cross-sections and heavy in 4596 cross-sections Fig. 4.
No difference in MMP-1 staining was observed be- tween the cross-sections that showed expansive and
constrictive remodeling in the different layers of the arterial wall total plaque, cap and shoulder of the
plaque and adventitia Fig. 4. In contrast to the macrophages, the extent of staining for MMP-1 was
more prevalent in the plaque than in the adventitia Table 2.Both MMP-2 and MMP-9 immunostaining
were more prevalent in plaques of the expansively remodeled segments versus constrictively remodeled
cross-sections Fig. 4. This enhanced staining in the expansively enlarged segments was evident in the cap
and shoulder region as well as at the medial border of the plaque. MMP-9 mostly co-localized with present
macrophages. MMP-2 showed more diffuse staining throughout the atherosclerotic plaque not specifically
co-localizing with macrophages. The differences in staining among groups for MMP-2 and MMP-9 were
not observed in either the media or the adventitia of the atherosclerotic cross-sections. MMP-2 and MMP-9
staining in the adventitia was much less prevalent when compared with the plaque Table 2.
The plaque area was larger in the segments with expansive remodeling, compared with segments that
showed constrictive remodeling Table 1. The plaque area may, therefore, be considered a confounder for the
interpretation of the differences in staining among groups. Zymography, in which equal amount of
proteins had been loaded, revealed more active MMP-2 in the segments with expansive compared with segments
with constrictive remodeling 340 9 319 vs. 199 9 180 adjusted pixel countsmm
2
, respectively P = 0.018 whereas the inactive MMP-2 did not differ 3030 9
1290 vs. 3080 9 1339, respectively. MMP-9 did not differ for expansively and constrictively remodeled
cross-sections 2884 9 1988 vs. 2597 9 1674, respec- tively.
4. Discussion
The extracellular matrix plays a role in maintaining structural integrity of the arterial wall. The matrix
remodels during the course of development, growth and wound healing. Comparable shifts in the synthesis
degradation balance occur in the pathogenesis of many diseases [21,22]. MMPs are natural matrix degrading
enzymes that are present in active as well as in inactive forms within the arterial wall. The role of MMPs in
plaque destabilization is well recognized: MMPs are more prevalent within high stress regions like the shoul-
ders of the plaque and may contribute to the break- down of the collagen and thinning of the fibrous cap
[18]. A role for MMPs in arterial wall remodeling is
supported by several studies. Abbruzzese et al. [23] recently showed that blocking MMPs may inhibit flow
dependent enlargement of the artery. In addition, inhi- bition of MMP-activity also reduces constrictive re-
modeling after balloon angioplasty [24]. Enhanced MMP-activity has been described extensively in in-
tracranial and abdominal aneurysms, suggesting an eti- ological role within this pathologic enlargement of the
artery [14,19,25,26].
Recently, we demonstrated in the femoral artery that the type of remodeling is related to histological markers
for plaque vulnerability: expansive remodeling is associ- ated with the presence of more inflammatory cells and
less collagen in the cap and shoulder of the plaque [11]. In the present study we investigated the prevalence of
macrophage staining in expansively and constrictively remodeled segments, in both the plaque and the adven-
titia of the atherosclerotic coronary artery. In addition, staining for the different matrix degrading MMPs were
evaluated.
The present study revealed that macrophages tend to be more prevalent in the plaques of enlarged segments
compared with shrunken segments. Differences among groups did not reach significance, however. The preva-
lence of macrophages in the adventitial layer was high but did not differ among groups. Immunohistochemical
stainings for MMP-2 and MMP-9, but not MMP-1, were more prevalent in the plaques with expansive
remodeling compared with segments with constrictive remodeling. In addition, active MMP-2 was increased
in expansively remodeled segments compared with con- strictively remodeled segments.
4
.
1
. Remodeling and plaque 6ulnerability in the coronary artery
The results of the present study, in which we investi- gated coronary artery segments, partly support our
previous observations in the femoral artery [11]. Expan- sive remodeling is associated with an enhanced inflam-
matory response in both the cap and shoulder of the plaque. Previous studies have shown that local inflam-
mation of the cap and shoulder of the plaque may predispose to plaque rupture [27 – 29]. It may be postu-
lated that plaques observed in expansively remodeled segments should be considered at risk for rupture. This
postulation is supported by recent intravascular ultra- sound studies that demonstrated that the type of re-
modeling is associated with the patients’ clinical syndrome. In these ultrasound studies, unstable angina
was more frequently associated with compensatory en- larged lesions, whereas shrunken segments were associ-
ated with stable syndromes [12,13].
