From the Western Vascular Society p27kip1 Knockout enhances collateralization in response to hindlimb ischemia Galit Ankri-Eliahoo, PhD,a Kevin Weitz, BS,a Timothy C. Cox, PhD,b and Gale L. Tang, MD,a,c Seattle, Wash Objective: The natural response to arterial occlusive disease is enlargement of collaterals; however, the molecular factors that control collateralization are not well understood. The gene p27Kip1 (p27) affects human response to arterial injury. Previous studies have shown that overexpression of p27 inhibits vascular endothelial and vascular smooth muscle cell (VSMC) proliferation and angiogenesis. To test the hypothesis that knockout of p27 would improve collateralization in reaction to L L ischemia, we performed in vivo and in vitro experiments using p27 knockout (p27 / ) and wild-type (wt) mice. L L Methods: Hindlimb ischemia was induced by left femoral artery ligation in p27 / and wt (C57BL/6) female mice. The mice underwent weekly laser Doppler perfusion imaging of the footpads until sacrifice on postoperative day 28 followed L L by microcomputed tomography scanning of both hindlimbs. VSMCs were isolated from p27 / and wt mice and used in migration and gel contraction assays in the absence and presence of the nonspecific matrix metalloproteinase (MMP) inhibitor BB94. MMP-2 and MMP-9 messenger RNA (mRNA) expression was measured by quantitative reverse L L transcription-polymerase chain reaction in p27 / and wt VSMCs. L L Results: p27 / mice reperfused more effectively than wt mice by laser Doppler starting from day 7 (ischemic/nonischemic ratio, 0.33 6 0.02 vs 0.25 6 0.02; P < .05) and continuing through day 28 (0.45 6 0.04 vs 0.31 6 0.04; P < .05). The L L gracilis collateral diameter was similar for the nonischemic hindlimbs of the p27 / and wt mice, and this collateral L L pathway increased similarly after ischemia as assessed by microcomputed tomography. However, the p27 / mice significantly enlarged a novel collateral pathway that bridged directly between the femoral artery proximal to the ligation site and the saphenous or popliteal artery distal to the ligation site more than wt mice (158 6 18.3 vs 82 6 22 mm; P < .001). L L p27 / VSMCs migrated more (79% 6 5% vs 56% 6 6%; P < .05) and caused more gel contraction (18% 6 5% of the initial L L area vs 43% 6 4%; P < .05) than wt cells. Migration and collagen contraction were abolished in p27 / and wt cells by MMP L L inhibition. p27 / cells expressed significantly more MMP-2 mRNA than wt cells did. Conclusions: Knockout of p27 enhances arterial collateralization in response to hindlimb ischemia through enlargement of a new collateral pathway. In vitro, knockout of p27 increases collagen gel contraction in addition to stimulating VSMC migration. We speculate that p27 may affect collateralization through its role in regulating MMP-2 expression. (J Vasc Surg 2016;63:1351-9.) Clinical Relevance: Atherosclerosis is the leading cause of mortality and morbidity in the United States. _ENREF_2 The adaptive response to the progressive occlusion of arteries is collateralization (arteriogenesis). As many patients with severe atherosclerosis are not good candidates for angioplasty or surgical bypass, therapies directed toward improving collat- eralization are needed. The molecular pathways controlling collateralization, however, are not well understood. The human response to arterial injury is affected by a genetic polymorphism in the gene CDKN1B (p27Kip1 or p27). We demonstrate that knockout of p27 improves collateralization. Study of the molecular partners of p27 will identify therapeutic candidates to enhance this process for patients. Collateralization (arteriogenesis) is the adaptive collaterals results in restoration of approximately 30% of response to the progressive occlusion of arteries caused original blood flow,1 which is frequently insufficient by atherosclerosis in both the coronaries and the peri- to avoid further interventions to improve distal blood pheral vasculature. At best, however, this enlargement of flow. Thus far, therapies directed toward enhancing From the Division of Vascular Surgery, University of Washingtona; the Additional material for this article may be found online at www.jvascsurg.org. Department of Pediatrics, University of Washington, and Center for Correspondence: Gale L. Tang, MD, VA PSHCS, Division of Vascular Developmental Biology and Regenerative Medicine, Seattle Children’s Surgery, Department of Surgery, University of Washington, Surgical Research Instituteb; and the Division of Vascular Surgery, Department Services 112, 1660 S Columbian Way, Seattle, WA 98108 (e-mail: of Surgery, VA Puget Sound Health Care System.c [email protected]). The content is solely the responsibility of the authors and does not neces- The editors and reviewers of this article have no relevant financial rela- sarily represent the official views of the Department of Veterans Affairs tionships to disclose per the JVS policy that requires reviewers to or the United States Government. decline review of any manuscript for which they may have a conflict Author conflict of interest: none. of interest. Presented as a long presentation at the Twenty-ninth Annual Meeting 0741-5214 of the Western Vascular Society, Coronado Bay, Calif, September Published by Elsevier Inc. on behalf of the Society for Vascular Surgery. 