**3. Results**

#### *3.1. MALP-2 Improved Perfusion Recovery and Collateral Growth in the Hind Limb Following FAL in Hypercholesterolemic Apoe-Deficient Mice*

Based on our previous findings [14–16], we hypothesized that MALP-2 is capable of promoting collateral growth. To analyze the functional e ffects of systemic MALP-2 application in this regard, the mouse FAL model was applied sequentially to two di fferent wild-type mice strains (C57BL/6 and BALB/c) and additionally to Apoe-deficient mice (Apoe-KO) on a high fat diet (HFD) for 12 weeks. Laser Speckle perfusion measurements were performed prior to and after surgery as well as on days 3 and 7 and, for Apoe-KO mice, on day 10. Following the left FAL, the ratio of left hind limb perfusion compared to that of the hind paw of the non-ligated right site dropped to less than 25% in all groups (Figure 1a). The perfusion recoveries of C57BL/6 and BALB/c wild-type mice which received MALP-2 or PBS (control) were found to be similar on day three and day seven post FAL (Figure 1a). However, MALP-2 significantly improved the perfusion recovery of hypercholesterolemic Apoe-KO mice on day three post FAL. The beneficial e ffect of MALP-2 on perfusion recovery was limited to early time points and returned to control conditions on day 10 post FAL (Figure 1a,b). Since the functional improvement of MALP-2 in the FAL model was limited to Apoe-KO mice on a HFD, we concluded that hypercholesterolemic conditions with compromised vascular functions are required for the observed beneficial e ffects of MALP-2; we therefore focused on this model in the following analysis.

The remodeling of the collateral arteries was verified by morphometry in cross sections of the left adductors 10 days after the FAL. The MALP-2 application significantly increased the collateral inner diameter as well as the collateral wall area, thus documenting enhanced collateral growth with MALP-2 (Figure 1c). Since collateral growth is critically influenced by hemodynamic forces, we analyzed the atherosclerotic arterial plaque load in the experimental Apoe-KO mice after 12 weeks of the HFD diet. As expected, we detected plaques in the aortic root and in the thoracoabdominal aorta. However, the atherosclerotic plaque load was not di fferent between the control and the MALP-2-treated group (Figure S2a,b). Plaques in the femoral artery were only detected in rare cases. In addition, we investigated the collateral arteries, which were found to be highly positive for Oil Red O, indicating lipid deposition in the collateral vascular wall in hypercholesterolemic Apoe-deficient mice (Figure S2c). As expected, this was not the case in parallel-performed control Oil Red O staining in collaterals from C57BL/6 mice (Figure S2c). However, atherosclerotic plaques were not detected, excluding the possibility that plaque morphology itself might influence the hemodynamics and thereby collateral remodeling and growth.

#### *3.2. MALP-2 Increased Pericollateral Macrophage Accumulation, Endothelial Cell Proliferation and Downstream Angiogenesis Following FAL*

In order to investigate the influence of MALP-2 on the vascular remodeling process, tissue from the adductor muscles was harvested from hypercholesterolemic Apoe-deficient mice 3, 7 and 10 days following the FAL. In the initial phase, the collateral growth is critically driven by pericollateral macrophage assembly and endothelial proliferation [4,6,18]. MALP-2 significantly increased the macrophage accumulation around the collateral artery on day three after the FAL compared to the control. There was no effect of MALP-2 at later time points (Figure 2a). Likewise, we detected significantly more proliferating endothelial cells in MALP-2-treated mice on days three and seven after the FAL. In general, no proliferating endothelial cells were detected on day 10 (Figure 2b). The stenosis or occlusion of a major arterial conductance vessel entails a reduced blood supply and subsequent ischemia in the downstream supply area. Angiogenesis with an increased capillary density is usually the counteracting adaptive process in tissue ischemia. Therefore, we investigated angiogenesis in the gastrocnemius muscle and found that MALP-2 increased the capillary density on day three and day seven. This was not different anymore on day 10 post FAL (Figure 2c). Our results indicated that the effects of MALP-2 on collateral growth occurred within the first seven days after FAL.

