Selective inactivation of p27^Kip1 in hematopoietic progenitor cells increases neointimal macrophage proliferation and accelerates atherosclerosis

Atherosclerotic plaque development is due in part to the recruitment of circulating blood leukocytes into the artery wall, the migration of vascular smooth myocytes (VSMCs) from the media toward the growing neointimal lesion, and hyperplastic cell growth.^1,2  The growth suppressor p27 coordinately inhibits cell proliferation and migration.³  Notably, p27 expression inversely correlates with macrophage and VSMC proliferation within human atherosclerotic tissue,⁴  and transforming growth factor β (TGF-β) present in human atherosclerotic tissue may mediate its growth-suppressive activity through p27.⁵  Furthermore, doubly deficient p27^–/–apoE^–/– mice display accelerated atherogenesis as compared with apoE^–/– mice with an intact p27 gene.⁶  This effect of p27 inactivation was dose dependent and correlated with augmented proliferation of arterial macrophages and VSMCs. In the present study, we used a bone marrow (BM) transplantation model to examine whether targeted inactivation of p27 in hematopoietic precursors is sufficient to increase arterial macrophage proliferation and exacerbate atherosclerosis development.

Female apoE^–/– mice (C57BL/6J; Taconic M&B, Ry, Denmark) irradiated with 2 doses of 5.1 Gy spaced 4 hours apart received transplants of 2 × 10⁶ BM cells obtained from the pooled femur of 2 male apoE^–/– or doubly deficient p27^–/–apoE^–/– mice (C57BL/6J).⁶  After 4 weeks on a standard diet (2.8% fat; Panlab, Barcelona, Spain), the mice that received transplants were fed an atherogenic diet for 2 months (15.8% fat, 1.25% cholesterol, 0.5% sodium cholate, S4892-S010; Ssniff, Soest, Germany). Plasma cholesterol level was determined as described.⁶ 

Fat-fed mice were killed and genomic DNA was isolated from BM, spleen, and thymus to quantify transplantation efficiency by polymerase chain reaction (PCR) amplification of a 184–base pair (bp) fragment of the Y chromosome–specific Zfy-1 gene. DNA amount was normalized by quantitative amplification of a 190-bp fragment of the ras gene. Genomic DNA (20 ng) was amplified in a LightCycler (Roche, Mannheim, Germany) using the SYBR Green I reaction mix (Roche). Serially diluted standard samples of male and female DNA were simultaneously amplified using the same reaction mixture. To quantify cytokine mRNA expression by real-time PCR, 2 μg DNaseI-treated total RNA obtained from snap-frozen aortic arch tissue (Ultraspec RNA isolation system; Biotecx, Houston, TX) was reverse transcribed with MuLV reverse transcriptase (RT) (Roche). Primer sequences and amplification programs are available upon request. Detection of fluorescent product was carried out at the last step of each cycle.

Aortic atherosclerosis was quantified as previously described.⁶ 

Aortic arch cross-sections were immunostained with antibodies against Mac-3 (1/100, sc-19991; Santa Cruz Biotechnology, Santa Cruz, CA), smooth muscle α-actin (SMα-actin) (using an alkaline phosphatase– conjugated antibody 1/100, clone 1A4; Sigma, St Louis, MO), and proliferating cell nuclear antigen (PCNA; 1/50, sc-7907; Santa Cruz Biotechnology). Immunocomplexes containing anti–Mac-3 and anti– PCNA antibodies were detected using the Ultra Streptavidin detection system (Level 2; Signet Laboratories, Cambridge, MA) and 3,3′-diaminobenzidine tetrahydrochloride dihydrate substrate (DAB; Sigma). SMα-actin expression was detected using Fast Red substrate (Sigma).

