Keywords

Introduction

Since the publication of this treatise on immunosenescence and its pathological consequences, the underlying basic science on immunopathology, vascular pathology continued to progress with several important results but also with some fallacious conclusions. For example, inflammation induced by a single pathogenic agent was shown to induce vascular lesions without increasing the concentration of circulating LDL: this is not really new, already Metchnikoff in the nineteenth century showed in his monography on inflammation all the symptoms this process can present. Hyperlipidemia which occurs in the organism triggers multiple mechanisms to fight the ensuing pathological process, this is an evolutionary acquisition. Therefore, the determination of LDL cholesterol as a marker of hyperlipidemia can certainly not be replaced by the determination of inflammatory markers. LDL cholesterol with its receptor identified by Brown and Goldstein, who received the Nobel Prize, remains an important event in the long road to fight vascular pathology. As a result of long evolution, the human organism reacts with multiple defense mechanisms, one of them, probably the oldest is the immunological process. Multiple studies on immunobiology showed that immune pathological processes regularly result in inflammatory reactions, but this does not mean that all the other symptoms of a specific pathology such as hyperlipidemia could be neglected. These reflections underline the importance of a second volume of this book on immunosenescence.

Historical Remarks

Although inflammation was recognized by Hippocrates and his school and defined by its cardinal symptoms (tumefaction, redness, heat, pain, and tissue dysfunction), its description in cellular terms had to await the birth of histochemistry. The invention of specific stains by Paul Ehrlich and others during the last decades of the nineteenth century helped to designate the leucocytes as the essential cellular elements of the inflammatory process. With the development of clinical chemistry and biochemistry, more and more molecular markers of the inflammatory process became available. Besides the increase of leucocytes in the blood and in tissues, the rate of red-cell sedimentation, followed by that of the so-called acute phase glycoproteins (haptoglobine, α1 acid glycoprotein or orosomucoid) and of C-reactive protein were routinely determined and considered as the hallmarks of the inflammatory process. This same period was dominated by the birth and expansion of humoral immunity, dominated by the French and German schools (Pasteur, Roux, Behring and others). Refinement of methodology, with the routine practice of passive hemagglutination and the ELISA-methodology, an important progress could be achieved, enabling the detection and quantification of low concentrations of nonprecipitating antibodies, cytokines, and other molecular players of the inflammatory process. Based also on refined methods of immunohistochemistry, it became difficult to maintain the strict distinction between degenerative and inflammatory processes, exemplified by the joint diseases. Arthrosis considered as a degenerative disease of articular tissue was more adequately designated osteoarthritis because of the inevitable development of the inflammatory process (Trentham 1984 for review). The possibility to create experimentally such articular pathology by immunizing with the major collagen component, collagen type II, of articular cartilage further blurred the frontiers between degenerative, inflammatory, or (auto)immune-mediated diseases. These conceptual and methodological advances prepared the way for the birth and expansion of cellular immunity, which dominated the field over the second half of the twentieth century. This slowly emerging and finally dominating methodological and conceptual advances reached also cardiovascular pathology. During the last decades of the twentieth century, several teams entered the field of cardiovascular pathology and introduced the above summarized methodology. These studies which will be described in this review revealed progressively the immuno-inflammatory nature of the athero-arteriosclerotic process also. Further progress came from the rapidly advancing field of cellular signaling. The study of receptors, agonists, antagonists, and message transmission pathways considerably improved the understanding of the details of cellular-molecular immuno-inflammatory processes underlying such chronic, degenerative diseases as cardiovascular pathology. As however the lipid-based concepts continued to dominate the field of atherogenesis, the final junction between these different avenues of approach for the understanding of this family of diseases was difficult to achieve. Of great help for open-minded scientists and physicians was the relatively rapid increase of the senior population over the last decades. Thanks to progress made in preventive and curative cardiovascular medicine, the fatal outcomes of cardiovascular pathologies were progressively postponed to later years of life, still remaining, however, the dominant cause of fatalities. This fact is certainly the motivation for experimental gerontologists to reassess the diverse contributing factors playing crucial roles in the development and progression of cardiovascular pathologies. Inflammation in particular was more recently recognized as an important factor in most age-related pathologies. A number of studies over the last decades clearly showed the importance of the inflammatory process as an omnipresent player in geriatric pathology. Cardiovascular pathology is no exception. Our laboratory actively participated in this progressive evolution of our concepts elaborated for the understanding of vascular diseases as relates to aging. We shall briefly review the successive stages of the abovementioned shifts of emphasis in the description and conceptualization of vascular pathology, essentially its importance as an age-related disease.

Arteriosclerosis and Atheromatosis

Progressive hardening, rigidification of the vascular wall, termed later arteriosclerosis, was recognized by early pathologists (Robert 1996, 1999a for review). These observations were based on autopsies and could not be easily generalized. The importance of nutritional factors in general and especially of cholesterol started to gain acceptance with the demonstration by Anitchkoff in the early decades of the twentieth century of the induction of lipid infiltration and plaque formation in rabbits kept for several weeks on a cholesterol-enriched diet (Olsson 1987 for review). This model became the most widespread in laboratories of experimental medicine and formed the basis of the lipid-hypothesis of atherogenesis. The development of atherosclerotic plaques, as observed in human blood vessels at autopsy, could therefore be attributed to nutritional factors, cholesterol, and saturated fat in particular. As calcified plaques and lipid infiltration were regularly associated in human blood vessels, the term proposed by the German physician Marchand athero-arteriosclerosis was progressively adopted as more adequate to describe the human disease. The vast majority of scientists engaged in this field adopted the lipid hypothesis further comforted by the characterization of lipoprotein-classes and during the second half of the twentieth century by the description of the LDL-recognizing receptor by Brown and Goldstein (Olsson 1987 for review). Some laboratories did, however, continue to explore avenues related to the immuno-inflammatory hypothesis.