**Figure 1.** MALP-2 improved the perfusion recovery and collateral growth in the hind limb following femoral artery ligation (FAL) in hypercholesterolemic Apoe-deficient mice. (**a**) Following the FAL, the perfusion recovery was determined by laser Speckle perfusion imaging for C57BL/6, BALB/c and hypercholesterolemic Apoe-KO mice (12 weeks on a high fat diet (HFD)) treated with MALP-2 or PBS (control) pre/pos<sup>t</sup> the FAL, after three and seven days and, in Apoe-KO mice, after 10 days. Data are expressed as the ratio of the ligated and the non-ligated hind limb. \*\* *P* < 0.01, *N* = 4–7. (**b**) Representative laser speckle perfusion images indicate the effect of MALP-2 compared to the control (PBS) on perfusion recovery in the ligated hind limbs of Apoe-KO mice pre/pos<sup>t</sup> the FAL and after 3, 7 and 10 days. (**c**) Representative haematoxilin-eosin staining of cross sections of collateral arteries in the adductor muscle of the ligated and the non-ligated hind limbs of hypercholesterolemic Apoe-KO mice treated with MALP-2 or PBS (control) 10 days after the FAL and the corresponding morphometric analysis of the collateral diameter and wall area. Scale bar = 10 μm. \* *P* < 0.05 vs. control, *N* = 6–14 collaterals.

**Figure 2.** MALP-2 increased pericollateral macrophage accumulation, endothelial cell proliferation and downstream angiogenesis following FAL. This shows the representative immunostaining of cross sections of collateral arteries in the adductor muscle and the calf muscle of the ligated hind limb in hypercholesterolemic Apoe-KO mice treated with MALP-2 and PBS (control) 3, 7 and 10 days after the FAL and the corresponding quantitative analysis. (**a**) CD68 staining to assess the accumulation of macrophages around the collateral (α-SMA indicates the media of the collateral wall). Scale bar = 25 μm. (**b**) Ki67 staining to determine the portion of proliferating CD31-positive collateral endothelial cells (white arrow heads). Scale bar = 25 μm. (**c**) CD31 indicates capillary density in the calf muscle. Scale bar = 50 μm. \* *P* <0.05, \*\* *P* <0.01 vs. control, *N* = up to 20 collaterals, n.d. = not detected.

To explore which factors were potentially involved in the process of MALP-2-induced collateral growth, we stimulated tissue pieces of the adductor muscles from C57BL/6 mice ex vivo with MALP-2. A real-time PCR analysis revealed increased expression levels for the established mediators of collateral growth such as CC-chemokine ligand 2 (*Ccl2*) [19] and granulocyte macrophage colony-stimulating factor (*Gm-csf*) [20] as well as, for the general inflammatory markers, interleukin 1β (*Il-1*β) and tumor necrosis factor-α (*Tnf-*<sup>α</sup>, Figure 3a). Likewise, the corresponding protein in the supernatant was found to be enhanced (Figure 3b). Das et al. recently reported that the axis of C-X-C motif chemokine ligand 2 (CXCL12, also known as stromal cell-derived factor 1) and its receptor C-X-C motif receptor 4 (CXCR4) is relevant for the injury-induced cardiac collateral growth in neonatal mice and could also be induced by exogenous CXCL12 in adult mice [21]. However, MALP-2 did not induce *Cxcl12* expression ex vivo in the adductor muscle tissue (Figure 3c) or in the cultured MyEnd endothelial cells (Figure 3d), suggesting that this process did not play a role in MALP-2-dependent collateral growth.

#### *3.3. MALP-2 Improved NO-Dependent Vascular Relaxation and Enhanced Endothelial Cell-Derived NO Release*

Since endothelial dysfunction may limit collateral growth itself or the beneficial effects of collateral vessels on tissue perfusion, we assessed the effect of MALP-2 on vascular relaxation. To this end, we isolated mesenteric arteries from C57BL/6 mice to perform wire myography. MALP-2 significantly improved acetylcholine (ACh)-induced relaxation of phenylephrine (PE)-preconstricted (10 μM) mesenteric arteries (Figure 4a). To test for differences in endothelium-derived NO release, we inhibited endothelium-dependent hyperpolarization by depolarizing the vessels with high potassium

buffers (60 mM K<sup>+</sup>) and by inhibiting cyclooxygenases using indomethacin. Under these conditions, the relaxing responses to ACh could be entirely attributed to NO [22]. The MALP-2 treatment resulted in significantly increased endothelium-derived NO responses (Figure 4b). This effect completely disappeared when the endothelial nitric oxide synthase (eNOS) was additionally blocked with L-NAME (Figure 4c). Furthermore, MALP-2 also significantly improved the relaxation response in thoracic aorta (data not shown). These results demonstrated a crucial role for endothelium-derived NO in MALP-2-dependent vascular relaxation. Moreover, in MyEnd cells, MALP-2 led to a fast and transient increase in the protein kinase B (also known as AKT) phosphorylation (Figure 5a) and eNOS phosphorylation (Figure 5b) and consequently to an increased NO release (Figure 5c).

