Repeated CXCR4 Blockade by Plerixafor Attenuates Transplant Vasculopathy in Murine Aortic Allografts

Transplantation remains the definitive treatment for many structural and ischemic heart diseases (1, 2). In 2022, a major advance in transplantation medicine was made with the first successful xenotransplantation to treat a patient with nonischemic cardiomyopathy (3). A major problem is the shortage of suitable human donor organs. However, even after successful heart transplantation (HTx), complications can limit long-term clinical results. The main culprit in transplantation, whether from a xenogeneic or human donor, remains the recipient’s immune system, attacking the graft. Cardiac allograft vasculopathy (CAV), the consequence of chronic rejection is the long-term limiting factor leading to luminal stenosis and subsequent graft failure (4).

Endothelial injury with upregulation of adhesion molecules leads to increased migration of T cells, dendritic cells and macrophages into the vessel wall. The release of proinflammatory cytokines such as IL-6, TNF-α, and IFN-γ promotes smooth muscle cell (SMC) proliferation, leading to neointima formation (5–7). Risk factors such as advanced age, sex, obesity, diabetes, or hypertension also contribute to the ongoing rejection and exacerbate proliferation (8, 9). Although some risk factors cannot be changed, it is clear that comorbidities should be kept under control to prevent avoidable complications and reduce the overall risk of chronic rejection. Another essential component of post-transplant care is immunosuppressive therapy. The current standard of care is a combination of a calcineurin inhibitor (cyclosporin or tacrolimus) with mycophenolate mofetil and cortisone (10). Statins, the pleiotropic effects of which have been shown to reduce the activity of cytotoxic T cells, may be added in support (11). Unfortunately, these medications are not able to prevent CAV. Ten years after HTx, cardiac allograft vasculopathy can be found in ∼50% of the patients, which makes it one of the most common causes of death after HTx (2).

Plerixafor (AMD3100), currently approved for stem cell mobilization, has been shown to expand the pool of hematopoietic stem cells (HSCs) but also the repertoire of WBCs including regulatory T cells (Tregs) in the peripheral blood (12, 13). The appropriate mechanism is that the cells are kept in the bone marrow (BM) through the interaction of the chemokine receptor CXCR4 on HSCs and Tregs, with its ligand CXCL12, which is abundantly expressed in the bone marrow (14, 15). Disruption of this interaction with a CXCR4 antagonist leads to an expulsion of cells from their BM niche into the periphery. In addition to its role in maintaining homeostasis, the CXCR4 is involved in leukocyte chemotaxis toward sites of inflammation (16), and the CXCR4/CXCL12 axis has been implicated in the modulation of cell adhesion molecules and angiogenesis (17–20). Cell adhesion plays a substantial role in the pathogenesis of CAV: both upregulation of cell adhesion molecules and repair of tissue damage by neoangiogenesis after ischemic injury promote inflammatory cell infiltration. Because no current treatment can effectively prevent or reverse CAV, further research into potential immunomodulatory drugs is still highly warranted.

The aim of our study was to investigate how a reversible CXCR4 blockade affects the chronic rejection in our mouse aortic transplant model. Therefore, we aimed to assess the degree of immunological processes and to analyze the effect of previously reported Treg mobilization via plerixafor in a transplant setting (12). The objective of our study was to evaluate the effects of a CXCR4 antagonism on chronic allograft rejection in a murine model of aortic transplantation.

All animal experiments were conducted in accordance with the demands of German law and authorized by the responsible government (Regierung von Unterfranken, Würzburg, Germany) under the license 55.2.2-2532.2-1060. Date of approval was January 7, 2020.

C57BL/6 (H2b) and CBA.J (H2k) mice were originally purchased from Janvier (Le Genest-Saint-Isle, France). C57BL/6 mice were used as donors, and CBA.J mice were used as recipients of the aortic allografts. All of the mice used in this study were aged between 8 and 12 wk at the time of experimental use and were maintained at the Preclinical Experimental Animal Center (PETZ) at the University of Erlangen–Nürnberg under specific pathogen-free conditions and treated in accordance with institutional and state guidelines. We used male and female littermate mice in equal numbers within each group. The mice were maintained under a 12-h light/12-h dark cycle with free access to water and standard mouse diet.

The procedure was performed using a modified technique initially described by Koulack et al. (21). In brief, the thoracic aorta of the donor was isolated, resected, and transferred to the recipient animal. The abdominal aorta of the recipient was clamped and then transected with sharp microvascular scissors. A size-matched piece of the donor aorta was inserted via a proximal and distal end-to-end anastomosis with single interrupted 11/0 sutures. Strain combinations CBA.J to CBA.J and C57BL/6 to CBA.J were used for syngeneic and allogeneic transplantation, respectively.

