Characteristics of the aortic elastic network and related phenotypes in seven inbred rat strains

Jacques Behmoaras, Mary Osborne-Pellegrin, Dominique Gauguier, Marie-Paule Jacob


Extracellular matrix (ECM) molecules such as elastin and collagen provide mechanical support to the vessel wall and are essential for vascular function. Evidence that genetic factors influence aortic ECM composition and organization was concluded from our previous studies showing that the inbred Brown Norway (BN) rat differs significantly from the outbred Long-Evans (LE) and the inbred LOU rat with respect to both thoracic aortic elastin content and internal elastic lamina (IEL) rupture in the abdominal aorta and iliac arteries. Here, we measured aortic elastin and collagen contents as well as factors that may modulate these parameters [insulin growth factor (IGF)-I, transforming growth factor (TGF)-β1, and matrix metalloproteinase (MMP)-2] in seven inbred rat strains, including BN and LOU. We also investigated whether IEL ruptures occur in strains other than BN. We showed that LOU, LE, BN, and Fischer 344 (F344) rats were significantly different for aortic elastin content and elastin-to-collagen ratio, whereas LE, Lewis, WAG, and Wistar-Furth (WF) were similar for these parameters. BN and F344 had the lowest values. BN was the only strain to present numerous IEL ruptures, whereas F344, LE, and WF presented a few and the other strains presented none. In addition, IGF-I and TGF-β1 levels in the plasma and aorta differed significantly between strains, suggesting genetic control of their production. Because inbred rat strains provide interesting models for quantitative trait locus analysis, our results concerning elastin, collagen, IEL ruptures, and cytokines may provide a basis for the search for candidate genes involved in the control of these phenotypes.

  • elastin
  • collagen
  • IEL rupture
  • IGF-I
  • TGF-β1
  • aorta
  • rat strain

extracellular matrix (ECM) components play a critical role in the integrity and the function of the vascular system. Elastic fibers in arteries are designed to maintain elastic function, permitting long-range deformability and passive recoil without energy output (11). This elastic function is complementary to that of collagen fibers, which provide tensile strength. In view of the low level of turnover of elastin and collagen in the adult and the high levels of cyclic mechanical stress to which the arterial wall is subjected, qualitative and/or quantitative alterations in elastic or collagen fibers may have serious functional repercussions. Mutations in key ECM genes such as elastin, fibrillin-1, and type III collagen genes result in human diseases characterized by severe vascular abnormalities, such as supravalvular aortic stenosis (SVAS), Marfan's syndrome, and Ehlers Danlos syndrome type IV, respectively. However, more minor quantitative changes in elastin and collagen networks of the arterial wall may be of importance in the predisposition to vascular disease later in life, but little is known of the genes that operate the fine control of these vital arterial ECM components.

The formation of elastic fibers is a highly complex, not completely understood process that, in addition to the initial synthesis of elastin molecules, involves their posttranslational modification, i.e., cross-linking, and their interactions with fibrillins and various other microfibril-associated molecules that are assembled into fibers in an as-yet-unknown temporal hierarchy (11, 17). Collagen fiber formation is better known and less complex than for elastic fibers but nevertheless involves posttranslational modifications (21). In addition, the quantity of elastic and collagen fibers in the adult results from an equilibrium between the processes of synthesis/assembly and that of degradation. In normal arteries, the degradation process of collagen and elastic fibers is very slow. Some proteases, such as matrix metalloproteinase (MMP)-2, are constitutively expressed, but their activity is very low due to the control of their activation and/or the action of tissue inhibitors (7).

