Transcutaneous vascular ultrasound in hypercholesterolaemic rabbits: a new method to evaluate endothelial function Katrin Hiss, Kerstin Steioff, Matthias Loehn and Hartmut Ruetten Aventis Pharma Deutschland GmbH, Cardiovascular Therapeutic Department, 65926 Frankfurt, Germany
Summary Measurement of endothelial function in patients with atherosclerosis and lipid disorders is an important tool for the risk evaluation of a cardiovascular event, such as acute myocardial infarction and stroke. The feasibility of measuring endothelial function non-invasively in animal models has been limited so far. Therefore, we compared the assessment of endothelial function by in vivo transcutaneous vascular ultrasound (TVU) with the classical method of ex vivo organ bath, using the carotid artery of hypercholesterolaemic and normocholesterolaemic rabbits. The assessments of endothelial function by both techniques were performed on the same segments of the carotid artery. Vascular ultrasound detected impaired endothelium-dependent vasorelaxation induced by acetylcholine in the common carotid artery of hypercholesterolaemic rabbits. These results strongly correlated with measurements of endothelial function of isolated carotid artery rings. Furthermore, atherogenic diet caused significant fatty streak formation in the aorta, as well as significant increase of C-reactive protein and cholesterol levels. Endothelial function, an early marker of cardiovascular risk, could be non-invasively assessed and graded by TVU measurements. It correlated highly with vasoreactivity of isolated vessels in an organ bath (r2 ¼ 0.68). We conclude that vascular ultrasound in hypercholesterolaemic rabbits is a valid method for evaluating endothelial function associated with atherosclerosis. Keywords Rabbit; hypercholesterolaemia; endothelial dysfunction; ultrasound;
endothelium-dependent relaxation
Atherosclerosis is the underlying cause of several cardiovascular diseases, comprising coronary heart disease, angina pectoris, myocardial infarction, and ischaemic stroke. Furthermore, recent studies have shown that the severity of endothelial dysfunction correlates with the risk for cardiovascular events (Schachinger & Zeiher 2001). Endothelial dysfunction associates with the morphological alterations of the vascular Correspondence: Katrin Hiss, DVM, Sanofi-Aventis, Cardiovascular Therapeutic Department, Industriepark Hoechst, Building H 821, 65926 Frankfurt, Germany. Email:
[email protected] Accepted 5 July 2005
wall, the main characteristic of atherosclerosis. In patients, endothelial dysfunction has been assessed by several methods: coronary diameter response to acetylcholine (Ach), forearm bloodflow response to Ach, and brachial artery flowmediated dilator response to postischaemic hyperaemia (Monnink et al. 2002). To understand better the pathogenesis of atherosclerosis and endothelial dysfunction, predictive animal models are needed and should meet several requirements. First, they should share the pathophysiology of the disease in humans and, secondly, the
r Laboratory Animals Ltd. Laboratory Animals (2006) 40, 80–86
Non-invasive vascular ultrasound in rabbits
determined functional and biochemical parameters should be comparable to the standard in patients. Dietary lipid manipulation in normal and genetically defective (e.g. Watanabe heritable hyperlipidaemia rabbits) rabbits was used to study atherogenesis for many years. Until recently, the non-invasive determination of endothelial function in these animals was not feasible and not validated. In this study, we investigated the use of transcutaneous ultrasound (TU) in the assessment of vascular function in vivo and compared the method with the wellestablished method of evaluating vascular reactivity in isolated arteries. Additionally, we measured the levels of C-reactive protein (CRP), a non-specific marker of inflammation, and serum cholesterol as significant biochemical parameters that are also measured in patients with atherosclerosis (Elgharib et al. 2003).
