Element analysis with a proton microprobe of early atherosclerotic lesions

Atherosclerosis is a progressive inflammatory vascular disease accompanied by a gradual build-up of cholesterol in the artery walls. The associated chronic inflammatory process leads to tissue damage in the vascular wall as a consequence of an excessive inflammatory response. Large calcified deposits of the intimal lesions are a prominent feature of advanced atherosclerotic lesions. The degree of these calcifications in the coronary arteries correlates highly with the severity of coronary plaque burden and is thus associated with increased cardiovascular risk. In lesions preceding atheromas calcifications are rarely observed, most likely because of insufficient sensitivity and/or specificity of the techniques used. It is generally thought that calcifications are composed of hydroxyapatite in advanced atherosclerotic lesions and that atherosclerotic calcification occurs passively as the result of precipitation of calcified material in cores of dying cells. However, recent findings might indicate that atherosclerotic calcification is an actively regulated process involving, among others, bone matrix regulatory proteins. Most information on atherosclerotic calcifications deals with advanced intimal lesions and chemical analysis was mainly performed on larger, solid crystals. Data about the distribution and composition of the calcifications at a very early stage of the atherosclerotic process are scarce, as is information whether trace-elements could be involved in the calcification process. Chemical bulk analysis of the mineral part of human aortas, with advanced atherosclerotic lesions, showed that the average mass ratio of calcium to phosphorus is close to the ratio of hydroxyapatite, i.e. 2.16, which is the main constituent of bone tissue1. Though, others found lower ratios of calcium to phosphorus in atherosclerotic plaques2, 3. Also traces of iron and zinc were occasionally observed in these bulk analyses. In a pioneering study of by Pallon et al.4 the co-localization of iron and zinc with calcium granules, 10 to 20 µm in diameter, was observed in early atherosclerotic lesions in human coronary arteries. These observations incited our interest in the pathological significance of calcium precipitations in the early phase of atherosclerosis and the role of trace-elements, such as iron and zinc. The pathophysiological background of the subject of this thesis is summarized in chapter 2. To investigate the distribution, composition (calcium-to-phosphorus mass ratio) and trace-element contents of calcifications in early atherosclerotic lesions a technique was applied with high sensitivity, specificity and resolution, i.e., micro-Proton Induced X-ray Emission in combination with Backscattering and Forward Scattering Spectroscopy5, 6. With this technique pixel wise element concentration maps can be created to visualize the distribution of elements, like calcium and phosphorus, in cryosections of artery wall tissues. The technique is very sensitive enabling element concentration maps of trace-elements like iron and zinc to be obtained. For this technique high energy protons are required and these are produced utilizing an acquired accelerator (SingletronTM, High Voltage Engineering Europe B.V., Amersfoort, the Netherlands). The accelerator is connected to a proton microprobe setup, which was optimized for the detection of trace-elements at hotspots of calcium. The proton-beam was miniaturization, as the size of the calcifications is expected to be in the range of micrometers or even smaller. The settings for the object and aperture diaphragms were optimized as well as for the quadrupoles. The profiles and sizes of the proton beam were measured. The measured beam size has been reduced to about 0.5 to 0.5 µm2 at best. The position of the beam relative to the position of the sample and its size were constant during the time needed for a measurement. A beryllium absorber with calibrated thickness was placed in front of an X-ray detector for the detection of characteristic X-rays (PIXE) corresponding to, for instance, phosphorus and calcium. A second ultra-LEGe detector with a larger surface area was introduced for the detection of X-rays (PIXE) corresponding to iron and zinc. The solid angles of the detectors were optimized for the detection of phosphorus, calcium and trace-elements. The detection limits were in the order of µg/g for the trace-elements. A new data acquisition hardware and software system with graphical user interface was introduced which remotely controlled the proton microprobe setup. This made it possible to acquire large datasets. New analysis software with graphical user interface was written, which had the advantage of monitoring the experiment in a flexible way and data analysis speed was improved dramatically (chapter 3). The proton microprobe was applied to obtain directly the distribution of calcifications in entire cross-sections of human coronary arteries with lesions preceding artheromas. This objective required that consecutive scans were made that cover the whole cross section of a coronary artery wall to obtain an overview of the calcium distribution in the entire artery wall. The composition of the micro-calcifications was deduced from the calcium-to-phosphorus ratio. In order to quantify the calcium-to-phosphorus ratio a specific X-ray absorber used for PIXE was gauged with a reference sample. Inside the sample, protons can lose their energy and X-rays can be