**Figure 3.** MALP-2 up-regulated inflammatory genes in the upper hind limb muscle. Tissue pieces of the adductor muscles of C57BL/6 mice were isolated and stimulated ex vivo with MALP-2 (1 μg/mL); *Ccl2*, *Gm-csf*, *Il-1*β and *Tnf-*α mRNA levels were analyzed after the indicated times by (**a**) real-time PCR and (**b**) the corresponding protein in the supernatant after 6 h by ELISA. *CXCL12* mRNA levels were analyzed (**c**) in tissue pieces of the adductor muscle of C57BL/6 mice ex vivo and in (**d**) MyEND cells following MALP-2 stimulation (1 μg/mL) after the indicated times by real-time PCR. \* *P* < 0.05, \*\* *P* < 0.01 vs. control, *N* = 4–6.

#### *3.4. MALP-2 Up-Regulated Endothelial Adhesion Molecules and Enhanced the Endothelial Adhesion of Monocytic Cells*

Arteriogenesis is a multi-faceted, highly coordinated process involving the endothelial adhesion of monocytes onto endothelial cells [4,6]. To explore the potential underlying mechanism responsible for the positive effects of MALP-2 on collateral growth after FAL, we conducted a series of in vitro experiments. In endothelial MyEnd cells, MALP-2 led to a strong transient increase in the mRNA levels of vascular cell adhesion molecule-1 (*Vcam-1*) after only 1 h (Figure 6a) and slightly delayed to

an increase in VCAM-1 protein levels (Figure 6b). The mRNA levels of the other major endothelial adhesion molecules, i.e., intercellular adhesion molecule-1 (*Icam-1*), *E-selectin* and *P-selectin*, were also increased between 1 and 3 h following the MALP-2 stimulation (Figure 6a). In addition, we investigated the mRNA expression of integrin receptors on monocytes/macrophages as counterparts to the endothelial adhesion molecules. Likewise, the mRNA levels of integrin α4β1 (very late antigen-4, *Vla4*), integrin αM (*Itgam*) and E-selectin ligand-1 (*Esl-1*) were slightly and transiently increased in the monocyte/macrophage cell line J774A.1 by MALP-2 (Figure S3). Consequently, the pretreatment of a monolayer of MyEnd cells with MALP-2 almost doubled the number of adherent J774A.1 cells (Figure 6c).

**Figure 4.** MALP-2 improved NO-dependent vascular relaxation in the mesenteric arteries of C57BL/6 mice. (**a**) The relaxation response to acetylcholine (ACh 0.001–10 μM) during phenylephrine-induced (PE, 10 μM) contraction in mesenteric arteries incubated with MALP-2 or PBS (control), *N* = 6. (**b**) The relaxation response to ACh (0.01–10 μM) during K<sup>+</sup>-induced (60 mM) contraction in mesenteric arteries incubated with indomethacin (10 μM, COX-inhibitor) and MALP-2 or PBS, *N* = 3. (**c**) The relaxation response to ACh (0.001–10 μM) in the presence of L-NAME (100 μM, NOS inhibitor) and indomethacin (10 μM). A.U.C. = area under the curve, \* *P* < 0.05 vs. control, *N* = 3.

**Figure 5.** MALP-2 enhanced the endothelial cell-derived NO release. MyEnd cells were stimulated with MALP-2 (1 μg/mL); (**a**) the AKT phosphorylation (p-AKT) as well as (**b**) the eNOS phosphorylation (p-eNOS) were analyzed after the indicated times by Western blot and (**c**) the NO release was analyzed with the Griess reagent. The numbers between panels indicate fold-change vs. unstimulated after normalization to total AKT or eNOS, respectively. β-Actin was used as the loading control. \* *P* < 0.05 vs. control, *N* = 4–5.