Materials and methods Animals and husbandry
Male New Zealand White (NZW) rabbits were obtained from Harlan Winkelmann (Borchen, Germany). The animals underwent an acclimatization period of 10 days after arrival at the facility. At the beginning of the study, animals were at the age of 3–4 months and weighed 2.5–3.5 kg. Rabbits were housed singly in plastic cages (82 66 65 cm, Easy Connectable 2-cage, Scanbur, Karlslunde, Denmark) and maintained under standard conditions: chow (Altromin, Lage, Germany) and tap water ad libitum, 45–55% relative humidity, 181C temperature and light cycle at 12/12 (300 Lux brightness). Cages, animal bedding, and food were sterilized and the entrance to the vivarium was restricted to persons wearing clean coats, surgical caps, masks, and overshoes. The research procedures and housing conditions were in accordance with the German Law of Animal Welfare. Experimental design
Rabbits were divided into two groups with different feeding protocols. Group 1 (n ¼ 6)
81
was fed a commercially available pellet diet for rabbits for 20 weeks. Animals of group 2 (n ¼ 8) were fed an atherogenic diet (3% cocos oil, 0.25% cholesterol) for four weeks. After four weeks, blood samples were collected from the animals of group 2 to determine the cholesterol levels. Only animals with cholesterol levels >10 mmol/L were included in the study (n ¼ 6) and continued the atherogenic feeding protocol for an additional 16 weeks. At week 19, rabbits of both groups were slightly sedated and functional measurements of the Arteria carotis were performed by TU. Animals were killed in week 20.
In vivo measurement of endothelial function of the A. carotis
At week 19, TU was performed under sedation (ketamine 50 mg/kg, Intervet, UnterschleiXheim, Germany and xylazine 10 mg/kg, Bayer Vital, Leverkusen, Germany), using a 10–5 MHz linear array transducer interfaced with a HDIs 3000 (ATL, Solingen, Germany). Images were printed on a video graphics printer (Sony Electronics, Tokyo, Japan). Animals were placed in dorsal recumbency on a heater blanket (35–371C) and the ventral neck was shaved. The transducer was lubricated with ultrasound gel (Heiland, Hamburg, Germany) and placed laterally to the trachea on the ventral neck. The right carotid artery (A. carotis communis dextra) was visualized on the monitor and the image was frozen at the largest cross-section. The maximal width of the cross-sectional diameter was determined perpendicular to the transducer. The measurements were performed at baseline and during systemic intravenous (via ear vein) infusion of Ach (5 mg/kg/min; Sigma-Aldrich, Steinheim, Germany) 2 min after the beginning of infusion. The transducer remained at the same position for the duration of measurement. The change of diameter after Ach infusion was stated as percentage of diameter at baseline (delta [D], %). The cross-sectional measurements were repeated three times at baseline and during Ach infusion; averaged data are reported. At the killing in week 20, blood was collected Laboratory Animals (2006) 40
82
K Hiss et al.
for determination of CRP and cholesterol. The aorta and the A. carotis were dissected. Ex vivo measurement of endotheliumdependent relaxation of the A. carotis
Animals were killed and three rings from the carotid artery of each animal were mounted on force transducers in organ baths gassed with carbogen at 371C and pH 7.4, as described previously (Lohn et al. 2002). After equilibration for 1 h, contraction was induced by KCl (60 mmol/L, Sigma-Aldrich, Steinheim, Germany) and, after washout, precontraction was induced by phenylephrine (1 mmol/L, Sigma-Aldrich, Steinheim, Germany). Thereafter, stepwise endothelium-dependent relaxation was induced by Ach (1 nmol/L to 1 mmol/L).
connective tissues and then cut longitudinally and stained with oil-red O (Sigma-Aldrich, Steinheim, Germany) to visualize the atherosclerotic plaque area (fatty streaks). Total and plaque-covered areas were quantified by planimetric analysis (Leica Imaging Systems, Cambridge, UK) of the photographed pictures and plaque area was expressed as a percentage of the total aortic area. Statistics
Values are given as mean7SEM. Statistical significances in mean values were tested by unpaired two-sided Student’s t-tests. A value of Po0.05 was considered statistically significant. All data were analysed with SigmaStat statistical software.
Measurement of cholesterol, triglyceride, and CRP levels
Results
Blood samples were obtained through puncture of the Arteria auricularis and collected in an ethylenediaminetetraacetate (EDTA) or serum tube (Sarstedt, Nu¨mbrecht, Germany). Plasma CRP levels were determined with a commercially available enzyme-linked immunosorbent assay (ELISA) (American Diagnostica, Pfungstadt, Germany). Serum cholesterol levels were determined with a commercially available enzymatic in vitro test (Roche Diagnostics, Mannheim, Germany) on an automated clinical chemistry analyser (Roche Diagnostics).
