Terbium Chelate Membrane Label for Time-Resolved, Total Internal

reflection fluorescence microscopy (TIRFM) was investigated. Swiss albino mouse 3T3 cells were cultured on silica and polystyrene substrates. Time-res...
0 downloads 0 Views 119KB Size
Download PDF
350

Bioconjugate Chem. 1998, 9, 350−357

Terbium Chelate Membrane Label for Time-Resolved, Total Internal Reflection Fluorescence Microscopy of Substrate-Adherent Cells Sam Phimphivong and S. Scott Saavedra* Department of Chemistry, University of Arizona, Tucson, Arizona 85721. Received September 11, 1997; Revised Manuscript Received March 6, 1998

A chelating agent conjugated to a lipid was synthesized by reacting the dianhydride form of diethylenetriaminepentaacetic acid sequentially with 4-aminosalicylate and dioleoylphosphatidylethanolamine. The product, DOPE-YAS-Tb, exhibits photophysical properties characteristic of a chelated Tb3+ ion bound to an organic triplet donor: an excitation maximum at 310 nm, narrow emission bands at 490, 545, 590, and 625 nm, and a lifetime of 1.57 ms. The suitability of DOPEYAS-Tb as a membrane-staining agent for morphological studies of cultured cells using total internal reflection fluorescence microscopy (TIRFM) was investigated. Swiss albino mouse 3T3 cells were cultured on silica and polystyrene substrates. Time-resolved detection was employed to reject shortlived background emission (autoemission from the cells and/or the polymer substrate), which allowed the long-lived Tb3+ emission to be selectively imaged. The results show that time-resolved TIRFM of cells stained with DOPE-YAS-Tb is an effective method of quantitatively examining the cell morphology in situations where background due to autoemission from cells and/or the substrate material is problematic.

INTRODUCTION

Total internal reflection fluorescence microscopy (TIRFM), in which an evanescent wave is used to selectively excite fluorophores near a solid-liquid interface, was first used by Axelrod to visualize cell-substrate contact regions by labeling the plasma membrane with an amphiphilic fluor (1). TIRFM has since been used widely to study cell adhesion phenomena. Examples include morphological studies of chick heart cells (2), Dicytostelium amoebae (3), bovine aortic endothelial cells (4, 5), and rat myotubes (6), Fc receptors in rat basophilic leukemia cells (7), the angiotensin-converting enzyme in endothelium (8), and the structural organization of interphase fibroblasts (9). A thorough review is provided in a recent paper by Burmeister et al. (10). Several groups have shown that digital TIRFM can be used to determine separation distances between the plasma membrane and the substrate in cell-substrate contact regions (5, 9-11), which allows the technique to be used for quantitative studies of cell adhesive interactions. For example, Burmeister et al. (10) examined relationships between contact area and adhesion strength in endothelial cells cultured on biomaterial surfaces and performed real-time measurements of the cellsubstrate separation distance as a function of applied shear stress. However, quantitative conversion of a digital TIRFM image to a map of contact region morphology can be compromised by extraneous emission from the cell and/ or the substrate (5). The emission spectra of conventional membrane labels such as DiIC18 and lipid-conjugated fluorescein and rhodamine overlap with cellular autofluorescence at wavelengths of less than 600 nm (12, 13), which can be problematic if the label density is relatively low and/or the separation distance is large. Furthermore, * Corresponding author. Phone: (520) 621-9761. Fax: (520) 621-8407. E-mail: [email protected].

the autofluorescence and Raman scattering background from many polymers, including those of biomedical interest, is usually more intense than that of a silica substrate. Methods for correcting steady-state fluorescence measurements on cells for autofluorescence have been described (13-15). Typically, correction factors are determined on a cell-by-cell basis by making a second measurement of emission intensity at a wavelength that is both spectrally resolved from the label emission and shown to be correlated with autoemission intensity within the label emission band. Time-resolved detection using an extrinsic label having decay kinetics much slower than those of the sources of background emission is an alternative approach. Since the radiative lifetime of cellular autofluorescence is less than 100 ns, it can be temporally discriminated from the emission of lanthanide metals such as Eu3+ and Tb3+, which have lifetimes in the millisecond range. Principles and applications of time-resolved fluorometry using lanthanide probes have been reviewed elsewhere (16, 17). Briefly, following excitation of a sample with a light pulse, photon detection is delayed to allow the short-lived background emission to decay to a negligible level. The detection system is activated thereafter to record only the signal from the long-lived label. We previously examined the use of two Tb chelates as staining agents for time-resolved TIRFM of substrateadherent cells (18). The results showed that elimination of short-lived cellular autofluorescence permitted selective detection of the long-lived label fluorescence. However, the chelates were found to stain not only the plasma membrane but also cytoplasmic regions, including the nucleus. Thus, interpretation of the TIRF image was somewhat ambiguous due to the nonselective localization of the label. Here we describe the synthesis of another Tb chelate, DOPE-YAS (dioleoylphosphatidylethanolamine conjugated to diethylenetriaminepentaacetyl-4-

