. This opens the possibility toinvestigate biological mechanisms temporally in a heretofore unattainable resolution. We have used AFM to investigate: (1) fundamentalissues in cell biology (secretion) and, (2) the pathological basis of a human thrombotic disease, the antiphospholipid syndrome (APS). Thesestudies have incorporated the imaging of live cells at nanometer resolution, leading to discovery of the “porosome,” the universal secretoryportal in cells, and a molecular understanding of membrane fusion from imaging the interaction and assembly of proteins between opposinglipid membranes. Similarly, the development of an in vitro simulacrum for investigating the molecular interactions between proteins andlipids has helped define an etiological explanation for APS. The prime importance of AFM in the success of these investigations will bepresented in this manuscript, as well as a discussion of the limitations of this technique for the study of biomedical samples.J. Cell. Physiol. 228: 1949–1955, 2013. � 2013 Wiley Periodicals, Inc.

The atomic force microscope (AFM) is a member of thescanning probe family of microscopies. The first member of thisfamily, the scanning tunneling microscope (STM) was inventedby Binnig et al. (1982). This instrument was capable of resolvingindividual atoms on hard, atomically flat surfaces. The AFMwasdeveloped several years later (Binnig et al., 1986), and as itsname implies was envisioned as an instrument able to measuresample forces at the atomic level. In contrast to the STM, whichfunctions through scanning a metal tip over a conductingsurface through a tunnel junction, the AFM, as originallydescribed by Binnig et al. (1986) is “a combination of theprinciples of the scanning tunneling microscope and the stylusprofilometer.” Moreover, since the AFM functions throughscanning a sharpened silicon probe over a surface whilemaintaining a constant force between tip and sample, it avoidsthe necessity of employing conducting samples, allowing theimaging of insulating samples (i.e., biological samples) at highresolution. In 1987 (1 year after the published introduction ofthe AFM), B.N.J. Persson posed the following question as thetitle of a published manuscript: “The atomic force microscope:can it be used to study biological molecules?” (Persson, 1987).Twenty-five years later the answer to this question has provento be a resounding “yes.” In the biological realm, consideringresolution, the AFM now sits squarely between super-resolution fluorescence microscopy and transmission electronmicroscopy. However, AFM enjoys a critical advantage overthese other high-resolution imaging modalities in the area ofsample preparation. In general, for conventional transmissionelectron microscopy, samples must be chemically or physicallyfixed (and embedded and sectioned in the case of thickersamples such as cells and tissues) and imaged under highvacuum conditions with operating voltages typically in the 60–120 kV range, prohibiting the viewing of live samples at highresolution (the recently described specialized environmentalphase-plate transmission electron microscope technique mayoffer an interesting approach to overcome the limitationsposed by conventional transmission electron microscopy for

imaging hydrated samples; see Inayoshi et al., 2012 for details).Likewise, super-resolution light microscopy techniques aretypically used on chemically fixed samples as well (reviewed inHuang et al., 2009; Schermelleh et al., 2010; Galbraith andGalbraith, 2011), though live cell applications have beenreported and are becoming more common for someapplications (Shroff et al., 2008; Jones et al., 2011; Shimet al., 2012). However, super-resolution microscopy stillrequires samples labeled with fluorescent probes forvisualization (Huang et al., 2009; Schermelleh et al., 2010;Dempsey et al., 2011; Galbraith andGalbraith, 2011). In strikingcontrast, high-resolution atomic force microscopy may beachieved with minimal sample preparation on unfixed,unlabeled specimens imaged in physiological fluid underambient pressure and temperature (Schneider et al., 1997;Muller and Engel, 1999; Montigny et al., 2006). Moreover, the

The authors declare no conflict of interest in the work presented inthis manuscript.

Additional supporting information may be found in the onlineversion of this article.

Contract grant sponsor: NIH;Contract grant numbers: R01 HL-61331, DK56212, NS39918.Contract grant sponsor: NCRR;Contract grant number: S10RR025498.

*Correspondence to: Douglas J. Taatjes, Department of Pathology,College of Medicine, University of Vermont, 89 Beaumont Avenue,Burlington, VT 05405. E-mail: douglas.taatjes@uvm.edu

Manuscript Received: 1 March 2013Manuscript Accepted: 5 March 2013

Accepted manuscript online in Wiley Online Library(wileyonlinelibrary.com): 22 March 2013.DOI: 10.1002/jcp.24363

PERSPECTIVE 1949J o u r n a l o fJ o u r n a l o f

CellularPhysiologyCellularPhysiology

� 2 0 1 3 W I L E Y P E R I O D I C A L S , I N C .

simple design of the AFM allows for the rapid exchange ofsolutions in the fluid cell, followed by the temporal visualizationof effects afforded by the addition of components to the fluidbathing the sample. These attractive features led us to explorethe efficacy of employing the AFM in questions pertaining tobasic cell biology and the pathobiology of disease. In this “Fromthe Bench” essay, we will describe how fluid-based atomicforce microscopy was used as a high-resolution imagingtechnique to uncover (i) existence of the “porosome,” theuniversal secretory portal in cells (Schneider et al., 1997; Choet al., 2002a,b, 2004, 2008; Jena et al., 2003; Jeremic et al., 2003;Siksou et al., 2007; Zhao et al., 2010; Drescher et al., 2011;Elshennawy, 2011; Craciun and Barbu-Tudoran, 2012;Japaridze et al., 2012; Lee et al., 2012; Okuneva et al., 2012), andto establish a molecular understanding of membrane fusionrequiring the assembly of proteins between opposing lipidmembranes (Cho et al., 2002c, 2005, 2009, 2011; Jeremicet al., 2004a,b, 2006; Cho and Jena, 2007; Shin et al., 2010); and(ii) a molecular mechanism to explain the etiology of theantiphospholipid syndrome, a human thrombotic disorder(Rand, 2002; Rand et al., 2008b, 2010b).

