Scanning ion conductance microscopy: a convergent high-resolution technology for multi-parametric analysis of living cardiovascular cells

Abstract

Cardiovascular diseases are complex pathologies that include alterations of various cell functions at the levels of intact tissue, single cells and subcellular signalling compartments. Conventional techniques to study these processes are extremely divergent and rely on a combination of individual methods, which usually provide spatially and temporally limited information on single parameters of interest. This review describes scanning ion conductance microscopy (SICM) as a novel versatile technique capable of simultaneously reporting various structural and functional parameters at nanometre resolution in living cardiovascular cells at the level of the whole tissue, single cells and at the subcellular level, to investigate the mechanisms of cardiovascular disease. SICM is a multimodal imaging technology that allows concurrent and dynamic analysis of membrane morphology and various functional parameters (cell volume, membrane potentials, cellular contraction, single ion-channel currents and some parameters of intracellular signalling) in intact living cardiovascular cells and tissues with nanometre resolution at different levels of organization (tissue, cellular and subcellular levels). Using this technique, we showed that at the tissue level, cell orientation in the inner and outer aortic arch distinguishes atheroprone and atheroprotected regions. At the cellular level, heart failure leads to a pronounced loss of T-tubules in cardiac myocytes accompanied by a reduction in Z-groove ratio. We also demonstrated the capability of SICM to measure the entire cell volume as an index of cellular hypertrophy. This method can be further combined with fluorescence to simultaneously measure cardiomyocyte contraction and intracellular calcium transients or to map subcellular localization of membrane receptors coupled to cyclic adenosine monophosphate production. The SICM pipette can be used for patch-clamp recordings of membrane potential and single channel currents. In conclusion, SICM provides a highly informative multimodal imaging platform for functional analysis of the mechanisms of cardiovascular diseases, which should facilitate identification of novel therapeutic strategies.

Figure 1.

Schematic illustration of scanning ion conductance microscopy as a tool to study tissues and cells at the macroscopic, microscopic and nanoscopic levels of organization. (Online version in colour.)

Figure 2.

Aortic valve architecture. (a) Surface topography of a live explanted porcine aortic valve demonstrating cell shape, size and alignment using scanning ion conductance microscopy (SICM) (A. Moshkov 2010, unpublished data). Effective pixel width 313 nm, scan duration 23 min. Scanning pipette had resistance of 100 MΩ and estimated tip diameter of 100 nm. (b) Glutaraldehyde-fixed sample of valve imaged using scanning electron microscopy, 2000×, showing cell shape and alignment similar to SICM image in (a). Scale bar, 5 µm. (Scanning electron microscope image courtesy of Dr Adrian H. Chester, Cardiovascular Science, Harefield Hospital, Imperial College London, London, UK.)

Figure 3.

Aorta cell alignment and architecture. (a) Intact hearts and attached thoracic aorta of 2-year-old landrace cross pig. (b) Representative SICM image of the inner part of the aorta where cells are organized in diffuse pattern. (c) SICM image of the outer part of the aorta shows a regularity of cell alignment, which indicates that this area is exposed to higher stress. Dashed arrows indicate blood flow direction. Effective pixel width in both images 625 nm, scan duration 13 min. Scanning pipette had a resistance of 100 MΩ and an estimated tip diameter of 100 nm. (A. Moshkov 2010, unpublished data.) (Online version in colour.)

Figure 4.

(a) Typical surface topography image of a healthy adult cardiomyocyte. Well-organized striation and Z-grooves can be observed. Effective pixel width 125 nm, scan duration 4 min. (b) Surface topography image of an adult cardiomyocyte from HOCM patients shows an absence of T-tubules in this 9 × 9 µm area of the cell. (c) Z-grooves ratio index quantification demonstrates a significant difference in HOCM compared with control cells (n = 5 ± s.e. in both control and HOCM patients, p < 0.05 Student's t-test). Scanning pipette had a resistance of 100 MΩ and an estimated tip diameter of 100 nm. Modified from Lyon et al. [37] with permission. (Online version in colour.)

Figure 5.

Neonatal rat ventricular myocytes were exposed to PE for 48 h to induce hypertrophy. (a) The 96 × 96 µm scan of cardiomyocyte after 48 h in culture under control conditions. The process on the right side of the cardiomyocyte appears to be cropped only owing to the angle of view. (b) Same size scan performed on a different cardiomyocyte exposed to 10 µmol l⁻¹ PE. (c) Average cell volume in control and hypertrophic cardiomyocytes (n = 15 ± s.d., p < 0.05 Student's t-test). Asterisk denotes significant difference compared with control. Scanning pipette had a resistance of 100 MΩ and an estimated tip diameter of 100 nm. Effective pixel width in (a,b) is 375 nm over the cell body and 750 nm over the empty area. (M. Miragoli & P. Novak 2010, unpublished data.) (Online version in colour.)

Figure 6.

