**2. Results**

### *2.1. Evaluation of a Training Regime in C57BL*/*6*

In order to maximize training intensity, we compared two training regimens: A treadmill to implement forced exercise and a free-to-access voluntary running wheel. During the treadmill protocol the mice were trained using an incremental protocol: initial running speed was 0.2 m/s, increased daily to the maximal speed of 0.3 m/s at day 7. The training frequency was twice per day, and each training lasted 30 min. A maximum distance of 1.08 km was covered each day. Distances traveled and running wheel speeds were recorded continuously. After an adaptation period a running distance of 4.36 ± 0.41 km/d with an average speed of 0.5 m/s of was recorded. We observed increased speed and distance values for the voluntary exercise mice when compared to those forced to exercise in a treadmill throughout the observation period (Figure 1a,b).

**Figure 1.** Daily running distance (**a**) and daily average speed (**b**) of C57BL/6 mice over time, during 7 days of adaptation in a treadmill receiving forced exercise (filled circles) or in a voluntary running wheel (open circles). Data are displayed as mean ± SEM. \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.005.

#### *2.2. Acceptance of Voluntary Training in ApoE*−/− *Mice after HFD and Post-Surgery*

To further increase a similarity to patients presenting PAD we included ApoE−/− mice fed with a high fat diet (HFD; 21% butter fat, 1.5% cholesterol) to the exercise study. These mice showed numerous plaques throughout the whole arterial system including the femoral artery and the aortic root as well as fat deposition in the collateral arterioles (Figure 2a–c).

Given the health constraints of ApoE−/− mice following 12 weeks of HFD, we intended to evaluate whether they were able to cover the same distance as healthy C57BL/6 mice pre- and post-FAL. During the adaptation the mean distance covered was 3.93 ± 0.28 km/d which was not significantly different to C57BL/6 mice (*p* = 0.36). Post-FAL the re-adjustment period was similar to C57BL/6 mice, with no significant difference, as visualized in the equal progression of running performance (Figure 2d). Both strains, before and after surgery, recovered quickly to their initial distance when they had access to a voluntary running wheel.

#### *2.3. Reperfusion Recovery after FAL in C57BL*/*6 and ApoE*−/− *Mice with and without Training*

In order to evaluate training effects on reperfusion recovery, 18-week-old ApoE−/− mice that had been fed with an HFD for 12 weeks and 12-week-old healthy C57BL/6 mice were subjected to the experimental protocols shown in Figure 3a,b. Both mouse strains were randomly subdivided into training or control groups. The control group was housed in motion-restricting cages. The training group was held in single cages containing a free-to-access running wheel starting 7 days prior surgery. FAL was performed on day 0 and perfusion of the hind-limb was measured by LDPI pre- and postoperatively and on postoperative days, d3, d7, and d14. Adductor muscle tissue was harvested from ApoE−/− mice at different time points following FAL. The ratio of hind-limb perfusion in the operated leg to that in the non-operated leg dropped to 10% on average in C57BL/6 mice (training: 11 ± 4%, *n* = 12, control: mean 8 ± 1%, *n* = 16; *p* > 0.05), whereas perfusion recovery increased similarly within 14 days in the control (69 ± 10%) and training group (69 ± 4%) (Figure 3c).

In contrast, exercising ApoE−/− mice showed a significantly faster approximation of perfusion with training (*n* = 19). A maximum of 78 ± 8% perfusion of the hind limbs could be reached within 7 days postoperatively, compared with the control group (*n* = 21) averaging 53 ± 3% (*p* = 0.012) (Figure 3d,e). Interestingly, immediately postoperatively ApoE−/− mice showed a baseline perfusion of 25% independent of a training adaptation that was significantly higher than that of C57BL/6 mice (*p* = 0.001) (Figure 3f).

**Figure 2.** Plaque development in (**a**) the femoral artery, (**b**) the aortic root, and (**c**) the collateral artery of ApoE−/− mice following 12 weeks of high fat diet, as visualized by Oil-red-O staining. (**d**) Daily running distances of C57BL/6 mice (BL6, circles) compared to ApoE−/− mice (ApoE, squares) over time, pre- and post-ligation of the femoral artery (FAL). Arrows indicate short term running breaks due to anesthesia during perfusion measurements.

**Figure 3.** Functional effects on hind limb perfusion in response to training. (**a**) Schematic of experimental setup in C57BL/6, (**b**) schematic of experimental setup in ApoE−/<sup>−</sup>, and (**<sup>c</sup>**,**d**) laser Doppler perfusion imaging in C57BL/6 and ApoE−/− as indicated. Data are expressed as ratio of the operated leg to the non-operated leg and represent mean ± SEM. Open symbols show the data of the training group whereas filled symbols represent the control group. As a statistical test, the unpaired t-test was used; \* *p* < 0.05. (**e**) Representative laser Doppler perfusion images indicate the effect of training in the operated hind limb when compared to the control group. (**f**) Postoperative perfusion ratio (R/L). Open symbols show the data of the training group whereas filled symbols represent the control group. As a statistical test, the one-way ANOVA was used; \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.005; n.s. not significant.

### *2.4. Increased Accumulation of Macrophages after Training in ApoE*−/− *Mice*

In order to investigate the beneficial influence of exercise training on the vascular remodeling process in ApoE−/− mice, adductor muscle tissue was harvested from these mice 21 days following FAL. Early perfusion benefits were not reflected by morphometric examination at the end of the experimental period. There was no difference of the size of the wall area between training and control groups on day 21 post-surgery (2.15 ± 0.53 mm<sup>2</sup> and 2.46 ± 0.53 mm2, respectively; *p* > 0.05; Figure 4a,b).

In the initial phase, collateral growth is critically driven by pericollateral macrophage assembly. Therefore, adductor muscle tissue was harvested from ApoE−/− mice 3 or 7 days following FAL and the macrophage number was quantified in the vascular nerve sheath of the collateral vessels. Seven days after FAL the number of macrophages in close proximity of growing collaterals was significantly higher in the training group with an average of 3.9 ± 0.8 compared to the control group with an average of 2.3 ± 0.4 (*p* = 0.042; Figure 4c,d).

**Figure 4.** Histological evaluation of collateral growth in cross sections of the adductor muscles of ApoE−/− mice in response to training. (**a**) Representative micrographs of collateral arteries for morphometry. Scale bar: 200μm. (**b**) Quantification of wall area 21 days after FAL of the training and control group. (**c**) Immunostaining to determine macrophage accumulation around collaterals 7 d after FAL of the training and control groups. Representative images of CD68 (green) and αSMA (red) immunostaining. Blue staining indicates nuclei and scale bars are 25 μm. (**d**) Quantification of macrophage number. Data are expressed as mean ± SEM of three collateral cross-sections per mouse of at least three mice per group (*n* ≥ 3). As a statistical test, the unpaired *t*-test was used; \* *p* < 0.05.
