Deletion of L-Selectin Increases Atherosclerosis Development in ApoE−/− Mice

Atherosclerosis is an inflammatory disease characterized by accumulation of leukocytes in the arterial intima. Members of the selectin family of adhesion molecules are important mediators of leukocyte extravasation. However, it is unclear whether L-selectin (L-sel) is involved in the pathogenesis of atherosclerosis. In the present study, mice deficient in L-selectin (L-sel −/−) animals were crossed with mice lacking Apolipoprotein E (ApoE −/−). The development of atherosclerosis was analyzed in double-knockout ApoE/L-sel (ApoE −/− L-sel −/−) mice and the corresponding ApoE −/− controls fed either a normal or a high cholesterol diet (HCD). After 6 weeks of HCD, aortic lesions were increased two-fold in ApoE −/− L-sel −/− mice as compared to ApoE −/− controls (2.46%±0.54% vs 1.28%±0.24% of total aortic area; p<0.05). Formation of atherosclerotic lesions was also enhanced in 6-month-old ApoE −/− L-sel −/− animals fed a normal diet (10.45%±2.58% vs 1.87%±0.37%; p<0.05). In contrast, after 12 weeks of HCD, there was no difference in atheroma formation between ApoE −/− L-sel −/− and ApoE −/− mice. Serum cholesterol levels remained unchanged by L-sel deletion. Atherosclerotic plaques did not exhibit any differences in cellular composition assessed by immunohistochemistry for CD68, CD3, CD4, and CD8 in ApoE −/− L-sel −/− as compared to ApoE −/− mice. Leukocyte rolling on lesions in the aorta was similar in ApoE −/− L-sel −/− and ApoE −/− animals. ApoE −/− L-sel −/− mice exhibited reduced size and cellularity of peripheral lymph nodes, increased size of spleen, and increased number of peripheral lymphocytes as compared to ApoE −/− controls. These data indicate that L-sel does not promote atherosclerotic lesion formation and suggest that it rather protects from early atherosclerosis.


Introduction
Endothelial activation and subsequent accumulation of leukocytes is a key event in early atherosclerosis [1]. The selectin family of adhesion molecules mediates initial rolling and tethering of inflammatory cells at sites of activated endothelium [2,3,4,5,6]. The family consists of the three closely homologous glycoproteins E-selectin (E-sel), P-selectin (P-sel), and L-selectin (L-sel), that all bind glycoproteins and glycolipids bearing sialyl Lewis X (sLeX) in a calcium-dependent manner [7,8]. Upon stimulation, E-sel is expressed on endothelial cells, while P-sel is expressed in both endothelial cells and platelets. L-sel, on the other hand, is constitutively expressed on the majority of leukocytes [6].
L-sel exhibits adhesive as well as signaling functions [9,10] and is particularly important for lymphocyte homing to secondary lymphoid organs [5,11]. Indeed, animals lacking L-sel display an altered size of secondary lymphoid tissues and increased numbers of peripheral lymphocytes [11,12,13]. Moreover, L-sel deficient mice show reduced leukocyte rolling along cytokine-stimulated endothelium in vivo. This is well documented in venules in the microcirculation and primarily depends on a lack of L-selmediated interactions between leukocytes regulating capture of cells from the free flow [14,15]. Indeed, whether functional L-sel ligand activity is regularly upregulated on inflamed endothelium is still under debate [16].
Since the selectins are known to regulate leukocyte recruitment in inflammation, they are interesting candidates to study in the context of atherogenesis. Indeed, mice deficient in E-and P-sel display attenuated development of atherosclerosis [2,17]. Moreover, lymphocyte recruitment to the aortic wall during atherosclerosis development is partially L-sel dependent [18]. However, there are no in vivo reports addressing the impact of L-sel for the development of atherosclerotic lesions in vivo. In this study, Lselectin deficient (L-sel 2/2 ) mice were crossed with Apolipoprotein E deficient mice (ApoE 2/2 ) to investigate the relevance of L-sel on both early and advanced stages of atherosclerosis.

L-selectin attenuates early, but not advanced atherosclerosis
The development of atherosclerosis was monitored in descending aortas of mice with or without L-sel. In 12 week old ApoE 2/2 L-sel 2/2 animals fed a HCD for 6 weeks, the percentage of the aorta occupied by atherosclerotic plaques was two fold higher than in age-and diet-matched ApoE 2/2 controls (2.46%60.54% vs 1.13%60.19%, respectively; p,0.05; Fig. 1A). The effect of L-sel deletion was even more pronounced in 6 month old animals fed a normal diet. Under these conditions, ApoE 2/2 L-sel 2/2 mice had 10.4562.58% of the descending aorta covered by plaques as compared to 1.8760.37% in ApoE 2/2 controls (p,0.05; Fig. 1B). In contrast, the atherosclerotic burden in 18 week old ApoE 2/2 L-sel 2/2 animals fed a HCD for 12 weeks (11.8061.86%) was similar to that of ApoE 2/2 controls (13.8962.06%; p = n.s.; Fig. 1C). There was no difference in plasma cholesterol levels between double knockout and control mice in any of the groups (p = n.s.; Table S1). Expression of E-sel did not differ in ApoE 2/2 controls and ApoE 2/2 L-sel 2/2 animals during atherosclerotic lesion formation (p = n.s.; Fig. S1A). P-sel expression was significantly increased in arteries of ApoE 2/2 mice after 6 weeks of HCD compared to ApoE 2/2 L-Sel 2/2 mice. (p,0.05; Fig. S1B) Vascular smooth muscle cell accumulation was similar in the two groups (p = n.s.; Fig. S1D). Staining for collagen exhibited a minor increase in plaques from ApoE 2/2 L-sel 2/2 animals as compared to ApoE 2/2 controls (p,0.05; Fig. S1C).

L-selectin does not influence leukocyte capture and rolling in atherosclerosis
Leukocyte capture and rolling were assessed using intravital microscopy. There was no difference between ApoE 2/2 and ApoE 2/2 L-sel 2/2 animals in primary leukocyte capture directly to the endothelium from the free flow (5.461.3 cells vs 5.561.3 cells, respectively; p = n.s.; Fig. 2A). Secondary capture mediated by interactions between leukocytes was low in ApoE 2/2 as well as in ApoE 2/2 L-sel 2/2 mice (1.2860.92 cells vs 0.1760.14 cells, respectively; p = n.s.; Fig. 2B). Correspondingly, there was no difference in the total number of cells rolling along the aortic endothelium in ApoE 2/2 controls and ApoE 2/2 L-sel 2/2 mice (p = n.s.; Fig. 2C). Total capture correlated with the number of rolling cells (Fig. 2D).
Increased number of circulating lymphocytes in L-sel 2/2 mice ApoE 2/2 L-sel 2/2 animals exhibited a 1.4 fold and 1.6 fold increased number of blood lymphocytes as compared to ApoE 2/2 controls after 6 and 12 weeks of HCD, respectively (p,0.05; Fig. 4A). Consistent with this observation, there was an increased number of CD8 + and CD19 + cells in ApoE 2/2 L-sel 2/2 mice irrespectively of the duration of HCD (p,0.05; Fig. 4B and C). Moreover, there was tendency towards an increased number of CD4 + cells in ApoE 2/2 L-sel 2/2 mice after 6 weeks of HCD (p = 0.24), which was significant after 12 weeks of this diet (p,0.05; Fig. 4D). The increased number of circulating lymphocytes in ApoE 2/2 L-sel 2/2 mice was associated with an increased number of naive T helper cells (CD4 + CD44 2 ; p,0.05; Fig. 4E) after 6 and 12 weeks of HCD. The number of activated T helper cells (CD4 + CD44 + ) did not differ after 6 weeks, but was lower in ApoE 2/2 L-sel 2/2 as compared to ApoE 2/2 mice after 12 weeks of HCD (p,0.05; Fig. 4F). No significant difference in the circulating leukocyte profile was observed after 6 and 12 weeks of HCD in any of the genotypes.
A significant reduction in size (p,0.05; Fig. S2A) and cellularity (p,0.05; Fig. S2B) of peripheral lymph nodes was observed in ApoE 2/2 L-sel 2/2 as compared to ApoE 2/2 mice after 6 and 12 weeks of HCD. In contrast, the spleen was 30% larger in L-sel deficient mice at both time-points (p,0.05; Fig. S3A). The increased spleen size was not associated with an altered cellularity, cell composition (p = n.s.; Fig. S3B and C) or an altered cytokine expression (p = n.s.; data not shown).

Mice
The development of atherosclerosis in ApoE 2/2 L-sel 2/2 mice was studied by two independent experiments. In the first set, L-sel 2/2 mice backcrossed to C57Bl/6 background for 9 generations were used. In the other set, L-sel 2/2 mice were purchased from Jackson Lab and backcrossed to C57Bl/6 background for 5 generations. L-sel 2/2 mice were cross-bred with ApoE 2/2 mice to generate ApoE 2/2 L-sel 2/2 and littermate ApoE 2/2 controls. In the first set, 6 week old males were fed a HCD (Clinton-Cybulski diet, 1.25% cholesterol, Research Diets #D12108) for 6 or 12 weeks. In the other set, male ApoE 2/2 Lsel 2/2 animals and controls were fed a normal diet for 6 months before analysis. All animal experiments were approved by the appropriate authorities.

Quantification of atherosclerosis development
Under anesthesia, blood was collected through right ventricular puncture. Mice were then perfused with ice-cold saline. Aortas were harvested for RNA isolation, en face, and histological analysis. The thoracic and abdominal part of the aortas were fixed overnight with 4% paraformaldehyde (PFA), washed 3 times with ice-cold PBS, and stained for 3 hours with Oil red-O (Sigma #O9755). Quantification of aortic plaque area was performed using AnalysisFIVE software manually by an investigator blinded to the genotypes.
Intravital microscopy 18 week old male mice fed with a HCD for 12 weeks were anesthetized with isofluran. The aorta was prepared as described previously [19]. Briefly, the abdomen was opened by a midline incision and the intestines were retracted. The peritoneum was then dissected to expose the abdominal aorta. The exposed tissue was superfused with a thermostated (37uC) bicarbonate-buffered saline solution. Microscopic observations were made using an intravital microscope (Leitz Biomed) with a water immersion objective (Leitz SW 256). Epi-illumination fluorescence microscopy (Leitz Ploem-o-pac, filter block M2 illuminated by a cooled infrared filtered lamp (Osram HBO 200W/4)) was started 2 minutes after labeling of circulating leukocytes with an intravenous injection of rhodamine 6G (0.3 mg/ml, 0.67 mg/kg). Images were televised and recorded on videotape using a VNC-703 video camera. Leukocyte rolling flux was determined as the average number of leukocytes rolling within a 10000 mm 2 area during 30 seconds within a total observation time of at least 180 seconds. Leukocyte capture was determined as the number of leukocytes that initiated rolling within a 10000 mm 2 area during 30 seconds [20]. Leukocyte capture in contact with or 50 mm downstream of rolling or adherent leukocytes were regarded as secondary, all other capture was regarded as primary.

Statistical analysis
The results were expressed as mean 6 S.E.M. Comparison of means was carried out by Student's t-test or ANOVA in case of multiple comparisons. For each experiment, P,0.05 was accepted as statistically significant.