Thus the control of the quantity of these vital arterial ECM components, i.e., elastin and collagen fibers, is likely to be complex and polygenic, for which the candidate gene approach is poorly adapted. Indeed, we have previously looked at elastin gene polymophism in the rat and shown that it accounted for <4% of the genetic variability in aortic elastin content in the rat aorta (20). A quantitative trait locus (QTL) approach is more suitable to the study of complex polygenic traits, such as aortic elastin and collagen contents, than the candidate gene approach. To perform QTL studies in rats, it is general practice to cross two inbred strains with highly contrasting phenotypes to obtain a wide variation of the quantitative trait of interest in the F2 or backcross generations. We have previously shown that the Brown Norway (BN) rat is characterized by a low aortic elastin content and a low elastin-to-collagen ratio compared with outbred Long-Evans (LE) (19) and inbred LOU strains (20). This lower aortic elastin content is partly explained by a lower elastin synthesis in the growing BN rat (19, 20). The BN rat also develops large numbers of internal elastic lamina (IEL) ruptures in its abdominal aorta, which are not observed in LE (3) or LOU. Crosses between BN and LOU are currently being used in our laboratory for identification of QTLs for both aortic elastin and collagen contents and for IEL rupture.

The objective of the present study was to establish, by biochemical and histological analyses, a profile of the aortic elastic network and related characteristics in a group of inbred, normotensive rat strains not hitherto studied for these phenotypes, which will be useful for further QTL identification by whole genome analysis. The strains were chosen in view of their common laboratory use in the cardiovascular field and their commercial availability and were more or less equally distributed throughout the rat phylogenetic tree (2). We thus investigated whether inbred strains other than BN and LOU present significant differences in their aortic elastin and collagen contents and IEL rupture and whether these phenotypes are linked. Because elastin and collagen synthesis is under the control of various growth factors such as insulin growth factor (IGF)-I (5, 18, 23) and transforming growth factor (TGF)-β1 (9, 13, 15), we also measured aortic and plasma levels of these cytokines. Finally, we chose to measure MMP-2 secreted by the aorta to test for interstrain differences in aortic ECM degradative activity.



Male inbred rats of the following strains, LE (LE/Cpb), Fischer 344 [F344/N (F344)], Wistar-Furth [WF/N (WF)], WAG (WAG/Rij), BN (BN/Rij), and Lewis [LEW/Han (LEW)], were supplied by Harlan (Gannat, France) and received in our laboratory at 10 wk of age. The male inbred LOU rats were provided by our own breeding stock.

Animal care complied with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Pub. No. 86-23, Revised 1985). The studies were carried out under authorization no. 006235 of the Ministère de l'Agriculture, France.

Blood pressure measurements.

Systolic arterial blood pressure and heart rate were measured at 2-wk intervals, starting at 10 wk of age, on conscious animals under standardized conditions routinely employed in our laboratory using a tail cuff and pulse transducer (blood pressure recorder 8006, Apelex) after 20 min at 32°C under a specially designed heating device.

Experimental protocol.

All determinations were performed on aortas of mature, adult, male rats (aged 18 wk). At this age, rats were weighed and anesthetized with pentobarbital sodium (60 mg/kg ip), and body length [nose-to-rump (N-R) length] was recorded. One milliliter of blood was then drawn from the jugular vein onto sodium citrate. Plasma was stored at −20°C until use for assay of TGF-β1 and IGF-I. The aortic arch and descending thoracic aorta were rapidly dissected out and cleaned of blood and periadventitial tissue in ice-cold saline. The arch was used immediately for ex vivo incubation in DMEM, whereas the descending aorta (down to the diaphragm) was frozen in liquid nitrogen and stored at −80°C until use for the determination of aortic elastin, collagen, and cell protein contents. The heart was removed, and its weight was recorded. A catheter was then inserted into the aorta at the level of the diaphragm for perfusion in situ of the arteries distal to the catheter. After the blood was washed out with heparinized saline, buffered formalin was perfused at 6 ml/min with a Braun perfusion pump. The vena cava was cut for outflow of perfusate. After 15 min of perfusion, the abdominal aorta and common iliac arteries were dissected out and stored in buffered formalin for quantification of IEL rupture and histological examination.

