Printer Friendly

The effect of garlic on arteriosclerotic nanoplaque formation and size.


Objective: In an in vitro biosensor model (PCT/EP 97/05212), the interplay between different lipoproteins in arteriosclerotic nanoplaque formation, as well as aqueous garlic extract (0.2-5.0 g/l from LI 111 powder) as a possible candidate drug against arterio/atherosclerosis were tested within the frame of a high throughput screening. Methods: The processes described below were studied by ellipsometric techniques quantifying the adsorbed amount (nanoplaque formation) and layer thickness (nanoplaque size). A thorough description of the experimental setup has been given previously. Results: Proteoheparan sulfate (HS-PG) adsorption to hydrophobic silica was monoexponential and after approximately 30 min constant. The addition of 2.52 mmol/l [Ca.sup.2+] led to a further increase in HS-PG adsorption because [Ca.sup.2+] was bound to the polyanionic glycosaminoglycan (GAG) chains thus screening their negative fixed charges and turning the whole molecule more hydrophobic. Incubation with 0.2 g/l aqueous garlic extract (GE) for 30 min did not change the adsorption of HS-PG. However, the following addition of [Ca.sup.2+] ions reduced the increase in adsorption by 50.8% within 40 min. The adsorption of a second [Ca.sup.2+] step to 10.08 mmol/l was reduced by even 82.1% within the next 40 min. Having detected this inhibition of receptor calcification, it could be expected that the build-up of the ternary nanoplaque complex is also affected by garlic. The LDL plasma fraction (100 mg/dl) from a healthy probationer showed beginning arteriosclerotic nanoplaque formation already at a normal blood [Ca.sup.2+] concentration, with a strong increase at higher [Ca.sup.2+] concentrations. GE, preferably in a concentration of 1 g/l, applied acutely in the experiment, markedly slowed down this process of ternary aggregational nanoplaque complexation at all [Ca.sup.2+] concentrations used. In a normal blood [Ca.sup.2+] concentration of 2.52 mmol/l, the garlic induced reduction of nanoplaque formation and molecular size amounted to 14.8% and 3.9%, respectively, as compared to the controls. Furthermore, after ternary complex build-up, GE similar to HDL, was able to reduce nanoplaque formation and size. The incubation time for HDL and garlic was only 30 min each in these experiments. Nevertheless, after this short time the deposition of the ternary complex decreased by 6.2% resp. 16.5%, i.e. the complex aggregates were basically resolvable. Conclusions: These experiments clearly proved that garlic extract strongly inhibits [Ca.sup.2+] binding to HS-PG. In consequence, the formation of the ternary HS-PG/LDL/[Ca.sup.2+] complex, initially responsible for the 'nanoplaque' composition and ultimately for the arteriosclerotic plaque generation, is decisively blunted.

Key words: Arteriosclerosis model, calcification, ellipsometry, aqueous garlic extract, lipoproteins, proteogly can receptor



The interfacial behaviour of lipoproteins and their protein components is of interest in numerous biophysical and biochemical processes, as well as in biomedical applications. The deposition of lipoproteins, notably low-density lipoprotein, at the surface of endothelial cell membranes and connective tissue matrices in blood vessels constitute one of the initial steps in arteriosclerosis (Malmsten et al. 2001). LDL, but also modified LDL, VLDL, IDL and Lp(a) are known risk factors for arteriosclerosis, particularly at high [Ca.sup.2+] concentrations (Brown et al. 1993; Anber et al. 1997; Sawamura et al. 1997). In contrast, HDL has been shown to be able not only to prevent arteriosclerosis, but also to reverse it (Gordon and Rifkind, 1989). Naturally, arteriosclerosis is much more complex than this, and after the initial lipoprotein deposition numerous processes, such as oxidation, glycosylation, calcification, posttranslational modification, and bacterial or viral infection may occur, involving not only the endothelium, monocytes and smooth muscle cells, but also cell surfaces and extracellular matrices (Fransson, 1987; Siegel, 1996; Ross, 1999). Nevertheless, the initial deposition step is a surface related phenomenon.

In order to learn more about the interfacial behaviour of lipoproteins, we performed this investigation on the adsorption of lipoproteins at two different surfaces, i.e. one hydrophobic, and one where the hydrophobic surface was further modified through adsorption of HS-PG, the latter substrate thus mimicking the surface receptors exposed to lipoproteins in the blood stream or on their paracellular pathway (Siegel et al. 1999a). In order to obtain information of some biological relevance, an important aspect of the present investigations was to perform these initial experiments at close to in vivo conditions. The results obtained with this simple model are correlated to the role of lipoproteins in the clinical situation. Furthermore, the use of the model system for screening of candidate drugs against arteriosclerosis in a biosensor application was tested for garlic (Siegel et al. 2001). The results of an acute intervention of GE (LI 111) into arteriosclerotic nanoplaque build-up utilizing the LDL plasma fraction at a 100 mg/dl concentration are highlighted and discussed in the present study, mainly on the background of a direct intervention of garlic into atherogenesis independent of any lipid lowering effect.

Materials and Methods

Preparations and Solutions

All experiments were carried out in a blood substitute solution. The normal blood substitute solution consisted of a Krebs solution simulating the extracellular ionic microenvironment of the proteoglycans and lipoproteins, and was of the following composition: [Na.sup.+] 151.16, [K.sup.+] 4.69, [Ca.sup.2+] 2.52, [Mg.sup.2+] 1.1, [Cl.sup.-] 145.4, HC[O.sub.3.sup.-] 16.31, [H.sub.2]P[O.sub.4.sup.-] 1.38 mmol/l (25[degrees]C, pH 7.27). Normally, the solution was aerated with a 95% [O.sub.2]/5% C[O.sub.2] gas mixture (carbogen). Krebs solutions containing LDL were gassed by 93% [N.sub.2]/7% C[O.sub.2] to hinder LDL oxidation (Malmsten et al. 2000).

Native proteoheparan sulfate (HS-PG) from the bovine aorta has a relative molecular mass of about 560 kDa, 11.6% of which is protein ([M.sub.r] [approximately equal to] 400 kDa). It has a few heparan sulfate (HS) side chains (88.4%) in covalent bonding (mol. wt. 35-38 kDa) with an average sulfate content of 0.5 mol sulfate/disaccharide unit (Schmidt and Buddecke, 1988; Siegel et al. 1996b).

Aqueous garlic extract (GE) was used in the concentration steps 0.2 g and 1 g powder suspended per litre [H.sub.2]O. For extraction, 100 g garlic powder (Kwai[R], LI 111, Lichtwer Pharma, Berlin) was eluted with 300 ml [H.sub.2]O and centrifuged (6000 X g) after 20 min (Siegel et al. 1998b). The extraction was repeated twice. About 900 ml supernatant was dried in a rotation evaporator and resuspended in 500 ml bidistilled water. The original garlic bulb powder LI 111 is standardized to 1.0-1.4% alliin, equivalent to 0.5-0.7% allicin. HPLC investigations of the suspension yielded a concentration of 3.58 mmol/l allicin and of 0.184 mmol/l ajoene (Siegel et al. 2001).

Blood samples from healthy volunteers who had fasted for at least 10 h and who did not use antioxidant vitamins or medicaments were collected in polypropylene tubes containing ethylene diamine tetraacetate (EDTA, final concentration 1 mg EDTA/ml blood) and processed within 2 h. During this period, they were kept in semi-darkness. Plasma was obtained by centrifugation at 2500 X g for 6 min. Lipoprotein fractions were isolated by sequential very fast ultracentrifugation using the Optima[TM] TLX ultracentrifuge with rotor TLA-120.2 and thick-wall polycarbonate tubes (Beckman Instruments Inc., Palo Alto, CA, USA) (Leonhardt et al. 1994a, b). Run conditions were full speed (120,000 rpm corresponding to 625,000 X g) and 18[degrees]C temperature. In the first step, the tubes were filled with 0.5 ml plasma which was overlayered with 0.5 ml medium of density 1.006 kg/1. The densities and periods chosen for flotation were 1.006 kg/1 and 30 min for VLDL, 1.063 kg/1 and 150 min for LDL, and 1.21 kg/l and 260 min for HDL. Density media were made oxygen-free by degassing and purging with argon. The flotated lipoproteins of six tubes were aspirated using capillary pipettes and pooled. They were purified from EDTA by gel filtration using size exclusion columns (#732-2010, Bio-Rad Laboratories, Hercules, CA, USA). For elution, Krebs buffer solution of pH 7.3, made oxygen-free for the native lipoproteins, was used.

The lipoprotein size was determined through photon correlation spectroscopy at a scattering angle of 90[degrees], with a Malvern Autosizer II (Malvern, USA). No concentration dependence was observed in the measurements, i.e., the samples were sufficiently diluted to remove obstruction effects. All measurements were performed in the same Krebs solution (pH 7.3) used for the ellipsometry experiments. The size distribution of LDL showed one single peak at 18 nm, and of HDL at 16 nm, respectively. Roughly 60% of the HDL particles had a size of 12-13 nm.

Prior to the ellipsometry investigations, the lipoprotein fractions were again run through a chromatography column (Econo-Pac Column 732-2010, Bio-Rad, Hercules, CA, USA) filled with Krebs buffer solution and finally applied to the experiment in their concentrations in question (Abletshauser et al. 2002). Upon acute in vitro-incubation, GE (0.5 g/l) dissolved in Krebs solution was added to the proteoglycan solution during adsorption to the methylated silica surface during the last 30 min adsorption time. In addition, the lipoproteins were pretreated with GE (0.5 g/l) for 1 h before they were added to the HS-PG solution. At the end of each experiment, GE (1 g/l) was administered once more to discover a possible disintegrating effect on ternary nanoplaque complexes (Siegel et al. 2003b).

Ellipsometry Measurements

The adsorption of the lipoproteins and of HS-PG was monitored by in situ ellipsometry, as detailed previously (Azzam and Bashara, 1989; Landgren and Jonsson, 1993; Malmsten, 1994). All measurements were performed with a Rudolph 436 thin film ellipsometer at 401 nm. In summary, the optical response of the bare surfaces was first investigated, whereafter the proteoglycan or lipoprotein was added and changes in the state of polarization of the reflected light were monitored. The adsorbed amount was determined according to de Feijter using a refractive index increment of 0.18 c[m.sup.3]/g (De Feijter et al. 1978). A constant bulk concentration of 0.1 mg/ml HS-PG (0.192 mmol/l in disaccharide units), of 58 mg/dl LDL or 29 mg/dl HDL were used throughout. Furthermore, the measurements were performed at different [Ca.sup.2+] concentrations around that in the biological system. The normal blood substitute solution consisted of a Krebs solution. The pH was kept between 7.25 and 7.35 by the bicarbonate/phosphate buffer, and by a continuous aeration of the cuvette solution with a 95% [O.sub.2]/5% C[O.sub.2] (lipoproteins: 93% [N.sub.2]/7% C[O.sub.2]) gas mixture (Aga, Sweden). The latter is included due to the necessity to keep the pH well regulated during these experiments in order to avoid both proteoglycan and lipoprotein degradation and calcium phosphate precipitation. The adsorption was performed at 25[degrees]C throughout. Further details can be found in previous publications (Malmsten et al. 1993; Siegel et al. 1996a).


The protein backbone of a syndecan molecule consists of small intracellular and transmembrane domains, and a large extracellular domain to which heparan/chondroitin sulfate GAGs are attached in variable manner. HS-PG deposition at a methylated silica surface is initiated by its transmembrane hydrophobic core domain (Malmsten et al. 1994; Siegel et al. 1998a). The GAG side chains are stretched out into the blood substitute solution because of their negative fixed charges giving rise to electrostatic repulsion (Malmsten et al. 1994; Siegel et al. 2001). Lipoprotein particles or cations may interact with the GAG chains, e.g. on occasion of their positive amino acid residues or simply their positive charges (Fig. 1). Thus, the situation reflects physiological conditions, since in the endothelial cell membrane the scenario differs only in one respect, that the short hydrophilic intracellular core domain is located in the cell interior. In this sense, the hydrophobic silica surface simulates the cell membrane overlaid with the glycocalyx quite perfectly. Furthermore, in this in vitro model any drug can be tested with respect to its antiarteriosclerotic potential (Siegel et al. 2001, 2003a).

Fig. 2 demonstrates HS-PG adsorption from a [Ca.sup.2+]-free Krebs solution to a methylated silica surface over time (Malmsten et al. 1993; Malmsten and Siegel, 1995). The adsorbed amount ([GAMMA]) is depicted on the ordinate. Since [Ca.sup.2+] ions screen the negative GAG chains and could be expected to reduce the charge density of the proteoglycan also by specific binding (Siegel et al. 1991; Malmsten et al. 1994), this ion species promoted HS-PG adsorption.

The adsorption of LDL at methylated silica was found to be quite different in time course as compared to HS-PG. After a fast initial adsorption, a desorption period of approximately 1 h followed (Fig. 3). After that time, the adsorption was constant. This adsorption behaviour could be explained by either competitive adsorption in a multicomponent mixture or so-called competitive spreading (Cohen Stuart, 1998). Furthermore, we were interested to learn whether [Ca.sup.2+] ions would influence LDL adsorption as could be demonstrated for HS-PG. Fig. 3 shows that [Ca.sup.2+] ions in a physiological concentration of 2.52 mmol/l (Ca1) did not change or slightly increased the LDL adsorption, whereas 10.08 mmol/l [Ca.sup.2+] (Ca2) accelerated the adsorption dramatically, probably by creating aggregates depositing at the silica surface. A similar aggregation behaviour was found for HDL.

In a second step, the physiological scenario was simulated. Lipoprotein binding to HS-PG preadsorbed from a [Ca.sup.2+]-free Krebs solution at methylated silica resulted in no additional adsorbed amount for LDL during the first 25 min (Fig. 4). [Ca.sup.2+] ions, however, significantly changed the HS-PG/LDL interaction. With [Ca.sup.2+]-induced LDL aggregation, the adsorbance increased dramatically. Therefore, [Ca.sup.2+] ions interfered decisively with the HS-PG/LDL interplay and formed big aggregates. This ternary complex, proteoglycan receptor--low density lipoprotein--calcium at the endothelial cell and interstitial matrix surfaces could be interpreted as primary lesion'and genuine nanoplaque build-up'on the molecular level (Siegel et al. 2001; Malmsten et al. 2001).




In this context, the known protective role of HDL was examined. After proteoheparan sulfate receptor adsorption from a [Ca.sup.2+]-free Krebs solution, HDL was introduced and subsequently, the LDL adsorption was examined. As can be seen, the addition of LDL in physiological concentrations resulted in no detectable lipoprotein deposition (Fig. 4). Most surprisingly, [Ca.sup.2+] ions had no effect at all, even in a concentration of 10.08 mmol/l. Moreover, the repeated addition of LDL did not change the adsorption likewise. Therefore, a close analogy is found between the clinically observed protective effect of HDL on the one hand (Ji et al. 1997) and the LDL/[Ca.sup.2+]-deposition observed in the present investigation on the other hand.



Fig. 5 shows in a typical experiment that the calcification of the proteoglycan receptor could be promptly halted by aqueous garlic extract (Lin et al. 2003). The incubation time for GE was only 30 min in the test experiment and the total concentration 0.2 g/l before the first [Ca.sup.2+] addition. Incubation with GE for half an hour during the second period of HS-PG adsorption led only to a small hump'in total adsorbed amount (see also Fig. 6). This indicative of GE binding to the proteoheparan sulfate receptor molecules. Already after [Ca.sup.2+] addition of 2.5 mmol/l, the adsorbed amount in the garlic experiment remained well below the control curve. First of all, the garlic curve in its initial phase followed the control curve but deviated downwards more and more from it with time to take finally even an opposite turn (falling e-function). Thus, the [Ca.sup.2+]-induced extra adsorption was strongly reduced within 40 min. This means that garlic hindered [Ca.sup.2+] ions to bind to their receptor sites thus preventing calcification. The reduction in calcification was 50.8% for Ca1 and 82.1% for Ca2, respectively.


Having detected this inhibition of receptor calcification, it could be expected that the build-up of the ternary nanoplaque complex is also affected by garlic (Siegel et al. 2002b; Vastag, 2002). While in Figs. 3-5 the molecular mechanisms were elucidated, through which garlic is implicated in its role as interference factor on the [Ca.sup.2+]-driven proteoglycan-LDL aggregation, we will now demonstrate direct evidence of the partial prevention of arteriosclerotic nanoplaque formation. We carried out complete [Ca.sup.2+] titration curves to demonstrate that plaque formation is diminished and shifted to higher [Ca.sup.2+] concentrations under garlic. With the experiments depicted in Fig. 6 we wanted to evaluate the applicability of the molecular model for nanoplaque formation and size in an in vitro biosensor application within the frame of a high throughput screening for candidate drugs against arteriosclerosis. The acute application of GE followed at two times: one half was added to the Krebs solution after 30 min during HS-PG adsorption, with the other half the LDL fraction dispersed in Krebs solution was preincubated for 60 min. The ternary complex build-up was pursued through increasing the [Ca.sup.2+] concentration. At the end of the experiment, control and test sample were again treated for 30 min with GE (1 g/l) after formation of the ternary nanoplaque complex.

Shown in Fig. 6 is the adsorption of LDL at an HS-PG coated surface. As can be seen, the adsorption is limited in the absence of [Ca.sup.2+], at least for a considerable time. In the presence of the physiological serum [Ca.sup.2+] concentration (Ca1: 2.52 mmol/l), the deposition was weakly decelerated in the control experiment (by -8.5%), but returned almost to the initial control value with twice the normal [Ca.sup.2+] concentration (reduction by -2.0%), whereas in the garlic experiment a stronger decrease in ternary complex deposition [in situ GE application: -22.0% (2.52 mmol/l [Ca.sup.2+]), -22.1% (5.04 mmol/l [Ca.sup.2+])] was measured (Fig. 7A). Already prior to the end of the 40 min Cal-application, the weak decrease of nanoplaque formation and deposition almost stopped, in parallel with a larger scatter of the measuring points (Fig. 6). The scatter reflects the heterogeneity in particle size of the aggregational ternary complexes. The shape of the curve orients itself by the [GAMMA]-mean values at the respective [Ca.sup.2+] concentrations, and was calculated using an algorithm for least-squares estimation of nonlinear parameters (Marquardt, 1963). The acute effect of garlic became apparent in a slight diminution of nanoplaque formation and a change of allosteric Hill kinetics towards simple saturation kinetics for higher [Ca.sup.2+] concentrations. Noteworthy is the final decline in ternary nanoplaque deposition upon first application of garlic at the end of the control experiment. This effect is highlighted in the test experiment preincubated with garlic, where the decrease was even more pronounced (see also Fig. 9). Therefore, GE was able to at least partially dissolve again nanoplaque complexes of recent formation.


These findings were confirmed by measuring the thickness of the interfacial layer. In the control experiment, the adsorbed layer thickness ([[delta].sub.el]) continued for the present its trend to slightly decrease after LDL application upon Ca1 addition, but then reversed still before the Ca2 addition and rose steeply with increasing [Ca.sup.2+] concentrations. Fig. 7 illustrates a significant garlic-induced decrease in formation and molecular size development of nanoplaques in function of the extracellular [Ca.sup.2+] concentration for normalized values of adsorbed amount and adsorbed layer thickness. It further demonstrates that beyond nanoplaque formation and deposition (A) also its dimensional build-up (B) is a [Ca.sup.2+] driven process. Both nanoplaque formation and size did not markedly change upon Ca1 and Ca2 additions, but prominently increased upon further Ca additions. Altogether, the augmentation in total adsorbed amount and layer thickness with increasing [Ca.sup.2+] concentration took a distinctly reduced course under garlic as compared to the control curve. And finally, the initial rise in adsorbed layer thickness at the end of the control experiment during the first GE incubation ran parallel to an equal increase in total adsorbed amount. On the other hand, a more or less marked reduction in size of the ternary nanoplaque complex appeared with the lipid fraction pretreated with garlic upon repeated garlic addition. This finding supports further evidence that garlic can disintegrate newly formed nanoplaques (Lin et al. 2003).

The investigations with low-density lipoprotein emphasized that the acute treatment with GE can partly prevent both the formation and the size development of ternary nanoplaques, and this even at high [Ca.sup.2+] concentrations (Siegel et al. 2002b; Vastag, 2002). Fig. 8 illustrates the extent of inhibition. There arose a similar curve profile for the reduction in adsorbed amount and layer thickness of ternary nanoplaques dependent on the [Ca.sup.2+] concentration used, whereby in the latter case the peak inhibition is merely shifted to higher [Ca.sup.2+] concentrations. The degree of inhibition varied between 14% and 31% for nanoplaque formation. Similar results were found for the reduction in adsorbed layer thickness, although nanoplaque size was only slightly diminished in normal serum [Ca.sup.2+] concentration. On the other hand, with in situ-application of garlic, a continuous diminution of nanoplaque size was measured with increasing [Ca.sup.2+] concentrations. In total, the inhibition ranged between 4% and 73%. In a normal blood [Ca.sup.2+] concentration of 2.5 mmol/l, the garlic-induced decrease in nanoplaque formation and molecular size amounted to 15% and 5% within 40 min, respectively, as compared to the controls. Furthermore, it should be stressed that the repeated addition of garlic at the end of the experiments led to a considerable decline in adsorbed amount and layer thickness.

Of profound therapeutic impact within this context is the question, whether nanoplaque complexes can be disrupted again. In previous measurements we could show that a ternary nanoplaque complex once created can be partially dissolved by HDL and garlic (Siegel et al. 2001). In those experiments, the incubation time for HDL and garlic was only 30 min each. Nevertheless, after this short time the deposition of the ternary complex decreased by 6.2% resp. 16.5%, i.e. the complex aggregates were basically resolvable. Thus, GE in a concentration relevant for man was even 2.5 times more effective in removing the ternary complex as compared with HDL. Both substances together disintegrated the nanoplaques by 22.7% within 1 h, a very short time span when contrasting it to in vivo conditions where such a plaque reducing process may last months or even years. Fig. 9 demonstrates a section of Fig. 6 in more detail representing the garlic-induced decrease in adsorbed amount after nanoplaque build-up in the present experiments. Both in the control and in the garlic test experiment, the arteriosclerotic nanoplaque neoformation increased further after addition of GE to the 17.64 mmol/l [Ca.sup.2+]-Krebs solution, however, was completely stopped already after 14 min. During the following 16 min, 25.4% of the formed nanoplaques in the control experiment were dissolved again upon application of 1 g/l GE for the first time. In the pretreated test experiment, the disintegration amounted to even 33.3% within the same period, and with a total GE concentration of 2 g/l. These results confirm impressively our evidence previously given.




Stretches of basic amino acid-rich residues within apoB, apoA and apoE constituting the predominant protein moiety of LDL and HDL display a high electrostatic binding affinity to proteoglycans. In the proteoheparan sulfate perlecan, the large protein core carries three HS chains of 30-70 kDa which are 100-170 nm in length and covalently attached to Ser-Gly consensus sequences clustered at the N terminus (Trescony et al. 1989). This major HS-attachment domain I is considered to be a putative lipoprotein binding site for reasons of electrostatic attraction (Siegel, 1996). Immediately adjacent, domain II contains four copies of LDL receptor repeats and one immunoglobulin repeat (NCAM module) (Timpl, 1993). Also such motifs in perlecan are thought to mediate LDL binding for reasons of core protein receptor module-lipoprotein ligand interactions. Interestingly, the expression of the proteoglycan in endothelial cells is modulated by LDL (Olgemoller et al. 1990). Full-blown arteriosclerotic plaques as clinically detected by spiral computer tomography and magnetic resonance imaging basically contain lipoproteins associated with proteoglycans of high diversity and versatility and complexly aggregated with high amounts of calcium (Chait et al. 1997). Therefore, the replica in nanoscale dimensions of the heterotrimeric aggregational complex consisting of proteoglycan receptor, lipoprotein and calcium presumably portrays the very earliest stages of an arteriosclerotic plaque, which remains undiscovered even with the most up to date clinical diagnostics. This so-called primary lesion impresses merely as endothelial dysfunction. Furthermore, this may also be related to inducible nitric oxide synthase (iNOS) that has recently been shown to be present in human arteriosclerotic lesions. As an atherogenic target, enzyme activity, expression and protein of iNOS were markedly reduced in the presence of the garlic constituents allicin and ajoene (Dirsch et al. 1998). In the current study, a direct molecular effect of GE on nanoplaque formation and size in a cell-free assay is highlighted.

The formation of the proteoglycan-lipoprotein-calcium complex (Figs. 4, 6, 9: ternary complex HS-PG/LDL/[Ca.sup.2+]) may be initiated by cooperative [Ca.sup.2+]-binding to the proteoglycanic GAG chain inducing a diminution of the axial rise of its left-handed [2.sub.1]-helix through a conformational change, so that the pitch of turns is compressed (Atkins, 1977; Siegel et al. 1996b). Simultaneously, the cooperativity of this binary complex may be amplified by allosteric lipoprotein docking to that shortened conformation supported by the higher charge density and fitting stereospecificity. Thus, the heterotrimeric complex would become rather stable. The adsorption to the silica surface may be promoted. This, however, is rather speculative and not known at present. It is equally possible that deposition through [Ca.sup.2+]-dependent complex formation between the lipoproteins and HS-PG is likely to occur and contribute to the observed effects.

Although much remains unclear regarding the mechanisms of lipoprotein depositions at proteoglycan-coated surfaces, particularly for HDL, it is clear that the use of this model system offers large possibilities for studying these processes on a molecular level. In particular, the [Ca.sup.2+]-promoted LDL deposition and the protective effect of HDL, even at high [Ca.sup.2+] and LDL concentrations, agree well with previous clinical observations regarding risk and beneficial factors for the early stages of arteriosclerosis (Voyiaziakis et al. 1998). The protection of HDL against calcified lipid particles in Fig. 4 has to be estimated all the more powerful because aggregation and increasing deposition of both LDL and HDL particles started always from a 10.08 millimolar [Ca.sup.2+]-Krebs solution even without proteoglycan receptor (Fig. 3). This HDL effect may be interpreted as an indication that HDL strongly counteracts primary lesion generation in the form of the heterotrimeric complex, proteoglycan-LDL-calcium. Furthermore, it was derived from these experiments that HDL has a higher affinity to its proteoglycan receptor by a factor of at least four compared to LDL (Abletshauser et al. 2002). This means that a person with an HDL plasma concentration of 50 mg/dl could tolerate an LDL concentration of 200 mg/dl if no further cardiovascular risk factor is present.

Altogether, we believe that the model system presented can be of some use in investigations, e.g. of the interplay between different lipoproteins in arteriosclerotic plaque formation, as well as in a high throughput screening of possible candidate drugs against arterio/atherosclerosis in a biosensor application (Siegel et al. 2001, 2002a, 2003b). In order to test its applicability, we investigated the efficacy of aqueous garlic extract on arteriosclerotic nanoplaque formation and size with a normal low-density lipoprotein concentration. Here, the results were presented with the LDL fraction after acute pretreatment of both proteoglycan receptor and lipoprotein for 30 (60) min. Therefore, possible effects of garlic could in no case be traced back to an influence on the lipid concentrations. We could show that neither adsorbed amount nor layer thickness of the ternary nanoplaque complex were elevated at a physiologic blood [Ca.sup.2+] concentration of 2.5 mmol/l (a distinct reduction was even the case) (Fig. 7) and that also at all augmented [Ca.sup.2+] concentrations, calcified atheromatous nanoplaques were reduced in formation and size in comparison to untreated controls of the same LDL fraction (Figs. 6-9). In these acute experiments, GE had been administered in situ at a constant concentration. Similar to HDL, compounds of GE were shown to bind both to the proteoglycan receptor and in or on the lipid particles. Successful experiments with statins (Abletshauser et al. 2002; Siegel et al. 2002a, 2003a, b) had directed us to pretreat proteoglycan receptors and lipoproteins in these experiments with in situ-application of GE before both biopolymers could interact with each other. The influence of garlic already appeared with the docking process which showed an acceleration in LDL-binding kinetics with simple exponential desorption and following adsorption. Without a detailed pharmacodynamic analysis we would like to abstain from an interpretation for the present (Derendorf et al. 1997). In agreement with previous studies in which under garlic a decline in atheromatous nanoplaques has been demonstrated (Siegel et al. 2001), the dissolving effect of GE on nascent, even calcified, nanoplaques was illustrated at the end of the test period (Figs. 6, 8. 9). Based on the background of recent clinical studies whereupon the [Ca.sup.2+]-content in the coronary vascular wall as measured by electron beam tomography is predictive of future cardiovascular events (Georgiou et al. 2001) this finding of a disintegration of calcified nanoplaques may have high clinical relevance (Lin et al. 2003).

Summarizing, it shall be stated that the pleiotropic effects of garlic described so far, such as lowering of LDL cholesterol and inhibition of its oxidation, enhancement of HDL cholesterol, reduction of serum triglycerides, lowering of fibrinogen concentration, decrease in blood pressure and promotion of blood supply and stimulation of fibrinolytic activity, inhibition of platelet aggregation and reduction of plasma viscosity (Siegel et al. 1999b), is joined by a further pleiotropic effect, that possibly has prophylactic and therapeutic importance under garlic for the lowering of cardiovascular risk with respect to myocardial infarction and stroke, independent of any decrease in LDL or increase in HDL concentrations (Shepherd, 1995; Schatz et al. 2001). According to our conception, garlic could acutely intervene in the deranged control circuit at the blood-endothelium-matrix interface and assist HDL in its vasculoprotective effect from an increased LDL uptake and formation of ternary proteoglycan receptor-lipoprotein-calcium nanoplaques as well as through their redissolving. In conclusion, these experiments clearly proved that aqueous garlic extract strongly inhibits [Ca.sup.2+] binding to heparan sulfate proteoglycan. Thus, the formation of the ternary proteoglycan receptor/LDL cholesterol/calcium complex, initially responsible for the 'nanoplaque' composition and ultimately for the arteriosclerotic plaque generation, is decisively blunted (Vastag, 2002; Lin et al. 2003). The similarity between and the concerted action of HDL and garlic extract in preventing and curing arteriosclerosis justifies the designation of garlic powder LI 111 as a 'phyto-HDL.'


The authors wish to thank Angela Becker (Berlin), Sigrid Nitzsche (Dresden) and Annika Dahlman (Stockholm) for their skilful technical assistance. This work was financed in part by the German Pharmaceutical Industry (PRO 99-02174, Freie Universitat Berlin), Lichtwer Pharma (Berlin, Germany) and by the Institute for Research and Competence Holding AB (IRECO, Stockholm, Sweden).


Abletshauser C, Klussendorf D, Schmidt A, Winkler K, Marz W, Buddecke E, Malmsten M, Siegel G (2002) Biosensing of arteriosclerotic nanoplaque formation and interaction with an HMG-CoA reductase inhibitor. Acta Physiol Scand 176: 131-145

Anber V, Millar JS, McConnell M, Shepherd J, Packard CJ (1997) Interaction of very-low-density, intermediate-density, and low-density lipoproteins with human arterial wall proteoglycans. Arterioscler Thromb Vasc Biol 17: 2507-2514

Atkins EDT (1977) Molecular architecture of the animal and some microbial extracellular polysaccharides. In: Proceedings of the first Cleveland symposium on macromolecules. Structure and properties of biopolymers (ed Walton AG): 61-85. Elsevier, Amsterdam

Azzam RMA, Bashara NM (1989) Ellipsometry and polarized light. North-Holland, Amsterdam

Brown BG, Zhao X-Q, Sacco DE, Albers JJ (1993) Lipid lowering and plaque regression. New insights into prevention of plaque disruption and clinical events in coronary disease. Circulation 87: 1781-1791

Chait A, Chang MY, Olin K, Wight T (1997) Interaction of oxidized LDL with arterial proteoglycans. Atherosclerosis 134: 197

Cohen Stuart MA (1998) Macromolecular adsorption: a brief introduction. In: Biopolymers at interfaces (ed Malmsten M): 1-25. Marcel Dekker, New York

De Feijter JA, Benjamins J, Veer FA (1978) Ellipsometry as a tool to study the adsorption behavior of synthetic and biopolymers at the air-water interface. Biopolymers 17: 1759-1772

Derendorf H, Mollmann H, Hochhaus G, Meibohm B, Barth J (1997) Clinical PK/PD modelling as a tool in drug development of corticosteroids. Int J Clin Pharmacol Therap 35: 481-488

Dirsch VM, Kiemer AK, Wagner H, Vollmar AM (1998) Effect of allicin and ajoene, two compounds of garlic, on inducible nitric oxide synthase. Atherosclerosis 139: 333-339

Fransson L-A (1987) Structure and function of cell-associated proteoglycans. Trends Biochem Sci 12: 406-411

Georgiou D, Budoff MJ, Kaufer E, Kennedy JM, Lu B, Brundage BH (2001) Screening patients with chest pain in the emergency department using electron beam tomography: a follow-up study. J Am Coll Cardiol 38: 105-110

Gordon DJ, Rifkind BM (1989) High-density lipoprotein: The clinical implications of recent studies. N Engl J Med 321: 1311-1316

Ji ZS, Dichek HL, Miranda RD, Mahley RW (1997) Heparan sulfate proteoglycans participate in hepatic lipase and apolipoprotein E-mediated binding and uptake of plasma lipoproteins, including high density lipoproteins. J Biol Chem 272: 31285-31292

Landgren M, Jonsson B (1993) Determination of the optical properties of Si/Si[O.sub.2] surfaces by means of ellipsometry, using different ambient media. J Phys Chem 97: 1656-1660

Leonhardt W, Pietzsch J, Julius U, Hanefeld M (1994a) Recovery of cholesterol and triacylglycerol in very-fast ultracentrifugation of human lipoproteins in a large range of concentrations. Eur J Clin Chem Clin Biochem 32: 929-933

Leonhardt W, Pietzsch J, Nitzsche S (1994b) Very-fast ultracentrifugation of human plasma lipoproteins: influence of the centrifugal field on lipoprotein composition. Clin Chim Acta 224: 21-32

Lin MC, Costello RB, Betz JM, Wong SS, Furberg C, Block E (2003) NIH workshop on mechanistic studies of cardiovascular effects of botanicals. Executive summary. Circulation (in press)

Malmsten M, Siegel G, Buddecke E, Schmidt A (1993) Cation-promoted adsorption of proteoheparan sulphate. Coll Surf B: Biointerfaces 1: 43-50

Malmsten M (1994) Ellipsometry studies of protein layers adsorbed at hydrophobic surfaces. J Colloid Interface Sci 166: 333-342

Malmsten M, Claesson P, Siegel G (1994) Forces between proteoheparan sulfate layers adsorbed at hydrophobic surfaces. Langmuir 10: 1274-1280

Malmsten M, Siegel G (1995) Electrostatic and ion-binding effects on the adsorption of proteoglycans. J Colloid Interface Sci 170: 120-127

Malmsten M, Siegel G, Wood WG (2000) Ellipsometry studies of lipoprotein adsorption. J Colloid Interface Sci 224: 338-346

Malmsten M, Siegel G, Becker A (2001) A model substrate for ellipsometry studies of lipoprotein deposition at the endothelium. J Colloid Interface Sci 240: 372-374

Marquardt DW (1963) An algorithm for least-squares estimation of nonlinear parameters. J Soc Indust Appl Math 11: 431-441

Olgemoller B, Schleicher ED, Schwaabe S, Guretzki H-J, Gerbitz KD (1990) High concentrations of low density lipoprotein decrease basement membrane-associated heparan sulfate proteoglycan in cultured endothelial cells. FEBS Lett 264: 37-39

Ross R (1999) Atherosclerosis--an inflammatory disease. New Engl J Med 340: 115-126

Sawamura T, Kume N, Aoyama T, Moriwaki H, Hoshikawa H, Aiba Y, Tanaka T, Miwa S, Katsura Y, Kita T, Masaki T (1997) An endothelial receptor for oxidized low-density lipoprotein. Nature 386: 73-77

Schatz IJ, Masaki K, Yano K, Chen R, Rodriguez BL, Curb JD (2001) Cholesterol and all-cause mortality in elderly people from the Honolulu Heart Program: a cohort study. Lancet 358: 351-355

Schmidt A, Buddecke E (1988) Cell-associated proteoheparan sulfate from bovine arterial smooth muscle cells. Exp Cell Res 178: 242-253

Shepherd J (1995) Fibrates and statins in the treatment of hyperlipidemia: an appraisal of their efficacy and safety. Eur Heart J 16: 5-13

Siegel G, Walter A, Ruckborn K, Buddecke E, Schmidt A, Gustavsson H, Lindman B (1991) NMR studies of cation induced conformational changes in anionic biopolymers at the endothelium-blood interface. Polymer J 23: 697-708

Siegel G (1996) Connective tissue: more than just a matrix for cells. In: Comprehensive human physiology, vol. 1 (eds Greger R, Windhorst U): 173-224. Springer-Verlag, Berlin, Heidelberg

Siegel G, Malmsten M, Klussendorf D, Walter A, Schnalke F, Kauschmann A (1996a) Blood-flow sensing by anionic biopolymers. J Auton Nerv Syst 57: 207-213

Siegel G, Walter A, Kauschmann A, Malmsten M, Buddecke E (1996b) Anionic biopolymers as blood flow sensors. Biosensors Bioelectron 11: 281-294

Siegel G, Malmsten M, Lindman B (1998a) Flow sensing at the endothelium-blood interface. Coll Surf A: Physicochem Eng Aspects 138: 345-351

Siegel G, Nuck R, Schnalke F, Michel F (1998b) Molecular evidence for phytopharmacological [K.sup.+] channel opening by garlic in human vascular smooth muscle cell membranes. Phytother Res 12: S149-S151

Siegel G, Malmsten M, Klussendorf D, Leonhardt W (1999a) Physicochemical binding properties of the proteoglycan receptor for serum lipoproteins. Atherosclerosis 144: 59-67

Siegel G, Walter A, Engel S, Walper A, Michel F (1999b) Pleiotrope Wirkungen von Knoblauch. Wien med Wschr 149: 217-224

Siegel G, Malmsten M, Klussendorf D, Michel F (2001) A receptor-based biosensor for lipoprotein docking at the endothelial surface and vascular matrix. Biosensors Bioelectron 16: 895-904

Siegel G, Abletshauser C, Malmsten M, Klussendorf D (2002a) The acute effect of fluvastatin on lipoprotein deposition in a model substrate for ellipsometry studies at an endothelial membrane equivalent. Desalination 148: 407-414

Siegel G, Malmsten M, Schneider W. Michel F (2002b) Inhibition of arteriosclerotic nanoplaque formation by Allium sativum. In: International Scientific Conference on Complementary, Alternative & Integrative Medicine Research, Abstract Book (eds Eisenberg DM, Folkman S) Boston: Harvard Medical School

Siegel G, Abletshauser C, Malmsten M, Klussendorf D (2003a) The effect of an HMG-CoA reductase inhibitor on arteriosclerotic nanoplaque formation and size in a biosensor model. Biosensors Bioelectron 18: 635-647

Siegel G, Abletshauser C, Malmsten M, Schmidt A, Winkler K (2003b) Reduction of arteriosclerotic nanoplaque formation and size by fluvastatin in a receptor-based biosensor model. Cardiovasc Res 58: 696-705

Timpl R (1993) Proteoglycans of basement membranes. Experientia 49: 417-428

Trescony PV, Oegema TR Jr, Farnam BJ, Deloria LB (1989) Analysis of heparan sulfate from the Engelbreth-Holm-Swarm (EHS) tumor. Connect Tissue Res 19: 219-242

Vastag B (2002) Health agencies update: garlic prevents plaque ... JAMA 288: 1342

Voyiaziakis E, Goldberg IJ, Plump AS, Rubin EM, Breslow JL, Huang LS (1998) ApoA-I deficiency causes both hypertriglyceridemia and increased atherosclerosis in human apoB transgenic mice. J Lipid Res 39: 313-321

G. Siegel (1), M. Malmsten (2), J. Pietzsch (3), A. Schmidt (4), E. Buddecke (4), F. Michel (5), M. Ploch (5) and W. Schneider (5)

(1) Institute of Physiology, Charite, Campus Benjamin Franklin, Berlin, Germany

(2) Institute for Surface Chemistry, Royal Institute of Technology, Stockholm. Sweden

(3) Institute for Bioinorganic and Radiopharmaceutical Chemistry, Research Center Rossendorf, Dresden, Germany

(4) Institute for Arteriosclerosis Research, University of Munster, Munster, Germany

(5) Lichtwer Pharma AG, Berlin, Germany


G. Siegel, Institute of Physiology, Biophysical Research Group, Charite, Campus Benjamin Franklin, Arnimallee 22, D-14195 Berlin, Germany

Tel.: ++49-30-84451685; Fax: ++49-30-84451684; e-mail:
COPYRIGHT 2004 Urban & Fischer Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Siegel, G.; Malmsten, M.; Pietzsch, J.; Schmidt, A.; Buddecke, E.; Michel, F.; Ploch M.; Schneider,
Publication:Phytomedicine: International Journal of Phytotherapy & Phytopharmacology
Geographic Code:4EUGE
Date:Jan 1, 2004
Previous Article:Isolation of linoleic acid as an estrogenic compound from the fruits of Vitex agnus-castus L. (chaste-berry).
Next Article:Characteristics, clinical effect profile and tolerability of a nasal spray preparation of Artemisia abrotanum L. for allergic rhinitis.

Related Articles
Tests prove garlic good for arteries.
Garlic's power.
Sankyo Ceases Clinical Studies of Pactimibe, Investigational Drug for ASCVD.
Billy Mays' death linked to cocaine use.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |