The hypothesis that lipoprotein association with perlecan is atherogenic was tested by studying atherosclerosis in mice that had a heterozygous deletion of perlecan, the primary extracellular heparan sulfate proteoglycan in arteries. We first studied the expression of perlecan in mouse lesions and noted that this proteoglycan in aorta was found in the subendothelial matrix. Perlecan was also a major component of the lesional extracellular matrix. Mice with a heterozygous deletion had a reduction in arterial wall perlecan expression. Atherosclerosis in these mice was studied after crossing the defect into the apolipoprotein E (apoE) and LDL receptor knockout backgrounds. At 12 weeks, chowfed apoE null mice with a heterozygous deletion had less atherosclerosis. However, at 24 weeks and in the LDL receptor heterozygous background, the presence of a perlecan knockout allele did not significantly alter lesion size. Thus, it appears that loss of perlecan leads to less atherosclerosis in early lesions. Although this might be attributable to a decrease in lipoprotein retention, it should be noted that perlecan might mediate multiple other processes that could, in sum, accelerate atherosclerosis. Much of atherosclerosis research in the 20th century focused on the cholesterol hypothesis. The evidence that plasma cholesterol-containing lipoproteins cause atherosclerosis is indisputable. However, as was nicely summarized nearly 50 years ago (1), how these lipid-containing particles accumulate within the artery and then produce inflammatory reactions is still unclear. A refinement to the "infiltrative theory" has been an attempt to define the biochemical interactions that lead to this process. Because the lipoproteins are initially found in the extracellular space, they are thought to bind to protein or carbohydrate components of the extracellular matrix. Among the molecules that have been most studied in this respect are proteoglycans (2).Within the artery are a number of classes of proteoglycans, complex proteins that contain highly negatively charged carbohydrates. There are several observations that support the proteoglycan-lipoprotein association hypothesis. 1 ) Dermatan/chondroitin sulfate proteoglycans are found in greater amounts in lesions (3, 4), and some classes of these proteins are found in regions that also contain apolipoprotein B (apoB) (5), the major protein component of LDL and remnant lipoproteins. 2 ) Complexes of proteoglycans and LDL have been isolated from arteries (6). 3 ) Prolonged incubations of proteoglycans and LDL will, in vitro, produce aggregates (7).Of all the proteoglycans, the most heavily charged and most avid LDL binding are the heparan sulfate proteoglycans (8). Heparin, a highly charged carbohydrate of this class, has been studied because it interacts with apoB-containing lipoproteins (9, 10). This reaction, however, is most evident in low ionic strength solutions; in several reports, the proteoglycan-LDL complexes are dissociated by physiologic saline (9). Moreover, increasing ...