**Figure 6.** MALP-2 up-regulated endothelial adhesion molecules and enhanced the endothelial adhesion of monocytic cells. (**a**) The MyEnd cells were stimulated with MALP-2 (1 μg/mL) and the VCAM-1, ICAM-1, E-selectin and P-selectin mRNA levels were analyzed after the indicated times by real-time PCR. \* *P* < 0.05, \*\* *P* < 0.01 vs. control, *N* = 6–8. (**b**) The MyEnd cells were stimulated with MALP-2 (1 μg/mL) and the VCAM-1 protein expression was analyzed after the indicated times by Western blot. β-Actin was used as the loading control. The numbers between panels indicate fold-change vs. unstimulated after normalization to β-Actin. \* *P* < 0.05 vs. control, *N* = 4–5. (**c**) Fluorescence images depicting calcein-AM-labeled J774A.1 cells on a MyEnd monolayer with or without pretreatment with MALP-2 (1 μg/mL) for 6 h with an additional adhesion time of 1 h and the corresponding quantitative analysis. Pictures before and after washing are shown. Scale bar = 100 μm, \*\* *P* < 0.01 vs. control, *N* = 3.

## **4. Discussion**

Atherosclerosis, as a chronic inflammatory arterial disease, contributes to the major mortality of cardiovascular diseases worldwide. On the one hand, this is due to acute events such as myocardial infarction and strokes [23], but on the other, this is due to progressive lumen stenosis, which is the main trigger for adaptive arteriogenesis [4–6,18].

In this regard, growing collaterals represent a naturally occurring adaptive bypass system to avoid tissue ischemia. Well-developed collaterals, despite significant stenosis or even the occlusion of major coronary or peripheral arteries, could be the reason why some patients stay asymptomatic over a long period of time [4]. However, collateral growth is usually not sufficient to protect patients against ischemia for all their lives and thus therapies supporting this process are desirable. The model used in this study was a model of hind limb ischemia in mice. Critical limb ischemia represents the most severe form of PAD in patients [24,25]. The highly deadly disease is characterized by pain during walking and even at rest, as well as non-healing ulcers in the lower extremities. If the extent of the femoral artery occlusion due to advanced atherosclerosis becomes too large for percutaneous or surgical interventions, limb amputation remains the only treatment option. Catheter-based angiographic interventions or surgical bypasses are basically emergency procedures for the revascularization of the main artery in order to restore limb perfusion. Similar to these interventions, novel therapies such as cell-based or molecular therapies normally do not promote collateral growth [24,25]. Studies addressing therapeutic arteriogenesis are rare. Some of those investigated the potential of GM-CSF, identified in a rabbit model [20], with different outcomes in patients with coronary artery disease [26] or PAD [27]. Finally, the therapeutic improvement of collateral growth in cardiovascular patients hardly plays a role in clinical practice at present. In the current study, we used the lipopeptide and TLR2/6 ligand MALP-2 to investigate therapeutic arteriogenesis. Over the past few years, we had already demonstrated the high potential of MALP-2 to promote vascular regeneration, such as angiogenesis [14] and endothelial regeneration after vascular wounding [15]. We now identified the possible application of MALP-2 to promote arteriogenesis and uncovered the potential underlying mechanisms. We found that MALP-2 functionally improved perfusion recovery in the hind limb by enhanced collateral growth. The increase in the collateral lumen diameter was driven by augmented pericollateral macrophage accumulation and enhanced endothelial cell proliferation. MALP-2-enhanced the NO release of endothelial cells and improved NO-dependent vasorelaxation as well as endothelial adhesion molecule expression and subsequent monocytic cell adhesion. We had already reported enhanced secretion of GM-CSF from endothelial cells of various origin following MALP-2 stimulation [14,15]. Since the beneficial effect of GM-CSF on collateral growth has already been proven in animal experiments [20] and clinical studies [27], it is conceivable that the observed beneficial effect of MALP-2 on collateral growth is dependent on growth factors such as GM-CSF as well. Of note, we did not see any beneficial effects of MALP-2 application in two commonly used wild-type mouse strains—neither in C57BL/6 mice nor in BALB/c mice, which have known differences in cardiovascular regeneration [28]. As we saw the functional and morphological changes upon MALP-2 treatment that were summarized above exclusively in Apoe-deficient mice on a HFD and not in wild-type mice, we concluded that hypercholesteremic conditions are required for the beneficial MALP-2 effects on arteriogenesis. This conclusion was supported by the observation—to our knowledge, for the first time—that the collaterals were already positive for Oil Red O in this model. The staining demonstrated lipid deposition in the vascular wall of the collaterals, indicating vascular dysfunction. Ultimately, the mouse model used—with compromised vascular function and advanced atherosclerotic plaque load in larger arteries— approximately reflects the situation of cardiovascular patients.

