Impact of Mouse Strain Differences in Innate Hindlimb Collateral Vasculature

【摘要】  Objective- To assess the importance of genetic background for collateral artery development.

Methods and Results- C57BL/6, BALB/c and 129S2/Sv mice were studied after femoral artery ligation by laser Doppler imaging, visible light oximetry, time-of-flight-magnetic resonance imaging, and treadmill testing; C57BL/6 and BALB/c also underwent electron paramagnetic resonance (EPR) oximetry, x-ray angiography, and histology. C57BL/6 had the least initial distal ischemia and most complete recovery. BALB/c had the most severe initial ischemia and poorest recovery. BALB/c had some vasodilatory reserve in their ligated limbs not seen in the other strains at 3 weeks. By in vivo TOF-magnetic resonance angiography, C57BL/6 had larger preexistent and developed collaterals. By x-ray angiography, C57BL/6 had a higher collateral-dependent filling score and number of visible collaterals immediately after femoral ligation and a higher number of visible collaterals at 1 week but not at 4 weeks. EPR oximetry and histology revealed hypoxia and tissue damage in regions of collateral growth of BALB/c but not C57BL/6 mice. In C57BL/6 BrdUrd uptake in the thigh was limited to larger vessels and isolated perivascular cells. Proliferative activity in collateral arterioles was similar in both strains.

Conclusions- Genetic differences in preexistent collateral vasculature can profoundly affect outcome and milieu for compensatory collateral artery growth after femoral artery occlusion.

The purpose of this study was to assess the importance of genetic background for collateral artery development. Genetic differences in preexistent collateral vasculature can profoundly affect outcome and milieu for compensatory collateral artery growth after femoral artery occlusion.

【关键词】  angiogenesis collateral circulation hypoxia mouse strains vascular biology

Introduction

Pre-existent interarterial anastomoses have been identified in the coronary, cerebral, and peripheral circulation of various species. 1 The number and size of these anastomoses varies between species and tissues, resulting in different degrees of protection after arterial occlusion. 1 Pre-existing collateral arterial connections have been found in human hearts without evidence for coronary artery disease. 1-3 Remodeling of preexistent arterioles into mature collateral arteries has been observed in dog hearts, 4 hindlimbs of rabbits, 5-7 rats, 8 and mice, 9 and this process has been suggested to be the dominant mechanism responsible for the restoration of blood flow after arterial occlusion. 10-12

Previously, we identified differences in perfusion recovery by laser Doppler imaging (LDI) after femoral artery occlusion in inbred strains of mice and attributed these to marginal differences in growth rates of collateral arterioles. 9 However, this study was limited by the fact that LDI and visible light oxygen spectrometry were the only comparative methods of in vivo assessment and histology was confined to isolated gracilis muscle collaterals with minimal surrounding tissue. Diameters of preexistent collaterals on the gracilis muscle were not significantly different between strains.

In this study we examined collateral artery development and related parameters after femoral artery ligation in C57BL/6, 129S2 and BALB/c mice using the most comprehensive approach ever attempted in this model, including magnetic resonance blood flow quantification and angiography, electron paramagnetic resonance (EPR) oximetry, functional testing, and histology on complete cross-sections of medial and caudal thigh muscles. Application of these techniques resulted in new insights about the importance of strain-dependent variations in the innate collateral circulation in this model.

Methods

Animals were handled in accordance with the American Physiological Society guidelines for animal welfare and the Bioethics Committee of the State of Hessen, Germany. We studied 11- to 13-week-old male BALB/c, 129S2/Sv, and C57BL/6 mice from Harlan (Borchen, Germany and Indianapolis, Ind) and Charles River Laboratories (Sulzfeld, Germany). Operations were performed under general anesthesia with intraperitoneal xylazine (25 mg/kg) and ketamine (150 mg/kg). Right (for stress testing both) femoral arteries were ligated (6/0 silk, Ethicon) just distal to the a. femoris profunda and proximal to the a. genus descendens. Before anesthesia for serial measurements, spontaneous movement of the right hindlimbs was scored as either "0" (dragging of foot), "1" (no dragging but no active plantar flexion), "2" (moderately to severely reduced plantar-flexion), or "3" (normal or only mildly abnormal use).