4
.
2
. Remodeling and MMPs The present study revealed that MMP-stainings
were more prevalent at the luminal border as well as the medial border of the plaque. Whether inflamma-
tion within the atherosclerotic plaque with subsequent release of matrix degrading enzymes is a causal factor
for expansive remodeling, remains speculative. Aneu- rysms are considered a pathologic subtype of expan-
sive
remodeling. These
pathologic forms
of enlargement have consistently been associated with
enhanced inflammation and protease activity within the arterial wall, which makes a causal relation more
plausible [14,19,25,26,30]. Mainly MMP-2 and MMP- 9 have been studied as proteases that play an impor-
tant role in matrix degradation with subsequent enlargement of the artery in aneurysm formation. The
present study supports the hypothesis that next to aneurysm formation, MMPs are involved in other
forms of atherosclerotic enlargement as well.
Adventitial macrophage staining was extensive in most cross-sections, irrespective of the vessel size. In
contrast with the macrophages in the plaque, the macrophages in the adventitia did not contain intra-
cellular fat and were moderately co-localized with MMP-1, MMP-2 or MMP-9. Thus, if MMP staining
represents active matrix degradation with subsequent arterial remodeling, then our observations suggest that
this remodeling process is probably not initiated from the adventitia but more likely from the atherosclerotic
plaque. Knox et al. [26] previously showed that in aortic aneurysms, proteolytic activity is particularly
prevalent on the luminal and neointimal part of the specimens, which supports our observations.
We did observe a discrepancy in staining for MMP- 1 versus MMP-2 and MMP-9 among constrictive and
expansive arterial segments. MMP-2 and MMP-9 staining differed among expansively remodeled and
constrictively remodeled
cross-sections, whereas
MMP-1 staining did not. It may well be that this difference in MMP-1 staining among enlarged and
shrunken segments is reflected in MMP-1 activity. The plaque area was larger when the segments
showed expansive enlargement. Thus, the increased immunostaining for MMP-2 and MMP-9 may be due
to the increased total plaque mass, rather than the expansive mode of remodeling. However, zymography
was performed with equal amounts of protein and revealed higher levels of active MMP-2 in expansively,
compared with constrictively, remodeled segments.
4
.
3
. Limitations In immunohistochemical staining, the presence of
MMPs do not establish their catalytic capacity, as the zymogen may lack activity. In addition, TIMPs may
block activated MMPs. Thus, the stain for the differ- ent MMPs may not reflect their capability to digest
the different matrix components necessary to induce arterial remodeling. Both remodeling and plaque infl-
ammation are local phenomena Fig. 1. Since adja- cent segments were already used for morphometric
measurements and immunostaining for localization purposes it was not possible to perform zymographic
studies of all investigated segments. Due to the dis- tance between the sites used for morphometry and
zymography, the measured MMP-activity may not be representative for the cross-section in which the vessel
size had been determined.
This is a cross-sectional study. Thus, it is difficult to establish causal relations between remodeling,
plaque vulnerability and inflammation within the arte- rial wall. Previously, others have shown that MMPs
do play a predominant role in plaque vulnerability [16,17]. Recent interventional studies showed that acti-
vated MMPs are obligatory in flow related remodel- ing and post angioplasty remodeling [23,24]. No
animal models have been described in which both re- modeling modes, expansive and constrictive remodel-
ing, occur. Thus far, studies in which alterations in de novo atherosclerotic remodeling is investigated in re-
sponse to MMP inhibition are lacking. Based on the present results it might be postulated that, next to
aneurysm formation, MMPs play a role in moderate forms of de novo atherosclerotic expansive remodel-
ing.
The cross-sectional study design also limits the choice of reference segments. ‘Constrictive’ and ‘ex-
pansive’ should therefore be considered as relative rather than absolute terms. A difference in vessel area
between the site of interest and adjacent reference sites is an arbitrary but widely used measure for sig-
nificant remodeling in cross-sectional studies [31 – 33].
In data analysis, the degree of staining was treated as a categorical variable. The cut off values for the
different groups were arbitrarily chosen, which may have influenced the outcome.
Expansively remodeled
segments contain
more plaque mass compared with constrictively remodeled
segments. Thus, although MMPs in immunohisto- chemical staining may differ among constrictively and
expansively remodeled segments, the values per unit plaque area may not differ. Plaque area may be a
confounder within the relationship between vessel area and MMP staining. Indeed, after correction for total
protein, MMP-9 and inactive MMP-2 did not differ among groups. However, even after correction for to-
tal protein, active MMP-2 was higher in expansively remodeled segments.
5. Conclusions