High resolution surface plasmon resonance imaging for single cells

Abstract

Background: Surface plasmon resonance imaging (SPRI) is a label-free technique that can image refractive index changes at an interface. We have previously used SPRI to study the dynamics of cell-substratum interactions. However, characterization of spatial resolution in 3 dimensions is necessary to quantitatively interpret SPR images. Spatial resolution is complicated by the asymmetric propagation length of surface plasmons in the x and y dimensions leading to image degradation in one direction. Inferring the distance of intracellular organelles and other subcellular features from the interface by SPRI is complicated by uncertainties regarding the detection of the evanescent wave decay into cells. This study provides an experimental basis for characterizing the resolution of an SPR imaging system in the lateral and distal dimensions and demonstrates a novel approach for resolving sub-micrometer cellular structures by SPRI. The SPRI resolution here is distinct in its ability to visualize subcellular structures that are in proximity to a surface, which is comparable with that of total internal reflection fluorescence (TIRF) microscopy but has the advantage of no fluorescent labels.

Results: An SPR imaging system was designed that uses a high numerical aperture objective lens to image cells and a digital light projector to pattern the angle of the incident excitation on the sample. Cellular components such as focal adhesions, nucleus, and cellular secretions are visualized. The point spread function of polymeric nanoparticle beads indicates near-diffraction limited spatial resolution. To characterize the z-axis response, we used micrometer scale polymeric beads with a refractive index similar to cells as reference materials to determine the detection limit of the SPR field as a function of distance from the substrate. Multi-wavelength measurements of these microspheres show that it is possible to tailor the effective depth of penetration of the evanescent wave into the cellular environment.

Conclusion: We describe how the use of patterned incident light provides SPRI at high spatial resolution, and we characterize a finite limit of detection for penetration depth. We demonstrate the application of a novel technique that allows unprecedented subcellular detail for SPRI, and enables a quantitative interpretation of SPRI for subcellular imaging.

Figure 1

SPR imaging through a microscope objective instrumentation schematic. Incident angle selected light is generated and patterned by a digital light projector that is then collimated, wavelength filtered, linear polarized, and then directed through an inverted microscope platform (shaded in grey). The reflected image is captured on a CCD camera after a switchable lens assembly selects for either the image plane or back focal plane.

Figure 2

Back focal plane (BFP) image of SPR angle reflectance data and demonstration of incident angle control of excitation light by digital light projector (DLP). A) The BFP is fully illuminated, at all incident angles available, by 620 nm excitation light off of a 45 nm gold coated coverslip under water media. The incident light is linearly polarized in the x-direction. The dark crescent shaped regions at the edges of the field are where the incident light passes through the objective lens at a high angle of incidence and is coupled into the surface plasmons of the metal film, thus reducing reflectance. The absorbance occurs at the sides of the field where the incident light is p-polarized, and not at the top and bottom of the field where the incident light is s-polarized. The red line traces the angular distribution of incident light onto the image plane from the center of the BFP (0°) to the periphery (60°). B) A line scan of the red line in A) in the p-polarized direction shows the angular dependence of reflectivity and SPR coupling. The green line depicts the angle of illumination that provides maximum SPR coupling (≈53.5° for 620 nm) to be used for SPR imaging. C) Same BFP image area as shown in A) except here the DLP is used to project a thin arc of light that contains the same radial incident angle (≈53.5° illumination angle shown) as depicted by the green line in B).

Figure 3

SPR and phase contrast images of five cell types (3T3, HepG2, Vero, A10 and A549) fixed under PBS buffer 72 h after plating on fibronectin coated substrates, and SPR image comparison with fluorescently stained α-vinculin. A) SPR images collected with 100X/1.65 NA objective using 590 nm incident light; phase contrast images acquired with a 20X/0.4 NA objective. The SPR image displays distinct changes in reflectivity for various intracellular components within the evanescent wave such as cell membrane, focal adhesions, and cell nucleus. B) SPR image of A10 cell collected as in A) but contrast adjusted on a linear scale to emphasize the focal adhesions. Subsequently, the sample was immunofluorescently labeled with α-vinculin and imaged with a 63X/1.3 NA objective. A manual intensity threshold of α-vinculin was then overlaid onto the SPR image for comparison. C) Imaging conditions same as in A). A region of intermediate intensity, putative extracellular deposited material, is detected along the cell periphery. A linescan in the SPR image is segregated into four distinct image features: fibronectin (FN) coated substratum, extracellular deposited material (ECM), the basal cell surface, and focal adhesions. The scale bar of 25 μm applies to all images in figure.

Figure 4

Spatial resolution of fluorescent (FL) and SPR imaging using fluorescent point-source nanospheres. A) Transparent coverslip a 0.17 μm diameter particle fluorescing at 515 nm peak emitting wavelength in epi-fluorescent mode. Line scan plot next to image is used to determine full width half-maximum (FWHM) at 0.29 μm for a 1.65 NA objective. B) SPR image of nanoparticle at 620 nm shows a FWHM of 0.3 μm in the x-direction (red) and 0.6 μm in the y-direction using the 1.65 NA objective. The scale bar of 2 μm applies to A and B. Nanoparticle measurements made under water media, and fluorescence emission collected with 530 nm bandpass filter.

Figure 5

Measurement of SPR evanescent wave (EW) penetration depth by geometric relationship of measured radii of polymer microspheres. A) Bright-field and SPR images (for 5 wavelengths) of a polymethylmethacrylate (PMMA) microsphere in water. B) Diagram of a microsphere at the SPR sensor interface showing that only a fraction of the bead lies within the EW. The equation shows the relationship between the EW penetration depth (d) and the radius of the sphere measured by bright field (r ₁) and SPRI (r ₂). The overlay shows a layer model of the interface; the water layer decreases thickness as the bead moves toward the surface and into the EW. C) SPR image and insert of a microsphere used to illustrate the image analysis procedure for measuring the value of r ₂, the SPRI detected radius. The standard deviation (σ) of background intensity is determined by the annulus-shaped ROI in the primary image, and in the image insert a value of 3σ is set as the threshold with pixels values above threshold colored red; the radius of the bead (r ₂ in blue) is determined from the area of circle (green dashed circle) computed from the threshold. D) Exponential decay from surface into media of the SPR generated EW calculated as field intensity versus distance from surface (nm). Two values are labeled: the 1/e decay at 37% field strength commonly referred to as the “penetration depth”, and the 5% field strength value, which is the theoretical detection limit. E) Extent of the EW field depth (L_p) measured for several polymer microspheres as a function of excitation wavelength, along with theoretical values of the penetration depth (1/e) and detection limit (5% field strength). Within the standard deviation of the measurement error, the decay values agree for all microspheres and the calculated penetration depth at 1/e field decay.