In Vivo Studies

Most serial measurements were performed under general anesthesia with intraperitoneal xylazine (15 mg/kg) and ketamine (85 mg/kg) at 37°C in a temperature controlled chamber or magnetic resonance spectrometer after 5 minutes equilibration time. EPR measurements and magnetic resonance angiograms were performed under isoflurane narcosis.

LDI and Reactive Hyperemia

Mouse feet were scanned with a laser Doppler imager (MLDI 5063; Moor Instruments Ltd, Devon, UK). Transient occlusion of blood flow to the limbs by inflatable cuffs was performed once on all animals to determine mean background flux values for each strain. Background values were subtracted from measured flux values before calculating right-to-left ratios. Three weeks after surgery, reactive hyperemia was assessed after 5 minutes of simultaneous bilateral cuff inflation to 300 mm Hg followed by sudden deflation.

Visible Light Oxygen Tissue Spectrometry

Hemoglobin oxygen saturation in hind feet was measured with a visible light spectrometer (AbTisSpec spectrometer; LEA Medizintechnik GmbH, Germany) using a specially designed circular probe.

Magnetic Resonance Flow Imaging

Blood flow was measured in 3 planes at calf-level in both hindlimbs using a spin labeling inflow sequence with a 7.05 Tesla Bruker PharmaScan horizontal bore magnet equipped with a 300 mT/m gradient system as described. 13,14

Magnetic Resonance Angiography

In the fourth week after ligation, mice magnetic resonance angiography with a field of view of 2.54 x 2.54 x 2.54 cm were obtained as described. 15 For general anesthesia, mice inhaled 1.6% isoflurane in 20% oxygen and air at 1.2 L/min via nasal cone. Three-dimensional maximum intensity projections viewable from multiple angles were analyzed and scored for collateral arterial vessel size and number by 2 independent blinded observers. The score differentiated "small," "medium," and "large" collaterals on a relative basis after the observers had overviewed the range of collateral sizes that occurred. Veins were differentiated from arteries by anatomic location and course.

Exercise Capacity

Exercise treadmill testing was performed on a rodent treadmill (University of Melbourne, Australia) after bilateral arterial ligation to ensure proper running without the possibility of compensation by the contralateral limb. Exercise treadmill testing began 7 days after surgery. Each test started with a 6° slope at a speed of 20 m/min for 15 minutes. Speed was increased by 5 m/min every 15 minutes until the mice were unable to keep pace. The same protocol was used for nonligated control animals.

EPR Oximetry

Tissue p O 2 was measured using 100 µg of oxygen-sensitive sterile charcoal powder, whose EPR spectra (line width) reflects the p O 2, 16 injected bilaterally into medial thigh or gastrocnemius muscles of BALB/c and C57BL/6 mice 2 weeks before surgery. Once implanted, the paramagnetic material remains at the site and is quite inert, permitting repeated tissue p O 2 measurements without invasion. Measurements were performed on a 1.2-GHz (L-band) EPR spectrometer while under inhalation narcosis with 1.25 to 1.5% of isoflurane with 26% oxygen. Right femoral artery ligations and left sided sham ligations were performed on day 0. Tissue p O 2 was measured serially for 14 to 21 days after surgery.

X-Ray Angiography

Seven or 28 days after right femoral artery occlusion the left femoral artery was ligated and angiograms were obtained (please see http://atvb.ahajournals.org).

Histology

BrdUrd uptake was evaluated on complete cross sections of medial and caudal thigh muscles at the level of the ligation site 7 days after surgery and continuous administration of BrdUrd.

Statistical Analysis

Please see http://atvb.ahajournals.org.

Results

Serial Measurements of Perfusion and Hemoglobin Oxygen Saturation in Hind Feet and Blood Flow at Calf Level

We used several complementary methods to assess collateral-dependent blood flow and related parameters to reduce the chance of experimental artifacts affecting the interpretation of our results and to gain insight into how these parameters relate to each other in this model.

One BALB/c mouse developed severe hindlimb necrosis on day 3 and had to be euthanized and a second one was found dead at day 14, reducing the numbers of animals in that strain available for measurements from 15 to 13 over time. BALB/c mice had the lowest relative LDI perfusion values immediately after femoral artery ligation and their recovery was the least complete ( Figure 1 A; Figures I and II, available online at http://atvb.ahajournals.org). C57BL/6 and 129S2 mice had higher relative perfusion immediately after surgery. C57BL/6 mice achieved almost full relative perfusion recovery at 14 days after surgery; 129S2 mice had a slower and less complete recovery.

Figure 1. A, Serial measurements of hind foot perfusion by LDI. B, MR flow at calf level. C, Hemoglobin oxygen saturation in feet. D, Spontaneous active use scores of right hindlimbs. E, Relative exercise tolerance. F, Cross-sectional hindlimb areas by MR at calf level. (N), mean±SEM, (only D is median/interquartile range). * P <0.05 vs BALB/c. P <0.05 vs 129S2 mice.

Large vessel blood flow at calf level as measured by MRI immediately after femoral artery ligation was higher in the C57BL/6J and 129S2 groups ( Figure 1 b; Figure III, available online at http://atvb.ahajournals.org). Thereafter, it completely and rapidly recovered in C57BL/6 mice, whereas 129S2 and BALB/c mice behaved virtually identical reaching a plateau at &60% of the nonligated side at 7 days.

Oxygen saturation obtained by visible light spectrometry mirrored the LDI findings, however, with faster initial recovery in all strains ( Figure 1 C).

For these studies animals of the same strain from 2 vendors were used. While there were no significant differences, there was a trend for faster recovery by LDI in C57BL/6J mice from Harlan compared with C57BL/6H mice from Charles River (data not shown). For the remaining studies only animals from Harlan were used.

Reactive Hyperemia

To assess, whether some of the observed differences in collateral-dependent perfusion are caused by differences in vasomotor tone the degree of reactive hyperemia was assessed 22 days after surgery. On the nonligated side, there was a uniform response in all strains with a 20% increase of LDI flux over baseline in the first minute (Figure IVA, available online at http://atvb.ahajournals.org). On the ligated side, no hyperemic response occurred in C57BL/6 and 129S2 mice (Figure IVB). However, BALB/c mice, with more severe ischemia by LDI at baseline, had a hyperemic response on the ligated side. Overall, the right-to-left ratios of the maximal flux values on each side mildly decreased in C57BL/6J and 129S2, whereas there was no significant increase in BALB/c mice.

Serial Assessment of Functional Capacity

All mice regained full functional recovery as assessed by active hindlimb movement score. C57BL/6 recovered most rapidly, followed by 129S2 and then BALB/c mice ( Figure 1 D).

On exercise treadmill testing, nonligated C57BL/6 mice were able to perform a workload of 70.9±3.5 J, 129S2 of 72.9±6.5 J, and BALB/c of 102.1±6.6 J ( P <0.05 versus 129S2 and C57BL/6). After bilateral femoral artery ligation, C57BL/6 and 129S2 mice recovered their exercise tolerance better than BALB/c mice ( Figure 1 E).

Serial Cross-Sectional Areas and Final Weights of Distal Hindlimb Musculature

Swelling, as assessed by MRI, only occurred in the ischemic hindlimbs of BALB/c mice 3 days after surgery, followed by a gradual decrease of average cross-sectional areas to lower levels than on the nonligated side consistent with tissue atrophy ( Figure 1 F; Figure III). Right-to-left ratios of calf muscle weights on day 28 confirmed the MRI findings at this stage: C57BL/6 0.96±0.05, 129S2 0.89±0.01, BALB/c 0.63±0.03 ( P <0.001 for BALB/c versus both other groups, not significant for C57BL/6 versus129S2). On the nonligated side, the corresponding muscle weights in C57BL/6 (239±13 mg) and BALB/c (224±5 mg) mice were higher than in 129S2 mice (170±4 mg, P <0.001).

Magnetic Resonance Angiography Assessment of Collaterals 3 Weeks After Surgery

Collateral vessels bridging the ligation site were identified subcutaneously, as well as in the quadriceps and occasionally more faintly in the medial or caudal muscle compartments of the thigh ( Figure 2 ). Pre-existing collateral vessels were visible on the nonligated side. Two blinded observers identified 1 to 2 "large" quadriceps collateral vessels on the ligated side in 4 or 5 of 5 C57BL/6J mice (mean for observer A 1.0, observer B 1.4), but in none of the other strains ( P <0.05). On the nonligated side 2/5 C57BL/6J mice but none of the other groups had "large" quadriceps collateral vessels (by both observers). "Medium" and "large" collateral vessels combined in the quadriceps on the nonligated side were identified by observer A in all (mean 1.0) C57BL/6J mice, 2/5 (0.4) BALB/c and 3/5 (0.6) 129S2 mice ( P =not significant). Observer B identified 1 to 2 large collaterals in 4/5 (mean 1.0) C57BL/6J mice, but none of the BALB/c ( P <0.05 versus C57BL/6J, Mann-Whitney Rank Sum test), and 3 of 5 129S2 mice (mean 0.8).

Figure 2. Magnetic resonance angiography in week 4 after surgery. Ligation site (*); collateral vessels are seen subcutaneously (black arrowhead with white disc) and in the cranial thigh (black arrowheads), faintly also in the medial or caudal thigh region. The deep femoral artery, a stem vessel for medial thigh collaterals, was visible in all mice (arrowhead with white circle).

Serial EPR Oximetry in Thigh (Region of Growing Collaterals) and Calf

Although there was a severe transient decrease of tissue oxygen tension (p O 2 ) on the ligated side in the medial thigh of BALB/c, there was no evidence for hypoxia in the medial thigh of most C57BL/6 mice ( Figure 3; Figure V, available online at http://atvb.ahajournals.org). In the calf, both groups had transient hypoxia; however, it was more severe and more prolonged in BALB/c mice. Immediately after femoral artery ligation, there was no overlap of oxygen saturation values in the thighs of BALB/c and C57BL/6 mice (Figure V), whereas there was some overlap between these strains by all other in vivo methods at that time point. ( Figure 1; Figure II; data not shown).

Figure 3. EPR oximetry. No decrease of mean p O 2 in the right (BL R) medial thigh muscles of C57BL/6 mice; however, transient severe decrease of p O 2 in thighs of BALB/c mice (BALB R). In the distal hindlimbs ( gastrocnemius ), BALB/c mice had significant hypoxia until 7 days after surgery; C57BL/6 only immediately after surgery (n=5 to 7 as indicated).

X-Ray Angiography

Overall location and course of collateral vessels were preserved among strains ( Figure 4; Figure VI, available online at http://atvb.ahajournals.org).

Figure 4. X-ray angiograms of mice 7 or 28 days after ligation of the right femoral artery and immediately after ligation of the left femoral artery. Femoral artery ligation sites are labeled (*). Collateral vessels are seen bridging the ligation site (see Table for quantification).

Quantification of X-ray Angiograms

Results are summarized in the Table. Number of grown collaterals and calculated conductance scores were higher in C57BL/6 than in BALB/c at 7 but not 28 days after surgery. At both times, C57BL/6 had higher collateral-dependent filling scores and number of visible preexistent collaterals on the acutely ligated (left) side ( Figure 4 ). Decreased collateral-dependent filling scores on day 28 are probably caused by a mildly higher viscosity of the gelatin-containing contrast agent.

Histology

Cross-sections of thigh muscles of C57BL/6 1 week after surgery and continuous BrdUrd administration revealed no uptake of BrdUrd by capillaries. However, some arteriolar/arterial vessels (of at least 27 µm diameter) had a large number of BrdUrd-positive nuclei in all of their wall layers ( Figure 5A and 5 B). Isolated BrdUrd-positive cells were found around growing vessels and rarely in the walls of large veins.

Figure 5. Histology of thigh muscles of C57BL/6 (A,B,E) and BALB/c mice (C,D,F). BrdUrd uptake (B,D) in C57BL/6 mice limited to wall cells of arterioles, and isolated perivascular cells (A,B). BALB/c mice in addition had large numbers of BrdUrd-positive cells in areas of tissue necrosis (C,D). A and C, red=BS-I lectin, blue=nuclei (DAPI). E and F, Hematoxylin & eosin-stained sections reveal evidence for myofiber destruction and regeneration (central nuclei) only in BALB/c (F), but not in C57BL/6 mice (E).

In BALB/c mice the same histological picture emerged in regions of normal appearing thigh muscle. In addition, multiple necrotic and regenerating areas with destroyed myofibers, myofibers with central nuclei, and large numbers of infiltrating BrdUrd-positive cells were seen ( Figure 5C and 5 D).

Interestingly, no significant difference in relative numbers of BrdUrd-positive cells in arterioles of C57BL/6 (0.56±0.07) and BALB/c mice (0.60±0.08) after 1 week of continuous administration of BrdUrd was found.

Discussion

Major findings of this study are that: (1) genetic mouse strain differences in preexistent interarterial connections exist; (2) these are of major importance for final recovery after femoral artery occlusion; (3) depending on collateral flow immediately after arterial occlusion, severe local hypoxia and necrosis can occur even in proximal hindlimbs where collaterals grow; (4) however, local resting hypoxia/ischemia is not essential for collateral growth as shown by C57BL/6 mice; (5) BrdUrd uptake in the thighs of C57BL/6 mice, and the preserved pattern of collateral growth in BALB/c and C57BL/6 support the concept that collateral arteries develop by remodeling of preexisting arterioles unrelated to de novo vessel growth processes. Furthermore, we demonstrate the feasibility and potential value of some newer techniques in this model, ie, assessment of collateral-dependent flow 13 and direct in vivo visualization of collateral arteries by high-field MRI, 15 evaluation of local tissue p O 2 by EPR oximetry, and assessment of reactive hyperemia by LDI.

Comparison of mice with the poorest (BALB/c) and the best (C57BL/6) recovery of collateral-dependent flow, perfusion, and function revealed that previously not recognized, genetically determined differences in preexisting interarterial anastomoses are of major importance in this model. These differences were suggested by assessment of distal blood flow and perfusion immediately after arterial occlusion and by in vivo imaging of collaterals by magnetic resonance angiography in the nonligated hindlimbs. However, as all of these in vivo methods can be affected by vasomotor tone, the final proof depended on angiographic assessment of collateral-dependent filling immediately after femoral artery ligation and perfusion-fixation. The observed differences in filling of collaterals and of the distal vasculature could be caused by differences in number or smallest diameters of innate collaterals. The very similar end results in terms of number and course of collaterals at later time points suggests that the differences immediately after ligation are most likely because of smaller minimal diameters of preexisting collateral connections in BALB/c mice. Because the contrast agent is too viscous to pass through capillary vessels, potential differences in capillary networks cannot explain the findings. The importance of preexisting collaterals has been previously suggested by studies in pigs and dogs, in which the initial collateral dependent perfusion after coronary artery occlusion correlated with final recovery of perfusion. 1

Differences in preexistent collateral vasculature can strongly impact the milieu for growing collaterals as demonstrated by the dramatic differences between BALB/c and C57BL/6 mice with regard to tissue oxygenation in regions of collateral growth: severe transient hypoxia and evidence for tissue destruction in regions of collateral growth was only consistently found in the thighs of BALB/c mice. The absence of any p O 2 decrease in thighs of most C57BL/6 mice with excellent collateral growth suggests that local hypoxia is not essential for collateral growth and proves that the lack of histological evidence of tissue damage in C57BL/6 is not primarily related to differences in tissue resistance to ischemia, as reported for the myocardium of inbred strains of rats, 17 but to differences in local tissue perfusion. Some previous studies did not find evidence for ischemia in the proximal hindlimbs of rabbits 18 and rats 19 after proximal femoral artery occlusion 18 and a more aggressive arterial ligation approach; 19 however, in these studies indirect markers for hypoxia were used, ie, upregulation of hypoxia inducible genes and ATP levels. By EPR oximetry and microspheres, evidence for hypoxia and ischemia in adductor and quadriceps muscles after femoral artery ligation in rabbits was found. 16 The presence of local tissue hypoxia and necrosis in areas of collateral growth in BALB/c mice suggests that the arterial vessels which become functional collaterals play an important role also for local muscle perfusion. This is further supported by evidence for multiple muscular side branches from many of these vessels, eg, on the surface of the gracilis muscle. 9,20 If these pre-existent interarterial connections are not large or not numerous enough, the intravascular pressure loss transmitted from their distal anastomoses after arterial occlusion apparently can result in an insufficient perfusion pressure for the surrounding muscle tissue.

One limitation of our study is that EPR oximetry can only be performed at rest. Thus, some intermittent hypoxia in the thighs of C57BL/6 mice with activity cannot be ruled out by our data. Furthermore, whereas the absence of measurable hypoxia in the thighs of most C57BL/6 mice after arterial ligation suggests that at least local hypoxia is not essential for collateral growth, it certainly does not rule out a modulating effect of hypoxia/ischemia on collateral growth.

Different from larger animal models, 5,10,12,21,22 in mice collateral conductance cannot be directly assessed as this would require not just measurements of collateral flow but also of proximal and distal blood pressure. For distal flow and perfusion measurements to reflect collateral conductance, peripheral vascular resistance should be minimized, which is difficult to achieve in serial studies of living mice. Also, the hemodynamic response to anesthesia could differ between strains of mice. 23 To address these concerns, measurements were performed at 37°C, which promotes peripheral vasodilation and results of most of the in vivo measurements were calculated as relative values using the nonligated leg as control. Furthermore, we performed LDI measurements after creating total ischemia by external cuff inflation around the hindlimbs. Surprisingly, on the ligated side the most ischemic stain, BALB/c, still had some capacity for perfusion increase over baseline, whereas the other 2 strains did not have any reactive hyperemia, suggesting that they were maximally vasodilated. On the nonligated side there was a uniform mild hyperemic response suggesting no differences in vasomotor tone. Thus, some of the differences in LDI perfusion, at least at later time points, may be caused by vasomotor tone differences on the ligated side between BALB/c and the other 2 strains of mice under the measurement conditions. A limitation of our reactive hyperemia study by LDI is the time resolution. We performed 1 minute scans and, therefore, may have underestimated the degree of early reactive hyperemia. However, spontaneous hindlimb function and exercise treadmill testing overall supported the validity of the in vivo resting measurements. X-ray angiography is not very sensitive for estimating collateral conductance, compared with measurements of flow, which may explain the lack of significant differences between BALB/c and C57BL/6 at 28 days.

The absence of evidence for hypoxia and angiogenesis in areas of collateral growth in C57BL/6 mice, combined with the absence of proliferative activity in small arteriolar vessels (<25 µm in diameter) in BALB/c and C57BL/6 on complete cross-sections of thigh muscles after cumulative BrdUrd administration, and the preserved pattern of collaterals among different strains of mice, support the concept that collateral growth occurs by remodeling of preexistent arteriolar vessels to collateral arteries independent of de novo vascular growth processes. Entirely ruling out any de novo formation of collateral vessels is difficult because of the incomplete visualization of small preexistent collateral arterioles. Interestingly, assessment of local tissue oxygenation in the areas of collateral growth by EPR oximetry appeared to be the best single in vivo method to discriminate between mice with good innate collateral circulation from mice with poor collateral circulation ( Figure 1; Figures II and V).

Because the histological analysis was based on complete cross-sections of medial and caudal thigh muscles, it provided a more reliable and global assessment of collateral growth than when the analysis is limited to small specimens of tissue around isolated collateral vessels as we have performed previously. 9 Perivascular BrdUrd-positive cells were noted, and in BALB/c mice areas of necrosis had many more BrdUrd-positive cells. Most likely these are leukocytes that proliferated in the bone marrow before homing to regions of collateral growth. This is supported by previous data in the same model in which we found many CD45 + cells accumulating around growing collateral vessels after bone marrow transplantation from mice expressing green fluorescent protein. 24 However, it cannot be entirely ruled out from our BrdUrd data that some perivascular cells proliferated in situ as recently proposed for tissue resident macrophages. 25

Previously, the significance of genetic strain-related variations of arterial networks was suggested by cerebral ischemia studies. The posterior communicating arteries in the circle of Willis were found to be less often patent in BALB/c than in BDF and CFW mice, and this was associated with an increased risk of infarction after creating focal or multifocal cerebral ischemia. 26 The higher susceptibility of C57BL/6 compared with 129/Sv mice to ischemia after transient bilateral common carotid artery occlusion was shown to be associated with poorly developed vascular connections in the circle of Willis of C57BL/6. 27

Even though in this study we could not confirm any significant differences in proliferative activity of thigh arterioles by BrdUrd uptake, it may not imply that vascular growth differences between inbred mouse strains do not exist, as our method may not have been sensitive enough. Collateral artery growth appears to be a strongly flow-dependent process. 28 Strain-related differences in flow-induced carotid artery remodeling have been described. 29,30 However, in a model of mycoplasma pulmonis -induced growth of tracheal microvessels, C57BL/6 mice responded with an increased number of capillaries and venules, whereas in C3H mice the diameters of arterioles, capillaries, and venules increased. 31 An additional problem in ischemia models, like the mouse femoral artery ligation model, is that little is known about the secondary effects of distal tissue necrosis on collateral growth. Conceivably, a major loss of collateral-dependent tissue caused by ischemia after arterial occlusion affects the peripheral vascular resistance which would impact blood flow through proximal collateral vessels and thus the driving forces for their growth.

