Variations in the tone of retinal arterioles plays an important role in auto-regulation of retinal blood flow. Our research aims to understand the physiological control of retinal arteriolar tone and how this control is altered in diabetes mellitus. Animal and clinical studies have shown that retinal vasoconstriction and reduced retinal blood flow precede the onset of diabetic retinopathy (Bursell et al. 1996). The studies summarised here provide evidence that this may result from down-regulation of the β1-subunit of the large conductance Ca²⁺ activated K⁺ (BK) channels. This reduces BK-channel activity, which normally hyperpolarizes and relaxes vascular smooth muscle cells. Understanding these changes may allow earlier diagnosis of retinopathy and suggest novel therapeutic strategies. Experiments were carried out using freshly isolated arterioles from male Sprague-Dawley rats (Scholfield & Curtis, 2000). Diabetes was induced by intra-peritoneal injection of streptozotocin and confirmed by measurement of blood glucose and glycosylated haemoglobin. Animals were terminally anaesthetized with CO₂ and retinal vessels isolated 3 months after injection. Tissues were compared with those from age-matched controls. Ca²⁺-measurements and imaging were carried out using isolated segments of retinal arteriole pre-loaded with fura-2 or fluo-4, respectively. Whole cell electrophysiology using the perforated patch technique (amphotericin B in the pipette) was applied to myocytes embedded within arteriole segments but electrically isolated by partial enzymatic digestion. Single channel recordings were made using inside-out patches. Data have been summarised as the mean±SEM. Myocytes in control arterioles generate spontaneous Ca²⁺-sparks, waves and oscillations (Tumelty et al. 2006). This opens up the possibility for feedback via Ca²⁺ -sensitive ion channels, since both Ca²⁺-activated K⁺ and Ca²⁺-activated Cl^– conductances are present in these cells (Scholfield et al, in press). Application of the BK blocker penitrem A (100 nM) decreased the diameter of pressurised retinal arterioles by more than 25% (P<0.05, paired t-test), indicating that BK activity normally limits constriction in these vessels. The BK current evoked by caffeine in voltage clamped myocytes from diabetic vessels was reduced by approximately 90% at 0 mV (P<0.05). This was selective for the BK current as there were no parallel decreases in the caffeine activated Cl^– current or the caffeine-evoked [Ca²⁺]_i transients. BK channel mediated spontaneous transient outward currents were also reduced in diabetic cells (P<0.001 at 0mV), even though the frequency of the Ca²⁺ sparks believed to activate these currents was unchanged and their amplitude (maximal change in normalised fluorescence) was actually increased from a control average of 0.42±0.03 to 0.92±0.06 in diabetic vessels (P<0.001). These data all suggest that Ca²⁺-activated BK activity was reduced by a mechanism downstream of the Ca²⁺ release process. Two observations suggest that this was not simply the result of decreased expression of the BK channels themselves. Quantitative PCR showed no change in mRNA levels for the pore forming α-subunit in diabetic arterioles and the amplitudes of Ca²⁺-independent BK currents elicited by depolarizing voltage steps were similar in control and diabetic arterioles (115±30 pA/pF and 114±26 pA/pF, respectively; NS). Single channel conductance was also unaltered but the Ca²⁺ sensitivity of single BK channels was markedly reduced. There was a rightward shift in the in the open probability vs. [Ca²⁺] relationship at +80mV for channels in inside-out patches from diabetic retinal arteriolar myocytes, with a parallel reduction in the Hill slope. When the voltage-dependence of channel activation was determined at a fixed [Ca²⁺] of 10µM, this was shifted to the right by >100mV in diabetic tissues. These observations are consistent with down-regulation of BKβ1 activity, since this subunit confers Ca²⁺-sensitivity to BK channel complexes (Brenner et al. 2000). Transcript levels for BKβ1 were appreciably lower in diabetic retinal arterioles and protein expression, as estimated using immunohistochemistry and confocal microscopy, was also reduced. The mean open times and the sensitivity of BK channels to tamoxifen were also decreased in diabetic cells, observations which are also consistent with a down-regulation of BKβ1-subunits. Overall, these results support the hypothesis that diabetes mellitus down-regulates expression of the BKβ1-subunit within retinal arteriolar myocytes and that this may explain the retinal vasoconstriction seen in the early phases of diabetic retinopathy.