Hyperlipidaemia and inflammatory response
Determination of aortic plaque size
The aorta, dissected from the aortic arch to the aorticoiliac bifurcation, was fixed in 10% formalin (Sigma-Aldrich, Steinheim, Germany). The aortas were freed of adherent
A significant hyperlipidaemia and hypercholesterolaemia was observed in rabbits of group 2, which received the atherogenic diet for 20 weeks. Furthermore, fatty streak formations were observed in the aorta along with a systemic inflammatory response, as indicated by an increase in CRP levels (Table 1). Ultrasonography of the carotid artery
The TU of the ventral neck generated clear pictures of the common carotid artery (Figure 1). The tracheal cartilages served as easily recognizable landmarks. Colour Doppler imaging that allowed visualization of the pulsatile blood flow detected the carotid artery. In some animals fed with the atherogenic diet, the generation of clear pictures of the carotid artery was more
Table 1 Body weights, fatty streak formation and serum levels of lipids and C-reactive protein (CRP)
Group 1 Group 2
Body weight (kg)
Cholesterol (mmol/L)
Triglycerides (mmol/L)
CRP (ng/mL)
Fatty streak formation (%)
3.670.2 3.570.2
0.8970.14 28.176.57*
0.4770.08 3.5171.42*
12279 3737157*
– 35.875.7*
Rabbits were fed a normal (group 1, n=6) or high-fat (group 2, n=8) diet for 20 weeks. *Po0.05 versus group 1
Laboratory Animals (2006) 40
Non-invasive vascular ultrasound in rabbits
83
Muscle, fat tissue A. carotis communis Trachea
Acetylcholine infusion B
A
1.65 mm
1.6 mm
Baseline
1.6 mm
1.7 mm
Ach infusion
Baseline
Ach infusion
Figure 1 Transcutaneous ultrasound of the A. carotis communis in rabbits. Schematic drawing of the experimental procedure and the anatomy of the ventral neck. Ultrasound images before and after acetycholine (Ach) infusion in normal fed (A) and atherogenic fed (B) animals. Arrows indicate the luminal diameter of the A. carotis communis
challenging, due to a greater extent of fat tissue in the ventral neck. Effect of hyperlipidaemia on vascular reactivity in vivo and ex vivo
First, we investigated the endotheliumdependent vasorelaxation to Ach determined by non-invasive TU. There was no difference in the cross-sectional carotid diameter at
baseline between the normo- and the hyperlipidaemic rabbits (Table 2). Infusion of Ach in the normal fed animals caused a vasodilation, as indicated by a significant increase in the cross-sectional diameter. In contrast, in the hyperlipidaemic animals, the cross-sectional diameter did not significantly increase during the infusion of Ach when compared with baseline (Table 2, Figure 2). Even a tendency towards Laboratory Animals (2006) 40
84
K Hiss et al.