Role of (Auto)Immune Factors

As mentioned in the introductory, historical section, immune-inflammatory factors were discovered as of crucial importance in some chronic diseases and especially in rheumatoid arthritis and osteoarthritis. Rapid progress in the field of extracellular matrix biology, in the characterization of collagen(s), major components of connective tissues (more correctly of extracellular matrix, ECM) led to the identification of collagen type II as a major component of articular cartilage (Comper 1996 for review). Immunization with purified collagen type II was shown to induce osteoarthritic pathology. Autoantibodies to collagen type II were demonstrated in patient’s sera (Trentham 1984 for review). Simultaneously, several teams showed that immunization with arterial extracts could induce in rabbits athero-arteriosclerotic lesions.

Induction of Athero-Arteriosclerotic Lesions by Immunization with Arterial Homogenates

Apparently, the first description of the production of athero-arteriosclerotic lesions in rabbits with homologous aorta-extracts was produced by a Hungarian team (Szigeti et al. 1960, 1968). These results were obtained without excess cholesterol administration. The severity of the lesions could be increased by prolonged protocols of immunization. The principally involved antigenic fraction was considered by these authors present in the saline-soluble fraction containing among other components a fraction with β-globulin mobility. Delayed-type tissue-allergic reactions could also be demonstrated. Similar lesions were created in rats by injection of rabbit sera immunized with aorta extracts. Total plasma lipids were shown to increase during immunization, similar to that found in cholesterol-fed animals (Szigeti et al. 1968 for review). Soon after these reports, the French team of Scebat and Renais reported the production of immuno-atherogenic reactions in rabbits by immunization first with heterologous (rat) aorta extracts and later with homologous aorta extracts (Renais et al. 1968; Scebat et al. 1966, 1967). In between White and Grollman (1964) produced periarteritis nodosa in rats also by immunization. The team of C.R. Minick at the New York Hospital (1966) also produced immuno-arteriosclerotic lesions by combining “allergic injury” and lipid-rich diet. Altogether these experiments illustrated the possibility of an immune-mechanism underlying the atherogenetic process. As however all the above cited experiments were carried out with aorta-homogenates, the principal antigen(s) involved in this immune-atherogenic process remained to be determined. Some of the authors opted for the β-lipoproteins (LDL) of the blood-serum, present in the soluble aorta-extracts also as the principal atherogenic antigen. The Hungarian team of S. Gero et al. (1959, 1960, 1967) induced lesions with anti-β-lipoprotein antibodies and proposed a variant of the immuno-atherogenic process based on these observations. The symposium organized by the French Atherosclerosis Society in 1964 in Bordeaux enabled the confrontation of these different views on the immune-factors involved in the atherogenic process. Our experiments, detailed in the next section, were also first presented at this meeting, proposing elastin as the main culprit as sensitizing antigen and as the target of the immune-pathological process underlying athero-arteriosclerosis (Robert et al. 1967, 1968, 1970b, 1971a).