Two independent experiments were performed using the above protocol. The first experiment comprised eight rats of each of the seven strains, and a second confirmatory experiment was performed on four rats of each strain. Because very similar results were obtained in both experiments, we present here only the results of the first group (n = 8), which are entirely representative. In addition, two extra rats of each strain were used for histological studies of the thoracic aorta. For histological examination of the abdominal aorta, a small ring was taken from the perfusion-fixed abdominal aortas assigned for “en face” observation. This sample was localized midway between the origin of the renal arteries and the iliac bifurcation.

Purification and quantification of aortic elastin.

Thoracic aortas were thawed and manipulated in a petri dish on ice under a dissecting microscope. Each segment was opened longitudinally, and its length was recorded using a grid in the eyepiece. Elastin was then quantified using a previously described method (20) based on that described by Wolinsky (24) as follows. After segments were delipidated in acetone-diethyl ether [1:1 (vol/vol)] and dried, their dry weight was recorded using a Sartorius R 160P balance (precision of 0.01 mg). Cell proteins were extracted by gentle agitation in 0.3% SDS for 12 h. The extracellular proteins other than elastin (including collagen) remaining in the aortic segments were solubilized by three 15-min extractions in 1 ml of 0.1 mol/l NaOH in a boiling water bath. Elastin was quantified by determining the dry weight of the residue.

Quantification of other aortic constituents: cell proteins and collagen.

Aortic total cell protein content was assayed in the SDS extract by the method of Lowry (14), and collagen content was determined by assaying the hydroxyproline present in the NaOH supernatant. NaOH supernatants were evaporated to dryness and hydrolyzed in 6 N HCl in vacuum-sealed vials for 24 h at 110°C. Hydroxyproline was determined in the hydrolysate using a colorimetric assay (22). Collagen was quantified from hydroxyproline values on the basis of the assumption that collagen contains 12.77% hydroxyproline by weight (10).

Ex vivo incubation of the aortic arch and measurement of gelatinase activity.

For the measurement of diffusible MMPs such as MMP-2, aortic arches were cut into small pieces and incubated for 24 h in 1 ml of DMEM (BioWhittaker) at 37°C in a 5% CO2-95% air atmosphere. At the end of the incubation, the conditioned medium was centrifuged, and the wet weight of tissue was recorded. The measurement of MMP-2 was performed using gelatin zymography as previously described (4). Conditioned media were electrophoresed in 0.2% SDS-10% polyacrylamide gels containing 1 mg/ml gelatin under nonreducing conditions. After electrophoresis, the proteins were renatured by exchanging SDS with 2.5% Triton X-100 (2 × 30 min, room temperature). The gels were then incubated at 37°C for 19 h in 50 mM Tris·HCl buffer (pH 7.8) and 10 mM CaCl2. The gels were stained with 0.5% Coomassie brilliant blue R-250 (Bio-Rad Laboratories) in 10% acetic acid-30% ethanol solution and then destained in 30% ethanol-10% acetic acid solution. Proteins with gelatinolytic activity (i.e., pro-forms activated by SDS during the electrophoresis and active forms) were visualized as areas of lytic activity on a blue background, and their apparent molecular weights were compared with prestained low-molecular-weight range markers (MWM). Gels were scanned by an imaging densitometer, and gelatinolytic activities were quantified using the NIH Image 1.60 program. The ratio of active MMP-2 was determined as (active MMP-2)/(pro-MMP-2 + active MMP-2).

TGF-β1 and IGF-I assays.

Proteins were extracted from aortic arches in PBS, 1% Triton X-100, and 2 mM EDTA (pH 7.2) as previously described (1). Biologically active TGF-β1 was assayed in plasma and aortic extracts by the TGF-β1 Emax Immunoassay system (Promega) before and after acidification of samples with 1.5 N HCl to measure the active form and total concentration, respectively. IGF-I was assayed in plasma and aortic extracts by a radioimmunoassay (Diagnostic Systems Laboratories).

Quantification of IEL ruptures.

En face preparations were made by cleaning formalin-fixed arteries of periadventitial tissue, opening them longitudinally, and pinning them out, luminal side up, using a dissecting microscope. The luminal surface was then stained with orcein and hematoxylin to stain the IEL and endothelial nuclei, respectively, to reveal ruptures in the IEL. After being stained, arteries were dehydrated, unpinned, cleared in xylene, and mounted on slides, luminal side up, for microscopic examination. For each artery, the total number of ruptures in the IEL, irrespective of size, was recorded.

Morphological studies of the thoracic and abdominal aorta.

All aortic samples were embedded in paraffin, and transverse sections of 6 μm were stained either with orcein or Sirius red to reveal the elastic and collagen fiber networks, respectively. Semiautomatic morphometric analysis was performed using Quancoul software on orcein-stained sections to determine media cross-sectional area (MCSA), medial thickness, and luminal diameter. The number of elastic lamellae (excluding the internal and the external elastic laminae) was counted manually using the objective at ×40 at 10 points around the medial circumference on each section, and a mean value was calculated for each section. This value was then corrected for medial thickness.

For the thoracic aorta, sections of the four different levels were analyzed for each rat, but because there were only 2 rats/group, statistical analysis could not be performed and the results may be considered only as indicative. In contrast, for the abdominal aorta, several sections of each abdominal aorta were analyzed, but in three rats of each strain, permitting statistical analysis.

Statistical analysis.

Data are expressed as means ± SD. One-factor ANOVA followed by the Scheffé's test was used for statistical analysis of results unless otherwise stated. Simple regression analysis was used to analyze correlations between the different parameters.


Hemodynamic values and weight of organs.

Table 1 provides general data concerning the seven strains at 18 wk of age: body weight, N-R length, systolic blood pressure, heart rate, heart weight index (g/100 g body wt), and dry weight of the thoracic aorta (expressed as mg/cm corrected for N-R length). The LOU strain had the lowest body weight (254 ± 26 g), whereas the LE strain had the highest (434 ± 21 g). Heart rate and systolic blood pressure remained stable between 10 and 18 wk of age, and so only mean values at 18 wk are presented (Table 1). Systolic blood pressure was highest in F344 rats (171 ± 14 mmHg) and lowest in LEW rats (145 ± 7 mmHg), whereas BN rats had the highest heart rate (413 ± 23 beats/min). Mean values of the heart weight index were within the normal range (0.236–0.292 g/100 g body wt), indicating no cardiac hypertrophy in any strain.

View this table:
Table 1.

General data concerning body weight, N-R length, systolic blood pressure, heart rate, heart weight index, and aortic dry weight corrected for body size in LOU, BN, F344, LE, LEW, WAG, and WF rats aged 18 wk

There were considerable differences in the aortic dry weight per unit length between strains, indicating differences in the aortic caliber. Because the differences in aortic caliber were largely a reflection of different body sizes, the dry weight per centimeter of the aorta was corrected for N-R length (Table 1). BN and F344 rats showed the lowest aortic dry weight compared with other strains both before (mg/cm aorta, results not shown) and after correction for body size (mg·cm−1·cm N-R length−1).

Aortic elastin, collagen, and cell protein content.

Figure 1, A and B, shows elastin content in the thoracic aortas of the seven strains aged 18 wk. When elastin was expressed as a percentage of the dry weight of the thoracic aorta (Fig. 1A), BN rats had the lowest values, significantly lower (P < 0.001) than those of the inbred LOU strain (by 8.4% dry wt) and of the inbred LE strain (by 5.4% dry wt), confirming our previous reports (19, 20). In addition, these results indicate that LE, LEW, WAG, and WF rats had similar elastin contents, which were all significantly different from that of the BN rat. Furthermore, the F344 rat showed the second lowest elastin content (35.0 ± 1.4%) after the BN rat, and there was also a significant interstrain difference for F344 versus LOU (P < 0.001) and F344 versus LE, LEW, WAG (P < 0.05), and WF (P < 0.01). When elastin was expressed as milligram per centimeter aorta and corrected for body size i.e., per centimeter N-R length (Fig. 1B), similar results were obtained, but the value for F344 was even lower, more closely resembling that of BN. These data clearly show that the BN and F344 strains present an elastin deficit compared with the other strains of rats studied here.

Fig. 1.

Quantity of elastin (A and B) and collagen (C and D) in the thoracic aorta of 18-wk-old rats of seven inbred strains expressed as a percentage of dry weight (A and C) or in milligrams per centimeter of aorta per centimeter of nose-to-rump length (NR; B and D) and the elastin-to-collagen ratio (E). LE, Long-Evans rats; LEW, Lewis rats; WF, Wister-Furth rats. ***P < 0.001, **P < 0.01, and *P < 0.05 vs. LOU rats; ###P < 0.001 and ##P < 0.01 vs. Brown Norway (BN) rats; §§§P < 0.001, §§P < 0.01, and §P < 0.05 vs. Fischer 344 (F344) rats.

In contrast to elastin, the percentage of dry weight represented by collagen was greater in BN than LOU (Fig. 1C). The collagen content of the BN rat was the highest (28.3 ± 2.0%), followed by the F344 rat (27.8 ± 1.9%). Thus, with the use of this mode of expression, the group of rats with the lowest elastin contents (BN and F344) have the highest collagen contents. However, when collagen was expressed as milligrams per centimeter of the aorta per centrimeter of N-R length (Fig. 1D), significant differences were observed between LOU versus LE and WAG rats (P < 0.01) but not amoung LOU, BN, F344, LEW, and WF rats.

The ratio of elastin to collagen in the thoracic segment of the aorta for each strain is presented in Fig. 1E. The elastin-to-collagen ratio was highest in LOU and WF rats (1.7 ± 0.10 and 1.6 ± 0.10, respectively) and lowest in F344 and BN rats (1.3 ± 0.10 and 1.1 ± 0.10, respectively).

Cell proteins were not significantly different between strains when expressed as a percentage of aortic dry weight (LOU, 14.8 ± 2.8; BN, 15.4 ± 2.7; F344, 16.1 ± 1.9; LE, 16.1 ± 1.5; LEW, 15.0 ± 2.1; WAG, 15.5 ± 1.3; WF, 16.1 ± 1.1).

Gelatinase activity and TGF-β1 and IGF-I assays.

Gelatinase activity of aortas was measured in conditioned medium by gelatin zymography (Fig. 2A). The only gelatinase detected was MMP-2, the major part of which was present in the pro-form (≥85%). F344 had the highest total MMP-2 activity expressed as densitometric units per milligram of the aorta (data not shown) and the highest ratio of active to total MMP-2 (Fig. 2B). F344 rats had also the highest plasma IGF-I concentration, whereas LOU rats had the lowest (Fig. 3A). The plasma IGF-I concentration correlated positively with aortic collagen, expressed as both a percentage of dry weight (r = 0.35, P = 0.011; Fig. 3E) and milligrams per centimeter of the aorta (r = 0.4, P < 0.01; data not shown), and correlated negatively with the elastin-to-collagen ratio (r = 0.3, P = 0.026; data not shown). The total TGF-β1 concentration in plasma was lowest in WAG rats and highest in BN rats (Fig. 3B). The variations between rat strains in aortic IGF-I and total TGF-β1 levels were different from those observed in plasma (Fig. 3, C and D). LEW, WAG, and WF rats had the highest aortic IGF-I levels, and LOU and BN rats had the lowest. The F344 rat had the most elevated total TGF-β1 level (123.4 ± 39.3 pg/mg aorta), followed by the BN rat (99.9 ± 33.4 pg/mg aorta). In addition, we found a positive correlation between aortic collagen content (percentage of dry wt) and the aortic TGF-β1 levels (Fig. 3F). Active TGF-β1 was detectable in aortic extracts but at a very low level (∼15% of total TGF-β1) in all strains.

Fig. 2.