S1043-1802(97)00160-2 CCC: $15.00 © 1998 American Chemical Society Published on Web 04/21/1998

Tb Chelate Lipid for TIRFM of Cultured Cells

aminosalicylic acid), and its use as a long-lived probe for time-resolved TIRFM of cells. The utility of this label in cell adhesion studies where background emission is problematic is demonstrated by performing TIRFM on stained cells cultured on polystyrene. Although this material is a ubiquitous cell culture substrate, in comparison to silica, it is a relatively poor material for TIRF studies due to inherently high levels of Raman scattering and autoluminescence (19). EXPERIMENTAL SECTION

Chemicals and Cell Culture Supplies. l,2-Di-cis9-octadecenoyl-sn-glycero-3-phosphoethanolamine (DOPE), Dulbecco’s modified Eagle’s medium (DMEM) without phenol red, trypsin, Hank’s balanced salt solution (HBSS) without phenol red, fetal bovine serum, penicillin (100 units/mL), streptomycin sulfate, and Tris hydrochloride were obtained from Sigma. Diethylenetriaminepentaacetic acid dianhydride (YdA), 4-aminosalicylate, sodium salt (pAS), pyridine, anhydrous dimethyl sulfoxide (DMSO), terbium(III) chloride hexahydrate, polystyrene (secondary standard, Mw ) 240 000), potassium bromide, and spectral-grade chlorobenzene were obtained from Aldrich. All chemicals were used as received. Deionized water was obtained from a Barnstead Nanopure system. Preparation and Characterization of DOPEYAS. YdA was reacted with pAS to produce the YAS monoanhydride, as previously described (20, 21). DOPEYAS was synthesized by reacting YAS monoanhydride with DOPE, following the general procedure of Nayar et al. (22). DOPE (53.7 µmol) was placed in a 100 mL round-bottom flask and dried under high vacuum for 2 h to remove residual chloroform. The dried lipid was redissolved in 10 mL of freshly distilled pyridine containing 537 µmol of YAS anhydride. The reaction mixture was stirred under a N2 atmosphere at 60 °C for 24 h. Excess YAS anhydride was hydrolyzed by adding 2 mL of 1:2:0.8 CHCl3/CH3OH/0.58% NaCl (v/v). Thin-layer chromatography of the unpurified product (using a mobile phase of 65:25:4 CH3Cl/MeOH/H2O and ninhydrin staining) showed a nearly complete conversion of DOPE to DOPE-YAS. Unreacted DOPE and DOPE-YAS were recovered by repeated extraction with a 1:1 mixture of CH3Cl/0.58% NaCl. The saline solution was buffered at pH 2 with 100 mM phosphate (note that DOPE-YAS is insoluble in water at pH 2). The CH3Cl fractions, which contained DOPE, DOPE-YAS, methanol, and pyridine, were pooled, dried on a rotary evaporator, and redissolved in 50 mL of 1:9 CH3Cl/CH3OH. DOPE-YAS was removed from this solution by repeated extraction with a 1:1 mixture of CH3Cl/0.58% NaCl. In this step, the saline solution was buffered at pH 8 with 100 mM phosphate (note that DOPE-YAS is slightly soluble in water at pH 8). The progress of the extraction was monitored by adding a few drops of the organic phase to a solution of TbCl3 dissolved in Tris buffer at pH 7.4. When DOPE-YAS was present in the organic phase (i.e., had not been quantitatively extracted), the interface between the droplets and the TbCl3 solution fluoresced green under illumination from a mercury lamp. The extraction was repeated until no fluorescence could be visually detected. The aqueous fractions were then pooled, washed three times with CH3Cl, and lyophilized to yield the DOPE-YAS product. The product was characterized by UV-visible and infrared absorbance spectrometry, using a HewlettPackard 8452A spectrophotometer and a Nicolet 510P