attenuated. These two effects influence the estimated values of the calcium-to-phosphorus ratio and, hence, corrective calculations were validated. To obtain the calcium-to-phosphorus ratio of a calcification the calcium and phosphorus concentrations of pixels belonging to the micro-calcification were represented in a scatter plot. The slope of a linear fit through the calcium and phosphorus concentrations in the scatter plot provides a reliable value of the calcium-to-phosphorus mass ratio of the micro-calcification (chapter 4). Tissue sections of coronary artery walls containing AHA (American Heart Association) type I to VII atherosclerotic lesions from patients who died from non-cardiac causes were the subject of investigation. Calcium deposits at a (sub)micrometer scale were observed in the thickened intima already in early pre-atheromatous stages of atherosclerosis (type II lesions), predominantly in the abluminal region of the intima. The occurrence of micro-calcifications increased in more severe stages of atherosclerosis. In type VII lesions big lumps of calcified material could be easily observed. The composition of the micro-calcifications, i.e. the calcium-to-phosphorus mass ratio, was investigated in type II, III, and IV lesions and micro-calcifications displayed a range of values (1.62 to 2.99). However, the calcium-to-phosphorus ratio was very homogeneous within a given hotspot: each micro-calcification area displayed a unique calcium-to-phosphorus ratio, which might be dependant on the local environment. The range of calcium-to-phosphorus mass ratios strongly suggests that the micro-calcifications consisted of amorphous calcium-phosphate crystals and was not restricted to hydroxyapatite. However, it can not be excluded that the hotspots consist of a mixture distinct calcium-phosphate compounds. For all the calcium hotspots described above also trace-elements were monitored. The detection limits for iron, copper and zinc were respectively 7, 5 and 5 µg/g. In none of the areas investigated copper could be found above the detection limit. The majority of the calcifications showed co-localisation with hotspots of iron and zinc. This finding might imply a pathogenic relationship between calcium precipitation on the one hand and iron and zinc on the other in these early atherosclerotic lesions in human coronary arteries. The pathophysiological significance and the underlying mechanism of this relationship are, however, incompletely understood. The involvement of bone matrix regulatory proteins during the calcification process was studied with immunohistochemistry. The occurrence of micro-calcifications preceded the expression of proteins actively involved in bone formation such as Matrix Gla Protein (MGP). Because MGP possesses the property to bind calcium ions and, hence, to inhibit the formation of insoluble calcium salts7, 8, the increased expression of MGP following the formation of micro-calcifications could point to an adaptive mechanism in the affected intima to mitigate or prevent further deposition and/or growth of calcified material. Increased expression of Bone morphogenetic protein-2 and Osteocalcin could only be observed in advanced stages of atherosclerosis when an atheroma had formed. Osteocalcin (OC or bone Gla protein) also exerts a mitigating effect on bone formation, but Bone Morphogenetic Proteins (BMP) are known to mediate ectopic bone formation. The same holds for Core binding factor-a1, showing a diffuse expression in advanced atherosclerotic lesions. This protein is known to regulate osteoblast and chondrocyte differentiation. The present findings infer that the formation of calcium phosphate granules is an early event in the atherosclerotic process and that the early deposition of calcium-rich material at the microscale is most likely not regulated by bone matrix regulatory proteins (chapter 5). In a carotid artery of an ApoE-/- mouse with an advanced atherosclerotic lesion (AHA type V lesion) also showed the presence of micro-calcifications. Atherosclerosis occurred in the low shear stress region. The high shear stress region remained unaffected. The calcifications were present both in the thickened intima and in the media of the low shear stress region. The calcium-to-phosphorus mass ratios ranged from 0.65 to 5.9, which is a much broader range than observed in human coronary arteries. These findings indicate that the composition differs between human and murine atherosclerotic calcifications. The question whether this implies that the nature of the atherosclerotic process is different between human and murine vascular tissue warrants further investigations (chapter 6). The findings described in this thesis are summarized and discussed in chapter 7. The main conclusions of the studies described in this thesis are that calcifications are present in very early atherosclerotic lesions as measured with a combination of proton-beam techniques. The composition of the micro-calcifications, i.e., the calcium-to-phosphorus mass ratio, in human coronary arteries displayed a range of values (1.62 to 2.99), strongly suggesting that the micro-calcifications consisted of amorphous calcium-phosphate crystals. The calcifications showed co-localisation with iron and zinc. This finding might imply a pathogenic relationship between calcium precipitation on the one hand and iron and zinc on the other. The present findings also indicate that the early deposition of calcium-rich material precedes the expression of bone matrix regulatory proteins.