The goal of the present study was to image sub-cellular Ca²⁺ transients in microvascular smooth muscle (MVSM) cells of intact, non-pressurised, retinal arterioles and to quantify the spatiotemporal coupling of these signals at an intercellular level.

Sprague Dawley rats (200-300g) were anaesthetised using CO₂ and killed by cervical dislocation. Arterioles were mechanically dispersed from fresh retinae using a fire-polished Pasteur pipette and MVSM cells loaded with 10 µM Fluo-4AM for 2 hours. Changes in [Ca²⁺]_i were imaged in MVSM cell arrays (9-17 cells) using confocal scanning laser microscopy in X-Y (2 images/s) or line scan mode (500 scans/s). Raw fluorescence data were extracted in numerical format using Image J (NIH) and cell-to-cell coupling was evaluated between adjacent cells using the cross-correlation coefficient (CCF; SPSS) applied to data derived from line scans parallel to the long axis of each vessel.

Under resting conditions X-Y scans revealed both discretely localised, spontaneous near-membrane Ca²⁺ events resembling Ca²⁺ sparks, and more global and prolonged Ca²⁺ transients. Line scan data was characterised by sporadic periods when global Ca²⁺ signals appeared to be entrained among up to six adjoining cells, interspersed with periods when little entrainment was obvious on visual inspection. A mean CCF of 0.38 ± 0.03 ( ± S.E.M.) was obtained in 56 MVSM cell pairs from 6 vessels, indicating that overall, intercellular coupling was weak. However, regions selected on the basis of having either: (a) one Ca²⁺ event which appeared to propagate across three or more adjacent MVSM cells, or (b) three or more temporally associated Ca²⁺ transients in a MVSM cell pair, had a mean CCF of 0.72 ± 0.04 (5 vessels; 7 regions). The lag time which gave the peak correlation for the propagating Ca²⁺ transients in these regions of strong coupling varied between 10-414 ms (mean 169.2 ± 26.1 ms). This delay was unrelated to the amplitude of the [Ca²⁺]_i rise in the lead cell (r² = 0.079; 5 vessels; 22 paired Ca²⁺ transients). When intercellular propagation was seen in a group of cells, the cell firing the initial Ca²⁺ transient could vary, suggesting the absence any fixed pacemaker.

We conclude that retinal MVSM cells exhibit both localised and global Ca²⁺ events whilst still embedded within their parent arterioles. These signals may occasionally synchronise in neighbouring cells. The mechanisms responsible for this coupling warrant further investigation.

We thank Fight for Sight and The Wellcome Trust for financial support.