Bioconjugate Chem., Vol. 9, No. 3, 1998 351

FT-IR spectrometer, respectively. Steady-state fluorescence spectroscopy of dissolved DOPE-YAS-Tb was performed using a Spex Fluorolog fluorometer (Spex Industries, Inc., Edison, NJ). Fluorescent lifetime measurements on samples of dissolved DOPE-YAS-Tb were performed using the time-resolved fluorescence spectroscopic system that was described previously (18). Substrates. Fused silica slides (Dynasil, Berlin, NJ) were cleaned by light mechanical scrubbing with cotton in 2% PCC-54 detergent (Pierce) and sonicated first in hot (ca. 60 °C) 2% PCC-54 for 15 min, then in methanol for 10 min, and finally in deionized water for 10 min. Slides were rinsed with deionized water between sonications and then dried overnight at 120 °C. Fused silica slides coated with polystyrene (PS) were prepared by spin coating at 1500 rpm, using an 8% (w/w) solution of PS dissolved in chlorobenzene, as described previously (23). After the slides were coated, films were dried at 70 °C for 24 h. The thickness and refractive index were determined by measuring prism coupling angles that produced waveguided modes in the film at 514.5 nm, as described previously (23). The typical thickness and refractive index of a PS film prepared in this manner were 0.84 µm and 1.59, respectively. These values differed by no more than 5% when measured at several different spots on a single slide. Among a batch of PScoated slides prepared on a given day, the variation in film thickness and refractive index was also no more than 5%. Cell Culture and Labeling. Cells were cultured as described previously (18). Briefly, Swiss albino mouse 3T3 cells (American Type Culture Collection, Rockville, MD) were cultured in 90% DMEM, supplemented with 10% fetal bovine serum, 100 µg/mL streptomycin, and 100 units/mL penicillin. Cells were grown, usually to confluence, in a 75 cm2 culture flask (Costar, Cambridge, MA) at 37 °C in a 5% CO2/95% air atmosphere. Cells were passaged by washing with HBSS, detached by incubation with 2 mL of a trypsin solution (diluted to 1:20 from the stock solution) for 5 min, and resuspended in supplemented medium at a density of ca. 105 cells/mL. Passage numbers 10-30 were used in experiments described herein. For fluorescence spectroscopy and microscopy, cells were plated at a density of approximately 103 cells/mL on either fused silica slides or PS-coated fused silica slides, with the planar substrate mounted in a Lucite chamber which also functioned as a TIRF cell. The plating time was 24 h on fused silica slides and 48-72 h for PS-coated slides. Longer times were required on PS because the cells adhered poorly to it and growth was relatively slow, which is well-known for PS that has not been oxidized (24). After plating, the adherent cells were washed with HBSS, fixed in 3.7% formaldehyde at room temperature for 7 min, and rinsed with deionized water. Labeling with Tb was performed using a two-step procedure. Cells were incubated with DOPE-YAS (16 µM in Tris buffer at pH 7.4) for 60 min, rinsed with deionized water, incubated with TbCl3 (16 µM in Tris buffer at pH 7.4) for 10 min, and then rinsed with deionized water. Fluorescence Microscopy of Cultured Cells. Fluorescence microspectroscopy and microphotography were performed using the instrumentation described previously (18). Briefly, the sample is excited in a TIR geometry at a 70° incidence angle with the 325 nm beam (about 1 mW) of a helium-cadmium laser (Liconix model 3207N). The beam is mechanically chopped at 400 Hz and adjusted to transverse electric (s) polarization with

352 Bioconjugate Chem., Vol. 9, No. 3, 1998

Phimphivong and Saavedra

Scheme 1a

a DOPE-YAS was synthesized with a two-step reaction. Diethylenetriaminepentaacetic dianhydride (YdA) was reacted with p-aminosalicylic acid (pAS) to produce diethylenetriaminepentaacetyl-4-aminosalicylic monoanhydride (YAS monoanhydride); YAS monoanhydride was then reacted with DOPE to produce DOPE-YAS.

a halfwave plate (Newport). Image acquisition was performed at the bottom port of a modified Nikon Diaphot inverted microscope using a CCD array (Tektronix model TK512CB, mounted in a Photometrics cryogenic dewar and cooled to -105 °C). Image acquisition was controlled via a Macintosh PC using IPLab software (Signal Analytics, Vienna, VA). All images were collected using a 40×, 0.55 NA, microscope objective lens. Time-resolved fluorescence imaging was performed using a liquid crystal shutter mounted on the CCD, as described previously (25). Spectroscopy of cultured cells was performed using the side port of the microscope. The light was focused on a rectangular aperture located at an intermediate image plane. The dimensions of the aperture were adjusted to limit light collection to a user-defined subregion of the field of view. Subsequently, the image was refocused onto the entrance slit of a monochromator (Spectra-Pro 275, Acton Research) and detected with a photomultiplier (9816-B, Thorn EMI) thermoelectrically cooled to ca. -30 °C (TE104RF housing, Products for Research) and operated in a photon counting mode (SR400, Stanford Research). For time-resolved spectroscopy, the counter was operated in a temporally gated mode, using the chopper TTL output as a timing signal. The “delay” prior to initiating counting and the “gate” during which counting takes place were set at 600 and 1200 µs, respectively. Lab Windows software running on a PC was used to control spectral acquisition. Spectra were not corrected for instrument response. Lifetime measurements were performed by holding the emission monochromator at 545 nm and operating the photon counter in a gate-scanning mode. The gate was set to 10 µs, and the delay was progressively increased in increments of 10 µs, integrating counts for 5 s at each delay setting. Lifetimes were recovered using a onecomponent model to perform a linear least-squares fit to a semilog plot of emission intensity as a function of delay time.

Image Processing. Fluorescence images were corrected for bias and dark signal using eq 1

Ic )

(Ir - Id)M If - I d

(1)

where Ic is the corrected image, Ir is the raw image, Id is the dark image, If is the flat field image, and M is the mean pixel value of If - Id. If was obtained by defocusing the image of the sample using the same microscope objective (40×) to obtain an even illumination of light. M was approximately 50% percent of the saturation level. RESULTS AND DISCUSSION

Preparation and Characterization of DOPEYAS. A chelate-capped lipid, DOPE-YAS, was prepared using a two-step reaction (see Scheme 1). The reaction between diethylenetriaminepentaacetic dianhydride (YdA) and p-aminosalicylic acid (pAS) in the first step produced YAS monoanhydride. In the second step, YAS monoanhydride was reacted with the amino headgroup of DOPE to yield the product. Like other polycarboxylated compounds, DOPE-YAS was isolated from the other reaction components on the basis of its solubility in organic solvents as a neutral species and in aqueous solutions as a charged species (see Experimental Section). The yield of the reaction (85%) was determined by titrating an aliquot of the CHCl3 fraction containing DOPE/DOPE-YAS (see Experimental Section) with aqueous Tb3+. The reaction product was characterized by UV-vis and infrared absorption spectroscopies. In the UV-vis region, DOPE and YdA do not absorb light appreciably at wavelengths above 250 nm, while pAS does. In buffer (pH 7.4), pAS absorption bands occur at 264 and 310 nm. The reaction product exhibited absorption bands centered at 272 and 308 nm. The similarity of the absorption spectra (Figure 1) indicates the presence of pAS in the

Tb Chelate Lipid for TIRFM of Cultured Cells

Bioconjugate Chem., Vol. 9, No. 3, 1998 353

Figure 1. Absorbance spectra of pAS (8 µM, solid line) and DOPE-YAS (84 µM, dashed line) dissolved in 0.1 M Tris buffer at pH 7.

Figure 2. Infrared spectra of pAS, DOPE, and DOPE-YAS (KBr pellet). Spectra were scaled for presentation.

reaction product. Figure 2 shows IR spectra of pAS, DOPE, and DOPE-YAS. In the pAS spectrum, the phenolic O-H stretch appears as a broad band at 36503200 cm-1 and overlaps the asymmetric (3372 cm-1) and symmetric (3290 cm-1) N-H stretching modes of the amine. This phenolic O-H stretch is also observed in the spectrum of the product, consistent with incorporation of the pAS moiety. In Tris buffer at pH 7.4 containing an equimolar amount of Tb3+, DOPE-YAS exhibits a fluorescent excitation maximum at 308 nm, similar to its absorption spectrum, and sharp emission maxima at 490, 545, 590, and 625 nm (uncorrected) which are characteristic of the Tb3+ ion (data not shown). The fluorescence emission decay is first-order with a lifetime of 1.57 ms (data not shown). These properties are very similar to those of YAS-Tb (18, 20, 21), which shows that the electronic properties of YAS-Tb are not significantly altered by conjugation to DOPE. The quantum yield of DOPEYAS-Tb was not measured. However, since the struc-

Figure 3. Images of a 3T3 cell cultured on a fused silica substrate for 24 h, fixed, and labeled with DOPE-YAS-Tb: (A) phase-contrast image, (B) steady-state TIRF image, and (C) time-resolved TIRF image. The image size is 160 µm × 118 µm.

ture of the chelating group is very similar to YAS-Tb, the quantum yield of DOPE-YAS-Tb is not expected to be significantly different from 0.1, the value reported previously for YAS-Tb (21). Fluorescence Microscopy. Cell Adhesion on Fused Silica. Shown in Figure 3 are phase-contrast and TIRF images of a 3T3 cell plated on a fused silica substrate, fixed 24 h later, and stained with DOPE-YAS-Tb. Under steady-state excitation (Figure 3B), the cell exhibits spatially distributed patches of green fluorescence. The brightest patches are in regions adjacent to the nucleus and at the cell periphery (identified from the

354 Bioconjugate Chem., Vol. 9, No. 3, 1998

Phimphivong and Saavedra

Figure 4. Steady-state (upper curve) and time-resolved (lower curve) TIRF spectra of a 3T3 cell cultured on a fused silica substrate for 24 h, fixed, and stained with DOPE-YAS-Tb.

phase-contrast image). A steady-state emission spectrum of this cell is shown in Figure 4. The emission bands at 490, 545, and 590 nm which can be attributed to the Tb label are superimposed on a broad background emission, which is due to cellular autofluorescence (18) that occurs over the entire 400-600 nm region. The corresponding time-resolved TIRF image of the same cell is shown in Figure 3C. Although the pattern is similar, the emission intensity is reduced relative to that of the steady-state image (Figure 3B). The corresponding time-resolved TIRF spectrum is displayed in Figure 4. It is apparent that gating the detection system essentially eliminates the broad-band autofluorescence present in the steady-state spectrum (Figure 4, upper curve), leaving the characteristic Tb3+ emission bands superimposed on a relatively flat background. Thus, the emission pattern in the time-resolved image (Figure 3C) arises almost entirely from the DOPE-YAS-Tb label. Elimination of the autofluorescence occurs because its lifetime is considerably shorter [in the nanosecond range (18)] than the delay period before the liquid crystal shutter was opened to expose the CCD. The percentage of DOPE-YAS incorporated into the membranes of fixed cells via the labeling procedure described above was not determined. The percentage of incorporated DOPE-YAS molecules that were complexed with Tb3+ was also not determined. However, on the basis of a formation constant of 1017 reported previously for YAS-Tb (21), the affinity of binding of Tb3+ to DOPE-YAS is likely to be very strong, and thus, the percentage of incorporated DOPE-YAS molecules that are complexed with Tb3+ is probably near unity. Furthermore, the fluorescence emission intensity from fixed cells incubated with 16 µM TbCl3 for 10 min followed by rinsing with deionized water (i.e., the cells were not stained with DOPE-YAS) was indistinguishable from the background intensity due to cellular autofluorescence. This result shows that, in the absence of DOPE-YAS labeling, fixed 3T3 cells have relatively little binding affinity for Tb3+. Cell Adhesion on a Polystyrene-Coated Fused Silica Substrate. The steady-state TIRF spectrum of an unstained 3T3 cell that was fixed 72 h after being plated on a fused silica substrate coated with polystyrene (PS) is shown in Figure 5A. The broad emission band with a maximum near 450 nm in the steady-state spectrum is due to autoemission from both the cell and the polymer film [the emission maximum of the PS itself was near

Figure 5. (A) Steady-state (upper curve) and time-resolved (lower curve) TIRF spectra of a fixed, unstained 3T3 cell cultured on a PS-coated silica substrate for 72 h. (B) Steadystate (upper curve) and time-resolved (lower curve) TIRF spectra of a 3T3 cell cultured on a PS-coated silica substrate for 72 h, fixed, and stained with DOPE-YAS-Tb.

520 nm (spectrum not shown)]. The time-resolved TIRF spectrum, also plotted in Figure 5A, shows that both sources of autoemission are essentially eliminated when time-resolved detection is implemented. Phase-contrast and TIRF images of a 3T3 cell plated on a PS-coated fused silica substrate, fixed 72 h later, and stained with DOPE-YAS-Tb are shown in Figure 6. The steady-state TIRF image shows spatially distributed patches of green fluorescence similar to those present in the images of the stained cell that adheres to fused silica (Figure 3). The steady-state emission spectrum of the cell shown in Figure 6 is plotted in Figure 5B and shows the Tb emission bands superimposed on the autoemission background which arises from the cell and the polymer film. The corresponding time-resolved TIRF image and spectrum of the same cell are shown in Figures 6C and 5B, respectively. A comparison of the two spectra shows that time-resolved detection essentially eliminates the background. Thus, with the background removed, the time-resolved image represents the actual spatial distribution of the Tb label in the cell. Steady-state and time-resolved TIRF spectra were also acquired from the three subcellular regions outlined by the boxes shown in Figure 6A. The spectra are plotted in Figure 7. In regions 1 and 2, the steady-state emission intensities at 545 nm are approximately 1.9 × 104 and 1.6 × 104 counts per second, respectively. However, the

Tb Chelate Lipid for TIRFM of Cultured Cells

Bioconjugate Chem., Vol. 9, No. 3, 1998 355

Figure 6. Images of a 3T3 cell cultured on a PS-coated silica substrate for 72 h, fixed, and labeled with DOPE-YAS-Tb: (A) phase-contrast image, (B) steady-state TIRF image, and (C) time-resolved TIRF image. The image size is 176 µm × 176 µm. The regions outlined by the boxes in panel A represent an area of 865 µm2 each.

time-resolved emission intensities, which more accurately reflect the relative number densities of DOPE-YAS-Tb in the two regions, are nearly equal at 545 nm. Thus, the difference observed in the steady-state mode is due primarily to differences in background intensity between the two regions. This comparison illustrates the fact that quantitative conversion of a digital TIRFM image to a map of contact region morphology can be compromised by extraneous emission from the cell and/or the substrate. The severity of the problem may be greater when the label emission intensity is much lower, such as in region 3, in which the mean cell-substrate gap is presumably larger than that in regions 1 and 2. Assessment. In a previous paper (18), we showed that background emission in TIRFM of substrate-adherent cells due to cellular autofluorescence could be rejected using a Tb chelate as a staining agent coupled with timeresolved detection. However, the Tb chelates examined in that study stained not only the plasma membrane but also cytoplasmic regions, including the nucleus, which made interpretation of the TIRF images somewhat ambiguous. Since the Tb chelate described in this study is a modified lipid, it was expected to exhibit greater selectivity for the membranes of substrate-adherent cells. The TIRF images shown in Figures 3 and 6 are consistent with this expectation. The spatial distribution of fluo-

rescence emission is similar to that observed for cells labeled with DiI (data not shown) and with TIRF images of substrate-adherent cells that have appeared in the literature (3-5, 9). Since the penetration depth of the evanescent wave used to generate the TIRF images shown in Figures 3 and 6 is less than 100 nm, the images presumably represent the ventral membrane topography of the cells. However, to identify particular regions as (for example) focal contacts, a complementary technique should be employed, such as immunofluorescence staining for a protein known to be present in these regions [e.g., vinculin (4, 26)]. Cell adhesion, growth, and motility on polymeric materials is a very active area of research because these materials are widely used in implantable devices (24, 2730). However, the use of TIRFM in studying cell adhesion behavior on biomedical polymers has been very limited to date, which is due in part to the high background emission characteristic of these materials. As noted above, converting a digital TIRFM image to a quantitative map of cell-substrate contact region morphology requires that the background emission be minimal or be removed (5). The results described herein show that the use of DOPE-YAS-Tb as a membrane-staining agent eliminates the background problem for polystyrene, which is a relatively poor substrate material for TIRFM