Immune-Atherosclerosis Obtained with Purified Aorta-Extracts

The most important step to follow up on the above summarized results showing that active and passive immunization with crude aorta extracts could induce vascular lesions similar to human athero-arteriosclerosis was the reproduction of such experiments with purified aorta extracts. These experiments were performed in our laboratory and will be summarized in this section. Human and porcine aorta extracts were prepared, using lesion-free portions by a fractional extraction procedure used previously for other ECM-rich tissues as cornea and skin (Robert and Parlebas 1965). The soluble macromolecules were extracted with a 1 M CaCl2 solution buffered to pH 8.0 with Tris/citrate (termed CTC-extract). The insoluble stroma remaining after several extractions in ice-cold buffer with the Ultra-turrax was suspended in 2.7% TCA and heated to 90 °C for 30 min to hydrolyze selectively insoluble collagen. The remaining stroma contains essentially the elastic fibers. Adhering microfibrillar fraction (structural glycoproteins) was extracted by tour mixing the washed residue (0.9% NaCl) in 8 M urea in presence of 0.1% mercaptoethanol. After centrifugation and washing in sterile 0.9% NaCl, the final residue analyzed as pure, insoluble elastin. Elastin was shown to be selectively hydrolyzed to large peptides when suspended in 1 M KOH in 80% (v/v) aqueous ethanol at 37 °C for about 30 min (Robert and Poullain 1963). This large peptide solution (average MW about 70 kDa) was used for immunization. The other CTC, TCA, and urea extracts were dialyzed and lyophilized. All these operations were carried out in sterile conditions, in the cold, on tissues delivered in dry ice. Rabbits (New Zealand-white or Fauve de Bourgogne) kept on rabbit-chow and fresh vegetables were immunized with 1–5 mg proteins in complete Freund’s adjuvant, two injections weekly as described (Robert et al. 1971a). After 4 weeks on this schedule the animals were left for 4 more weeks and received a final injection (i.v. or i.p. without Freund’s adjuvant) with aluminum hydroxide as adjuvant. This was followed by testing animals for delayed hypersensitivity directly or using guinea pigs sensitized to the same antigens, as described (Robert et al. 1971a; Jacob et al. 1984). Control animals received saline injections with or without Freund’s complete adjuvant. Some rabbits received a cholesterol-enriched diet (1 g cholesterol in 7.5 ml peanut-oil homogenized with bran and barley). Titration of immune-sera was carried out with passive hemagglutination using glutaraldehyde-treated sheep erythrocytes. The antigens were fixed on treated erythrocytes using either a water-soluble carbodiimide or diazotated benzidine (Bing et al. 1967). The histological, histochemical, and electronmicroscopic procedures were described (Robert et al. 1971a). The CTC-extract contained a number of saline-soluble proteins and glycoproteins as shown by immune-electrophoresis and immune-diffusion according to Ouchterlony. The urea-extract contained several glycoproteins characterized by their size and glycan composition (Robert et al. 1967, 1971a). The κ-elastin solution had the typical amino-acid composition of purified elastin (Robert and Poullain 1963). Purified elastin before and after urea extraction was also examined by electron microscopy (Robert et al. 1971b) in order to demonstrate that urea-extraction largely eliminated the microfibrillar components. The amino-acid and glycan composition of the aorta fractions used as sensitizing antigens was described (Robert et al. 1971a). As shown by immune-diffusion, antisera to the CTC-extract showed strong precipitation lines to several soluble macromolecules, not further characterized. They also showed faint precipitation lines to the urea- and κ-elastin fractions. Antisera to the urea-extract showed one strong and several faint precipitation-lines to the urea extract and one to three faint precipitation lines to κ-elastin. Anti-κ-elastin antisera also gave faint precipitation lines to the κ-elastin solution as well as to 2–3 faint lines to the urea extract. This precipitation lines, observed with antiCTC and anti-urea extract antibodies to the κ-elastin solution can be considered as a first indication of the presence of elastin peptides in the soluble fractions of human and porcine aortas. Up to this time, elastin was considered as strictly insoluble. As shown by the passive hemagglutination tests, peak values of titrable antibodies were obtained at about 100 days after immunization, both to the urea extract and to κ-elastin. The titers fall than rapidly, suggesting absorption of antibodies by the proper tissue-antigens of the animals. Relatively high titers were obtained, about 10−5 of the urea extract and about 5 × 10−3 to κ-elastin (lowest hemagglutinating dilution of antisera). κ-elastin from human aorta gave higher titers than from porcine aorta. Cholesterol-fed animals showed low titers to all antigens tested (well below 10−2). Delayed hypersensitivity reactions were regularly observed on the immunized rabbits as well as on guinea pigs sensitized to the same antigens. This can be taken as an indication that besides the humoral immune-reaction, immunizations with aorta-extracts did trigger also cellular immune-reactions. Total serum cholesterol did increase significantly 1-month after the onset of immunization (from 0.36 ± 0.07 to 0.88 ± 0.01 g/l in elastin immunized animals) and returned later to preimmunization values. β-lipoprotein (LDL) determined according to Burstein and Samaille (1959) did increase continuously with time in immunized as well as in control rabbit sera. This selective increase of circulating LDL might be attributed, besides immunization, to the diet and the aging of the animals. There were, however, significant differences between animals immunized with different aorta extracts. The strongest increase of LDL as compared to the starting levels (before immunization) were found in animals immunized with the urea-extracts (137–200% increase) followed by those immunized with elastin (without adjuvant!), 145% increase on the average. Immunization with elastin in complete adjuvant produced a less important increase (about 70%) compared to similar increase in nonimmunized control animals. The macroscopic and microscopic observations of the aorta of immunized rabbits revealed conspicuous modifications (Figs. 1, 2, and 3). The most important modifications were observed on the aortas of animals immunized with pure elastin: lipidic-sclerous infiltrations covering most of the intimal surface, from the origin of aorta to the renal arteries, most obviously on the cross and around the ostia. Calcified plaques imitating egg-shell appearance, sometimes with aneurismal dilations, were observed (Fig. 1). Migrating smooth muscle cells (SMCs) were observed in the intima, accompanied by fragmentation of elastic lamellae (Figs. 2, 3, and 4). Ultrastructural and histochemical studies confirmed the strong calcification of the fragmented elastic fibers. At some places necrotic modifications of the media were observed with SMCs undergoing lysis. In all cases of immune-lesions a striking difference was observed with lesions produced by the high cholesterol diet. The cholesterol-induced lesions were similar to those described by a number of authors in other cholesterol-fed rabbits (Table 1). Differences were observed between the two strains of rabbits used. White rabbits were more resistant to immune-induced lesions than the fauve de Bourgogne strain as observed after several 1-month cycles of immunization followed by 1-month without. In our first experiments (Robert et al. 1971a), about 40% of all the immunized animals presented the above summarized lesions with 25% presenting wide-spread macroscopic and microscopic lesions as those shown on Figs. 1, 2, and 3. There were, however, important differences in the frequency and severity of lesions according to the antigen used (Table 1). Immunization with the soluble extract (CTC-extract) gave only infrequent lesions (1 out of 7 animals immunized) and only at the ultrastructural level. Immunization with elastin produced lesions both at the microscopic and macroscopic levels in more than 70% of the immunized animals. Immunization with the urea extract gave macroscopically detectable lesions, only in about 15% of animals, and microscopic lesions in about half of the immunized animals. These results suggested the presence of the most pathogenic antigen, supposed to be elastin, in all extracts, but in highly variable proportions: lowest concentrations in the CTC extract, somewhat higher concentration in the urea extract, and the highest in the purified elastin fraction. Further studies largely confirmed these observations (Jacob et al. 1984). The most severe microscopic and macroscopic lesions were seen in elastin immunized animals, followed by the urea extract. Such lesions included vacuolized, calcified endothelial cells (Fig. 2) above strongly dislocated, lyzed, and calcified elastin lamellae. Table 1 shows a comparison of most frequent lesions seen in the aorta of animals immunized with the different aorta extracts. One of the crucial observations was the inverse relationship between the three aorta extracts used for immunization as far as the induction of precipitating antibodies and lesion-induction is concerned. The soluble (CTC) extract induced high titers of precipitating antibodies and only infrequent and relatively mild lesions. Purified elastin, whether used in its insoluble, fibrous form or as soluble, κ-elastin, did induce regularly the most severe lesions but gave only low titer hemagglutinating or precipitating antibodies, the urea extract occupying an intermediary position. The low titer of circulating antibodies could best be explained by their adsorption on elastin fibers and was confirmed by the fixation of immunofluorescent anti-rabbit-IgG antibodies on elastic fibers (Jacob et al. 1984; Robert et al. 1967, 1968, 1970b). The predominant localization of the lesions at the level of elastic fibers was certainly the most conspicuous observation: fragmentation and calcification. Degenerescent modifications of endothelial cells and SMCs were also a constant observation. Extracellular calcium-containing crystals could also be observed occasionally. The macroscopic and microscopic ultrastructural lesions were quite similar to those observed in human athero-arteriosclerosis but quite distinct from those produced by the cholesterol-rich diet. Although a more or less important and transitory increase in β- lipoproteins (LDL) was observed in immunized animals, we also did observe unexpectedly an increase of antielastin antibody titer in the sera of cholesterol-fed animals, although the titers remained relatively low (≤10−2). Further experiments on the elastin receptor (ER) of vascular cells confirmed the role of elastin peptides in the lipid-induced lesions also (See later). Some years later, these experiments were repeated in our laboratory, using only κ-elastin as the sensitizing antigen (Jacob et al. 1984). These experiments will now be described.