20-23, 2014. http://dx.doi.org/10.1016/j.jvs.2014.12.047 1351 JOURNAL OF VASCULAR SURGERY 1352 Ankri-Eliahoo et al May 2016 collateralization have been disappointing in larger, ran- Femoral A. domized clinical trials,2-4 most likely because the molecular factors that control collateralization are not well Superficial Bridge Collaterals understood.5 Epigastric A. Ligature A genetic polymorphism in the gene for the cell cycle inhibitor CDKN1B, also known as p27Kip1 (p27), affects Artery 1 Gracilis the human response to arterial injury. Patients who are ho- Collateral 1 mozygous for the minor variant A allele of a common Popliteal A. single-nucleotide polymorphism of the promoter region of p27 have decreased coronary in-stent restenosis and Artery 2 60% less chance to experience vein graft failure than pa- Gracilis tients who are heterozygous or homozygous for the C Collateral 2 allele.6,7 The A allele in the promoter was associated with 20-fold increase in reporter gene expression.7 Overexpres- Artery 3 sion of p27 using an adenoviral vector resulted in decreased Gracilis blood flow recovery in a murine hindlimb ischemia model,8 Collateral 3 suggesting a possible role for p27 in collateralization. If it is confirmed, genetic variation in p27 may explain why some patients have abundant collaterals and remain asymptom- atic whereas others manifest severe disease. Fig 1. Schematic diagram showing the arterial anatomy of the To better understand the role of p27 in collateraliza- À/À mouse hindlimb. Indicated on the diagram are the femoral area tion, we tested the collateralization response of p27 position (just proximal to the ligation); artery positions 1 (proximal mice to hindlimb ischemia. We hypothesized that reduced saphenous artery), 2 (midsaphenous artery), and 3 (distal saphe- expression of p27 would improve collateralization in nous artery); gracilis collateral positions 1 (upstream collateral), 2 response to ischemia. We also tested the effect of p27 on (midcollateral), and 3 (downstream collateral); and bridge collat- vascular smooth muscle (VSMC) migration and gel eral positions. contraction in the presence and absence of matrix metallo- proteinase (MMP) inhibition. We further compared messenger RNA (mRNA) expression levels of MMP-2 blanket to 37C under 1.5% isoflurane anesthesia delivered À/À and MMP-9 in p27 and wild-type (wt) cells to provide with 1 L/min O2, were scanned with a MoorLDI2 laser a possible mechanism by which p27 inhibits collateral ar- Doppler perfusion imager (Moor Instruments, Wilming- tery remodeling. ton, Del) immediately postoperatively and on postopera- tive days 7, 14, 21, and 28. Flux values were averaged METHODS for the entire footpad and then expressed as a ratio of ischemic (left)/nonischemic (right). Animal care and procedures Pressure perfusion fixation. All mice subjected to Animal care. All animal experiments were performed hindlimb ischemia underwent laparotomy under 2% isoflur- with approval of the Institutional Animal Care and Use ane anesthesia with 1 L/min O2 on postoperative day 28. Committees of the University of Washington and the VA A 22-gauge silicon intra-aortic catheter was used to deliver Puget Sound Health Care System. p27-deficient female 10 mL of vasodilator solution (1 mM nitroprusside, mice (p27À/À, backcrossed at least nine generations to 10 mM adenosine, 2.5 U/mL heparin in saline) after a nick C57BL/6 background, B6.129S4-Cdkn1btm1Mlf/J, in the inferior vena cava was made to allow outflow.10 weights 22-27 g) and C57BL/6 female mice (wt, weights Pressure perfusion fixation was then carried out at 18-21 g) were obtained from the Jackson Laboratories 100 mm Hg with methyl Carnoy solution. After 2 minutes (Bar Harbor, Me). Female mice were used because they are of fixation, contrast material (60% barium sulfate, liquid E- known to collateralize more poorly than male mice.9 All Z-Paque) was infused through the intra-aortic catheter mice were maintained under standard conditions of 12- until the hindlimb arteries were filled (approximately hour light and dark cycles, with access to chow and water 200 mL per mouse). The hindlimbs then underwent im- ad libitum. mersion fixation with methyl Carnoy overnight, followed Hindlimb ischemia by femoral artery ligation. The by dehydration with 70% EtOH. The hindlimb specimens 3- to 5-month old p27À/À (n ¼ 10) and wt (n ¼ 9) mice then underwent microcomputed tomography (microCT) fl were anesthetized using iso urane with 1 L/min O2. scanning. Buprenorphine (0.05 mg/kg) was given subcutaneously MicroCT scanning and analysis. Hindlimbs (n ¼ 5 for analgesia before ligation. The left femoral artery was p27À/À and n ¼ 4 wt sets of nonischemic and ischemic ligated with 6-0 silk suture just distal to the superficial hindlimbs) were imaged at the Seattle Children’s Research epigastric artery (Fig 1).