Method

Although many implementations of the technology exist, in ageneral sense, AFM can be thought of as a tactile sensing systemin which a fine probe usually composed of silicon nitride orsilicon, and available in a variety of geometries, is affixed to theend of a flexible cantilever, which is in turn connected to apiezoelectric device (Montigny et al., 2006; Quinn et al., 2012a).Insomuch as three principle AFM imaging modes have beendeveloped (contact, non-contact, and tapping (AC mode)),most biomolecule imaging has employed AC (alternatecontact) mode, as this mode is believed to effect the leastphysical damage on delicate biomedical samples (Fig. 1). As thecantilever is made to oscillate at its resonance frequency,physical contact with the sample surface results in a decrease inthe oscillation amplitude. The decreased cantilever amplitude

is in turn detected by a finely tuned infrared laser which isfocused to reflect off the back of the cantilever and then sensedby a photodiode. Finally, the force engendered between theraster scanning cantilever tip and the sample is maintained at aconstant value by a feedback loop mechanism, resulting inthe generation of a topographic map of the sample surface(Katan and Dekker, 2011; Quinn et al., 2012a). Although ACmode imaging often results in a somewhat poorer resolutioncompared to contact mode, the potential attenuation in sampledamage is usually deemed an acceptable trade-off. Accordingly,in the examples from our lab to follow, AC imaging of samplesin fluid (biological buffers) was our method of choice toapproach molecular imaging of important cell biological/pathobiological questions. The AFM imaging assembly may beused as a stand-alone instrument, or may be affixed onto aspecially designed rigid stage, which in turn is mounted on thebase of an inverted light microscope (Quinn et al., 2012b). Theimmediate advantage of the light microscope component ofthe system for biomedical samples lies in the ability to opticallyimage the sample and thereby choose microscopically areas ofinterest for further AFM examination. The optical microscopemay be outfitted with contrast enhancement features such asphase contrast optics or fluorescence filters to improve theability to distinguish sample features of interest. The entireAFM/optical microscope assembly is then situated on avibration damping platform and housed within anenvironmental isolation station (Supplementary Figs. 1 and 2)to provide an optimal disturbance-free environment necessaryfor high-resolution imaging.

AFM in the discovery of the porosome: The universalsecretory portal in cells

Cellular cargo destined for secretion are packaged and storedin membranous sacs or vesicles, and following a secretorystimulus, they dock and fuse with the cell plasma membraneto release intravesicular contents outside the cell. It waspreviously believed that the secretory vesicle membrane

Fig. 1. Schematic depiction of the principles of AFM (see text for details; modified from Asylum Research Instruments original drawing).

JOURNAL OF CELLULAR PHYSIOLOGY

1950 T A A T J E S E T A L .

completely merges with the cell plasma membrane duringsecretion. However, observation from electron micrographsdemonstrating the accumulation of partially filled secretoryvesicles in cells following secretion contradicted this commonlyheld belief. Likewise, the fact that even single-cell organismsincluding bacteria possess a well-developed secretoryapparatus challenged the veracity of the all-or-none form ofcomplete vesicle merger in mammalian cells. To address thisconundrum, the only definitive approach would be thedirect observation of cell secretion at nanometerresolution in real time in live cells.This dreamwas realized inthe mid 1990s through live cell AFM, when secretory portals,or porosomes, at the cell plasma membrane in live pancreaticacinar cells were first discovered (Schneider et al., 1997).Subsequently, the porosome has been demonstrated to bepresent in every cell type thus far examined, includingendocrine cells and neurons (Fig. 2) (Cho et al., 2002a,b, 2004,2008; Jena et al., 2003; Jeremic et al., 2003; Siksou et al., 2007;Zhao et al., 2010; Drescher et al., 2011; Elshennawy, 2011;Craciun and Barbu-Tudoran, 2012; Japaridze et al., 2012;Lee et al., 2012; Okuneva et al., 2012). These studies havedemonstrated that secretory vesicles “transiently” dock and

fuse at the base of the porosome complex to release theircontents to the outside milieu of the cell, a concept supportedby other investigations of cell secretion and synaptictransmission (Aravanis et al., 2003; Taraska et al., 2003; Thornet al., 2004). The porosome discovery, made possible in largepart through the use of high-resolution, live cell AFM hasestablished a new paradigm in cell secretion.

