Action of Tropoelastin and Synthetic Elastin Sequences on Vascular Tone and on Free Ca2 Level in Human Vascular Endothelial Cells
The elastic properties of extensible tissues such as arteries and skin are mainly due to the presence of elastic fibers whose major component is the extracellular matrix protein elastin. Pathophysiological degradation of this protein leads to the generation of elastin peptides that have been identified in the circulation in the ng/mL to µg/mL range. Similar concentrations of an elastin peptide preparation (-elastin) were previously demonstrated to induce among other biological actions a dose- and endothelium-dependent vasorelaxation mediated by the elastin/laminin receptor and by endothelial NO production. To determine the elastin sequence(s) responsible for vasomotor activity and to learn more about possible signaling pathways we have compared the action of different concentrations (10-13 to 10-7 mol/L) of recombinant human tropoelastin eight synthetic elastin peptides and a control peptide (VPVGGA) on both rat aortic ring tension and [Ca2 ]i of cultured human umbilical vein endothelial cells. No vasoactivity could be detected for VPVGGA and for the elastin-related sequences VGVGVA PGVGVA and GVGVA. Tropoelastin VGV PGV and VGVAPG were found to induce an endothelium- and dose-dependent vasorelaxation and to increase endothelial [Ca2 ]i whereas PVGV and VGVA produced these effects only at low concentration (10-11 mol/L). A likely candidate for mediating the elastin peptide–related effects is the elastin/laminin receptor since the presence of lactose strongly inhibited the vasoactivity associated with these compounds. Our results show that although the flanking amino acids modulate its activity VGV seems to be the core sequence recognized by the elastin receptor.
Elastin the extracellular matrix protein responsible for the major part of tissue elasticity in vertebrates functions as an insoluble component of elastic fibers in many extensible organs such as skin lung and arteries.The alteration of this protein mediated by elastase-type enzymes leads to severe pathologies such as emphysema and arteriosclerosis and diverse skin diseases.3As a consequence of either physiological (such as aging) or pathological elastin degradation soluble elastin peptides are released into the surrounding extracellular space and have been detected in the circulation at a concentration range of 10-6 to 10-2 mg/mL。
Numerous studies have demonstrated that elastin peptides have diverse biological properties. In particular elastin peptides have been shown to stimulate fibroblast adhesion to elastin fibersto regulate cellular proliferationand to be chemotactic for monocytes and (similar to tropoelastin) fibroblasts.Moreover-elastin a mixture of peptides resulting from the degradation of elastin in KOH has recently been shown to induce dilation in aortic rings from adult rats but not from younger and older animals. These peptides have also been shown to trigger intracellular events such as an increase in both Ca2 influx and [Ca2 ]i in leukocytesand in endothelial cells (authors' unpublished data 1997). The effects of tropoelastin and elastin peptides are mediated by their binding to a multifunctional high-affinity receptor which on most cells including endothelial cells is a 67- to 69-kD peripheral membrane protein also able to bind laminin. One exception however is a 59-kD elastin binding protein in Lewis lung carcinoma cells. The 67-kD receptor has been shown to also bind lactose-type carbohydrates to a lectin binding site which has the effect of inducing both the release of elastin from its binding site and the release of the receptor itself from the cell membrane.
Several biologically active sequences in elastin have been identified. In early studies the repetitive sequence VGVAPG present in bovine and human tropoelastin was found to be chemotactic for both monocytes and fibroblasts.2This same peptide and related sequences GFGVGAGVP and GLGVGAGVP were shown to be chemoattractants for aortic endothelial cells.Other repeated elastin sequences were also found to be biologically active including AGVPGFGVG and GFGVGAGVP which are chemotactic for fibroblasts and (VPGVG)n which induces calcification in the rat.Furthermore recent results indicate that the elastin peptide VGVAPG and five related shorter or frame-shifted elastin sequences (VGV VGVA GVGVA PGVGVA and VGVGVA) act in leukocytes on intracellular free Ca2 metabolism elastase release released myeloperoxidase activity or hydrogen peroxide or superoxide anion production.
We have previously demonstrated that elastin induces a dose- and endothelium-dependent vasorelaxation on NA-precontracted rat aortic rings mediated by the elastin/laminin receptor and involving both NO synthase and cyclooxygenase pathways. To determine the active sequence(s) of elastin on vasomotor activity we have investigated the action of eight synthetic elastin (VGV PGV VGVA GVGVA PGVGVA and VGVAPG) and related (PVGV and VGVGVA) peptides on NA-contracted aortic rings and on [Ca2 ]i of cultured HUVECs. The activity of these peptides was also compared with the activity of rTE. The peptide VPVGGA a rearrangement of the amino acids in the active peptide VGVAPG served as a control. Our results show that rTE and several elastin-related peptides induce both endothelium-dependent vasorelaxation and an increase in endothelial [Ca2 ]i. The core sequence of this activity seems to be VGV whose effect is modulated by the flanking amino acids.
Preparation of -Elastin
Elastin peptides (elastin) were obtained by hydrolysis of highly purified bovine ligamentum nuchae elastin with 1 mol/L KOH in 80% (vol/vol) aqueous ethanol. By adjusting the time of hydrolysis peptides of different molecular mass can be obtained. Peptides used in these experiments had an average molecular mass of 75 kD.
Vasorelaxant Activity of Elastin-Related Compounds
The vasoactivity of elastin-related peptides was found to be compound and concentration dependent on aortic rings with intact endothelium (Table 1 category A). Compared with the control the peptides VGV PGV VGVAPG and rTE induced dose-dependent vasorelaxation in the NA-precontracted rat aortic rings with maximum values of 32% 21% 26% and 41% respectively (Fig 1A). Nevertheless for each concentration for which several peptides were found to be active no significant difference in activity could be detected between these peptides except at 10-9 mol/L where rTE was more active than VGV (Fig 1A). Moreover the peptide PGV was slightly active (14% vasorelaxation) at 10-12 mol/L (Fig 1A). The sequences VGVA and PVGV induced a relaxation of the aortic rings only at a concentration of 10-11 mol/L with values of 14% and 19% respectively and no activity difference could be detected between these two peptides (Fig 1B). Finally the peptides PGVGVA VGVGVA (for which the induced aortic ring tensions presented a slightly higher variance) and GVGVA as well as the control peptide VPVGGA produced tension variations of 12% maximum compared with the control with all values not statistically significant .
Figure 1. Concentration-response curves showing the effects of elastin-related compounds after a 15-minute exposure on rat aortic ring tone. The rings were precontracted and were then perfused for 105 minutes with 10-6 mol/L NA (control). In some rings increasing doses of elastin peptides were also added to the perfusion. The peptides added were as follows: rTE VGV VGVAPG and PGV (A); PVGV and VGVA (B); and VPVGGA GVGVA VGVGVA and PGVGVA (C). Each point represents the mean of five to nine experiments. The results are expressed as tension (mg) (mean±SEM). *P<.05 vs control tension by least significant difference test. #P<.05 vs corresponding VGV-induced tension by least significant difference test. P<.065 (least significant difference test) and P<.05 (U test) vs corresponding control tension.
By contrast on aortic rings without endothelium the 10 compounds (rTE VGV PGV VGVA PVGV GVGVA VGVAPG VGVGVA PGVGVA and VPVGGA) had no effect on the tone of NA-precontracted aortic rings for the two tested concentration ranges: 10-11 mol/L and 10-8 to 10-7 mol/L the active concentrations determined above (Table 2). All the tested compound-induced tension values were not different from the corresponding control tension values (Table 1 category B).
Table 2. Effect of Elastin-Related Compounds on Tone of Rat Aortic Rings Without Endothelium.
10 -11mol/L 10-8mol/L (10-7 mol/L for PGV)
Control 1104±8 1081±12
VPVGGA 1107 ±9 1091±11
VGVAPG 1111±10 1087±24
VGVGVA 1112±17 1101±29
PGVGVA 1112±5 1096±3
GVGVA 1099±22 1073±29
PVGV 1106±15 1090±18
VGVA 1101±10 1083±14
PGV 1106±3 1085±10
VGV 1089±8 1102±37
rTE 1118±3 1103 ±
Aortic rings were precontracted with 10-6 mol/L NA. Each value represents the mean of four experiments. The results (mean±SEM) are expressed as tension (mg). Each peptide was applied for 15 minutes at 10-11 mol/L and then for 15 minutes at 10-8 mol/L (10-7 mol/L for PGV). All the tensions measured after 30 minutes (15 minutes after addition of 10-8 to 10-7 mol/L of the compounds except for the control) were significantly different (two-way ANOVA P<.05) from the corresponding tensions measured after 15 minutes (15 minutes after addition of 10-11 mol/L of the compounds except for the control). For each compound and for each concentration tested (at each time) no difference with the corresponding control ring tension could be detected.
The effect of the elastin/laminin receptor antagonist 'lactose' was then tested on the endothelium-dependent vaso-relaxant action of the active elastin-related sequences (VGV PGV VGVA PVGV VGVAPG and rTE) for concentrations demonstrated to induce a submaximal vasorelaxation. VGV VGVAPG and rTE were used at 10-8 mol/L. Since only one concentration was found to be active in the range of the maximal effect for PGV (10-7 mol/L) VGVA and PVGV (10-11 mol/L) this concentration was used for these compounds. The direct effect of lactose (10-4 mol/L) was first studied on the NA-precontracted rat aortic rings with endothelium. Lactose (10-4 mol/L)–induced tension (mg) was identical to the control tension after 15 minutes but was lower than the control tension after 30 minutes and lower than the control tension and the initial lactose-induced tension after 45 minutes. At these times the slight vasorelaxation induced by lactose was in the range of 5% compared with the control value (Fig 2and Table 1 category C). After assessment of a second normalization of the tension values to balance the effect of lactose itself the effect of lactose on the vasorelaxant effect induced by the active compounds (rTE PGV VGV PVGV VGVA and VGVAPG) was investigated. As shown in Tables 1(category D) and 3 the presence of lactose had a significant inhibitory effect on the elastin peptide–induced vasorelaxation and the inhibition was independent of the active peptide used. These results indicate that lactose strongly inhibited the vaso-relaxant action of all the active elastin-related compounds.
Figure 2. Action of lactose (Lac 10-4 mol/L) on aortic ring tone. The rings were precontracted with NA (10-6 mol/L) and then the tension was recorded in the presence or absence of Lac for 45 minutes. Each point represents the mean of 9 (absence of Lac) to 17 (presence of Lac) experiments. The results are expressed as tension (mg) (mean±SEM). *P<.05 vs corresponding control tension without Lac by least significant difference test. P<.05 vs Lac-induced tension after 15 minutes by least significant difference test.
Action of Elastin-Related Compounds on HUVEC [Ca2 ]i
To investigate the intracellular mechanisms explaining the vasorelaxant action of rTE and certain elastin-related sequences we studied the effect of these compounds on [Ca2 ]i in adherent HUVECs loaded with the fluorescent dye fura 2. We first verified that the -elastin–induced [Ca2 ]i increase previously demonstrated on pooled suspended HUVECs loaded with indo 1 or on individual adherent HUVECs loaded with fluo 3 was also occurring in the fura 2–loaded adherent HUVECs stimulated with 10-3 mg/mL (10-8 mol/L) elastin (Fig 3A). In these conditions the-elastin–induced [Ca2 ]i increase exhibited slow kinetics and an amplitude (average raw increase 2.06-fold or 1.73 times the control value) similar to those previously observed in pooled suspended HUVECs loaded with indo 1 (average raw increase 2-fold). The control peptide VPVGGA (10-7 mol/L) had no effect on HUVEC [Ca2 ]i compared with cells with no peptide added (Fig 3Band 3Cand Table 4). The effects of the elastin-related compounds on HUVEC [Ca2 ]i were then investigated at concentrations in the range of 10-8 to 10-7 mol/L for the peptides found to have no activity on the vascular tone and at active concentrations for the peptides found to induce a vasorelaxation. Only rTE (10-9 to 10-7 mol/L) VGV (10-8 to 10-7 mol/L) PGV (10-8 to 10-7 mol/L) VGVA (10-11 mol/L) PVGV (10-11 mol/L) and VGVAPG (10-8 to 10-7 mol/L) ie all the peptides found active on the vascular tone induced an increase in HUVEC [Ca2 ]i compared with the control cells (Fig 4Aand 4Cand Tables 1[category E] and 4). Nevertheless no statistically significant difference (U test) between the [Ca2 ]i increases induced by the different active compounds could be detected. The compounds found to have no activity on the vascular tone (VPVGGA PGVGVA VGVGVA and GVGVA) also exhibited no significant activity on HUVEC [Ca2 ]i (Fig 3Band Tables 1[category E] and 4).
Figure 3. Action of 10-3 mg/mL -elastin (A) and the control peptide VPVGGA at 10-7 mol/L (B) on [Ca2 ]i in cultured HUVECs compared with the control experiment (C). Each trace is representative of six seven and three experiments respectively.
Table 4. Effect of rTE and Elastin-Related Sequences on [Ca2 ]i in HUVECs in the Absence or Presence of 10-4 mol/L lactose.
% of [Ca2 ]i Variation in Absence of % of [Ca2 ]i Variation in Presence of
Lactose Compared With Control Lactose Compared With Control
ND indicates not done. Concentrations were as follows: 10-7 to 10-8 mol/L (VGV PGV GVGVA VGVGVA PGVGVA VPVGGA and VGVAPG) 10-7 to 10-9 mol/L (rTE) and 10-11 mol/L (VGVA and PVGV). Each value represents the mean of three to seven experiments. The results (mean±SEM) first expressed as concentration (nmol/L) were converted to the percentage of [Ca2 ]i variation compared with control values after assessment of the statistical tests.
1 1.P<.05 vs control cells (U test);
2 2.P<.05 vs cells incubated with the same elastin-related compound in the absence of lactose (two-way ANOVA).