Effects of some anesthetic agents on skin microcirculation evaluated by laser Doppler perfusion imaging in mice

Microcirculation is the final link between the cardiovascular system and cellular interfaces and, ultimately, molecular processes. Many studies have investigated the effect of anesthetics on peripheral and systemic microcirculation in humans [1–5], especially their effects on microvascular perfusion, aiming to ensure adequate tissue oxygenation and nutritional supply. Mice are an ideal model to study anesthetic action due to their easy manipulation, well-established behavioral and homeostatic responses to anesthesia, and well-known genetic background. Outbred mouse strains are widely used in toxicology and pharmacology, and CD1 mice have been employed in anesthesia research [6–8] on the assumption that most characteristics of interest have a polygenic inheritance and are related to phenotypic variation in a genetically heterogeneous population [9, 10]. Moreover, anesthesia is required for most in vivo studies using mouse microcirculatory models, and the use of diverse anesthetic agents in translational research can interfere with experimental results [11, 12]. As an example, pentobarbital [13, 14], midazolam-medetomidine [15] and isoflurane [16] have been used in preclinical studies on peripheral arterial disease to evaluate the effects of new angiogenetic therapies. Microcirculatory responses to the most popular inhalation (halothane, isoflurane) or injectable anesthetics (propofol-fentanyl, barbiturates and ketamine) have been investigated in rats at the level of intestinal [17], cremaster or dorsal muscle microcirculation [18–20] using invasive dorsal microcirculatory chambers or intravital microscopy. So far, few data have been reported regarding the microvascular effects of the popular laboratory-animal anesthetic agents acepromazine and dexmedetomidine. Acetylpromazine maleate is an α-adrenergic receptor antagonist broadly used for sedation and balanced anesthesia in animals [21]. Concurrent administration of acepromazine reduces the required dose of isoflurane while potentiating peripheral vasodilation and lowering blood pressure in dogs [22]. The combination of acepromazine with ketamine and xylazine is recommended for a safe and reliable surgical anesthesia in mice, although it is associated with marked hypotension [23, 24]. Dexmedetomidine hydrochloride is a selective α₂-adrenoceptor agonist with preferential affinity for α₂A and α₂B receptors [21]. Perioperative administration of dexmedetomidine hydrochloride reduces the required doses of isoflurane, thiopental and propofol in humans and animals, and it reduces the activation of the sympathetic nervous system during surgery, preventing harmful hemodynamic events such as acute kidney injury [25]. Reliable techniques for measuring perfusion in accessible tissues such as skin may have significant potential to improve our understanding of microvasculature regulation under anesthesia. Laser Doppler perfusion imaging (LDPI) is a noninvasive technique allowing real-time quantification of skin perfusion in two-dimensional color-coded images. Enhancement of the measured area provides a better evaluation of blood flow heterogeneity, allowing for the identification of subtle changes in skin perfusion induced by anesthesia and indicating circulatory status in other areas [26, 27]. Although LDPI offers a simple and accurate estimate of peripheral perfusion, a standard method for the study of microcirculatory changes related to anesthesia in mice is lacking. In the present study, we reviewed several biological variables, such as gender, environmental variables and operational variables, such as body temperature, skin district and recording conditions, to develop a LDPI protocol to evaluate the effects of some anesthetic agents on microcirculation in mice. Our LDPI protocol is a potentially valuable research tool to detect in vivo real-time microcirculatory changes in preclinical experiments in mice.

Anesthetics modulate microcirculation mainly via autonomic sympathetic and parasympathetic nerves on vascular smooth muscle. Phenothiazine tranquilizers as well as α₂-agonists exert their hemodynamic effects mainly by interacting with α-adrenergic-receptors. Phenothiazines cause vasodilation predominantly by blocking α₁ receptors but are also dopamine receptor antagonists [28]. While D₁-like dopamine receptors induce relaxation of resistance arteries [29, 30] D₂-like dopamine receptors are typically present on postganglionic sympathetic neurons, where their excitation leads to a reduction of the neural release of norepinephrine, inducing a passive fall in vascular resistance and heart rate [31]. Dexmedetomidine is a selective α₂-adrenoceptor agonist that shows a dose-dependent, preferential affinity for α₂A and α₂B receptors [21], evoking a biphasic blood pressure response: a short hypertensive phase mediated by the α₂B receptors, followed by hypotension mediated by the α₂A receptors [32, 33]. The peripheral hemodynamic effects of phenothiazines and of α₂-agonists thus differ: while acepromazine causes significant hypotension in isoflurane-anesthetized animals [34], dexmedetomidine [35, 36] increases peripheral vascular tone, counteracting the isoflurane-induced vasodilation and reduction in arterial blood pressure [22]. LDPI permits a noninvasive, real-time measurement of microvascular blood flow using two-dimensional color-coded images of skin perfusion. The use of laser doppler flowmetry technique to detect the sympathetic tone during general anesthesia in humans has been reported [3, 37, 38], and translational approaches using LDPI in microvascular perfusion mouse models offer the advantages of being easy and fast [39]. Moreover, various anesthetics alter blood flow in rodents [40–43], so anesthetic regimens used in mouse microcirculatory models should be taken into account, and they should not adversely affect the vascular bed to be examined. The main finding in this study is that LDPI is able to evaluate in real time the anesthesia-induced changes in mouse peripheral microcirculation. The hemodynamic effects recorded during the different anesthetic protocols were as expected based on previous clinical and animal studies, although the analyzing techniques were different [2–5, 22]. Because LDPI scans are disrupted by motion [44] and single or combined sedatives have lacked restraining effects in mice, we chose to perform our study under 1.5% isoflurane anesthesia to record reference perfusion values. Isoflurane produces only minor effects on murine hemodynamic status [6, 7]. Costantinides et al. (2011) [45] reported that 1.5% isoflurane produces stable body temperature, mean arterial pressure (MAP) and heart rate (HR) values in mice, comparable to those observed in awake animals, so they recommended it for physiological and pharmacological studies of cardiac function and to facilitate translational research in non-invasive imaging platforms. In the present study, isoflurane anesthesia yielded a reproducible and stable effect on peripheral blood perfusion over time (P = 0.1). Acepromazine increased isoflurane plantar perfusion, as reported by Lemke et al. [22], and reduced vascular tone and arterial pressure due to its α-blocking action. After dexmedetomidine administration, a rapid and intense decrease in plantar perfusion was followed by a longer phase of increased perfusion, in agreement with the typical biphasic hemodynamic effect of this class of sedatives. In all of the mice, the increased perfusion recorded 15 minutes after dexmedetomidine administration did not surpass the perfusion brought about by isoflurane (P = 0.6), and it was noticeably lower than the perfusion recorded following acepromazine administration (P = 0.01), which did not increase significantly even after atipamezole injection (P = 0.9). Special care was taken to avoid methodological bias. To date, the skin has been used as a model of microcirculation to investigate vascular mechanisms in cardiovascular [46–48] or kidney diseases [49] and diabetes [50, 51]. Autonomic innervation of microvessels in the region of interest [3, 28, 52–54], somatic stimulation of cutaneous arterial sympathetic nerve activity [55], positioning and body temperature [56] are all crucial factors affecting skin blood flow measured by LDPI. Glabrous skin areas are highly innervated by noradrenergic sympathetic vasoconstrictor nerves [3, 28], which are regulated by α-adrenoceptors [57] in several species [52–54]. For these reasons, we chose plantar region to investigate blood flow changes brought about by anesthetic drugs, also avoiding hair clipping, which might affect cutaneous arterial sympathetic nerve activity and alter LDPI measurements [55]. In our setting, precise hind plantar surface positioning was achieved to keep the site of interest highly symmetrical and precisely perpendicular to the laser beam (Figure 1) [58–60]. Moreover, body temperature was monitored by a rectal probe and adjusted between 35.5-36.5°C by an infrared lamp. Our experiments were performed in a temperature-controlled room [29], and we started LDPI recordings after each animal had acclimatized. The effects of sex hormones on vascular tone continue to be a matter of debate [61, 62]. Stucker et al. (2001) [63] reported in an LDPI study only a tendency toward higher perfusion values in men than in women, stating that moderate gender differences in skin perfusion between study groups should be tolerated. Similarly, Kunkel et al. (2007) [64] found that foot skin perfusion in normal human subjects was independent of gender. In our experience, no significant differences between males and females were found in the peripheral blood flow in either control or treated animals. In accordance with the manufacturers’ technical instructions, the room lighting should be kept to a minimum brightness. We set our lighting discrimination between background and the site of measurement to the default threshold level of 6.2 volts, adjusting the backscattered laser light intensity in the range of 7–9 volts, obtaining an optimum quality of data. To compare perfusion in different images, a user-defined color scale was adopted during the acquisition process, ranging from 0 to 5 volts (perfusion output value of 0 volts was calibrated to 0% perfusion, whereas 5 volts was calibrated to 100%). The average perfusion in each region of interest (ROI) was normalized to the wall plantar surface area to reduce bias related to unavoidable anatomical and position variance. To further minimize any data divergence, the hind paw perfusion value for each animal at each time point was calculated as the average value of both hind paw ROIs.

LDPI is able to evaluate noninvasively and in real time the skin microcirculation changes induced by general anesthesia in mouse models. LDPI could be useful for studying the effects of anesthetics on peripheral microcirculation and to avoid the inconsistent use of anesthetic agents in cardiovascular translational research. Standardization of an appropriate LDPI procedure is needed in preclinical studies to avoid bias in experimental results.