Matrix metalloproteinase (MMP)-2 activity secreted in DMEM by the aorta was measured by gelatin zymography (A), and the MMP-2-to-total MMP-2 ratio (B) was calculated for the seven strains. A one-factor ANOVA followed by the Bonferroni/Dunn test was used here for statistical analysis. Ref, rat lung conditioned medium. ***P < 0.001 and **P < 0.01 vs. LOU rats; ##P < 0.01 vs. BN rats; §§P < 0.01 vs. F344 rats; ++P < 0.01 vs. WAG rats.

Fig. 3.

Insulin growth factor (IGF)-I (A and C) and total transforming growth factor (TGF)-β1 (B and D) levels in the plasma (A and B) and aorta (C and D) in the seven strains and simple regression analyses between collagen (%dry weight) and plasma IGF-I concentration (E) or aortic total TGF-β1 concentration (F). ***P < 0.001, **P < 0.01, and *P < 0.05 vs. LOU rats; ###P < 0.001, ##P < 0.01, and #P < 0.05 vs. BN rats; §§§P < 0.001, §§P < 0.01, and §P < 0.05 vs. F344 rats; $$P < 0.01 and $P < 0.05 vs. LE rats; ++P < 0.01 vs. WAG rats.

Quantification of IEL ruptures.

Results are presented as the total number of IEL ruptures observed in the abdominal aorta and both iliac arteries taken together. Results showed that the BN rat is the only strain to present significant numbers of IEL ruptures in the abdominal aorta and iliac arteries (58 ± 10; Fig. 4A). F344, LE, and WF rats presented a few small ruptures (2.5 ± 2.7, 5.1 ± 1.9, and 1.1 ± 0.5, respectively), but no ruptures were observed in the other strains (Fig. 4A). The IEL ruptures in the F344 rat were predominantly in the iliac arteries, whereas in the LE and WF rats they were in the abdominal aorta. If we exclude BN rats from the regression analysis, because it presents a very different phenotype compared with the other six strains for IEL ruptures, the IEL phenotype was negatively correlated to the thoracic aorta elastin content (r = 0.34, P = 0.014; Fig. 4B) and elastin-to-collagen ratio (r = 0.40, P = 0.005; data not shown) in all the other rats studied.

Fig. 4.

A: numbers of internal elastic lamina (IEL) ruptures in abdominal and iliac arteries (RIEL AA + IAs) in the seven strains. B: simple regression analysis between RIEL AA + IAs and elastin (%dry weight) in six inbred rat strains (excluding BN rats). ###P < 0.001 vs. BN rats.

Morphological studies of the thoracic and abdominal aorta.

All parameters measured on thoracic and abdominal aortic cross sections, which are linked to aortic caliber (MCSA), differed considerably between strains but were generally proportional to the size of the rat (Table 2). This is illustrated by the fact that, when corrected for body size (N-R length), these differences were no longer significant. No gross differences were visible between strains for either elastin or collagen networks at low magnification. At higher magnification, in the thoracic aorta stained with orcein, only the BN and the F344 rats exhibited distinctive characteristics of the medial elastic network, with the other five strains appearing rather similar. Although variable, as shown in Fig. 5, the elastic lamellae in the BN rat often appeared thinner than in the other strains, with less visible interlamellar elastic fibers, whereas the F344 rat presented lamellae of normal thickness but that, not infrequently, appeared disorganized and ruptured (Fig. 5A). The mean number of elastic lamellae in the thoracic aortic media, after correction for medial thickness, did not greatly differ between strains (Table 2).

Fig. 5.

A and B: elastic fiber networks in orcein-stained transverse sections of thoracic (A) and abdominal aortas (B) in LOU, BN, and F344 rats. C: “en face” preparations of abdominal aortas (LOU and BN rats) and iliac artery (F344 rat). Ruptures in the IEL appear as dark gray transverse bands due to the absence of the IEL (see BN and F344 rats).

View this table:
Table 2.