356 Bioconjugate Chem., Vol. 9, No. 3, 1998

Phimphivong and Saavedra

more recently by the National Science Foundation (Grants CHE-9403896 and CHE-9726132). LITERATURE CITED

Figure 7. Steady-state and time-resolved TIRF spectra (upper and lower curves, respectively, in each pair of spectra) acquired from three subregions of the cell shown in Figure 6. The numbers on each pair of spectra correspond to the labeled subregions shown in Figure 6A.

studies due to its intrinsic autoluminescence and Raman scattering (19). This technique may therefore be useful in TIRFM studies of cell adhesion on other types of polymers, although we have not yet tested materials other than PS. Finally, we note that the TIRFM experiments described above were performed on fixed cells. The degree to which live cells internalize DOPE-YAS was not investigated. Internalization is undesirable since it produces fluorescence emission from regions of the cell other than the plasma membrane. In the presence of significant emission originating from the cell interior, the use of digital TIRFM to quantitatively assess cellsubstrate separation distances would be problematic (10). ACKNOWLEDGMENT

This work was supported initially by a Biomedical Engineering Grant from the Whitaker Foundation and

(1) Axelrod, D. (1981) Cell-substrate contacts illuminated by total internal reflection fluorescence. J. Cell Biol. 89, 141145. (2) Gingell, D., Todd, I., and Bailey, J. (1985) Topography of cell-glass apposition revealed by total internal reflection fluorescence of volume markers. J. Cell Biol. 100, 1334-1338. (3) Todd, I., Smellor, J. S., and Gingell, D. (1988) Mapping cellglass contacts of Dicytostelium amoebae by total internal reflection aqueous fluorescence overcomes a basic ambiguity of interference reflection microscopy. J. Cell Sci. 89, 107114. (4) Truskey, G. A., Burmeister, J. S, Grapa, E., and Reichert, W. M. (1992) Total internal reflection fluorescence microscopy (TIRFM) II. Topographical mapping of relative cell/substratum separation distances. J. Cell Sci. 103, 491-499. (5) Burmeister, J. S., Truskey, G. A., and Reichert, W. M. (1994) Quantitative analysis of variable-angle total internal reflection fluorescence microscopy (VA-TIRFM) of cell/substrate contacts. J. Microsc. 173, 39-51. (6) Bloch, R. J., Velez, M., Krikorian, J. G., and Axelrod, D. (1989) Microfilaments and actin-associated proteins at sites of membrane-substrate attachment within acetylcholine receptor clusters. Exp. Cell Res. 182, 583-596. (7) Weis, R., Balakrishnan, K., Smith, B., and McConnell, H. M. (1982) Stimulation of fluorescence in a small contact region between rat basophil leukemia cells and planar lipid membrane targets by coherent evanescent radiation. J. Biol. Chem. 257, 6440-6445. (8) Nakache, M., Gaub, H. E., Shreiber, A. B., and McConnell, H. M. (1986) Topological and modulated distribution of surface markers on endothelial cells. Proc. Natl. Acad. Sci. U.S.A. 83, 2874-2878. (9) Lanni, F., Waggoner, A. S., and Taylor, D. L. (1985) Structural organization of interphase 3T3 fibroblasts studied by total internal reflection fluorescence microscopy. J. Cell. Biol. 100, 1091-1102. (10) Burmeister, J. S., Olivier, L. A., Reichert, W. M., and Truskey, G. A. (1998) Application of total internal reflection fluorescence microscopy to study cell adhesion to biomaterials. Biomaterials (in press). (11) Gingell, D., Heavens, O., and Mellor, J. S. (1987) General electromagnetic theory of total internal reflection fluorescence: The quantitative basis for mapping cell-substratum topography. J. Cell Sci. 87, 677-693. (12) Aubin, J. E. (1979) Autofluorescence of viable cultured mammalian cells. J. Histochem. Cytochem. 27, 36-43. (13) Roederer, M., and Murphy, R. F. (1986) Cell-by-cell autofluorescence correction for low signal-to-noise systems: Application to epidermal growth factor endocytosis by 3T3 fibroblasts. Cytometry 7, 558-565. (14) Corsetti, J. P., Sotirchos, S. V., Cox, C., Cowles, J. W., Leary, J. F., and Blumburg, N. (1987) Correction of cellular autofluorescence in flow cytometry by mathematical modeling of cellular fluorescence. Cytometry 9, 539-547. (15) Steinkamp, J. A., and Stewart, C. C. (1986) Dual laser, differential fluorescence correction method for reducing cellular background autofluorescence Cytometry 9, 566-574. (16) Soini, E., and Lovgren, T. (1987) Time-resolved fluorescence of lanthanide probes and applications in biotechnology. CRC Crit. Rev. Anal. Chem. 18, 105-155. (17) Holzwarth, A. F. (1995) Time-resolved fluorescence spectroscopy. Methods Enzymol. 246, 335-362. (18) Phimphivong, S., Kolchens, S., Edmiston, P. L., and Saavedra, S. S. (1995) Time resolved, total internal reflection fluorescence microscopy of cultured cells using a Tb chelate label. Anal. Chim. Acta 307, 403-417. (19) (a) Rabolt, J. F., Santo, R., and Swalen, J. D. (1979) Raman spectroscopy of thin polymer films using integrated optical techniques. Appl. Spectrosc. 33, 549-551. (b) Ives, J. T., and Reichert, W. M. (1987) Protein adsorption on the surface of a