Fig. 1
figure 1

Aorta of a rabbit immunized with κ-elastin from human aorta. Isolated or confluent fibrous calcified lesions covering most of the intimal surface (From Robert et al. 1971a)

Full size image
Fig. 2
figure 2

Intimal portion of rabbit aorta immunized to urea extract of porcine aorta. Electron microscopy, 3,200x. Notice the granular infiltrate of strongly modified SMCs, fragmented, lyzed and calcified elastic lamellae and vacuolized and calcified endothelial cells (From Robert et al. 1971a)

Full size image
Fig. 3
figure 3

Modified, fragmented, lyzed, and calcified elastic lamellae (LE) in the thoracic aorta of rabbits immunized with human κ-elastin. Fibrillar deposits (arrows) and strong perielastic calcification (double arrow; Electron microscopy by Grosgogeat Y. in collaboration with Robert et al. 1971a)

Full size image
Fig. 4
figure 4

Fragmentation of elastic lamellae in aortas of rabbits immunized with κ-elastin (right, K) as compared to a control aorta (left, C). E elastic fibers (From Jacob et al. 1984)

Full size image
Table 1 Macroscopic, microscopic and ultrastructural lesions of the aortas of rabbits immunized with aorta extracts
Full size table

Extension of Elastin-Immune-Pathology to Pulmonary Vessels, Metabolic Effects

The above described experiments were repeated using this time only κ-elastin of high molecular weight purified by gel filtration (≥ 70 kDa, designated κ1 – elastin) from bovine ligamentum nuchae-elastin as the immunizing antigen with New Zealand white rabbits, carried out essentially as in the first experiments (for methodological details, See the original publication, Jacob et al. 1984). This time however, besides the aorta, detailed investigations were carried out on the small arteries in the lung parenchyma also. The macroscopic, microscopic and ultrastructural investigations were completed by experiments on the biosynthetic capacity of aorta-wall explant cultures using 14C-lysine and 14C-glucosamine incorporation. Another innovation as described in more detail in the next section was the determination of elastase-type endopeptidase activity in aorta extracts from control and immunized animals. The presence of such activity in human aorta extracts was previously described in our laboratory (Hornebeck et al. 1975; Robert et al. 1974b) as will be detailed below. This activity could explain the presence of soluble elastin-peptides in the buffer-soluble aorta extracts, suspected in our first experiments. The biological properties of elastin peptides started to be studied in several laboratories and will be described in the following sections. These relatively recent results extended the horizon of our investigations and triggered a more complete search for pathological effects produced by elastin immunization as well as by elastin peptides. This time all rabbits immunized to κ-elastin developed typical vascular lesions at the macroscopic and microscopic level similar to those seen in our first experiments described in the previous section. The histological study of the lungs of the immunized rabbits revealed the presence of granulomatous lesions, absent in the control animals (Figs. 5 and 6). The most conspicuous modifications were, however, the lysis of elastic laminae in the large vessels as well as in the walls of small lung-vessels (Figs. 4 and 5). This time, the loss of continuity of elastic fibers was quantitatively assessed by computerized image-analysis. A pronounced fragmentation of elastin fibers could be demonstrated as shown on Fig. 4. No such fragmentation was seen in vessels of control animals injected only with complete Freund’s adjuvant or with BSA in complete Freund’s adjuvant. A curious observation was the more pronounced fragmentation of the elastic lamellae in the outer segment of the media than in its inner segment, near the intima. The SMCs of elastin-immunized animals showed striking morphological changes and random orientation, comparable to those seen in our first experiment (Figs. 2 and 3). Their number was also decreased. The average length of elastic lamellae were about 13-times shorter in elastin-immunized aortas as compared to controls receiving adjuvant injections with or without BSA.