Measurement of contraction by SICM in cluster of (i) human embryonic stem cell-derived cardiomyocytes (hESCMs) and (ii) neonatal rat ventricular myocytes. (a) (i) hESCMs stain with myosin heavy chain. (ii) Topographical 32 × 32 µm image of cluster of hESCM using SICM. (b) (i) Contraction of hESCM cluster in the presence of (i) doxorubicin and esmolol (ii) resulting in changes in pipette vertical displacement of SICM. As expected, the presence of doxorubicin affects cardiac contraction; this condition is restored by esmolol (ii). (c) Technical scheme of SICM/Ca²⁺ dynamics for concurrent measurement of contraction and intracellular Ca²⁺ transient. Simultaneously, the light emission of the stained cell loaded with Fluo-4 AM was detected by a custom-made photomultiplier tube apparatus. (d) Overlapped traces of Ca²⁺ transient (normalized at % dF/F) and contraction (vertical displacement). (i) Control cluster of cardiomyocytes denote spontaneous firing (approx. 60 b.p.m.). (ii) Same as (i) but in the presence of taurocholic acid that affects calcium transient amplitude and contraction (p < 0.05, Student's t-test). Scanning pipette had a resistance of 100 MΩ and an estimated tip diameter of 100 nm. Topography image in (a) was recorded in the conventional distance-modulated mode with pixel number set to 1024 × 256. Scan duration was 23 min. Modified from Gorelik et al. [51,52] with permission. (Online version in colour.)

Figure 7.

(a) Illustration of the use of SICM for electrophysiological measurement. (i) A scan of a region of neonatal rat ventricular myocytes monolayer with highest thickness = 12 µm. (ii) Representative resting V_m measured with SICM (n = 20). (iii) Scan of a monolayer of cardiac myofibroblasts. Note that the highest thickness (6 µm) corresponds to the region above the nuclei. (iv) SICM allowed successful measurement of V_m in the region of the cell surface above the nucleus (n = 20). Pipette had a resistance of approximately 20 MΩ and an estimated tip diameter of approximately 500 nm. Effective pixel width in topography images was 400 nm, scan duration 20 min. (b) Whole-cell recording in neonatal rat ventricular myocytes using SICM. (i) Resistance of the pipette used for whole-cell recording. The distribution of the seal resistance R_SEAL measured after obtaining stable gigaseal configuration (solid squares). (ii) Schematic of a patch-pipette performing whole-cell recording in a neonatal rat ventricular myocyte. (iii) Example of a whole-cell action potential recording in a neonatal rat ventricular myocytes in the current-clamp mode showing spontaneous action potential firing. Values were corrected for liquid junction potential (n = 42). Pipettes used for whole-cell patch-clamp recording had resistance in the range of 6–9 MΩ and estimated diameter of 1.7–1.1 µm diameter. (M. Miragoli 2009 & P. Novak 2010, unpublished data.) (Online version in colour.)

Figure 8.

L-type Ca²⁺ channel distribution in the cardiac myocytes sarcolemma: mapping of ion channels by the high-resolution scanning patch-clamp technique. (a) Experimental topographic image of a representative rat cardiomyocyte sarcolemma. Z-grooves, T-tubule opening and characteristic sarcomere units are marked. (b) Functional schematic of sarcomere units showing the position of the probed region (Z-groove, T-tubule opening and scallop crest). Probabilities of forming a gigaseal as a function of surface position shown in parentheses. (c) Statistical distribution of L-type Ca²⁺ channels with the highest density near the T-tubule opening. (d) Cell-attached Ba²⁺ current transients at voltages of +20, ±0, −20 mV. (e) Several current transients elicited at 0 mV from one patch and ensemble average of 12 transients showing typical L-type inactivation kinetics. Scanning pipette had a resistance of 100 MΩ and an estimated tip diameter of approximately 100 nm. Topography was recorded in the conventional distance modulated mode with pixel number set to 1024 × 256. Scan duration was 20 min. Modified from Gu et al. [55], with permission. (Online version in colour.)

Figure 9.

Principle of the SICM/FRET technique and its use to study βAR localization in cardiomyocytes. (a) SICM image (32 × 32 µm) of an adult rat cardiomyocyte acquired using a nanopipette from the top of the cell. The sample is positioned on an inverted epifluorescent microscope, so that recordings of cellular fluorescence can be performed. (b) Inset shows a 10 × 10 µm scan of the cardiomyocyte surface with characteristic structural features (cell crests, Z-lines and T-tubule openings). Effective pixel width was 156 nm, scan duration 4 min. The cells are expressing a FRET-based cAMP sensor Epac2-camps, which reports changes in intracellular cAMP levels after local cell surface stimulation via an SICM nanopipette with β1AR or β2AR selective ligands applied either into a T-tubule opening or onto the cell crest. Binding of cAMP to the sensor causes a change in its conformations, which results in a longer distance between the fluorophores (CFP and YFP) and lower FRET signal. (c) Stimulation of β1ARs in both T-tubular (red line) and cell crest region (black line) results in a decrease of FRET, which reflects an increase in cAMP levels. In contrast, β2AR induces cAMP signals only when stimulated in the T-tubule, but not on the cell crest (n = 9). Scanning pipette had a resistance of 100 MΩ and an estimated tip diameter of 100 nm. Modified from Nikolaev et al. [18] with permission. (Online version in colour.)