Whole-cell extracts were prepared from cultured NIH3T3 fibroblasts (American Type Culture Collection, Manassas, VA), BM-derived macrophages,⁷  and from the pooled aortic arch tissue from 4 mice that received transplants. Lysis buffer (50 mM Tris(tris(hydroxymethyl)aminomethane)– HCl [pH 7.5], 1% Nonidet P-40 (NP-40), 0.25% sodium deoxycholate, 150 mM NaCl, 150 mM EDTA [ethylenediaminetetraacetic acid], 1 mM PMSF [phenylmethylsulfonyl fluoride], 1 mM orthovanadate, 1 mM NaF, 0.1% sodium dodecyl sulfate [SDS]) was supplemented with protease inhibitor Complete Mini cocktail (Roche). Western blot analysis was carried out with antitubulin (1/200, sc-3035) and anti–Mac-3 (1/200, sc-19991) antibodies (Santa Cruz Biotechnologies) as previously described.⁶ 

Because reconstitution of apoE^–/– mice with wild-type BM markedly reduces atherogenesis,⁸  we performed our studies with apoE-null marrow. Female apoE^–/– mice were sublethally irradiated and reconstituted with BM from male apoE^–/– and p27^–/–apoE^–/– doubly deficient donors (apoE^–/– → apoE^–/– and p27^–/–apoE^–/– → apoE^–/– mice, respectively). Animals were challenged for 8 weeks with a high-fat, cholesterol-rich diet beginning 1 month after transplantation. apoE^–/– → apoE^–/– and p27^–/–apoE^–/– → apoE^–/– mice had comparable body weight and developed similar hypercholesterolemia (Table 1). Likewise, no statistically significant differences were observed in transplant efficiency in BM and thymus when comparing both groups. However, repopulation efficiency in the spleen was significantly higher in p27^–/–apoE^–/– → apoE^–/– mice, consistent with previous transplantation studies with p27-null BM.⁹ 

Aortic tissue was examined to quantify the extent of atherosclerosis by 2 independent approaches. First, we examined en face preparations of the aorta stained with Oil Red O (Sigma) to mark lipid-laden lesions. Computerized planimetric analysis revealed a 1.85-fold increase in mean lesion area in p27^–/–apoE^–/– → apoE^–/– versus apoE^–/– → apoE^–/– mice (P < .00005; Figure 1A). Second, quantification of the area of atheroma (intima) and media in aortic arch cross-sections in another set of mice disclosed a 2.1-fold increase in the intima-to-media ratio in p27^–/–apoE^–/– → apoE^–/– compared with apoE^–/– → apoE^–/– counterparts (P < .05; Figure 1B). Thus, reconstitution of apoE-null mice with p27-deficient BM enhances diet-induced atherosclerosis.

The cellular composition of atherosclerotic lesions was examined in aortic arch cross-sections by immunohistochemistry with anti–Mac-3 and anti–SMα-actin antibodies to visualize macrophages and VSMCs, respectively, which disclosed an overtly similar histopathology of the atheromas in both groups of mice. SMα-actin was predominantly expressed in the media (Figure 1C-D). In contrast, macrophages prevailed in the atheroma in both groups (Figure 1C-E). We estimated the relative abundance of macrophages in aortic tissue by Western blot analysis using the anti–Mac-3 rat monoclonal antibody. To verify the specificity of this antibody, we first examined cultures of fibroblasts and macrophages obtained from wild-type and p27-null mice, which revealed 2 macrophage-specific bands of similar intensity in wild-type and p27-null cells (Figure 2A, left). As shown in Figure 2A (right), Mac-3 expression was increased in the aorta of fat-fed p27^–/–apoE^–/– → apoE^–/– mice versus apoE^–/– → apoE^–/– counterparts. Averaged over 3 independent experiments, densitometric analysis normalized by the internal tubulin loading control revealed a 2-fold increase in aortic Mac-3 expression in p27^–/–apoE^–/– → apoE^–/– versus apoE^–/– → apoE^–/– mice. Because the main cellular components of the atherosclerotic lesions in our studies are macrophages (Figure 1C-E), increased Mac-3 expression in the aorta of fat-fed p27^–/–apoE^–/– → apoE^–/– mice is consistent with larger plaques in these animals.

To examine whether p27 inactivation in the transplanted marrow increases arterial cell growth in recipient mice, we examined in aortic tissue the expression of the proliferation marker proliferating cell nuclear antigen (PCNA).^10-12  To ascertain the identity of PCNA-immunoreactive cells, consecutive sections were examined by either double SMα-actin/Mac-3 and SMα-actin/PCNA immunohistochemistry (Figure 1C-D), or by single immunostaining with anti-PCNA and anti–Mac-3 antibodies (Figure 1E). SMα-actin/PCNA colocalization was scant, even in intimal SMα-actin–immunoreactive cells (Figure 1C, bottom photomicrograph). In contrast, Mac-3/PCNA coexpression was abundant in the atheromas of both apoE^–/– → apoE^–/– and p27^–/–apoE^–/– → apoE^–/– mice (Figure 1C-E). PCNA immunoreactivity was higher in the atheroma versus the media in both groups, and a tendency toward more PCNA-expressing cells was noted in the atheromas of p27^–/–apoE^–/– → apoE^–/– compared with apoE^–/– → apoE^–/– mice (Figure 2B). Moreover, the proliferative capacity of cultured p27-null macrophages was significantly higher than that of wild-type counterparts (data not shown). Overall, these findings are in agreement with previous studies showing that p27 inactivation enhances hematopoietic progenitor cell proliferation⁹  and implicating p27 as a critical macrophage growth suppressor.^13,14  Because macrophages were the most abundant cells in our study, we suggest that macrophage p27 safeguards against the inflammatory/proliferative response induced by dietary cholesterol in apoE-null mice. These results are consistent with our studies demonstrating that whole-body inactivation of p27 enhances atherosclerosis.⁶  Similarly, both systemic¹⁵  and macrophage-specific^16,17  p53 deficiency enhances diet-induced atherosclerosis. Thus, macrophage expression of growth suppressors seems to play a crucial role in atherosclerosis.

We next examined cytokine expression in aortic tissue by quantitative RT-PCR (Figure 2C). CCL2/MCP-1 (monocyte chemoattractant protein 1) and CCL5/RANTES (regulated on activation, normal T cell expressed and secreted) mRNA were significantly increased in p27^–/–apoE^–/– → apoE^–/– versus apoE^–/– → apoE^–/– mice. In contrast, differences in interleukin-4 (IL-4) and interferon γ (IFN-γ) mRNA expression did not reach statistical significance. Because CCL2/MCP-1 and CCL5/RANTES promote macrophage infiltration and atheroma formation,^18-20  increased macrophage infiltration in response to augmented aortic CCL2/MCP-1 and CCL5/RANTES expression in p27^–/–apoE^–/– → apoE^–/– mice might contribute to accelerated atherosclerosis in these animals. It should be noted that lymphocyte recruitment and proliferation are also implicated in atherogenesis,^1,2  and that p27 is a key regulator of lymphocyte proliferation in different physiopathologic conditions.^21-27  Thus, enhanced lymphocyte activation may also contribute to the phenotype of p27^–/–apoE^–/– → apoE^–/– mice.

In summary, we have shown that targeted p27 inactivation in hematopoietic progenitors enhances arterial macrophage proliferation and inflammatory response resulting in accelerated atherosclerosis. It is noteworthy that retrovirus-mediated constitutive p27 overexpression impaired the engraftment of transplanted cells in our murine model (V. A., A. D.-J., M. A., and A. B, unpublished results, October 2002). Thus, future studies using an inducible, macrophage-specific promoter are warranted to examine whether increasing p27 expression in macrophages may reduce atheroma development.

We thank M. J. Andrés-Manzano for preparing the figures, P. Carbonell for technical assistance, J. Cubells for animal care, and C. Caelles for L929 cells.