Tb Chelate Lipid for TIRFM of Cultured Cells thin-film polymer integrated optical waveguide. Appl. Spectrosc. 42, 68-72. (20) Bailey, M. P., Rocks, B. F., and Riley, C. (1984) Terbium chelate for use as a label in fluorescent immunoassays. Analyst 109, 1449-1450. (21) Saavedra, S. S., and Picozza, E. G. (1989) Time-resolved fluorometric detection of terbium-labeled deoxyribonucleic acid separated by gel electrophoresis. Analyst 114, 835-838. (22) Najar, R., and Schroit, L. (1985) Generation of pH-sensitive liposomes: Use of large unilamellar vesicles containing N-succinyldioleoylphosphatidylethanolamine. Biochemistry 24, 5967-5971. (23) Saavedra, S. S., and Reichert, W. M. (1990) Prism coupling into polymer integrated optical waveguides with liquid superstrates. Appl. Spectrosc. 44, 1210-1217. (24) Steel, J. G., Dalton, A. B., Johnson, G., and Underwood, P. A. (1993) Polystyrene chemistry affects vitronectin activity: An explanation for cell attachment to tissue culture polystyrene but not unmodified polystyrene. J. Biomed. Mater. Res. 27, 927-940. (25) Edmiston, P. L., Kolchens, S., and Saavedra, S. S. (1992) Temporally gating a slow-scan CCD with a liquid crystal shutter. Appl. Spectrosc. 47, 250-253. (26) Burridge, K., Faith, K., Kelly, T., Nuckolls, G., and Turner, C. (1988) Focal adhesions: Transmembrane junctions be-

Bioconjugate Chem., Vol. 9, No. 3, 1998 357 tween the extracellular matrix and the cytoskeleton. Annu. Rev. Cell Biol. 4, 487-525. (27) Burmeister, J. S., Vrany, J., Reichert, W. M., and Truskey, G. A. (1996) Effect of vitronectin amount and conformation on the strength of endothelial cell adhesion to HEMA/EMA copolymers. J. Biomed. Mater. Res. 30, 12-22. (28) Horbett, T. A., and Schway, M. B. (1988) Correlations between mouse 3T3 cell spreading and serum fibronectin adsorption on glass and hydroxyethylmethacrylate-ethylmethacrylate copolymers J. Biomed. Mater. Res. 22, 763793. (29) Williams, S. K., Rose, D. G., and Jarrell, B. E. (1994) Microvascular endothelial cell sodding of ePTFE vascular grafts: Improved patency and stability of the cellular lining. J. Biomed. Mater. Res. 28, 203-212. (30) Schakenraad, J. M., Busscher, H. J., Wildevuur, C. R. H., and Arends, J. (1986) The influence of substratum surface free energy on the growth and spreading of human fibroblasts in the presence and absence of serum. J. Biomed. Mater. Res. 20, 773-784.

BC9701609