In order to optimize the application route of MALP-2, we tested different variants. Initially, our intention was to choose an application route to bring MALP-2 as close as possible to the pre-existing collaterals after ligation. Therefore, we injected MALP-2 divided into small quantities into the Musculus adductor near to the collaterals. However, at the sites of injection, the tissue was affected in such a manner that subsequent histological analyses were not possible anymore. In addition, we tried to

inject directly into the femoral artery proximal to the ligation. This application route proved di fficult due to the small dimensions of the vessel. Since we observed increased mortality after the operation, we refrained from using this method. In the end, we chose the widely used intravenous application route (tail vein) for the MALP-2 injection, knowing that the lipophilic substance would be partially absorbed by the endothelium and that only small amounts would enter the target area of the collaterals. Although our approach was successful, there is still room to improve application strategies to bring MALP-2 into close proximity to the collaterals, e.g., in a biodegradable intra-arterial matrigel deposit or similar.

The potential limitations of our study are the same as those that generally apply for experimental studies in mice. The ligation of the femoral artery induces the growth of pre-existing collateral arteries and is therefore widely accepted as a reliable model for arteriogenesis. However, the vascular dimensions and related hemodynamic forces are di fferent to the situation of cardiovascular patients. To substantiate our findings for a potential therapeutic use in promoting collateral growth, experiments in higher animals are needed. In regard to therapeutic angiogenesis, this has been already done in a sheep model of tissue engineering [16]. Moreover, we used just one single dose of MALP-2 (1 μg/mouse) as this was proved to be e ffective in a previous in vivo study by our group [15]. Dose-response experiments would maybe reveal an even more e ffective dose. However, based on the data already published, our local animal authorities did not approve dose-finding experiments in this study.

Seemingly, TLR2/6 signaling is particularly suitable in promoting vascular regeneration and adaptation. This is not only documented by our studies [14–16]. Indeed, other TLR2 ligands, such as bacterial peptidoglycan [29] or the proteoglycan versican as an endogenous ligand [30], have been shown to induce angiogenic factors. Likewise, endogenous lipid oxidation productions are capable of promoting angiogenesis [31]. The common principle of our studies is a single bolus injection of MALP-2 to transiently increase inflammation, which could be considered an immunological mechanism to promote regeneration and adaptation. In contrast, long-term application of MALP-2 led to increased circulating inflammatory markers and increased atherosclerosis [32].

In summary, we identified a novel property of the lipopeptide and TLR2/6 ligand MALP-2 to restore blood flow recovery by enhanced collateral growth with possible implications for therapeutic arteriogenesis (Figure S4).

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4409/9/4/997/s1, Figure S1: Characterization of MyEnd cells, Figure S2: Vascular lipid deposition in hypercholesterolemic Apoe-KO 7 days after FAL and MALP-2 or PBS (control) injection, Figure S3: MALP-2 up-regulates mRNA expression for integrin receptors in J774A.1 cells, Figure S4: Mode of action of MALP-2 in arteriogenesis.

**Author Contributions:** Conceptualization, K.T. and K.G.; methodology and validation, K.T., C.S., A.-K.V., R.C., S.K., J.S., R.O., K.G.; formal analysis and investigation, K.T and K.G; writing—original draft preparation, K.G.; writing—review and editing, K.T., W.S., T.S.-R., B.S.; supervision, project administration and funding acquisition, K.T. and K.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Anna-Maria and Uwe Karsten Kühl foundation (T188/30462/2017), the B. Braun foundation (BBST-D-18-00021) and the von Behring-Röntgen foundation (62-0002).

**Acknowledgments:** We thank Daniela Beppler, Silke Brauschke, Brigitte Matzke and Christina Reschke for excellent technical assistance.

**Conflicts of Interest:** The authors declare no conflict of interest.