Discussion
Accumulation of leukocytes in the arterial wall is an important pathogenic event in atherogenesis. It is well documented that the selectin family of adhesion molecules mediates initial attachment of leukocytes to activated endothelium, representing the first step of leukocyte emigration into sites of inflammation [2,6]. Correspondingly, L-sel may play a role in the migration of leukocytes to atherosclerotic lesions and data have been presented supporting that lymphocyte recruitment during atherosclerosis development is partially L-sel dependent [18]. Thus, we hypothesized that deletion of L-sel might attenuate the development of atherosclerosis due to inhibition of leukocyte rolling and capture. To test this hypothesis, we compared atherogenesis in ApoE 2/2 L-sel 2/2 mice with that of ApoE 2/2 controls [21]. Interestingly, the data show that L-sel does not promote atherosclerotic lesion formation in ApoE 2/2 mice. On the contrary, genetic deficiency in L-sel resulted in a significant increase in lesion formation, at least during early stages of the disease. Indeed, after 6 month of normal diet, atherosclerosis was relatively advanced in the absence of L-sel, while plaque burden was still low in the ApoE 2/2 control group and comparable to the ApoE 2/2 control animals after 6 weeks of HCD. Thus, both feeding protocols induce an early stage of the disease; however, the absence of L-sel results in a strong increase of atherosclerosis. In line with this, there was no decrease in leukocyte rolling between ApoE 2/2 L-sel 2/2 and ApoE 2/2 control mice in the atherosclerotic aorta. These observations indicate that other members of the selectin family are sufficient to maintain leukocyte-endothelium interactions under conditions of L-sel deletion [17,22,23]. In line with this interpretation, expression of E-sel and P-sel were not upregulated in the absence of L-sel during atherosclerotic lesion formation. Previous data reveal that the effect of combined deficiency of P-and E-sel has an effect on rolling and recruitment in inflammation which is much stronger than that seen in mice deficient in L-sel [22,23]. Blockage of P-sel also virtually abolishes interactions between leukocytes and endothelium in the atherosclerotic aorta and inhibition of E-sel stabilizes leukocyte rolling under these conditions [2] supporting that E-and P-sel are key mediators of initial leukocyte attachment in arteries. Combined deficiency in E-sel and P-sel also strongly reduces the formation of atherosclerotic lesions [17]. In contrast, as previously indicated [14], L-sel-dependent secondary capture does not increase rolling on atherosclerotic endothelium. Nonetheless, L-sel increases rolling in venules in the microcirculation [11,14,24], which has been shown to be dependent mainly on interactions between leukocytes. Ligands for L-sel are only expressed by endothelium in secondary lymphoid tissues and, under certain circumstances, also by chronically inflamed systemic endothelium [18,25]. Interestingly, data from a previous study  Table 1. mRNA expression (normalized to S12 expression) of different cytokines in atherosclerotic plaques is not affected by L-sel after 6 and 12 weeks of HCD.  suggested that L-sel dependent accumulation of lymphocytes in arteries occurs almost exclusively from the adventitial side of the vessel suggesting that L-sel influences recruitment from the vasa vasorum [18]. This apparent role of L-sel could be mediated by both direct interactions between leukocytes and endothelium as well as secondary capture interactions. Thus, it is possible that Lsel influences rolling and recruitment in other parts of the vascular wall than in the arterial lumen. A strong argument against this  interpretation is that plaques from ApoE 2/2 L-sel 2/2 mice exhibited similar numbers of macrophages and T cells as compared to lesions from ApoE 2/2 controls. Ideally, the cellular composition of the plaque should be examined in the descending aorta, i.e. at that site of the aorta in which significant differences in plaque size were noted. As the atherosclerotic alterations in the descending aorta are focal, it is virtually impossible to cut the descending aorta at the same site and find a plaque of similar size to assess and compare its composition. For this reason, cellular plaque composition was studied in the aortic sinuses, i.e. at a site where plaques of similar size could consistently be detected. It has been observed that even if plaque size is similar at the level of the aortic root, differences in plaque composition can be detected [26].
No difference in cytokine expression in atherosclerotic vessel walls from ApoE 2/2 L-sel 2/2 and ApoE 2/2 mice was detected indicating a similar extent of local inflammation without or with Lsel. The enhanced aortic cytokine levels in animals treated with HCD for 12 weeks as compared to those treated for 6 weeks is consistent with a more advanced stage of atherosclerosis in these mice and did not differ between strains. Hence, alterations in local inflammation do not seem to account for the atheroprotective actions of L-sel. In line with these observations, plasma cytokine levels were similar in ApoE 2/2 L-sel 2/2 and ApoE 2/2 mice. In plasma of ApoE 2/2 L-sel 2/2 mice fed a HCD for 6 weeks the levels of the chemotactic cytokine MCP-1 were elevated compared to ApoE 2/2 mice, which most likely reflects the increased plaque burden in these animals.
Atherosclerosis does not only affect the wall of blood vessels, but also provokes changes at the systemic level [27,28]. Deletion of Lsel resulted in abnormal systemic leukocyte distribution, which could potentially affect atherosclerosis development [29]. Both size and cellularity of peripheral lymph nodes were decreased in ApoE 2/2 L-sel 2/2 mice as compared to ApoE 2/2 controls, which is consistent with the observation that migration of naive lymphocytes into peripheral lymph nodes is impaired in L-sel deficient mice [11,13]. Likely due to compensation for this impaired migration into tissues, L-sel deficiency results in increased numbers of circulating lymphocytes. Hence, it is possible that the enhanced atherosclerosis in ApoE 2/2 L-sel 2/2 mice is driven by more abundant circulating proatherogenic cells, in particular because migration of these cells into lesions appears not to be impaired by lack of L-sel [2,30].