Histomorphometric data of transverse sections of TA and AA in the seven strains

In contrast to the thoracic aorta, in the abdominal aorta, apart from the obvious presence of ruptures in the IEL in some sections from the BN rat, it was not as easy to discern differences in the elastic lamellar network between strains, because the abdominal elastic network is much less well organized than in the thoracic segment (Fig. 5B). However, thinner elastic lamellae and sparser interlamellar elastic fibers were observed in some sections of the BN aorta. Furthermore, the mean number of medial elastic lamellae, expressed per 100 μm of medial thickness, was found to be significantly lower in the BN rat than in all the other strains, including the F344 rat (Table 2). Examples of en face preparations of the abdominal aorta and iliac artery, with different degrees of IEL ruptures, are illustrated in Fig. 5C.


Inbred rat strains may be useful in the development of physiological and biochemical models to study the genetic basis of ECM composition and IEL rupture in the rat aorta.

Taken together, our data concerning aortic ECM analysis demonstrates a deficit in elastin content not only in the BN but also in the F344 rat (whether expressed as a percentage of dry weight, elastin mg·cm aorta−1·cm N-R length−1, or elastin-to-collagen ratio), which has not been previously described. When aortic collagen is expressed as milligrams per centimeter of the aorta per centimeter N-R length, no significant differences were observed among BN, F344, and LOU rats, suggesting that the elastin deficit observed in BN and F344 rats is not simply the result of an increase in other aortic components, especially collagen, but does indeed reflect an absolute deficit in this component.

Elastin and collagen synthesis are known to be modulated by hemodynamic factors. For this reason, we measured arterial pressure in the seven rat strains to check for differences in pressure that could explain the observed differences in elastin and collagen contents. We confirmed that all strains are normotensive. Because F344 rats had the highest systolic blood pressure and BN rats has the highest heart rate, the elastin deficit in these two strains is not readily explained by hemodynamic factors. In addition, we have previously shown that the BN rat presents a deficit in elastin synthesis that is independant of hemodynamic factors because it is maintained in BN arterial smooth muscle cells in culture (20).

Assays of plasma and aortic IGF-I and TGF-β1 allowed us to demonstrate significant differences between the strains, suggesting a genetic component in the determination of the levels of these cytokines. Plasma but not aortic IGF-I was positively correlated to aortic collagen. This agrees with the results of a previous study (16) demonstrating the effects of elevated circulating IGF-I on the collagen content in skin, trachea, and lung tissue. Although IGF-I has been reported to increase elastin gene transcription in aortic smooth muscle cells (5, 23), we did not observe any correlation between elastin content and plasma or tissue IGF-I levels in the strains studied. Aortic levels of TGF-β1 also correlated positively with the percent collagen, because rats with the highest percent collagen (BN and F344) had elevated aortic TGF-β1 levels, but not with elastin. Thus, although IGF-I and TGF-β1 are known to stimulate both elastin and collagen synthesis (9, 13, 15, 18), the results obtained here suggest that they are more potent stimulators of collagen than elastin expression in the rat aorta.

Because the absolute and relative quantities of collagens and elastin in aorta are also dependant on degradation rates, we measured secreted MMPs from the thoracic aorta. It is tempting to relate the observed high level of active to total MMP-2 in the F344 rat with its low elastin content and its disorganized and ruptured medial elastic network, but this relation does not hold across strains because the WF rat also had high MMP-2 levels associated with a high elastin content and a normal elastic fiber network and the BN rat, with low elastin and numerous IEL ruptures, had low MMP-2 levels. Indeed, active-to-total MMP-2 levels did not correlate with any phenotype when all rats are considered.

In previous studies, we have shown that the inbred BN rat differs significantly from the outbred LE rat and the inbred LOU rat with respect to both aortic elastin content and IEL rupture in the abdominal aorta (3, 19, 20). Our present results show clearly that the BN rat is very different with respect to IEL rupture from all the other strains considered here. Although F344, LE, and WF present a few very small IEL ruptures in the aorta or iliac arteries, they are all very different from BN in regard to both the number and size of these ruptures. This striking difference between BN and other strains concerning the aortic IEL rupture phenotype may be related to the fact that BN is the strain that presents the greatest genetic divergence from other strains and is closest to the wild rat (2).