The results of this study highlight the need to consider the genetic background and the innate collateral circulation when designing and interpreting hindlimb ischemia studies. Differences in innate collateral circulation may make it impossible to interpret differences in final recovery of collateral-dependent flow and related parameters as caused by differences in collateral growth rates. However, genetic analysis of mouse strain-dependent variations in pre-existent collateral vessels, for example by strain intercross studies, might offer the opportunity for a better understanding of the mechanisms involved in collateral artery development.

Acknowledgments

The study was supported by the Max-Planck Society and European Commission grant QLG1-CT-2001-01172. We acknowledge Siegfried Langsdorf, Christof Bingel, Sandra Martin (Bad Nauheim, Germany), and Jianfeng Qin (Dartmouth) for technical assistance, and Drs Michael Simons and Nicholas Shworak (Dartmouth) for helpful comments.

【参考文献】
  Helisch A, Schaper W. Arteriogenesis: the development and growth of collateral arteries. Microcirculation. 2003; 10: 83-97.

Baroldi G, Mantero O, Scomazzoni G. The collaterals of the coronary arteries in normal and pathological hearts. Circ Res. 1956; 4: 223-229.

Fulton WFM. Morphology of the collateral circulation in the human heart. In: Schaper W, Schaper J, eds. Arteriogenesis. Boston: Kluwer; 2004: 298-314.

Schaper W, Jageneau A, Xhonneux R. The development of collateral circulation in the pig and dog heart. Cardiology. 1967; 51: 321-335.

Ito WD, Arras M, Scholz D, Winkler B, Htun P, Schaper W. Angiogenesis but not collateral growth is associated with ischemia after femoral artery occlusion. Am J Physiol. 1997; 273: H1255-H1265.

Scholz D, Ito W, Fleming I, Deindl E, Sauer A, Wiesnet M, Busse R, Schaper J, Schaper W. Ultrastructure and molecular histology of rabbit hind-limb collateral artery growth (arteriogenesis). Virchows Arch. 2000; 436: 257-270.

Hoefer IE, van Royen N, Buschmann IR, Piek JJ, Schaper W. Time course of arteriogenesis following femoral artery occlusion in the rabbit. Cardiovasc Res. 2001; 49: 609-617.

Herzog M, Sager H, Khmelevski E, Deylig A, Ito W. Collateral arteries grow from preexisting anastomoses in the rat hindlimb. Am J Physiol Heart Circ Physiol. 2002; 283: H2012-H2020.

Scholz D, Ziegelhoeffer T, Helisch A, Wagner S, Friedrich C, Podzuweit T, Schaper W. Contribution of arteriogenesis and angiogenesis to postocclusive hindlimb perfusion in mice. J Mol Cell Cardiol. 2002; 34: 775-787.

Unthank JL, Nixon JC, Lash JM. Early adaptations in collateral and microvascular resistances after ligation of the rat femoral artery. J Appl Physiol. 1995; 79: 73-82.

Hershey JC, Baskin EP, Glass JD, Hartman HA, Gilberto DB, Rogers IT, Cook JJ. Revascularization in the rabbit hindlimb: dissociation between capillary sprouting and arteriogenesis. Cardiovasc Res. 2001; 49: 618-625.

Yang HT, Ren J, Laughlin MH, Terjung RL. Prior exercise training produces NO-dependent increases in collateral blood flow after acute arterial occlusion. Am J Physiol Heart Circ Physiol. 2002; 282: H301-H310.

Wagner S, Helisch A, Bachmann G, Schaper W. Time-of-flight quantitative measurements of blood flow in mouse hindlimbs. J Magn Reson Imaging. 2004; 19: 468-474.

Heil M, Ziegelhoeffer T, Wagner S, Fernandez B, Helisch A, Martin S, Tribulova S, Kuziel WA, Bachmann G, Schaper W. Collateral artery growth (arteriogenesis) after experimental arterial occlusion is impaired in mice lacking CC-chemokine receptor-2. Circ Res. 2004; 94: 671-677.

Wagner S, Helisch A, Ziegelhoeffer T, Bachmann G, Schaper W. Magnetic resonance angiography of collateral vessels in a murine femoral artery ligation model. NMR Biomed. 2004; 17: 21-27.