AFM has unraveled the molecular assembly ofmembrane-associated SNAREs

The fusion of membrane-bound secretory vesicles at theporosome base is mediated by calcium and a specialized setof three soluble N-ethylmaleimide-sensitive factor (NSF)-attachment protein receptors called SNAREs (Malhotraet al., 1988; Trimble et al., 1988; Oyler et al., 1989; Bennettet al., 1992). In neurons for example, targetmembrane proteinsSNAP-25 and syntaxin (collectively called t-SNAREs) presentat the base of porosomes at the presynaptic membrane,together with a synaptic vesicle-associated membrane protein(VAMP) or v-SNARE, form part of the conserved proteincomplex involved in membrane fusion and neurotransmission.In the presence of Ca2þ, AFM studies first demonstrated that t-SNAREs and v-SNARE in opposing membrane bilayers interactand self-assemble in a ring or rosette conformation, to form aconducting channel (Fig. 3) (Cho et al., 2002c). Such self-assembly of t-/v-SNARE rosettes or rings occurs only when therespective SNAREs are membrane-associated (Choet al., 2002c), and the final size of the SNARE ring complex isdependent on the curvature of the opposing lipid membrane(Cho et al., 2005). Other imaging techniques, including electronmicroscopy and electron density maps, together with 3-Dtopography, have further corroborated the AFM imaging

Fig. 2. Porosomes in the exocrine pancreas (A,B) and neurons (C,D). A: An electron micrograph of a single porosome at the apicalplasma membrane (PM) of a pancreatic acinar cell is shown. Notethe porosome membrane (POM, yellow arrowhead) associated withthe membrane of a secretory vesicle (ZGM). A circular ringstructure (blue arrowhead) forms the neck of the porosomecomplex. B: An atomic force micrograph of the apical end of a livepancreatic acinar cell, demonstrating the presence of fourporosomes (one indicated by the yellow arrowhead). Porosomes inthe exocrine pancreas range in size from 100 to 180nm in diameter.C: An electron micrograph of a neuronal porosome (redarrowheads) in association with a synaptic vesicle (SV) at thepresynaptic membrane (Pre-SM) of the nerve terminal, isdemonstrated. Note a central plug in the porosome complex. D: Anatomic force micrograph of a neuronal porosome at the presynapticmembrane in a live cell. Note the central plug (red arrowhead). Theneuronal porosome is an order of magnitude smaller (10–12 nm indiameter) than the porosome in the exocrine pancreas. Compositeof images originally published in: Proc Natl Acad Sci 94:316–321(1997); Biophys J 85:2035–2043 (2003); Cell Biol Int 28:699–708(2004); J Microsc 232:106–111 (2008). �Bhanu Jena.

Fig. 3. AFM micrographs of NSF-ATP induced disassembly ofthe t-/v-SNARE ring complex. Representative AFM micrograph oft-/v-SNARE complexes formed when large (top panel A) or small(bottom panel A) t-/v-SNARE ring complexes are formed due to theinteraction of large and small v-SNARE reconstituted vesiclesinteracting with a t-SNARE reconstituted lipid membrane (ScaleBar¼ 250nm). B: Disassembly of large t-/v-SNARE complex(Scale Bar¼ 250nm). C: High resolution image of an intactt-/v-SNARE ring complex, and one that has been disassembled (D).Images originally published in: J Am Chem Soc 128:26–27 (2006).�Bhanu Jena.

JOURNAL OF CELLULAR PHYSIOLOGY

H I G H R E S O L U T I O N A T O M I C F O R C E M I C R O S C O P Y 1 9 5 1

studies demonstrating the SNARE ring complex (Choet al., 2009, 2011). These structural AFM imaging studies,combined with functional electrophysiological measurementshave greatly advanced our understanding of membrane-directed SNARE ring complex assembly and membrane fusionin cells.

AFM in delivering insights into the pathobiology of thethrombotic antiphospholipid syndrome

The antiphospholipid Syndrome (APS) is a perplexingthrombotic disorder characterized by peripheral thrombosis

and recurrent pregnancy losses accompanied by laboratoryevidence for the presence of autoantibodies in the blood thatrecognize proteins that bind to anionic phospholipids (ratherthan recognizing phospholipids directly, as its misleading namemay suggest) (Rand, 2002). The major antigenic target of theautoantibodies generated in this disorder was identified tobe b2-glycoprotein I (b2GPI), a phospholipid binding proteinthat is a member of the complement control superfamily(de Groot and Urbanus, 2012). One of us (Rand) has positedthat APS results from an antibody/co-factor protein-mediateddisruption of the anticoagulant annexin V (AnxA5) shieldexisting on the surfaces of endothelial cells and placentaltrophoblasts, leading to the unveiling of plasma membraneanionic phospholipids which serve as binding sites forcoagulation enzymes involved in the thrombotic cascade. Datato support this premise came from studies showing (1) asignificant reduction of AnxA5 protein on the surfaces ofplacental villi from APS patients, and on placental villi treatedwith antiphospholipid (aPL) antibodies in vitro (Randet al., 1994); and (2) exposure of both cultured placental villitrophoblasts and HUVECs to aPL antibodies resulted indecreased surface expression of AnxA5 and an acceleratedplasma coagulation at the membrane surface (Rand et al.,1997). What was missing was direct observation ofantibody/cofactor disruption of the AnxA5 protein shield.Such studies would require high-resolution temporal imagingof macromolecular component interactions in a fluid

Fig. 4. AFM images demonstrating the formation of an AnxA5 2Dcrystal lattice on a planar lipid layer. Amplitude image (A) depictsthe erupted (pebbled appearance) PC/PS surface following additionof AnxA5. The inset shows a 10.0mm square area of incompletelycoalesced planar lipid vesicles. The hashed boxed area in the insetis indicated by the arrow in the main figure and represents anucleation site. Amplitude images (B, C, and D) at progressivelyhigher magnifications from the same region as the inset in A showAnxA5 crystal formation from nucleation sites. The open arrows inC indicate the furrows where the forming AnxA5 crystal (AnxA5)has not yet completely covered the underlying planar lipid layer (L).D: The furrow areas from C are shown in greater detail at highermagnification. Amplitude image (E) and height image (F) showcompleted AnxA5 2D crystalline lattice at medium and highmagnification, respectively. The zoomed height image (F)demonstrates AnxA5 trimers. Amplitude images were processedwith 1� convolution, and the height image was zero-flattened andlow-pass filtered. Panels D and F were originally published in Am JPathol 163:1193–1200 (2003).

Fig. 5. AFM images showing the effect of aPL mAbs on previouslyformed AnxA5 2D crystal. A,C: AnxA5 surface in the presence ofmAbs CL15 and CL1, respectively, prior to the addition of b2GPI.Following the addition of b2GPI, CL15 grossly disrupts the AnxA5crystal (B), forming large furrows (arrows), represented by darkareas. D: CL1 partially and mildly (asterisk) disrupts the AnxA5crystal lattice. Note that in (C) the addition of mAb CL1 to theAnxA5 layer does not alter the structural integrity of the lattice, buta few globular structures appear (closed arrows); after addition ofb2GPI, the globules enlarge and the crystal structure is minimallydisrupted (D). Images (B, C, D) are height images with off-line zero-flatten and low-pass filtering applied; image (A) is an amplitudeimage with no processing. Images originally published in Am JPathol 163:1193–1200 (2003).

JOURNAL OF CELLULAR PHYSIOLOGY

1952 T A A T J E S E T A L .

environment. At the time, it was known that AnxA5 binds toanionic phospholipids (such as phosphatidylserine), forming atwo-dimensional (2D) crystal lattice which can be viewed byAFM (Reviakine et al., 1998). Accordingly, we developed asimple in vitro simulacrum in which an AnxA5 2D crystal latticewas formed on a planar lipid layer (phosphatidylcholine/phosphatidylserine mixture) coating an atomically flatmuscovite mica disk (Fig. 4). To this AnxA5 2D crystal lattice,antibodies from patients with APL, together with the proteincofactor b2GPI, were added while simultaneously andtemporally imaging with AFM (Quinn et al., 2012a,b). Theresulting images (Fig. 5) demonstrated for the first time aPLantibody/b2GPI cofactor-mediated disruption of an AnxA5crystal lattice (Rand et al., 2003; Taatjes et al., 2008; Quinnet al., 2012a). Although AFM quantitative measurements in thex, y dimension are artificially enlarged due to the finite size ofthe probe tip, z dimension measurements on height-modeimages are extremely accurate (Taatjes et al., 1999). We haveused this feature to analyze the change in height as molecularcomponents were added to the mica disk. As shown inSupplementary Figure 3, the planar lipid layer is approximately4.5 nm in thickness, while the formed AnxA5 2D crystal layer isapproximately 2.5 nm in thickness. When the b2GPI cofactorprotein is added to the AnxA5 layer on the lipid, no alterationin component thickness occurs; however, following theaddition of a specific aPL antibody, the thickness of thecomponents is reduced down once again to the level of the lipidlayer (approximately 4.5 nm), indicating disruption/degradationof the AnxA5 2D crystal (Montigny et al., 2006; Quinnet al., 2012a).

This exciting ability afforded by AFM to temporally imagemacromolecular interactions at high resolution in a hydratedenvironment prompted us to next investigate the effects ofcandidate pharmacological agents that may possibly interferewith the aPL antibody/b2GPI cofactor-mediated disease

process. Hydroxycholoroquine (HCQ), an amphiphilicantimalarial compound, has been used clinically for many yearsas an effective immunosuppressive treatment for systemiclupus erythematosus (SLE) (Rothfield, 1988); interestingly, SLEpatients treated with HCQ were found to have a reduced riskfor thrombosis (Petri, 1996). Moreover, HCQ was found tosignificantly reduce the extent of thrombosis in an animalmodel of injury-induced thrombosis in APS (Edwardset al., 1997). In response to these published data we used AFMto investigate the potential effect of HCQ on: (1) formed aPLantibody/b2GPI complexes on planar lipid layers (Randet al., 2008a), followed by (2) AnxA5 2D crystal shield onplanar lipid layers in the presence of aPL antibody/b2GPIcomplexes (Rand et al., 2010a). The high resolution imagesobtainable with AFM revealed a profound effect of HCQ inboth experimental protocols. First, the addition of HCQ to theimaging buffer resulted in an erosion of aPL antibody/b2GPIcomplexes (Fig. 6) from an average height of 17.7� 5.8 nmto 9.3� 2.4 nm (40 complexes measured; P< 0.001) (Randet al., 2008a,b). Second, in the presence of HCQ, AnxA5formed 2D crystal islands around aPL antibody/b2GPIcomplexes as previously described; however, quite strikingly asecond layer of AnxA5 was then observed to form around andover the aPL antibody/b2GPI complexes (Fig. 7), in effect“patching” the damaged 2D crystal areas, thereby potentiallystabilizing the anti-thrombotic characteristic of the surface.

The described AFM data were supported by diverseexperimental techniques, including ellipsometry, biochemicalbinding assays, plasma coagulation times, and confocalmicroscopy, yet it was the direct high resolution visualization ofmacromolecular interactions afforded by fluid cell AFM thatprovided the most compelling answers towards the puzzlepresented by APS and its treatment. Clearly, however,these molecular mechanistic results derived from an in vitrosimulacrum must be addressed at the cellular level. For

Fig. 6. AFM height images (three-dimensional view with 50 degree pitch) showing the effect of hydroxychloroquine (HCQ) on three separateaPL mAb/b2GPI complexes. A,E,I: Addition of b2GPI to buffer covering the phospholipid bilayers resulted in the formation of distinctstructures raised above the bilayers. B,F,J: Subsequent addition of aPL mAb to the phospholipid-bound b2GPI shown in panels A, E, and Iresulted in large aggregates of increased height (indicated by white color) composed of aPLmAb/b2GPI complexes over the lipid bilayers. C,G,K: At 12min after the addition of HCQ to the supernatant fluid covering the bilayer, the immune complexes, shown in panels B, F, and J, weresignificantly eroded. D,H,L: The immune complexes, shown in panels B, F, and J, showed further disintegration 30min after the addition ofHCQ. All images show complexes electronically zoomed from an original 30mm� 30mm scan. Images were minimally processed to removescan lines. The color shades represent height variations in the images, with the darker colors (rust) indicating lower heights and the lightercolors (white), the higher profiles. Images originally published in: Blood 112:1687–1695 (2008).

JOURNAL OF CELLULAR PHYSIOLOGY

H I G H R E S O L U T I O N A T O M I C F O R C E M I C R O S C O P Y 1 9 5 3

instance, to determine if a similar HCQ-induced AnxA5“patching” phenomenon occurs at the plasma membranesurface, studies must be undertakenwith live cultures of humantrophoblastic cell lines and endothelial cells. Such live cell-based AFM imaging investigations have been plentiful, yet oftensuffer from diminished resolution, mainly a result of probestiffness/soft-sample interactions (You et al., 2000; Murphyet al., 2006; Heinisch et al., 2012; Aifantis et al., 2013).Alternatively, cells can be chemically fixed and imaged eitherin fluid or after drying (Supplementary Fig. 4), resulting inimproved imaging parameters and resolution (Franciset al., 2010a,b). Other methodologies, such as partial removalof cell membranes through semi-drying to reveal underlyingsubcellular cytoplasmic structures also yield higher resolutionimages by AFM (Wang et al., 2012; Lee et al., 2013). However,to image temporal interactions on cell surfaces at themolecular level will require further developments in probetechnology and the adherence of cells to various substrata.Murphy et al. (2006) have already reported on the complexitiesof live AFM imaging of two architecturally distinct cell types(fibroblastic and epithelial) using both contact and AC mode.Moreover, the cell surface as visualized by AFM is structurallyquite heterogeneous, rendering the structural identification ofspecific macromolecular components (such as proteins)problematic. The combination of AFM instruments with otherimaging modalities, such as fluorescence microscopy (Eben-stein et al., 2009), confocal microscopy (Kassies et al., 2005),and TIRF (Shaw et al., 2006), together with advances influorescent dye technology and the understanding of cantile-ver/surface interactions will likely be necessary to achievefurther advances in high-resolution live cell AFM imaging.

Acknowledgments

Wewould like to gratefully thank ourmany colleagues over theyears who contributed to these studies, and Nicole Bouffardfor assisting with preparation of the figures. We apologizeto those whose stellar work in biological AFM was notmentioned due to space constraints. The work described inthis manuscript was supported by grants from the NIH (R01HL-61331; J.H.R. and DK56212, NS39918; B.P.J.), and NCRR(S10RR025498; D.J.T.).

Literature Cited

Aifantis KE, Shrivastava S, Pelidou SH. 2013. Capturing the elasticity and morphology of livecell fibroblast cell cultures during degradation with atomic force microscopy. J Microsc249:62–68.

Aravanis M, Pyle JL, Tsien RW. 2003. Single synaptic vesicles fusing transiently andsuccessively without loss of identity. Nature 423:643–647.

Bennett MK, Calakos N, Schller RH. 1992. Syntaxin: A synaptic protein implicated in dockingof synaptic vesicles at presynaptic active zones. Science 257:255–259.

Binnig G, Rohrer H, Gerber Ch, Weibel E. 1982. Surface studies by scanning tunnelingmicroscopy. Phys Rev Lett 49:57–61.

Binnig G, Quate CF, Gerber Ch. 1986. Atomic force microscope. Phys Rev Lett 56:930–933.Cho W-J, Jena BP. 2007. N-ethymaleimide sensitive factor is a right-handed molecular

motor. J Biomed Nanotech 3:209–211.Cho S-J, Quinn AS, Stromer MH, Dash S, Cho J, Taatjes DJ, Jena BP. 2002a. Structure and

dynamics of the fusion pore in live cells. Cell Biol Int 26:35–42.Cho S-J, Jeftinija K, Glavaski A, Jeftinija S, Jena BP, Anderson LL. 2002b. Structure and

dynamics of the fusion pores in live GH-secreting cells revealed using atomic forcemicroscopy. Endocrinology 143:1144–1148.