Because BN also has by far the lowest aortic elastin content and elastin-to-collagen ratio of the strains studied here, it is reasonable to suspect that these two phenotypes may be linked to IEL rupture. Excluding the BN rat, a significant but weak negative correlation was observed in the six other strains between aortic elastin content or elastin-to-collagen ratio and IEL rupture, suggesting that the aortic elastin deficit is, at the most, a minor contributory factor that may be permissive, but it certainly does not explain the very high levels of IEL rupture in the BN rat. Recent studies have identified QTLs for IEL rupture on chromosomes 5 and 10 (6, 8), findings that may lead to identification of candidate genes involved in this phenotype.

The results of our histological studies in the thoracic aorta are coherent with our biochemical data, because only the BN and F344 rats showed obvious differences, albeit minor, in the structure of the medial elastic network. The aspect of thinner elastic lamellae and reduction of interlamellar elastic fibers in BN could be the morphological counterpart of the decreased elastin content. In F344, although lamellae appeared of normal thickness, their disorganization and occasional rupture could also explain the observed biochemical deficit, which was nevertheless less important than in BN.

It is noteworthy that the aortic elastin deficit in BN and F344, which is in part due to decreased elastin synthesis in BN (19, 20) does not appear to result in a phenotype similar to that observed in the mouse model of heterozygosity for the elastin gene (ELN+/−). The important deficit (∼50%) in elastin synthesis in these mice results in an increased number of thinner elastic lamellae in the aorta and increased smooth muscle (12). BN and F344 showed no increase in elastic lamellar number and no increase in smooth muscle. Indeed, these strains showed the lowest values of thoracic and abdominal aortic MCSA and thoracic aortic dry weight (mg/cm), when expressed per centimeter of body length, suggesting a degree of arterial hypotrophy compared with the other strains. For arterial hypotrophy, we used N-R length as the correcting factor here, because it takes into account somatic size and not adiposity.

Our observation of a decreased number of elastic lamellae per unit of media thickness in the BN abdominal aorta is of great interest in view of the fact that spontaneous aortic IEL rupture only occurs in the abdominal segment (3, 19). An inappropriately low number of lamellar units could render the aorta less resistant to hemodynamic stress and lead to IEL rupture in the abdominal part. It is of relevance here to recall that the human abdominal aorta is exceptional in that it possesses far fewer lamellar units compared with aortas of similar caliber in other mammalian species (25), and this has been considered determinant for the high susceptibility of this segment to arterial disease such as atherosclerosis and aneurysms. However, in view of the large numbers of IEL ruptures present in these 18-wk-old rats, we cannot exclude that the observed decrease in the number of medial elastic lamellae reflects wall remodelling as a consequence of IEL rupture and is not a primary event.

Taken together, these data suggest that LOU and LE rats are significantly different from both BN and F344 rats for elastin content and elastin-to-collagen ratio in the aorta. The striking IEL rupture phenotype of the BN rat is not found in any of the other strains and does not appear to be linked to any other parameter measured. Significant variations in aortic and plasma levels of IGF-I and TGF-β1 regulating the synthesis of ECM components were observed between strains, suggesting that genetic factors influence these cytokines. Because inbred rat strains provide interesting models for QTL analysis, our results concerning elastin, collagen, IEL ruptures, and cytokines may provide a basis for the search for candidate genes involved in the control of these phenotypes.


This study was suppported by Institut National de la Santé et de la Recherche Médicale and the European Community (QLK6-CT-2001-00332). D. Gauguier holds a Wellcome senior fellowship in basic biomedical science.


We thank Liliane Louédec for carrying out the blood pressure and heart rate measurements and Philippe Zizzari (Institut National de la Santé et de la Recherche Médicale U549, Centre Paul Broca, Paris, France) for the plasma IGF-I assays.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract