A low‐cost, scalable technique to study distal coronary arteriole function

Although coronary heart disease is most often tied to obstructions of flow by atherosclerotic plaques in large epicardial arteries, about 10–20% of patients with chest pain considered to be cardiac in nature have normal coronary angiograms (Cannon et al. 1992). In these cases, coronary disease is often linked to contractile dysfunction, for example in relation to metabolic syndrome or obesity (Abel et al. 2008, Pries et al. 2008). Conversely, if the problem is large-artery atherosclerosis, the adverse condition may be partially offset by adaptive responses of the microvessels, such as increased vasodilation (Abel et al. 2008). In both the cases, microvessel function constitutes an important part of the pathology. Despite the essential function of coronary arterioles in the heart, current methodology prevents vessels smaller than 40–50 μm to be studied ex vivo (Goebel et al. 2003). In humans and most animal models, this is two to four times the dimension at the most distal end. According to Poiseuilles law, which defines vascular resistance as the product of vessel length and blood viscosity divided by radius squared four, resistance generated by the regions that have been most studied is 16–256 times less than that by the regions which have thinnest lumen. Taken together with the extensive work specialization of larger coronary vessels (Kuo et al. 1995), it is almost certain that the distal coronary arteriole harbours important vasoregulatory features not yet discovered. In vivo microscopy provides access to a much larger span of the coronary vascular tree than ex vivo techniques, including the distal coronary arteriole (Kuo et al. 1995, Goebel et al. 2003). Although such approaches have so far seen limited use in the literature, we believe imaging techniques such as intravital multi-fluorescence microscopy (Schramm et al. 2007) will provide important information about coronary circulation in the future. Nevertheless, access to the same vascular segments in an ex vivo setting will be essential, as regulation can then be studied and understood isolated from the multitude of neuronal, humoral and local stimuli present in the animal. In addition, the vessels can be exposed to precisely known concentrations of agonists, antagonists or drugs, and the environment can be manipulated to a greater extent than in living tissue. In response to this challenge, we have adapted the agarose preparation, originally developed for renal vessels (Loutzenhiser & Loutzenhiser 2000), to study diameter changes and signalling in coronary vasculature. Small arteries and arterioles are prepared with relatively little effort and modest investments in equipment and training. The preparation provides access to many segments of the coronary microcirculation, including the distalmost coronary arteriole. Its main area of use is to map signalling pathways, compare arteriole-performance and study drug effects. All chemicals were bought from Sigma-Aldrich®(St. Louis, MO, USA) unless otherwise specified. All experiments conformed to Good Publishing Practice in Physiology (Persson & Henriksson 2011). Four male Wistar rats weighing 200–300 g were bought from Möllegaard breeding farm, Denmark (Taconic), and eight male C57Bl6/J mice weighing 20–25 g were bred at Dyrestallen, Institute of Biomedicine, University of Bergen. All animals had free access to standard rat food and tap water. The Roswell Park Memorial Institute (RPMI) media used during diameter measurements contained (in mmol L−1) NaCl 130.90, KCl 5.37, MgCl2 2.13, NaH2PO4 1.67, HEPES 5.62, glucose 5.55, Na pyruvate 1.00, NaHCO3 4.17, CaCl2 1.98, RPMI vitamins (product no. R7256) and amino acids (product no. R7131). During the isolation procedure, CaCl2 was omitted. Agarose/water solution was prepared by gradually adding Seaprep agarose (Lonza, Rockland, ME, USA) to stirred room temperature water until the concentration was 40 mg mL−1. The solution was stirred at 80 °C until all the agarose was completely dissolved. Aliquots of 8 mL (mice) and 50 mL (rats) were stored at 4 °C for up to 1 week. At the day of the experiment, the agarose aliquot was added to the same volume of 2× RPMI buffer without CaCl2 and warmed to 60 °C and mixed, producing a final agarose concentration of 20 mg mL−1. Roswell Park Memorial Institute buffer was added to 0.5 mg mL−1 collagenase from Clostridium histolyticum (product no. C5138, activity: 396 μ mg−1, batch 086K8619), 0.5 mL mL−1 protease from Streptomyces griseus (product no. P8811, activity: 5.9 μ mg−1, batch 027K1514) and 0.01 mg mL−1 trypsin inhibitor from Glycine max (product no. T6522). The solution was filtered using a non-adhesive 0.2-μm sterile filter. Animals were anaesthetised with an intraperitoneal injection of 50 mg kg−1 pentobarbital (Haukeland Hospital Pharmacy, Bergen, Norway). In rats, the upper body was infused with 50 mL (8 mL min−1) warmed (37 °C) RPMI agarose solution through a caudal aorta incision using a blunted hypodermic needle (21 G) secured with a ligature. The agarose was fed using an automatic syringe piston as described before (Helle et al. 2013). In mice, a Terumo I.V venflon was inserted into the aorta through the branching iliac arteries, and >8 mL RPMI agarose (37 °C) was infused by hand. In both species, vena cava was cut to vent excess perfusate, and animals were killed by clipping the diaphragm while anesthetized. Immediately after agarose infusion, the heart was transferred to 4 °C RPMI and incubated for 10 min to solidify the intravascular agarose. The right ventricular wall was dissected from the heart, and 100-μm-thick slices were cut using a Thomas slicer (Thomas Scientific, Swedesboro, NJ, USA) and incubated in the RPMI enzyme solution for 80 min at 35–37 °C. After removing enzymes, slices were gently shaken in a 15-mL tube and transferred to a BSA-coated Petri dish with RPMI medium (room temperature). Agar-filled coronary arterioles were seen free-floating or protruding from small chunks of myocytes (Fig. 1a), in which case they were cut loose with a BSA-coated No. 15 scalpel. If there was less amount of cardiomyocytes and connective tissue, as in Figure 1b, vessels were used without further manipulation. Arterioles identified positively by the presence of capillaries were harvested by a 150-μm ø BSA-coated pressure-controlled glass pipette. Vessels were transferred to a new glass-bottom Petri dish (P35G-1.5-14-C; MatTek, Ashland, MA, USA) with 3 mL RPMI without CaCl2 and stuck by self-adhesion to the glass. The ability of the vessel to self-adhere was dependent on low concentration of cell debris and other pollutants (such as BSA) in the medium. Before recording, temperature (from room temperature: 2–5 °C min−1) and Ca2+ concentration (3 min steps: 10−5, 10−4 and 10−3 m) were increased gradually. Comparing lumen diameters with microsphere, experiments previously performed at our laboratory indicates that the agarose method produces maximally dilated vessels (Hansen et al. 2005). Venules were rarely seen in the suspension, but could be discriminated from their cousins by much thinner vessel wall, protruding nuclei (in mice) and a spiral-ling pattern of the SMCs (in contrast to the parallel pattern shown in Figure 1c upper left) when encountered. A chamber insert was placed in the Petri dish to hold the tubing and restrict volume to 0.8 mL. The Petri dish was then placed on a warmed microscope stage connected to a perfusion inline heater (35–36 °C, controlled by Warner TC344-B) with a gravity-fed flow of 1−2 mL min−1. Vessels were stimulated by 10−9, 10−8, 10−7 and 10−6 m ANG II followed by washout and 50 mm KCl, all periods lasting 3 min. Phenylephrine (PE) doses were 10−7, 10−6, 10−5 and 10−4 m. Each vessel was stimulated with PE or ANG II once, thereafter KCl. In all experiments, depolarization was used as a measure of viability. If the vessel did not visibly contract when stimulated with 50 mm K+ at the end of the protocol, all measurements from that arteriole were discarded (including the preceding agonist stimulation). The switching between perfusion fluids was done automatically with a Valvebank8® (AutoMate Scientific, Berkeley, CA, USA). Image time series (2–4 images min−1) were acquired using an UltraView IIIu camera mounted at an Olympus IX70 microscope using 10 or 20× magnification. Lumen area and length were measured at every time point using the ‘Area/Perimeter’ (only area was used) and ‘Polygon Length’ tools in Cell^B from Olympus. Mean lumen diameter was then calculated at every time point using the formula (mean lumen diameter) = (lumen area)/(lumen length). Fifty microgram of Fura-2 acetoxymethyl (AM) ester was dissolved in 10 μL of dimethylsulphoxide (DMSO) (both Invitrogen, Carlsbad, CA, USA). The solution was added to 6 mL of RPMI medium and stored away from light to avoid photobleaching. Isolated coronary arterioles were incubated in 3 mL of the Fura-2/RPMI solution for 90 min at room temperature, before they were transferred to a BSA-coated Petri dish with clean RPMI medium for 30 min to allow de-esterification of Fura-2. The vessels were then placed in glass-bottom Petri dishes (glass thickness 0) with the same chamber insert as described above for the diameter recordings. The vessels were excited with alternating lights of 340 and 380 nm, and fluorescence was collected at 510 nm. Images were captured with a frequency of 2 Hz and analysed with ImageMaster 5 software (Photon Technology International, Birmingham, NJ, USA). Regions of interest (ROIs) were defined, and background fluorescence from a nearby region not covering the vessels was subtracted from each wavelength before the 340/380 ratio was calculated. The peak and plateau changes in ratio from baseline are reported. Stimulation was performed by switching the perfusate to 50 mm KCl, as described above. Student's t-tests were used to compare the mean lumen diameter after depolarization. Two-way repeated-measures anova was used to compare PE dose responses in rats and mice. All tests were performed in SigmaPlot version 12.3 (Systat Software, San Jose, CA, USA). P < 0.05 was considered significant. All values are reported as mean ± SEM. The surgical procedure was straightforward and did not pose technical difficulties in neither mice nor rats. Gentle flushing of the slices (5 mL medium from a Pasteur pipette) produced small holes after about 80 min of enzyme treatment, indicating that slices were suitably digested. We were able to isolate about 5–15 arterioles from each slice of the ventricle. Only vessels positively identified as distal by attached capillaries were used in the experiments. Morphology was well preserved in the preparation (Fig. 1a, b). Figure 1a shows a terminal coronary arteriole ending into two capillaries, one going upwards and the other leading into some remnants of the heart muscle (arrows). The vessel in Figure 1b also displays a branching capillary (arrow), but does not contain the terminal part as seen in Figure 1a. Baseline lumen diameter was tested in three vessels for 30 min during perfusion with RPMI and did not visibly deviate from the lumen tracing at t = 0, indicating a stable baseline lumen diameter. The distal coronary arteriole lumen diameter for all the mice was 21.9 ± 1.3 μm (n = 10). These vessels were divided randomly in two groups stimulated with PE and ANG II, which had baseline lumen diameter of 20.0 ± 1.4 μm (n = 6) and 24.3 ± 1.9 (n = 4) respectively (NS). In rat vessels stimulated with ANG II, the lumen diameter was 25.2 ± 1.2 μm (n = 4). During contractions, the outline between lumen and vessel wall was readily visible (Fig. 1c, d) and could be traced to calculate mean lumen diameter (Fig. 1e–f). After washout, vessels reverted to about 90–95% of baseline diameter, which permitted a second exposure to KCl to test vitality. The incomplete return to baseline diameter is in line with earlier experiments with renal microvessels (Dahl et al. 2011) and may explain why depolarization in some instances produced smaller responses than the preceding agonist stimulation. Figure 1g–i shows baseline and Figure 1j–l peak Ca2+ fluorescence in a Fura-2-loaded arteriole (upper right) situated on a larger vessel. Three vessels (two from rat and one from mouse) were excluded from the study due to absent contraction after depolarization. Vessels from both species contracted when stimulated with PE and KCl (Fig. 2a). The dose–response curves in mouse and rat vessels were not significantly different (P = 0.069), while depolarization induced stronger contractions in mice (P < 0.05). To test whether other agonists than PE could be used, mice vessels were exposed to ANG II, which induced reproducible contractions as well (Fig. 2b). Arterioles from mice (n = 8) stimulated with 50 mm KCl elicited a sharp increase in Fura-2 ratio, followed by a sustained phase which returned to baseline values after washout of the compound (Fig. 3a). At baseline, the Fura-2 ratio was 0.74 ± 0.02, and the peak ratio increase from baseline was 0.21 ± 0.03 (Fig. 3b). For more than a decade, the agarose preparation has provided renal physiologists with easy access to the numerous segments of renal microcirculation. Cardiac physiologists, on the other hand, have lacked a low-threshold alternative to pressurized arteriole preparations and none to study the most distal region of the coronary arteriole ex vivo. Because of this, many studies on coronary vasoregulation have relied upon whole-organ blood flow only. As only the integral action of structurally and functionally diverse (Cohen & Kirk 1973) arteriole segments is taken into account in these cases, important details necessary to understand the complex regulation of coronary microcirculation may have been overlooked. Important vasoactive compounds have been shown to have dramatically different effects on coronary vessels of different sizes, for example, vasopressin causes vasoconstriction in arterioles smaller than 100 μm in diameter, but has no or a somewhat vasodilatory effect on those larger than 100 μm in diameter (Lamping et al. 1989). This and more findings suggest that regulatory mechanisms in vessels of different sizes are strikingly different (Haegeman & Fiers 1978). One much studied coronary hormone is PE, which for 50 years has been recognized to produce coronary vasodilation (Berne 1958), although the direct effect on coronary resistance is actually vasoconstriction. The apparent vasodilation is a result of increased myocardial metabolism (Crystal et al. 1991), which is tightly linked to coronary blood flow (CBF) (Williams & Most 1981). As such, the contractions observed in agarose-infused arterioles are consistent with the CBF responses from earlier reports and suggest that a significant portion of coronary resistance is generated in the distal coronary arteriole. Determining the exact magnitude of the resistance generated by agarose vessels should be performed with care, however, as the tensile of the intravascular is to allow reproducible contractions, not to intravascular that PE blood flow in muscle also by effect on arteries et al. We are not of findings in coronary arteries, but we it that this hormone may have vasoregulatory effects in vessels from the coronary arterioles to the large circulation in the is to ANG II, but the between In the larger arteries, ANG II produces a increase in while more contractions are produced in resistance vessels (Cohen & Kirk The responses seen in mice in the present are in line with these although we did not contractions as This could be due to species or that the vessels used in the present study are from more distal regions of the or that the nature in vivo is by a which is not present in the ex vivo This to study ANG II vasoregulation ex vivo may be as prevents the effects of et al. a which has also been observed in the (Helle et al. We were able to about 10 arterioles from each dissected right all positively identified as by attached with this can the of animals used for experiments. the that diameter and regulatory such as signalling with can be measured in vessels from the same animal. This is as it that regulatory (Hansen et al. and in have also been performed on agarose preparation but the amount of arterioles that to diameter in to these One of the agarose preparation, which may be a or on the is the of a in after In a preparation, it can be to intravascular away from the pipette. In the agarose preparation, the a and the length of the vessel. A of the agarose preparation is the of which the preparation to The of flow can in some be an however, as it may other In a study on it was that agonists are able to during certain in the This example may be for coronary circulation as as in coronary metabolism have been (Helle et al. et al. 2013). of the agarose preparation is of In this study and studies with renal baseline lumen diameter only about to baseline after agonist the time with other preparations (Dahl et al. 2011). we not whether the is the of the agarose or the contractile of the vessel During the of the agarose in contractions were absent in renal vessels after of with that the vasopressin was both present and et al. The was a trypsin in the collagenase which the of the vasopressin before experiments the was was by adding a small amount of inhibitor (Hansen et al. 2005). are not in enzyme and should be in if is encountered. The of the are by to the vessel. is essential that the enzyme is for the tissue, and batch of During vessels are with 2–5 mL medium using a Pasteur mL If small holes or arteriole and response to K+ depolarization should be of the enzyme solution should be prepared and stored at °C to enzyme at a stage if further experiments are The concentration should be if or contractile responses are as this can contractions significantly greater than the background agarose concentration of 20 mg mL−1 was used in the present which is when contractile stimuli or For vessels or agonist we 15 mg agarose mL−1. vessels more enzyme before they are with normal In with of vascular vessels may be as they are more For groups with different of it may be to an and enzyme concentration which is for In the present experiments, RPMI was used, but other the environment will work as Ca2+ on the other hand, should be low during and If present at a during these often of the vessels and of the diameter One for this is that pathways, which may be during the of the et al. 2003). In the present capillaries were used as a of distal This was necessary features are after and enzyme We believe that lumen diameter, wall and are features to vessels as In contrast to the agarose-infused most other vascular preparations to measure and vascular This may be if for example the infused agarose is by epicardial coronary arteries during in however, during the of blood when the heart is incubated at 4 °C for of the agarose. wall in the final preparation will be a between the vascular of the vessel and the tensile The final preparation displays a low in baseline lumen diameter, when measured to the capillary work in Taken these a low in In the agarose preparation is a low-threshold alternative to preparations of the coronary arteriole. a numerous vessels from vascular segments may be This the distalmost region of the coronary which previously has not been ex vivo. responses dependent on such as vasodilation or the are not vessels on the other hand, well to study contractile responses and much of the a to study vasoregulation in the distal coronary arteriole. to and at Institute of Biomedicine, Bergen, for the rats used in the and mice for to for on the of this The study has been by from the This no part in or of the present