Cho S-J, Kelly M, Rognlien KT, Cho JA, Horber JKH, Jena BP. 2002c. SNAREs in opposingbilayers interact in a circular array to form conducting pores. Biophys J 83:2522–2527.

Cho W-J, Jeremic A, Rognlien KT, Zhvania MG, Lazrishvili I, Tamar B, Jena BP. 2004.Structure, isolation, composition and reconstitution of the neuronal fusion pore. Cell BiolInt 28:699–708.

Cho W-J, Jeremic A, Jena BP. 2005. Size of supramolecular SNARE complex: Membrane-directed self-assembly. J Am Chem Soc 127:10156–10157.

Cho W-J, Ren G, Jena BP. 2008. EM 3D contour maps provide protein assembly at thenanoscale within the neuronal porosome complex. J Microsc 232:106–111.

Fig. 7. Dynamic AFM imaging of HCQ-induced formation of AnxA5 patch layer over aPL IgG/b2GPI complexes. A: Addition of b2GPI to aplanar lipid bilayer formed on mica resulted in formation of b2GPI clusters (marked 1–5). After the addition of HCQ (B), AnxA5 was added,which at first crystallized frommultiple nucleation sites on the phospholipid bilayer (C), and then formed a secondary layer of AnxA5 over theb2GPI clusters in the presence of antiphospholipid antibody IS4 (D). This AnxA5 patch subsequently enlarged over time (E,F) progressivelycovering the adjacent primary layer of AnxA5 crystal, leaving none of the surface uncovered by AnxA5. Subsequent addition of EDTAremoved all of the AnxA5 crystal (due to the Ca2þ requirement for AnxA5 binding to phospholipid), leaving the b2GPI clusters intact (G,H).For clarity, the original b2GPI clusters are labeled 1–5 in all of the panels. This figure is a modified version of figure originally published inBlood 115:2292–2299 (2010).

JOURNAL OF CELLULAR PHYSIOLOGY

1954 T A A T J E S E T A L .

Cho W-J, Shin L, Ren G, Jena BP. 2009. Structure of membrane-associated neuronalSNARE complex: Implication in neurotransmitter release. J Cell Mol Med 13:4161–4165.

ChoW-J, Lee J-S, Ren G, Zhang L, Shin L, Manke CW, Potoff J, Kotaria N, Zhvania MG, JenaBP. 2011. Membrane-directed molecular assembly of the neuronal SNARE complex. J CellMol Med 15:31–37.

Craciun C, Barbu-Tudoran L. 2012. Identification of new structural elements within‘porosomes’ of the exocrine pancreas: A detailed study using high resolution electronmicroscopy. Micron 44:137–142.

de Groot PG, Urbanus RT. 2012. The significance of autoantibodies against b2-glycoproteinI. Blood 120:266–274.

Dempsey GT, Vaughan JC, Chen KH, Bates M, Zhuang X. 2011. Evaluation of fluorophoresfor optimal performance in localization-based super-resolution imaging. Nat Meth8:1027–1036.

Drescher DG, ChoW-J, Drescher MJ. 2011. Identification of the porosome complex in thehair cell. Cell Biol Int Rep 18:31–34.

Ebenstein Y, Gassman N, Kim S, Weiss S. 2009. Combining atomic force and fluorescencemicroscopy for analysis of quantum-dot labeled protein-DNA complexes. J Mol Recognit22:397–402.

Edwards MH, Pierangeli S, Liu X, Barker JH, Anderson G, Harris EN. 1997.Hydroxychloroquine reverses thrombogenic properties of antiphospholipid antibodies inmice. Circulation 96:4380–4384.

ElshennawyWW. 2011. Image processing and numerical analysis approaches of porosome inmammalian pancreatic acinar cell. J Am Sci 7:835–843.

Francis LW, Gonzalez D, Ryder T, Baer K, Rees M,White JO, Conlan RS, Wright CJ. 2010a.Optimized sample preparation for high-resolution AFM characterization of fixed humancells. J Microsc 240:111–121.

Francis LW, Lewis PD,WrightCJ, Conlan RS. 2010b. Atomic forcemicroscopy comes of age.Biol Cell 102:133–143.

Galbraith CG, Galbraith JA. 2011. Super-resolution microscopy at a glance. J Cell Sci124:1607–1611.

Heinisch JJ, Lipke PN, Beaussart A, Chatel SEK, Dupress V, Alsteens D, Dufrene YF. 2012.Atomic force microscopy—Looking at mechanosensors on the cell surface. J Cell Sci125:4189–4195.

Huang B, Bates M, Zhuang X. 2009. Super-resolution fluorescence microscopy. Annu RevBiochem 78:993–1016.

Inayoshi Y, Minoda H, Arai Y, Nagayama K. 2012. Direct observation of biological moleculesin liquid environmental phase-plate transmission electron microscopy. Micron 43:1091–1098.

Japaridze NJ, Okuneva VG, Osovreli MG, Surmava AG, Lordkipanidze TG, Kiladze MT,Zhvania MG. 2012. Hypokinetic stress and neuronal porosome complex in the rat brain:The electron microscopic study. Micron 43:948–953.

Jena BP, Cho S-J, Jeremic A, Stromer MH, Abu-Hamdah R. 2003. Structure and compositionof the fusion pore. Biophys J 84:1–7.

Jeremic A, Kelly M, Cho S-J, StromerMH, Jena BP. 2003. Reconstituted fusion pore. Biophys J85:2035–2043.

Jeremic A, Kelly M, Cho J-H, Cho S-J, Horber JKH, Jena BP. 2004a. Calcium drives fusion ofSNARE-apposed bilayers. Cell Biol Int 28:19–31.

Jeremic A, Cho W-J, Jena BP. 2004b. Membrane fusion: What may transpire at the atomiclevel. J Biol Phys Chem 4:139–142.

Jeremic A, Quinn AS, ChoW-J, Taatjes DJ, Jena BP. 2006. Energy-dependent disassembly ofself-assembled SNARE complex: Observation at nanometer resolution using atomic forcemicroscopy. J Am Chem Soc 128:26–27.

Jones SA, Shim S-H, He J, Zhuang X. 2011. Fast, three-dimensional super-resolution imagingof live cells. Nat Meth 8:499–505.

Kassies R, Van der Werf KO, Lenferink A, Hunter CN, Olsen JD, Subramaniam V, Otto C.2005. Combined AFM and confocal fluorescence microscope for applications in bio-nanotechnology. J Microsc 217:109–116.

Katan AJ, Dekker C. 2011. High-speed AFM reveals the dynamics of single biomolecules atthe nanometer scale. Cell 147:979–982.

Lee J-S, Jeremic A, Shin L, ChoW-J, Chen X, Jena BP. 2012. Neuronal porosome proteome:Molecular dynamics and architecture. J Proteom 75:3952–3962.

Lee J-S, HouX, BishopN,Wang S, Flack A, ChoWJ, ChenX,MaoG, Taatjes DJ, Sun F, ZhangK, Jena BP. 2013. Aquaporin-assisted and ER-mediated mitochondrial fission: Ahypothesis. Micron 47:50–58.

Malhotra V, Orci L, Glick BS, Block MR, Rothman JE. 1988. Role of an N-ethylmaleimide-sensitive transport component in promoting fusion of transport vesicles with cisternae ofthe Golgi stack. Cell 54:221–227.

Montigny WJ, Quinn AS, Wu X-X, Bovill EG, Rand JH, Taatjes DJ. 2006. Atomic forcemicroscopy in the study of macromolecular interactions in hemostasis and thrombosis:Utility for investigation of the antiphophospholipid syndrome. In: Jena BP, Horber KHeditors. Forcemicroscopy: Applications in biology andmedicine. Hoboken,NJ: JohnWiley& Sons, Inc. pp 267–286.

Muller DJ, Engel A. 1999. Voltage and pH-induced channel closure of porin OmpF visualizedby atomic force microscopy. J Mol Biol 285:1347–1351.

Murphy MF, Lalor MJ, Manning FCR, Lilley F, Crosby SR, Randall C, Burton DR. 2006.Comparative study of the conditions required to image live human epithelial and fibroblastcells using atomic force microscopy. Microsc Res Tech 69:757–765.

Okuneva VG, Japaridze ND, Kotaria NT, Zhvania MG. 2012. Neuronal porosome in the ratand cat brain. Cell Tissue Biol 6:69–72.

Oyler GA, Higgins GA, Hart RA, Battenberg E, Billingsley M, Bloom FE, Wilson MC. 1989.The identification of a novel synaptosomal-associated protein, SNAP-25, differentiallyexpressed by neuronal subpopulations. J Cell Biol 109:3039–3052.

Persson BNJ. 1987. The atomic force microscope: Can it be used to study biologicalmolecules? Chem Phys Lett 141:366–368.

Petri M. 1996. Thrombosis and systemic lupus erythematosus: The Hopkins Lupus Cohortperspective. Scand J Rheumatol 25:191–193.

Quinn AS, Wu X-X, Rand JH, Taatjes DJ. 2012a. Insights into the pathophysiology of theantiphospholipid syndrome provided by atomic force microscopy. Micron 43:851–862.

Quinn AS, Rand JH, Wu X-X, Taatjes DJ. 2012b. Viewing dynamic interactions of proteinsand a model lipid membrane with atomic force microscopy. In: Taatjes DJ, Roth J, editors.Cell imaging techniques: Methods and Protocols, Methods in Molecular Biology, Vol. 931.New York: Springer Science þ Business Media. pp 259–293.

Rand JH. 2002. Molecular pathogenesis of the antiphospholipid syndrome. Circ Res 90:29–37.

Rand JH,WuX-X, Guller S, Gil J, Guha A, Scher J, LockwoodCJ. 1994. Reduction of annexin-V (placental anticoagulant protein-I) on placental villi of women with antiphospholipidantibodies and recurrent spontaneous abortion. Am J Obstet Gynecol 171:1566–1572.

Rand JH,Wu X-X, Andree HA, Lockwood CJ, Guller S, Scher J, Harpel PC. 1997. Pregnancyloss in the antiphospholipid-antibody syndrome—Apossible thrombogenicmechanism. NEngl J Med 337:154–160.

Rand JH, Wu X-X, Quinn AS, Chen PP, McCrae KR, Bovill EG, Taatjes DJ. 2003. Humanmonoclonal antiphospholipid antibodies disrupt the annexin A5 anticoagulant crystalshield on phospholipid bilayers: Evidence from atomic force microscopy and functionalassay. Am J Pathol 163:1193–1200.

Rand JH,Wu X-X, Quinn AS, Chen PP, Hathcock JJ, Taatjes DJ. 2008a. Hydroxychloroquinedirectly reduces the binding of antiphospholipid antibody-b2-glycoprotein I complexes tophospholipid bilayers. Blood 112:1687–1695.

Rand JH, Wu X-X, Quinn AS, Taatjes DJ. 2008b. Resistance to annexin A5 anticoagulantactivity: A thrombogenicmechanism for the antiphospholipid syndrome. Lupus 17:922–930.

Rand JH, Wu X-X, Quinn AS, Ashton AW, Chen PP, Hathcock JJ, Andree AAM, Taatjes DJ.2010a. Hydroxychloroquine protects the annexin A5 anticoagulant shield from disruptionby antiphospholipid antibodies: Evidence for a novel effect for an old antimalarial drug.Blood 115:2292–2299.

Rand JH, Wu X-X, Quinn AS, Taatjes DJ. 2010b. The annexin A5-mediated pathogenicmechanism in the antiphospholipid syndrome: Role in pregnancy losses and thrombosis.Lupus 19:460–469.

Reviakine I, Bergsma-Schutter W, Brisson A. 1998. Growth of protein 2-D crystals onsupported planar lipid bilayers imaged in situ by AFM. J Struct Biol 121:356–361.

Rothfield N. 1988. Efficacy of antimalarials in systemic lupus erythematosus. Am J Med85:53–56.

Schermelleh L, Heintzmann R, Leonhardt H. 2010. A guide to super-resolution fluorescencemicroscopy. J Cell Biol 190:165–175.

Schneider SW, Sritharan KC, Geibel JP, Oberleithner H, Jena BP. 1997. Surface dynamics inliving acinar cells imaged by atomic force microscopy: Identification of plasma membranestructures involved in exocytosis. Proc Natl Acad Sci USA 94:316–321.

Shaw JE, Oreopoulos J, Wong D, Hsu JCY, Yip CM. 2006. Coupling evanescent-wavefluorescence imaging and spectroscopy with scanning probe microscopy: Challenges andinsights from TIRF-AFM. Surf Interface Anal 38:1459–1471.

Shim S-H, Xia C, Zhong G, Babcock HP, Vaughan JC, Huang B, Wang X, Xu C, Bi G-Q,Zhuang X. 2012. Super-resolution fluorescence imaging of organelles in live cells withphotoswitchable membrane probes. Proc Natl Acad Sci USA 109:13978–13983.

Shin L, Cho W-J, Cook J, Stemmler T, Jena BP. 2010. Membrane lipids influence proteincomplex assembly-disassembly. J Am Chem Soc 132:5596–5597.

Shroff H, Galbraith CG, Galbraith JA, Betzig E. 2008. Live-cell photoactivated localizationmicroscopy of nanoscale adhesion dynamics. Nat Meth 5:417–423.

Siksou L, Rostaing P, Lechaire JP, Boudier T, Ohtsuka T, Fejtova A, Kao HT, Greengard P,Gundelfinger ED, Triller A, Marty S. 2007. Three-dimensional architecture of presynapticterminal cytomatrix. J Neurosci 27:6868–6877.

Taatjes DJ, Quinn AS, Lewis MR, Bovill EG. 1999. Quality assessment of atomic forcemicroscopy probes by scanning electron microscopy: Correlation of tip structure withrendered images. Microsc Res Tech 44:312–326.

Taatjes DJ,WadsworthMP, Quinn AS, Rand JH, Bovill EG, Sobel BE. 2008. Imaging aspects ofcardiovascular disease at the cell and molecular level. Histochem Cell Biol 130:235–245.

Taraska JW, Perrais D, Ohara-Imaizumi M, Nagamatsu S, Almers W. 2003. Secretorygranules are recaptured largely intact after stimulated exocytosis in cultured endocrinecells. Proc Natl Acad Sci USA 100:2070–2075.

Thorn P, Fogarty KE, Parker I. 2004. Zymogen granule exocytosis is characterized by longfusion pore openings and preservation of vesicle lipid identity. Proc Natl Acad Sci USA101:6774–6779.

TrimbleWS, CowanDW, Scheller RH. 1988. VAMP-1: A synaptic vesicle-associated integralmembrane protein. Proc Natl Acad Sci USA 85:4538–4542.

Wang S, Lee J-S, BishopN, Jeremic A, ChoWJ, Chen X, MaoG, Taatjes DJ, Jena BP. 2012. 3Dorganization and function of the cell: Golgi budding and vesicle biogenesis to docking at theporosome complex. Histochem Cell Biol 137:703–718.

You HX, Lau JM, Zhang S, Yu L. 2000. Atomic force microscopy imaging of living cells:A preliminary study of the disruptive effect of the cantilever tip on cell morphology.Ultramicroscopy 82:297–305.

ZhaoD, Lulevich V, Liu F, LiuG. 2010. Applications of atomic forcemicroscopy in biophysicalchemistry. J Phys Chem B 114:5971–5982.

JOURNAL OF CELLULAR PHYSIOLOGY

H I G H R E S O L U T I O N A T O M I C F O R C E M I C R O S C O P Y 1 9 5 5