Fig. 5
figure 5

Histological appearance of rabbit lung arterioles in control (a) and in κ-elastin immunized (b) rabbits (Modified from Jacob et al. 1984)

Full size image
Fig. 6
figure 6

Granulomatous lesion with vasculitis in the lungs of κ-elastin immunized rabbit (Unpublished results from Chantal Lafuma in the author’s laboratory)

Full size image

This proportion when calculated separately for the internal part of the media gave a ratio of only four times shorter elastin segments for the inner part of the media and a much more pronounced fragmentation, 40-times as compared to controls in the outer half of the media. In the lungs of the elastin-immunized animals, intense granulomatous lesions were seen as well as a strong elastolysis in the walls of small lung-vessels (Figs. 5 and 6.). The granulomatous lesions contained giant cells and eosinophil leucocytes. On the ultrastructural level, the findings were similar to those described in our first experiments (Robert et al. 1971a). Electron microscopy confirmed the alterations of SMCs and elastic lamellae which were fragmented, disorganized with an increased density of the microfibrillar components (Fig. 3). Endothelial cells were enlarged, vacuolated, and calcified as in the first experiments (Fig. 2). One of the striking observations was the loss of continuity of elastic fibers or close contact between SMCs and elastic lamellae, (Fig. 7) clearly seen in control aortas. The increased titer of antielastin antibodies in the κ-elastin immunized rabbits could be confirmed by passive hemagglutination, the titers were, however, lower than in the first experiments (10−2–10−3). The adsorption of these antielastin antibodies to purified, micronized elastin fibers could be confirmed by immuno-peroxidase staining using sheep anti-rabbit Fb antibodies. No such reaction was seen with sera of animals immunized to BSA in Freund’s complete adjuvant. The same reaction, adsorbed autologous IgG on aorta elastic fibers, could be demonstrated in the κ- elastin immunized rabbits also.

Fig. 7
figure 7

Strong adhesion of a fibroblast to a purified and micronized elastic fibril under the electron microscope. No such adhesion with SMCs from elastin-immunized animals was observed (Modified from Perdomo et al. 1994)

Full size image

Figure 8 shows the results of radioactive lysine and glucosamine incorporation by aorta explant cultures, results being expressed as cpm per mg DNA. Incorporation was strongly decreased in the aorta explants from the κ-elastin immunized animals as compared to control animals injected with the complete adjuvant alone. Another important observation was the increase of elastase-type endopeptidase activity of the aorta extracts of κ-elastin immunized animals (Table 2). N-succinyl-ala-PNA was used as the chromogenic substrate, relatively specific for elastase-type endopeptidases (Bieth, 1978). These experiments largely confirmed and extended our first observations (Robert et al. 1971a) on the pathological modifications of large and small elastic vessels in animals immunized to human, porcine, or bovine elastin-peptides. All treated animals showed characteristic macroscopic, microscopic, and ultrastructural modifications with conspicuous fragmentation of elastic lamellae, modifications of endothelial cells, and SMCs and also strong calcification. Further experiments were carried out in order to specify the individual aspects of these modifications at the cellular and molecular levels.

Fig. 8
figure 8

Incorporation of 14C–lysine (a) and 14C–glucosamine (b) in proteins of arterial wall extracts. Bars with lines: control group (CFA); bars with points: κ1-elastin immunized rabbit group (κ1-CFA). Ordinates: radioactivity incorporated in each extract, expressed as cpm/μg DNA. Abscissa: MgCl2 2 M MgCl2 extract, GUA1, first guanidinium chloride extract, COLL collagenase extract, GUA2 second guanidinium chloride extract, ELAST elastase extract. The bars represent the SEM for the average value of each extract (For details See Jacob et al. 1984)

Full size image
Table 2 Elastase-type endopeptidase activity of soluble extracts of aortas from rabbits immunized or not with κ-elastin
Full size table

Cellular-Molecular Mechanisms Involved in the Immune-Inflammatory Vascular Pathology

The above-described experiments confirmed the possibility of the induction in rabbits of a macro and microvascular pathology reminiscent of arterio-atherosclerosis by immunization with elastin. Peptides derived from highly purified fibrous elastin, κ-elastin was shown to be the most efficient (if not the only) inducing antigen. The humoral immune reaction could be confirmed by the passive hemagglutination test and by immunodiffusion. There was, however, an inverse relationship between antibody titers and efficiency of lesion-induction when consecutive aorta extracts were used. The ultrastructural studies revealed severe modifications of cells and ECM components, mainly of elastin. Among the most conspicuous modifications was intra- and extracellular calcification, obvious even at macroscopic examination of longitudinally opened aortas of elastin-immunized rabbits. It remained to elucidate the underlying cellular and molecular mechanisms. Some of these experiments will now be described.

Degradation of Elastic Fibers

One of the most conspicuous modifications produced by immunization with elastin was the pronounced degradation of elastic fibers of the large and small blood vessels. It could be shown that this is at least partially the result of an autoamplifying vicious circle mediated by the action of elastin peptides on the ER. As shown on Fig. 9, the addition of elastin peptides at concentrations shown to be present in the blood serum (See later) produced a pronounced upregulation of the production of elastase-type endopeptidases (Robert et al. 1986) essentially of MMP-2 and MMP-9 (Archilla-Marcos and Robert 1993). This increase of elastase production could be inhibited by lactose and melibiose, antagonists of the ER. It could also be shown during these studies that aging, both chronological and in vitro (increasing passage number), produced an increased expression of elastase-type endopeptidases (Fig. 10). These experiments, repeated over the years, suggested a mechanism for the progressive degradation of vascular and pulmonary elastic fibers. Using an ELISA-procedure, the concentration of circulating elastin peptides could be determined in a large number of normal and pathological human sera (Bizbiz et al. 1997; Fülöp et al. 1989a). Their level followed a Gaussian curve with an average concentration in the μM range, largely exceeding the KD value of the ER, shown to be in the nanomolar range (Fülöp et al. 1989b; Robert et al. 1989). In elastin-immunized animals, these processes appear to be exaggerated by the action of soluble immune complexes on monocytes and lymphocytes able to trigger an increased release of elastin degrading endopeptidases. The presence of antielastin antibodies in all human sera tested suggests a similar mechanism in humans also (Fülöp et al. 1989a, b). The selective and irreversible adsorption of elastases on the surface of elastic fibers largely limits the possibility to counteract this process by elastase inhibitors (Robert et al. 1974a).

Fig. 9
figure 9

Increase of elastase-type endopeptidase activity in fibroblast cultures in presence of increasing concentrations of κ-elastin (0, 10 and 100 μg/ml). Addition of melibiose, an ER antagonist inhibits this upregulation of elastase activity (Modified from Archilla-Marcos and Robert 1993)

Full size image
Fig. 10
figure 10

Elastase type endopeptidase activity determined with N-suc-ala -3Pna as substrate. (a) in human aorta extracts. Abscissa: age in years, ordinates: log elastase activity per cell (DNA); (b) in successive passages of aorta smooth muscle cells. Abscissa: passage number, ordinates: elastase activity given as ng equivalent of pancreatic elastase; (c) in successive passages of human skin fibroblast cultures. Abscissa: passage numbers, ordinates: nM substrate hydrolyzed/106 cells; (d) effect of human lipoproteins added to SMC-cultures (For other details See text and Robert et al. 1986)

Full size image

Calcification of Elastin

The intense calcification of elastin fibers in the immunized animals, seen macroscopically and on the electron-microscopic preparations, could be confirmed by specific histochemical methods (Van Kossa staining) and direct chemical determinations. A partial explanation of this strong affinity of elastin for calcium came from physicochemical studies of Dan Urry (1980) showing that the β-turns of elastin represent high affinity fixation sites for calcium. Earlier studies by Max Burger (1947) and Lansing (1959) revealed the progressive Ca-fixation in elastic blood vessels in a diffuse fashion, and still present in elastin purified by heating to 100 °C in 0.1 N NaOH, the standard procedure for purifying elastin (Robert et al. 1985 for review). Moreover it was shown that calcium fixation on elastin strongly potentiates its affinity for lipids (determined by using 14C-cholesterol (Jacob et al. 1983; Hornebeck and Partridge 1975)) and vice versa, lipid fixation potentiates Ca-fixation. As β-lipoproteins (LDL) increased during the first phase of elastin-immunization, this also could contribute to Ca-retention in elastin. We have to mention here the demonstration by the team of J.L. Beaumont the role of anti-LDL antibodies in the development of atherosclerosis (Beaumont 1965, 1969, 1970; Beaumont et al. 1965; Beaumont and Beaumont 1968). This team demonstrated also the importance of anti-heparin antibodies in atherogenesis (Beaumont and Lemort 1974; Buxtorf et al. 1981; Lorenzelli-Edouard et al. 1980). The work of Bihari-Varga and Gero pointed also on the potential role of LDL and anti-LDL antibodies on the atherogenic process (1966; Bihari- Varga et al. 1984, 1986). This team pioneered also the recognition of the acid polysaccharides (glycosaminoglycans, proteoglycans) of the vascular wall in the retention of lipoproteins (Bihari-Varga and Gero, 1966). These two, essentially postsynthetic processes, calcification, lipid fixation confirmed by direct analysis of lipid classes in purified human aorta elastin by Claire et al. (1976) largely explained the progressive loss of elasticity and increased susceptibility to degradation of vascular elastic fibers.

Role of the Elastin Receptor

The demonstration of an elastin-recognizing cell-membrane receptor started in our laboratory with the observations on the capacity of cells to strongly adhere to micronized elastin fibrils, as shown by time-lapse video microscopy (Hornebeck et al. 1986). Addition of elastin peptides to cells (fibroblasts and vascular SMCs) was supposed to compete with fibrous elastin and inhibit elastin fiber fixation on cells. The opposite effect was observed; low concentrations of κ-elastin strongly increased the speed of adherence of cells to elastic fibers (Groult et al. 1998). This effect could be attributed to the induced synthesis of a membrane glycoprotein of 120 kDa termed elastonectin (Hornebeck et al. 1986). These initial experiments were followed by a series of other experiments aimed to the exploration of the physiopathological roles of the elastin-receptor. Using immune-histochemical procedures, the presence of the elastin-receptor could be demonstrated on vascular cells, endothelial cells, SMCs, fibroblasts, as well as on WBC-s, monocytes, PMN-s, and lymphocytes (Faury et al. 1995, 1998a; Jacob et al. 1987a; Perdomo et al. 1994; Péterszegi et al. 1997c). Most importantly, the elastin-receptor could be demonstrated on monocytes and lymphocytes inside the human atherosclerotic plaques obtained by endarterectomy (Péterszegi et al. 1997a). A variety of tumor cells did also exhibit the ER (Timar et al. 1995). All these cells, mobile and sessile, found in the vascular wall might therefore contribute to the ER-mediated upregulation of elastolytic protease production. It could be shown also during a long collaboration with T. Fülöp and his team that the activation of the ER contributes by still other mechanisms to the progression of vascular lesions. One of these mechanisms is the elastin-peptide triggered release of superoxide (Fülöp et al. 1989b) suggesting a free-radical (ROS-mediated) contribution to the vascular-cellular lesions. Another significant observation was the demonstration of accelerated oxidation of LDL by activation of the ER (Fülöp et al. 2005). It also could be shown that aging accelerated these harmful effects mediated by the ER. In cells (monocytes, PMN-s) obtained from old individuals, the normal transmission pathway of ER, as shown to function on young cells (Fülöp et al. 1990a; Varga et al. 1988, 1989) did no more function as shown by the inefficiency of pertussis toxin to block the transmission pathway mediated by a Gi protein in young cells. Free-radical release was, however, maintained and even amplified. These observations suggested an uncoupling of ER in old cells from its normal transmission pathway. This uncoupling results in the exacerbation of the harmful effects mediated by the receptor and loss of its physiologically useful functions, as NO-mediated vasodilation (Faury et al. 1997; Fülöp et al. 1992) and inhibition of cholesterol synthesis by monocytes (Varga et al. 1997). Most of these experiments as well as others pertaining to the role of ER in the malignant process were reviewed recently (Labat-Robert and Robert 2007). ER-mediated mechanisms play an important role in the inflammatory process also, which accompanies the immune-induced pathologies. Elastin peptides were shown to be chemotactic to WBC-s, monocytes especially (Antonicelli et al. 2007 for review). Activation of the ER on PMN-s and monocytes triggers the release of proinflammatory enzymes and cytokines (Antonicelli et al. 2007 for a review). ER-triggered reactions include endothelial iNOS upregulation and NO-dependent vasodilation (Faury et al. 1998a, b). As mentioned above, ER-NOS coupling decreases with age with a steady age-dependent loss of vasodilation by elastin peptides (Faury et al. 1997). Besides the ER-triggered proinflammatory reactions, soluble immune-complexes formed by the reaction of antielastin antibodies and circulating elastin peptides might well contribute also to the harmful effects produced by immunization with elastin. These rapidly summarized experiments substantiated the claim for an important pathogenetic role of the elastin–antielastin system in the genesis of athero-arteriosclerosis (Fülöp et al. 2001; Robert 1999a, b).

Extension to Human Pathology

The first experiments on immune-atherosclerosis were performed on human sera with the demonstration of antielastin antibodies present in all sera tested, using the passive hemagglutination method (Stein et al. 1965). Although the titers were relatively low (from 1/2 dilution to 1/512 in these first experiments), it could be shown that sera from 68 atherosclerotic persons were in the lower range (1/2–1/32 dilution still giving agglutination) attributed to the enhanced adsorption of antielastin antibodies on degrading elastin fibers offering a larger surface for antigen–antibody complex formation. Such complexes were demonstrated on fibrous elastin (Jacob et al. 1984, 1987b). The anti-elastin antibodies appeared at about 20 years and disappeared after 80 years (Stein et al. 1965). Surface area measurements on purified elastin using radioactive Krypton (Robert et al. 1970a, 1971c) revealed a relatively large specific surface of elastin fibers, favoring interaction (adsorption) with soluble molecules. These results were further confirmed by microcalorimetric adsorption studies (Robert et al. 1971c). Later experiments (Fülöp et al. 1989a) confirmed the presence of antielastin antibodies, by direct titration of IgG and IgM type antibodies, in a larger number (265) of human sera, normal and pathological, mainly athero-arteriosclerosis of the legs, ischemic heart disease, stroke, diabetes, hyperlipidemia type II/b and IV, and hypertension. No obvious correlation was found with age. IgG-type antielastin antibodies were elevated in obliterative arteriosclerosis of the legs and ischemic heart disease. No such modifications were seen in the IgM-type antibodies. Both types of antibodies were, however, decreased in type IV hyperlipidemia. Figure 11 shows the age-dependent evolution of the above-summarized determinations of antielastin antibodies in human sera. More details on the specificity and sensitivity of the ELISA method used can be found in the original publication (Fülöp et al. 1989a). Circulating elastin peptides were also determined by an ELISA-procedure on about 1,500 human sera (Bizbiz et al. 1997; Fülöp et al. 1990b). These last experiments performed in collaboration with epidemiologists on a study on the role of vascular aging on brain aging (the EVA-study, Bizbiz et al. 1996) confirmed also the correlation between carotid artery wall thickness and plaques with the elastin-elastase parameters. Elastase-type endopeptidases were also demonstrated in a large number of human sera (Table 3; Hornebeck et al. 1983; Bihari-Varga et al. 1984; Bizbiz et al. 1996). Monocyte-macrophage elastase-type activity was shown to be upregulated by cholesterol and by proinflammatory cytokines (Rouis et al. 1990). It should be mentioned, however, that after screening a large number of synthetic elastin peptides by monoclonal and polyclonal antibodies, it could be shown that according to their structure they presented a variable reactivity to antibodies used for their detection and quantification (Wei et al. 1993). Further confirmation of the role of the elastin-immune system and of the central importance of ER-mediated processes came from the above-mentioned demonstration of ER-exhibiting mononuclear cells in freshly excised human endarterectomy samples (Péterszegi et al. 1997a). It also could be shown that human helper (CD4) and memory (CD45R+) lymphocytes when cultured in presence of elastin peptides exhibit inducibly the ER as shown by flow-immuno-cytofluorimetry (Péterszegi et al. 1997c). Addition of increasing concentrations of elastin peptides upregulated in a dose-dependent fashion the expression of a serine-elastase identical to PMN-elastase (shown by inhibition with monoclonal anti-PMN-elastase antibody) followed by cell death (Péterszegi et al. 1999; Fig. 12). The crucial role of the ER could be demonstrated in these experiments also by the inhibition of cell death in presence of the antagonists of the ER, lactose, and melibiose (Robert 1999a, b). It also could be shown that circulating lymphocytes isolated from the blood of elderly patients suffering from denutrition and dementia exhibited an increased elastase and cathepsin G activation mediated by the ER (Péterszegi et al. 1997b). Another finding in favor of the above hypothesis, connecting the immune-hypothesis to the lipid-theory of atherogenesis, came from the demonstration of an increased concentration of elastase-type endopeptidases in aorta-extracts from cholesterol-fed animals (Jacob et al. 1982) confirmed by in vitro experiments showing that addition of human LDL and VLDL (but not HDL) to vascular SMC-cultures strongly increased the production of elastase-type endopeptidases (Fig. 10; Bourdillon et al. 1984).

Fig. 11
figure 11

Correlation between anti-elastin peptide antibody concentration (1-g/ml) of IgG type and the age of subjects. Males (r = 0.19, n = 117)(-), females (r = 0.01, n = 166)(- -)

Full size image
Table 3 Elastase-type endopeptidase activity of normal and pathological human sera
Full size table
Fig. 12
figure 12

Upper figure (a) Human lymphocytes cultured in presence of increasing concentrations of κ-elastin (in log conc.μg/ml on the abscissa). : modification of cell proliferation and of elastase production as a function of κ-elastin concentration. (b) increase of cell death with increasing κ-elastin concentration, as % of total cells. (c) Electron microscopy of a normal lymphocyte from the above cultures (control) and a lymphocyte with apoptotic bodies (κ-elastin) (Modified from Péterszegi et al. 1999)

Full size image

The pharmacological consequences of the above-described immuno-atherosclerotic process were also investigated. It could be shown that immunization of rabbits with elastin strongly increased SMC-membrane permeability, as shown by the increase of ouabaine-insensitive 22Na+ efflux, the 86Rb efflux (indicator of K+ efflux), and the 45Ca++ influx. Passive permeability to Na+ and K+ as well as the sodium pump were enhanced. Administration of porcine calcitonine largely prevented these modifications, as well as the development of the athero-arteriosclerotic plaques (Jacob et al. 1987b). Treatment with calcitonin largely prevented also the calcification of elastin fibers in these animals (Jacob et al. 1987b). The inhibition of Ca-influx by calcitonine might well be the key factor in these experiments.

Among the not completely elucidated consequences of the above-summarized immune mechanisms remain the effect of possible age-dependent modifications of immune functions on elastin induced athero-arteriosclerosis. Age-dependent modifications of the human and animal immune systems were extensively investigated over the second half of the twentieth century (Robert and Robert 1973 for a review). The described age-dependent modifications of the immune systems might well influence the outcome of the immune-atherosclerotic process also. As the emphasis of our research and the interest of other teams shifted to the age-dependent modifications of receptor function (Joseph and Roth 1990; Robert 1998; Roth 1979 for reviews), the interpretation of the observed pathological modifications shifted progressively towards postgenetic (epignetic, posttranscriptional) mechanisms (Robert and Labat-Robert 2000 for review). Several experimental facts pleaded in favor of the progressive preponderance of such mechanisms. Although the selective uptake of cholesterol by the vascular elastin fibers were first demonstrated in vivo in animals (Jacotot et al. 1973; Robert et al. 1984b; Szigeti et al. 1972), qualitative and quantitative determinations of lipid classes strongly associated with purified human elastin carried out on human aorta samples (Claire et al. 1976) confirmed the same strong affinity of human elastin also to lipids, attributed to factors inherent to the hydrophobic nature of elastin. Circulating elastin peptides present in all human sera examined were shown to exhibit a number of relevant pharmacological properties (Robert et al. 1984a) important also for lipid-mediated processes during atherogenesis.

Some remarks on circulating elastase-type endopeptidases. Their upregulation could be, at least partially, attributed to circulating elastin peptides as described in previous sections of this review. The presence of identical or very similar peptide sequences in elastin of several species might also contribute to the explanation of the incomplete elimination of elastin-recognizing immunocompetent cells during human development. And finally, the demonstration of increased elastin mobilization with cholesterol feeding alone (Jacob et al. 1982, 1983) might have justified to some extent the neglection of the immune-mediated mechanisms in human atherogenesis. With the rapid increase of human longevity, this approach might, however, reach its limits, producing a revival in the interest of immune-mediated processes in human vascular diseases.

Conclusion

The above summarized experiments and conceptions support the contention for an immuno-inflammatory mechanism involved in the development of athero-arteriosclerosis. Although most studies were conducted on animal models, the findings on humans reviewed in the preceeding sections strongly support the validity of the elastin-antielastin antibody-based process as of importance also for human pathology (Robert et al. 1967, 1968, 1970b, 1974c, 1984c; Robert and Robert 1975). The connection of this process to the ER-mediated pathways via elastin degradation and elastin peptide liberation in the circulation is of importance for the rationalization of the role for both immune-mediated and receptor-mediated mechanisms. There is also the connection to the lipid-mediated processes, essentially by the upregulation of elastase production by the atherogenic lipoproteins, increasing therefore the local liberation of elastin peptides. Not all elastin peptides are, however, equivalent in this respect. Synthetic hexapeptides made on the pattern of several exons of the elastin-gene were used for screening with monoclonal antibodies (Wei et al. 1993) and for triggering the ER on endothelial cells (Faury et al. 1998b). These and other studies revealed the sequences recognized by the ER, corresponding to the pattern GXXPG comprising the most studied peptide, VGVAPG. Using the endothelial cell model, it could be shown, however, that even tripeptides as VGV can already trigger the ER (Faury et al. 1998b). Although all the above-cited studies enlarged the scope of the initially proposed immune-theory of atherogenesis, its essential features remained valid, close to the original hypothesis (Robert et al. 1967, 1968). Further experiments are clearly needed essentially to explore the relevance of notions acquired during the study of the ER and their physiopathological consequences as well as their connection to known risk factors of atherogenesis. It seems probable that such studies could open new vistas for original pharmacological innovations.