Grinberg OY, Hou H, Grinberg SA, Moodie KL, Demidenko E, Friedman BJ, Post MJ, Swartz HM. pO2 and regional blood flow in a rabbit model of limb ischemia. Physiol Meas. 2004; 25: 659-670.

Baker JE, Konorev EA, Gross GJ, Chilian WM, Jacob HJ. Resistance to myocardial ischemia in five rat strains: is there a genetic component of cardioprotection? Am J Physiol Heart Circ Physiol. 2000; 278: H1395-H1400.

Deindl E, Buschmann I, Hoefer IE, Podzuweit T, Boengler K, Vogel S, van Royen N, Fernandez B, Schaper W. Role of ischemia and of hypoxia-inducible genes in arteriogenesis after femoral artery occlusion in the rabbit. Circ Res. 2001; 89: 779-786.

Luo F, Wariaro D, Lundberg G, Blegen H, Wahlberg E. Vascular growth factor expression in a rat model of severe limb ischemia. J Surg Res. 2002; 108: 258-267.

Gruionu G, Hoying JB, Pries AR, Secomb TW. Structural remodeling of mouse gracilis artery after chronic alteration in blood supply. Am J Physiol Heart Circ Physiol. 2005; 288: H2047-H2054.

Schaper W, Flameng W, Winkler B, Wusten B, Turschmann W, Neugebauer G, Carl M, Pasyk S. Quantification of collateral resistance in acute and chronic experimental coronary occlusion in the dog. Circ Res. 1976; 39: 371-377.

Lloyd PG, Yang HT, Terjung RL. Arteriogenesis and angiogenesis in rat ischemic hindlimb: role of nitric oxide. Am J Physiol Heart Circ Physiol. 2001; 281: H2528-H2538.

Zuurbier CJ, Emons VM, Ince C. Hemodynamics of anesthetized ventilated mouse models: aspects of anesthetics, fluid support, and strain. Am J Physiol Heart Circ Physiol. 2002; 282: H2099-H20105.

Ziegelhoeffer T, Fernandez B, Kostin S, Heil M, Voswinckel R, Helisch A, Schaper W. Bone marrow-derived cells do not incorporate into the adult growing vasculature. Circ Res. 2004; 94: 230-238.

Khmelewski E, Becker A, Meinertz T, Ito WD. Tissue resident cells play a dominant role in arteriogenesis and concomitant macrophage accumulation. Circ Res. 2004; 95: E56-E64.

Barone FC, Knudsen DJ, Nelson AH, Feuerstein GZ, Willette RN. Mouse strain differences in susceptibility to cerebral ischemia are related to cerebral vascular anatomy. J Cereb Blood Flow Metab. 1993; 13: 683-692.

Fujii M, Hara H, Meng W, Vonsattel JP, Huang Z, Moskowitz MA. Strain-related differences in susceptibility to transient forebrain ischemia in SV-129 and C57black/6 mice. Stroke. 1997; 28: 1805-1810.

Pipp F, Boehm S, Cai WJ, Adili F, Ziegler B, Karanovic G, Ritter R, Balzer J, Scheler C, Schaper W, Schmitz-Rixen T. Elevated fluid shear stress enhances postocclusive collateral artery growth and gene expression in the pig hind limb. Arterioscler Thromb Vasc Biol. 2004; 24: 1664-1668.

Harmon KJ, Couper LL, Lindner V. Strain-dependent vascular remodeling phenotypes in inbred mice. Am J Pathol. 2000; 156: 1741-1748.

Korshunov VA, Berk BC. Strain-dependent vascular remodeling: the "Glagov phenomenon" is genetically determined. Circulation. 2004; 110: 220-226.

Thurston G, Murphy TJ, Baluk P, Lindsey JR, McDonald DM. Angiogenesis in mice with chronic airway inflammation: strain-dependent differences. Am J Pathol. 1998; 153: 1099-1112.

作者单位：Department of Experimental Cardiology (A.H., S.W., S.W., M.H., T.Z., U.B., W.S.), Max-Planck-Institute for Physiological & Clinical Research, Bad Nauheim, Germany; Angiogenesis Research Center & Section of Cardiology (A.H., M.D., U.B., J.D.P.), Dartmouth Medical School, Lebanon, NH; and EPR