Table 2 Diameter of the common carotid artery
Group 1 1.6770.05 Group 2 1.6170.03
1.8570.09 1.6370.05*
10.472.8 1.472.9*
The luminal diameter of the right common carotid artery was assessed at baseline and during systemic intravenous infusion of acetycholine (Ach) (5 mg/kg/min). The change in diameter (D) during Ach infusion is expressed as a percentage of the diameter at baseline. *Po0.05 versus group 1
10−6
−20 −40 −60 −80 −100 −120
Figure 3 Ex vivo vasoreactivity in isolated carotid arteries. Endothelium-dependent vasorelaxation to acetycholine (Ach) of phenylephrine preconstricted segments of the common carotid artery is shown in normal () and atherogenic (&) fed rabbits. Po0.05 versus normal fed animals
10.0 7.5 5.0
∗
2.5
2 G ro up
G ro up
1
0.0
Figure 2 In vivo vasoreactivity of the A. carotis communis. The change of luminal diameter of the ACC (D, %) before and during infusion of acetycholine (Ach) (5 mg/kg/min) was assessed after 20 weeks on normal (group 1, n ¼ 6) or high-fat (group 2, n ¼ 8) diet. Po0.05 versus group 1
vasoconstriction was observed in some hyperlipidaemic animals when challenged by Ach infusion. The contractile responses to KCl (60 mmol/L) of the carotid arteries were not different between the groups (data not shown). Phenylephrine (1 mmol/L)-induced contractions were higher in western-diettreated animals, but this difference was not statistically significant (data not shown). In animals treated with the atherogenic diet, a pronounced endothelial dysfunction was observed over the entire range of concentrations of Ach tested (Figure 3). Maximal endothelium-dependent relaxation was significantly reduced to 47.4712.1% in phenylephrine precontracted carotid arteries of animals treated with the atherogenic diet compared with carotid arteries of control animals (10072.4%). We concluded that a Laboratory Animals (2006) 40
−20 Ex vivo vasoreactivity (%)
12.5
∆ (%)
Relaxation (%)
Carotid Carotid diameter diameter pre-Ach (mm) post-Ach (mm) D (%)
Concentration Ach (M) 10−7
10−8 0
−40
Regression coefficient r2 = 0.68 P = 0.002
−60 −80 −100 −120 −140 −10
−5
0 5 10 15 In vivo vasoreactivity (%)
20
Figure 4 Ultrasound assessed vasoreactivity related to measurements in isolated arteries in an organ bath. Vasoreactivity of the common carotid artery (ACC) assessed by in vivo transcutaneous ultrasound was correlated (regression coefficient r ¼ 0.68) to ex vivo endothelium-dependent vasorelaxation to acetycholine (Ach) in the same segment of the isolated ACC (n ¼ 12). Data are shown at the dose of 5 mg/kg/min Ach for in vivo measurements and at 12 mmol/L for ex vivo measurements
pronounced endothelial dysfunction is present in animals treated with the atherogenic diet for five months. Importantly, the carotid vasorelaxation measured by TU correlated highly with the results of ex vivo determination in the organ bath (regression coefficient r2 ¼ 0.68, P ¼ 0.002) (Figure 4).
Discussion In our study, we showed that an atherogenic diet given to rabbits for five months resulted
Non-invasive vascular ultrasound in rabbits
in a significant increase of triglycerides, cholesterol and CRP (a marker of systemic inflammatory response), as well as a significant deterioration of endothelial function. The comparison between a noninvasive and an ex vivo method to determine vascular reactivity revealed a strong correlation between the results of each methodology. Endothelial dysfunction is generally defined as a decrease in the capacity of the endothelium to dilate blood vessels in response to physical and chemical stimuli. The determination of the vascular reactivity is an important tool in the diagnosis of endothelial dysfunction. Several studies have shown that the evaluation of endothelial dysfunction of peripheral arteries is a good and independent predictor for coronary heart disease (Leng et al. 1996, Chambless et al. 1997, O’Leary et al. 1999, Sonoda et al. 2004), and more importantly for cardiovascular events (Schachinger & Zeiher 2001). Ideally, methods for the measurement of endothelial function should be accurate and reproducible, safe and non-invasive to perform, and be able to allow an evaluation of risk for cardiovascular events and success of therapy. Whereas in patients a number of non- or minimal-invasive tests, such as forearm bloodflow, are available to detect the presence and severity of extracoronary atherosclerotic lesions, there is a lack of non-invasive and repeatable methods for vascular function measurements in animal models of endothelial dysfunction. Most of the methods used in patients, e.g. brachial artery flow-mediated dilation or strain-gauge plethysmography, are not feasible in animals. Others, for example, positron emission tomography (PET), are very expensive and not widely available. Until now, animals have been killed prior to determining endothelial function accurately in isolated arteries, and the information on endothelial function has been restricted to a single time point. TU offers the option of investigating endothelial function repeatedly, and enables the investigator to document the progression of the pathology at different time points in the same animal. The advantages of this
85
technique are firstly to understand the functional alterations during the progression of atherosclerosis, and secondly to reduce the animal numbers per group. However, the use of TU is limited to functional measurements because the resolution of the utilized transducer is not sufficient for monitoring and analysing the morphology of the vascular wall. Another limitation of this method is the compensatory response of the circulatory system to the systemic infusion of Ach. Indeed, Ach led to a decrease of systemic blood pressure and a reflex tachycardia that was similar in both groups (data not shown). However, local intraarterial application of Ach and simultaneous ultrasound of the vessel are not feasible and would require an invasive intervention. In contrast, intravascular ultrasound (IVUS) is an accurate imaging modality for the assessment of arterial wall morphology and function. The use of ultrasound in the determination of vascular morphology and function is well established, but the method has been restricted to invasive applications. In 1990, Nishimura and colleagues first validated the method of IVUS in an in vitro model and obtained an excellent correlation with histopathology (Nishimura et al. 1990). Since then, this method has been used in patients as well as in animal models, and provides an accurate assessment of vascular morphology and function (Mano et al. 1996, Okabe et al. 2003). Though there are advantages in accuracy and validity, this method is nevertheless an invasive intervention and involves stress for the animal. Also of note is that endothelial damage has been observed in the vessel subsequent to IVUS passage (Verheye et al. 2004). Our findings suggest that TU of peripheral arteries may provide a new tool for evaluating endothelial function and for monitoring vascular response to therapy in hyperlipidaemic rabbits, while preventing additional stress for the animal. References Chambless LE, Heiss G, Folsom AR, et al. (1997) Association of coronary heart disease incidence with carotid arterial wall thickness and major risk Laboratory Animals (2006) 40
86
factors: the Atherosclerosis Risk in Communities (ARIC) Study, 1987–1993. American Journal of Epidemiology 147, 483–94 Elgharib N, Chi DS, Younis W, Wehbe S, Krishnaswamy G (2003) C-reactive protein as a novel biomarker. Reactant can flag atherosclerosis and help predict cardiac events. Postgraduate Medicine 114, 39–44 Leng GC, Fowkes FG, Lee AJ, et al. (1996) Use of ankle brachial pressure index to predict cardiovascular events and death: a cohort study. British Medical Journal 313, 1440–4 Lohn M, Dubrovska G, Lauterbach B, Luft FC, Gollasch M, Sharma AM (2002) Periadventitial fat releases a vascular relaxing factor. FASEB Journal 16, 1057–63 Mano T, Masuyama T, Yamamoto K, et al. (1996) Endothelial dysfunction in the early stage of atherosclerosis precedes appearance of intimal lesions assessable with intravascular ultrasound. American Heart Journal 131, 231–8 Monnink SH, van Haelst PL, van Boven AJ, et al. (2002) Endothelial dysfunction in patients with coronary artery disease: a comparison of three frequently reported tests. Journal of Investigative Medicine 50, 19–24 Nishimura RA, Edwards WD, Warnes CA, et al. (1990) Intravascular ultrasound imaging: in vitro
Laboratory Animals (2006) 40
K Hiss et al.
validation and pathologic correlation. Journal of the American College of Cardiology 16, 145–54 O’Leary DH, Polak JF, Kronmal RA, et al. (1999) Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. New England Journal of Medicine 340, 14–22 Okabe T, Asakura Y, Asakura K, Kawamura A, Ogawa S (2003) Usefulness of residual percent plaque area after percutaneous coronary intervention in predicting peristent positive remodelling. American Journal of Cardiology 92, 1399–403 Schachinger V, Zeiher AM (2001) Prognostic implications of endothelial dysfunction: does it mean anything? Coronary Artery Disease 12, 435–43 Sonoda M, Yonekura K, Yokoyama I, et al. (2004) Common carotid artey intima-media thickness is correlated with myocardial flow reserve in patients with coronary artery disease: a useful non-invasive indicator of coronary atherosclerosis. International Journal of Cardiology 93, 131–6 Verheye S, De Meyer GR, Krams R, et al. (2004) Intravascular thermography: immediate functional and morphological vascular findings. European Heart Journal 25, 158–65