Plerixafor (TargetMol TMO-T1776L, Wellesley Hills, MT) was obtained from Hölzel Biotech (Köln, Germany) and dissolved in PBS solution. An ALZET osmotic pump (Durect Corporation, Cupertino, CA) was used for continuous application in the isograft control, allograft control, and P-A groups. With an average pumping rate of 0.21 µl/h (± 0.01), the mice received a total of 101 µl (± 1.65) of saline or 1 mg plerixafor per kg body weight (BW), respectively, over a span of 14 d, assuming an average weight of 25 g. Groups P-A 1 × 1 and P-A 1 × 5 received one dose of 1 mg/kg BW and 5 mg/kg BW, respectively. Group P-A 4 × 1 received a pulsed regimen consisting of four single doses of 1 mg/kg BW on days 0, 7, 14, and 21 after transplantation.

Transplantation (Tx) was performed on day 0 with simultaneous intra-abdominal pump insertion for the isograft control, allograft control, and P-A groups. To prevent osmotic irritation, the pumps were removed via a small abdominal incision in short general anesthesia on day 15 after drug delivery was completed. Successful drug delivery was confirmed by measurement of residual pump volume. Groups P-A 1 × 1 and P-A 1 × 5 received one i.p. injection on day 0 after successful operation. Pulsed application of plerixafor in group P-A 4 × 1 was also performed i.p. on days 0, 7, 14, and 21 after Tx. Specimen harvest was performed on day 14 for analysis of intragraft gene expression and cell populations by FACS in peripheral blood and bone marrow. Specimen harvest on day 30 was used for histological and immunohistochemical analysis, as well as FACS measurements (Fig. 1).

Recovery of aortic grafts was carried out on day 30 under total anesthesia. This time point has proved to show sufficient formation of vasculopathy for comparative analysis, because a delayed harvest after 60 d showed only marginally increased vessel occlusion in the vessel in fully allogeneic aortic allografts compared with those retrieved on day 30 as shown in a previous study (22). The grafts were flushed with saline and flash-frozen in optimal cutting temperature medium (Tissue-Tek, Sakura, Netherlands) in liquid nitrogen for morphometric analysis of 7-μm cryostat sections. A minimum of 10 transverse sections of each graft were analyzed. The sections were prepared with Elastica-van-Gieson stain and analyzed by two independent examiners blinded to the experimental conditions at an original magnification of ×40 by using a conventional light microscope. A digitized image of each section was captured, and the intraluminal area and the internal and external elastic lamina were manually circumscribed and measured as previously described (22). All image analyses were performed on a color monitor using cellSens software 1.18 (Olympus, Hamburg, Germany).

To quantitatively assess cell populations, flow cytometry and cell sorting were performed on the BDLSR Fortessa (BD Biosciences, San Jose, CA). PBMCs were prepared from blood drawn from the inferior vena cava during the harvest procedure. After erythrocyte lysis and centrifugation, the cells were suspended in PBS supplemented with 0.5 M EDTA and 2% FBS, centrifuged a second time, and finally suspended in PBS for FACS analysis. Bone marrow was extracted from both femur and tibia bones via centrifugation, pooled, and suspended in PBS solution and filtered for FACS analysis. The cells were incubated with the respective Abs for HSC and Treg analysis according to the manufacturer’s instructions and fixed with BD CellFix (BD Biosciences). The eBioscience FOXP3/transcription factor staining buffer set (Thermo Fisher Scientific, Waltham, MA) was used for permeabilization and fixation in Foxp3 staining. The cells were then washed with FACS buffer and underwent flow cytometry the following day. Subsequent analysis was carried out with FlowJo version 10.8.1 (BD Biosciences).

Identification of HSCs is usually defined by a characterization of lin⁻(CD3, CD11b, CD45R, Ter-119, Gr1)/Sca-1⁺/c-kit⁺ cells. However, the CBA mouse strain, used as recipients in our study, shows a lack of Sca-1 in hematopoietic progenitor cells. To characterize HSC adequately, independent of Sca-1, CD201 and CD27 were used. HSC are thus defined as lin⁻/c-kit⁺/CD201⁺/CD27⁺ (23). T regulatory cells were identified as CD3⁺/CD4⁺/CD25^high/CD127^low/Foxp3⁺ (Fig. 2A). The values are given as percentages of viable cells.

The following Abs and conjugates were used for FACS analysis: CD3-FITC, CD11b-FITC, CD45R-FITC, Ter-119-FITC, Gr-1-FITC, c-kit(CD117)-PE, CD201-APC, and CD27-PE/Cy7 for staining of hematopoietic stem cells and CD3-BV510, CD4-FITC, and CD25-PE or CD25-PE/Cy7 for staining of T regulatory cells were obtained from BioLegend (Amsterdam, Netherlands). Foxp3-APC was obtained from Thermo Fisher Scientific. All Abs in this study are listed in Supplemental Table I with the appropriate RRID number (https://rrid.site) for better traceability.

The following Abs and conjugates were used for immunohistochemical analysis: anti-CD4 and anti-CD8 were purchased from BD Bioscience (Heidelberg, Germany), anti -CXCR4 from Thermo Fisher Scientific. Anti-macrophage F4/80 was obtained from AbD Serotec (Düsseldorf, Germany). Anti-SIGLEC H for plasmacytoid dendritic cells (pDCs) was purchased from HycultBiotech (Uden, Netherlands) and smooth muscle actin (SMA) for SMCs and anti-neutrophile Ab from Abcam (Cambridge, UK). Endothelial cells were detected with CD31-FITC (BD Biosciences). We also used mouse anti-rat IgG Cy3 (Dianova, Hamburg, Germany) or Alexa Fluor 555 goat anti-rabbit (Life Technologies, Carlsbad, CA) as secondary Ab to detect the primary Abs staining. All Abs are listed in Supplemental Table I with the appropriate RRID number (https://rrid.site) for better traceability.

Standard staining protocol was used for 7-μm cryostat sections on gelatin-coated slides. After air drying and fixation in acetone for 10 min, the slides were rehydrated in staining buffer and preincubated in staining buffer (1× TBS buffer, with 0.05% Tween 20, pH 7.4) containing 5% heat-inactivated serum (Invitrogen, San Diego, CA) for 15 min. After incubation of primary Abs in a humidified chamber for 1 h at room temperature, the slides were washed three times with staining buffer. Detection of Ag was revealed with appropriate secondary Ab followed by incubation with the CD31-FITC for 1 h. After three final washes, the slides were mounted using Vectashield hard set mounting medium with DAPI (Vector Laboratories, Burlingame, CA) for nucleus staining and analyzed by epifluorescence microscopy (Olympus, Hamburg, Germany). Quantification of the intragraft cellular infiltrate at day 30 after transplantation was performed by computerized image analysis using cellSens software version 1.18 (Olympus, Hamburg, Germany). The positively stained area in the neointima in relation to the total neointima area of each section was analyzed using an original magnification of ×100.

On day 14 after transplantation aortic grafts were removed to analyze cytokine expression inside the graft. This time point proved to show peak expression of cytokines for comparative analysis (22). Grafts were flushed with sterile saline and stored in RNAlater (Qiagen, Hilden, Germany) at −80°C until further treatment. RNA was isolated from whole grafts. RNA isolation and cDNA synthesis were performed according to standard protocols. TaqMan PCR was performed in triplets by using the StepOne real-Time PCR system and TaqMan Gene Expression Master Mix (Applied Biosystems Forster City, CA). Oligonucleotide sequences have been published in previous studies (24–26) or were designed new with primer3 output: CXCR4: forward primer 5′-GGACCGGTACCTCGCTATTG-3′, reverse primer 5′-GGATCCAGACGCCCACATAG-3′, and probe 5′-ACGCCACCAACAGTCAGAGGCCA-3′; CXCL12: forward primer 5′-AGCCAACGTCAAGCATCTGA-3′, reverse primer 5′-TTTTCTTCTCTGCGCCCCTT-3′, and probe 5′-TGCCCTTCAGATTGTTGCACGGCT-3′; and HO1: forward primer 5′- CACGCATATACCCGCTACCT-3′, reverse primer 5′-CCAGAGTGTTCATTCGAGCA-3′, and probe 5′-TCCCGAACATCGACAGCCCC-3′. All oligonucleotides were synthesized by Eurofins Genomics (Ebersberg, Germany). To generate PCR standards, the respective PCR product was cloned into a TOPO cloning vector (Invitrogen, Karlsruhe, Germany). The exact identity of the cloned amplicons was backed up by sequence analysis. Standard curves with known concentrations of template copy numbers were used to determine the expression of the amplified target. In each experimental setup, we also analyzed the 18 S rRNA expression as housekeeping gene by TaqMan PCR. The samples were normalized against the expression of the housekeeping gene (the ratio between the copies of the target gene and copies of the 18 S rRNA divided by 100,000). The results are given as relative copy numbers.

The mice were bled from the vena cava, and blood was collected into 1.5-ml Eppendorf tubes, incubated for 1 h at room temperature, and then centrifuged at 10,000 rpm for 6 min. The recovered sera were stored at −20°C. Upon thawing, the sera were assayed for CXCL12 using the mouse SDF-1 α CXCL12 ELISA kit (catalog no. EMCXL12, Invitrogen, Karlsruhe, Germany) according to the manufacturer’s instructions. The OD was measured in triplicate at wavelengths of 450 and 540 nm with an ELISA reader (Multiscan FC, Thermo Fisher Scientific).

The results are given as the means per group with a SD, which was derived from the mean per graft. The data were analyzed using a two-tailed ANOVA followed by Bonferroni correction. A p value of ≤0.05 was considered as significant. A p value of ≤0.01 was considered as highly significant.

Due to previous results from Fu et al. (27) and preliminary dose-finding studies (data not shown), we decided to compare different application strategies with various dosages as described above and depicted in Fig. 1. During the dosage finding and the data analysis of this very study, we found no beneficial effects of continuous plerixafor treatment via the osmotic pump over 14 d. We therefore excluded the plerixafor pump group from the main data presentation to show results from the injection groups more clearly. The results from the plerixafor pump groups can be viewed in the supplemental data.

Looking at lin⁻/c-kit⁺/CD201⁺/CD27⁺ HSCs in the bone marrow on day 14 (Fig. 2B), the treated allograft groups, like the isograft control, show lower HSC values in the BM compared with the untreated allograft control. The ongoing immune process in allograft controls led to increased HSCs numbers in the bone marrow on day 14. Plerixafor treatment mobilized these cells back into the periphery. Fittingly, measurements of HSCs in the peripheral blood showed a significant increase in the group of pulsed injection with a substantial increase on day 14 compared with untreated allograft (0.0450 ± 0.0024%; p < 0.01 [pulsed] versus 0.0068 ± 0.0018% [allograft control]) (Fig. 2B). On the day of graft recovery, 30 d after transplantation, the levels of HSCs in the peripheral blood of treated animals returned to levels similar to untreated animals. Additionally, the HSC content in the blood of untreated mice that received an aortic allograft decreased clearly. In comparison, HSCs in the bone marrow saw an increase in all treated groups (apart from 5 mg/kg BW) on day 30, which could be explained as a reactive production after depletion through CXCR4-mediated mobilization (Fig. 2B).

With regard to alterations of T regulatory cells in the bone marrow and the peripheral blood, we could observe that Tregs were raised significantly in syngeneic grafts and in the groups of pulsed injection, as well as the one time injection with a concentration of 5 mg/kg BW in the bone marrow on day 14 (0.473 ± 0.067% [one-time 5 mg/kg BW], 0.273 ± 0.15% [pulsed]; p < 0.01 versus 0.045 ± 0.017% [control]) (Fig. 2B). Similarly, in both groups, raised concentrations were observed in the periphery on day 14. A minor, yet significant increase was observed after the single injection of 1 mg/kg BW, too (0.68%±0.11%; p < 0.05 [one-time 1 mg/kg BW], 1.04 ± 0.54%; p < 0.05 [one-time 5 mg/kg], 0.93 ± 0.23%; p < 0.01 [pulsed]; versus 0.47 ± 0.19% [allograft control]). On day 30 after transplantation, the data of Tregs in the BM appeared inconsistent, but we could detect decreased Tregs in the bone marrow in the P-A 1 × 5 and P-A 4 × 1 groups, as well as in the isograft control group compared with Tregs on day 14. Tregs in the peripheral blood return to concentrations comparable to the untreated allograft control group on day 30 after transplantation yet remain elevated when compared with syngeneic controls (Fig. 2B).

In our study, the animals treated with plerixafor over 14 d continuously (1 mg/kg BW per day) did not yield high levels of peripheral stem cells when compared with control and other treatment regimens. Although Treg levels were elevated 14 d after pump removal in the bone marrow, no change was observed in the peripheral blood on day 14 or day 30 when compared with untreated allograft controls (Supplemental Fig. 1A, 1B).

Although not specifically evaluated in human trials, in past animal studies a common side effect described after extended plerixafor administration is splenomegaly (HMP ASSESSMENT REPORT FOR Mozobil, European Medicines Agency, Doc.Ref.: EMEA/CHMP/303556/2009 from https://shlk.io/P3J8prar). To assess spleen size, the organ was removed along with the graft on day 14 or 30 after transplantation and weighed before further analysis. In our study, spleen weight did not increase in any of the injection groups (Fig. 2C).

One of the effects of plerixafor is also an egress of neutrophils from the BM and their appearance among others in the blood; thus, we stained for neutrophils in the spleen as well as in the aortic grafts. In the aortic grafts, we could not see any differences between the untreated allograft and the treated groups (Fig. 2D). In the spleen, there was a not analyzable difference within the groups independent of plerixafor treatment. We did not include these data in the article.

The effects of a blockade of the CXCR4 receptor on mRNA expression of cytokines, growth factors, and adhesion molecules within the grafts was analyzed on day 14 after transplantation (Fig. 3A). In summary, the expression of proinflammatory genes was reduced in the plerixafor-treated groups compared with the untreated allografts, whereas the expression of anti-inflammatory genes generally increased, especially in the pulsed injection group, as shown in Fig. 3B, which presents the main results in a cartoon (Fig. 3B).

To give more detail (Fig. 3A), a notable reduction of endothelial cells’ adhesion molecules like E-selectin and ICAM-1, as well as L-selectin, was achieved in the groups receiving injection of 5 mg and repeated injection of 1 mg plerixafor per kg of body weight. Regarding proinflammatory cytokines, IFN-γ was significantly reduced in every plerixafor-treated group. Granzyme B, commonly found in granules of cytotoxic T cells, was substantially reduced only in the groups that received an injection of 5 mg or repeated injection of 1 mg of plerixafor. Simultaneously, the immunomodulatory ligand PD-L1 (programmed death-ligand 1) was elevated in these groups.

Both of these groups also had altered expression of immunoregulatory genes like CTLA4 and Foxp3. TGFβ and IL-10, often secreted by T regulatory cells in their ability to counteract pathological immune response, were both significantly elevated in the group receiving repeated treatment. IL-6 acting either as proinflammatory cytokine or as an important factor in the differentiation of various T cell subsets was also significantly upregulated in this very group. In all treated groups, vascular endothelial growth factor (VEGF), a crucial mediator of hypoxic damage and neo-angiogenesis was significantly upregulated at 14 d post-transplantation. To account for the role of hypoxia HO1 (heme oxygenase 1), iNOS (inducible NO synthase) and eNOS (endothelial NO synthase) were analyzed. An upregulation of the former two genes in untreated allograft became even more pronounced in plerixafor-treated animals. Again, pulsed treatment showed a significant increase of both HO1 and iNOS. Although significant, slightly fewer changes were observed regarding eNOS. Animals treated with a pump over 14 d had no relevant changes in gene expression levels in comparison with allograft controls (Supplemental Fig. 2A–C).

Analyzing the effect of plerixafor on intimal proliferation of aortic allografts (Fig. 4), there was no improvement concerning the occlusion in the group receiving plerixafor as a single injection of 1 mg or 5 mg/kg BW compared with allografts of untreated mice. An insignificant increase was seen in allografts of mice receiving only one injection of 1 mg plerixafor/kg BW (53.01 ± 13.29 [5 mg/kg BW single injection]; 66.45 ± 10.8% [1 mg/kg single injection]; all p > 0.05 versus 53.13 ± 12.41% [control]). Pulsed injection, however, leads to a significant reduction when compared with the untreated allografts (33.65 ± 8.84%; p < 0.05 [pulsed application] versus 53.13 ± 12.41% [control]) (Fig. 4F). As expected, no intimal proliferation was present in transplanted isografts (Fig. 4A).

SMC accumulation is one important aspect of the development of a neointima. SMA was analyzed to quantify the amount of SMCs in the neointima. Administration of plerixafor over 14 d with an osmotic pump had no relevant effect on SMC proliferation (Supplemental Fig. 3A). On day 30 after transplantation, we observed a reduction of SMA in all groups receiving plerixafor irrespective of treatment regimen (15.16 ± 7.58% [1 mg/kg BW single injection]; 10.67 ± 3.05% [5 mg/kg BW single injection], all p < 0.05; 11.49 ± 6.28%, p < 0.01 [pulsed application] versus 18.57 ± 9.82% [control]) (Fig. 5A, panels I–V). The group receiving the pulsed injection showed a highly significant reduction of SMCs proliferation, fitting the data of intimal proliferation.

In addition to smooth muscle cell migration, heightened infiltration of immune cells is yet another important feature leading to an increase of neointimal volume. Inflammatory-type cells like CD4⁺ T cells, CD8⁺ T cells and macrophages, as well as pDCs, were quantified in the graft on day 30 after transplantation (Fig. 5B). Interestingly, the image that presented itself was a mixed one. In all treated groups, no matter the regimen, pDCs were decreased (Fig. 5B, panels XIV–XVII and XVIII; Supplemental Fig. 3A). Regarding the levels of macrophages, only minor changes, all of them on an insignificant level, were observed (Fig. 5B, panels IX–XIII and XVIII). When looking at CD8⁺ T cells, the only group showing a change at all was again the one receiving repeated injection. However, in contrast to CD4⁺ T cells, the CD8⁺ T cell levels were clearly elevated after 30 d when compared with the control group (4.98 ± 2.34% [1 mg/kg BW repeated injection]; p > 0.05 versus 3.29 ± 1.41% [control]) (Fig. 5B, panels V–VIII and XVIII). The number of CD4⁺ T cells was effectively cut in half in the specimen receiving a one-time injection of 5 mg, as well as a repeated injection of 1 mg. Both of these values are highly significant (1.5 ± 1.74% [5 mg/kg BW single injection]; 1.5 ± 1.21% [1 mg/kg BW repeated injection]; both p < 0.01 versus 3.84 ± 2.31% [control]). No change of CD4⁺ T cell infiltration was observed in the other test group (Fig. 5B, panels I–IV and XVIII). No significant change of CD4⁺, CD8⁺, and macrophage infiltration was observed in the animals having received plerixafor via pump over 14 d (Supplemental Fig. 3A).

To know whether plerixafor treatment also affects CXCR4 in the aortic allografts, we also conducted intragraft CXCR4 immunofluorescence staining. As shown in Fig. 5C, the intragraft protein content of the C-X-C chemokine receptor 4 was significantly reduced in all groups that received plerixafor treatment, although to varying degrees (4.58 ± 2.68% [1 mg/kg single injection] 1.69 ± 1.69% [5 mg/kg BW single injection]; 3.28 ± 1.91% [pulsed application]; all p < 0.01 versus 8.55 ± 5.72% [control]) (Fig. 5C, panels I–V). In contrast, intragraft gene expression levels of CXCR4 on day 14 did not show any significant alterations in comparison with allograft controls. All allograft groups showed a similar heightened CXCR4 expression when compared with syngeneic controls (Fig. 5D). However, the intragraft gene expression of the ligand CXCL12 was downregulated in groups with plerixafor compared with the untreated allograft and reached levels of the syngeneic control group (Fig. 5D). Only CXCR4 expression levels in the pump groups had increased by ∼50% on day 14 (Supplemental Fig. 3C).

Correspondingly, using ELISA, we determined the presence of CXCL12 in the serum on days 14 and 30. Serum levels of CXCL12 were reduced in all treated groups on day 14 with narrow differences versus control in the single 5-mg group (12.63 ± 1.62 [pg/ml; 5 mg/kg BW single injection] versus 13.09 ± 0.71 [pg/ml; control]) but significant lower in the other plerixafor-treated groups (10.35 ± 1.78 [pg/ml; 1 mg/kg single injection], and 8.21 ± 1.22 [pg/ml; pulsed application] both p < 0.01 versus [control]). On day 30, the CXLC12 level in the 1-mg single-injection group was unexpectedly increased compared with control (14.20 ± 1.13 [pg/ml; 1 mg/kg single injection] p < 0.01 versus 9.53 ± 2.63 [pg/ml; control]), but no difference was seen in the 5-mg and pulsed-application groups (11.04 ± 1.49 [pg/ml; 5 mg/kg BW single injection] and 11.81 ± 1.52 [pg/ml; pulsed application] versus [control]) (Fig. 5E). Although continuous application of plerixafor did not affect CXCL12 serum levels, CXCL12 expression inside of the graft was insignificantly decreased (Supplemental Fig. 3B, 3C).

To counteract the chronic rejection process in heart transplantation, we investigated the application of plerixafor, a novel agent in transplantation medicine, and its immunomodulatory effects on cardiac allograft vasculopathy. We sought to determine whether plerixafor is able to alter the effects of chronic inflammation in the context of cardiac transplantation and the development of CAV. We used the well established mouse aortic allograft model with different applications of plerixafor.

Our study provides evidence that plerixafor diminishes the effects of chronic allograft rejection in aortic grafts and may be further investigated as a possible treatment option for CAV. Observations of luminal occlusion 30 d after transplantation in the aortic allografts suggest a beneficial effect, especially from a pulsed injection of plerixafor, which is underscored by gene and protein expression data as well as FACS analysis. Immunological mechanisms during allograft rejection alone led to an increase in HSCs in the bone marrow. In our experiments, this increase was abolished regardless of the dose by plerixafor. All types of plerixafor application resulted in nearly the same amount of HSCs being washed out of the bone marrow on day 14 after aortic transplantation. However, corresponding higher numbers of HSCs were detected only in the peripheral blood of the pulsed application group on day 14 and an insignificant increase in the group receiving a single high dose. Only the group receiving a single low dose of plerixafor did not result in a significant mobilization of HSCs. Altogether, these observations in the blood were abolished by day 30. When regarded in light of human treatment protocols, a timed and repeated injection regimen is key to a promising HSC mobilization. Plerixafor is licensed to be administered daily for up to 4 days to achieve sufficient mobilization. Treatment with plerixafor alone has been used in experimental setups and achieved only moderate mobilization results (28). It is important to note that HSC mobilization with plerixafor in humans is only carried out in combination with G-CSF, which prevents a full comparison (29). The supercompensation of HSC production observed at day 30 in the bone marrow could be attributed to plerixafor’s ability to induce greater HSC division, as described in studies investigating hematopoietic recovery after stem cell transplantation (30, 31).

Plerixafor mobilizes not only HSCs but also leukocytes, and it can therefore be administered alone to boost leukocyte counts (lymphocytes, monocytes, and neutrophils) among patients with myelokathexis, a disorder affecting the circulation of leukocytes. It has been shown to boost the expulsion of leukocytic cells, leading to heightened neutrophil and lymphocyte counts (32). The role of neutrophils in the context of transplantation is complex, because they can both promote inflammation, whereas specific subsets, in contrast, may act protective on the graft (33). Although neutrophils may be heightened under plerixafor therapy, we could not detect any increased accumulation of neutrophils in the aortic grafts. One possible reason could be that neutrophils are not activated by plerixafor (34).

Another side effect of plerixafor treatment may be splenomegaly. This was not observed in either of the injection groups, although we noticed an increase in spleen weight at both time points in the group receiving plerixafor by osmotic pump for 14 d (data not shown).

There are other cell lines that emerge from the BM through plerixafor. Kean et al. (12) described the intense mobilization of T regulatory cells by plerixafor in a rhesus macaque model of leukapheresis. Although this mobilization was also observed in our study, it was only observable in the injection groups on day 14 after transplantation and disappeared over time in all plerixafor-treated groups. Although elevated levels of Foxp3, a commonly used marker for Tregs, detected by PCR in the groups receiving a single injection of 5 mg/kg BW and a repeated injection of 1 mg/kg BW, correlated with the described modulation of T regulatory cells, we assume that the Treg mobilization at the beginning is not sufficient to induce a sustainable graft tolerance, because there was no difference in circulating Tregs in the blood of the plerixafor-treated groups compared with the allograft control at day 30 post-transplantation. Furthermore, the Treg analyses in the aforementioned study (12) were performed at a maximum of 6 h following plerixafor administration, rendering direct comparison with our data unfeasible.

The chronic rejection in human organ transplantation after HTx as the ongoing alloimmune response is the main feature that induces SMC migration, immune cell infiltration with subsequent neointima formation, and is detrimental to long-term graft survival (35). Our data show that a pulsed injection of plerixafor on days 0, 7, 14, and 21 results in a reduction of inflammation and thus a reduction of neointimal development in this model of CAV 30 d post-transplantation. Immunohistochemical analysis showed a significantly reduced amount of C-X-C chemokine receptor 4 within the graft on day 30 after transplantation. This observation supports the notion that a blockade of CXCR4 not only inhibits the homing of leukocytes to the bone marrow but also inhibits their recruitment into the graft. The pleiotropic and complex CXCR4 function makes it difficult to determine the exact underlying mechanism (16), but blocking CXCR4 in our aortic transplant model directly results in a reduced infiltration of pDCs and CD4⁺ T cells into the allografts with a decrease in inflammation. Although there was no change in intragraft gene expression of CXCR4 on day 14 in any group, by day 30, treated groups exhibited a reduction in protein expression of CXCR4. Combining this observation with results from previous studies, it seems that the infiltration of cells occurs mainly between days 14 and 30 after transplantation (36). Proliferation and migration of smooth muscle cells, which is crucial for the formation of the neointima, is also influenced by the interaction of CXCR4 with its ligand CXCL12 (also known as SDF-1) (37). Blockade of the ligand by Thomas et al. (38), similar to the blockade of the receptor in our setup, resulted in a reduced development of cardiac allograft vasculopathy. This underscores the role of the CXCL12/CXCR4 axis in smooth muscle cell proliferation and aligns with our results, because treatment with plerixafor markedly reduced neointimal smooth muscle cell immigration. Furthermore, IFN-γ, a principal cytokine implicated in the proliferation and migration of SMCs and T cells (7, 39), predominantly from the luminal side of the graft, shows increased expression in chronic allograft rejection. However, this expression was significantly reduced in the plerixafor-treated groups. This was supported by reduced levels of the cell adhesion molecules L-selectin, E-selectin, and ICAM-1, which are critical for lymphocyte adhesion on the sites of inflammation and explain the observation that particularly fewer CD4⁺ T cells and dendritic cells were found in the allograft (40).

Cell-mediated immunity has been shown to play a key role in the process of chronic rejection, in which alloreactive CD4⁺ T cells promote a prolonged vascular inflammatory and fibroproliferative response (5, 41). In conclusion, the reduced gene expression especially in the group receiving a single injection of 5 mg and with a pulsed application corresponded to a lower infiltration of T helper cells in the graft in both groups. Adding to this, a reduced infiltration of plasmacytoid dendritic cells was also observed.

Plasmacytoid dendritic cells are characterized by their high secretion of type I IFNs, have been confirmed to play an important role in a variety of immunological responses, and have even been implicated in the pathogenesis of autoimmune diseases (42, 43). Previous studies have also shown that pDCs influence the development of atherosclerosis by promoting early inflammation (44, 45). In contrast, results from a murine model of allograft transplantation suggest that pDCs are a key mediator of the regulatory response associated with allograft rejection, thereby prolonging graft survival (27). However, we assume restricted immunological response of pDCs due to plerixafor because we clearly found fewer pDCs in the neointima of treated mice compared with allograft control and consider it to be a positive effect because it leads to lower neointima formation. In a study conducted by Fu et al. (27), only a high dose of plerixafor in combination with the mTOR inhibitor rapamycin led to a regulatory response of pDCs.

Surprisingly, the macrophage infiltration was not affected regardless of treatment dose. The CD8⁺ T cells showed a substantial increase in infiltration in animals that received repeated injections of 1 mg/kg body weight of plerixafor. This upregulation would be consistent with further observations by Kean et al. (12) that plerixafor also mobilized CD8⁺ T cells in the periphery. This observation was made despite a significant lower expression of granzyme B on day 14. In case of pulsed plerixafor application, the last administration was carried out on day 21 after Tx. The elevated immigration of CD8⁺ T cells on day 30 in this group compared with the animals having received a single injection 30 d prior to analysis is potentially attributable to these different intervals since the last drug administration. Furthermore, because the infiltration of cells in this mouse aortic transplant model occurs mainly between day 14 and day 30 after transplantation, a mobilization of CD8⁺ T cells between days 14 and 30 is thus potentially more likely in this time frame. Nevertheless, the occlusion of grafts treated with repeated injections of plerixafor is markedly diminished, thereby indicating that the advantageous effects of plerixafor are more pronounced than the heightened infiltration of CD8⁺ T cells.

Another CAV promoting factor is the VEGF. CXCR4–CXCL12 interaction normally leads to an increase in VEGF-mediated angiogenesis, as shown in a tumor growth model (46). VEGF upregulation in turn promotes inflammation through neo-angiogenesis and enhanced endothelial permeability and has been shown to play a detrimental role in the development of CAV (47–49). Although a blockade of CXCR4 should result in a reduction of VEGF expression, all treated groups showed elevated levels of VEGF expression (50). Additional intragraft gene expression analysis of HO1 (heme oxygenase 1), iNOS (inducible NO synthase), and eNOS (endothelial NO synthase), which are upregulated in states of hypoxia, confirms the detrimental amplification of hypoxic damage in allogeneic grafts. Plerixafor amplifies the expression of these protective genes, especially in the pulsed injection group. All enzymes are primarily characterized as protective but also promote chronic inflammation, because an upregulation is a sign of tissue injury as well as than attempt to protect the transplant (51–53). Although the allogeneic immune reaction exacerbates hypoxic damage, we hypothesize that the additional increase (when compared with untreated allografts) observed in the plerixafor-treated groups might be protective. Pharmacological intervention targeting the processes leading to hypoxic graft damage, like inhibition of VEGF expression, may prove to be another point of attack for the future.

Although the modifying effects of plerixafor on chronic rejection were confirmed in this model, these effects alone were not sufficient to counteract the allograft rejection entirely. Our measurements showed a downregulation of inflammatory processes in some aspects, yet an increase in others. Previous studies have confirmed that a CXCR4 blockade in combination with low-dose immunosuppressants such as mycophenolate mofetil has a superadditive immunosuppressive effect and can induce long-term acceptance of solid organ allografts in rat and porcine models of heart transplantation (54–56). In another model, Fu et al. (27) showed that mTOR inhibition through rapamycin in combination with plerixafor prolonged graft survival through plasmacytoid dendritic cell-mediated immune regulation of T regulatory cells. Therefore, plerixafor treatment alone might not be sufficient to adequately inhibit CAV development.

An important limit of the current study lies in the time points used for analysis. Because this study aimed to investigate the general effects of plerixafor in a setting of chronic rejection, we chose two well established time points for the analysis of the process: days 14 and 30 after transplantation (22, 36). In summary, a pulsed injection of plerixafor on days 0, 7, 14, and 21 seemed most beneficial to counteract chronic allograft rejection. It is worth mentioning that two fixed points of analysis and differing treatment schemes resulted in different intervals since last time of drug administration. Although the general effects of plerixafor on chronic rejection were observable this way, to better analyze and compare the effects of plerixafor on different cell subsets including the underlying mechanistic (e.g., cell mobilization) processes, further studies are required. In relation to this study, analysis relative to the last time of drug administration would be beneficial.

In summary, a pulsed injection of plerixafor on days 0, 7, 14, and 21 seemed most beneficial to counteract allograft rejection. Although the effects of treatment cannot fully induce allograft tolerance, results from flow cytometry, quantitative PCR and immunohistochemistry support the beneficial effects of a repeated CXCR4 antagonism as confirmed by morphometric graft analysis. We can conclude that plerixafor has an attenuating effect in this model of allograft rejection. Nevertheless, further investigation of the modulatory properties of plerixafor still appears highly warranted, particularly with respect to potential synergistic effects in combination with other immunosuppressive agents.

The authors have no financial conflicts of interest.

We thank Valentina Rabinovich (Cardiac Surgery) and Lina Meretuk and Lena Tarantik (Core Unit Cell Sorting and Immunomonitoring Erlangen) for the excellent technical assistance. Furthermore, many thanks to Professor Stephan von Hoersten and the staff of the animal facility of the University of Erlangen-Nürnberg for their expert care of animals used for this study. The current work was performed in fulfillment of the requirements for obtaining the doctor of medicine degree by Frank Theil.