Limited Impact of Imatinib in a Murine Model of Sclerodermatous Chronic Graft-versus-Host Disease

Background Sclerodermatous chronic Graft-versus-Host Disease (scl-cGVHD) is one of the most severe form of cGVHD. The Platelet-derived Grotwth Factor (PDGF) and the Transforming Growth Factor-β (TGF-β) play a significant role in the fibrosing process occurring in scl-cGVHD. This prompted us to assess the impact of the PDGF-r and c-Abl tyrosine kinase inhibitor imatinib on scl-cGVHD. Methods To assess the impact of imatinib on T cell subset proliferation in vivo, Balb/cJ recipient mice were lethally (7 Gy) irradiated and then injected with 10x106 bone marrow cells from B10.D2 mice on day 0. Fourteen days later, 70x106 carboxyfluorescein succinimidyl ester (CFSE)-labeled splenocytes from B10.D2 mice were infused and imatinib or sterile water was administered for 5 days. To induce severe scl-cGVHD, Balb/cJ mice were injected i.v. with 10.106 bone marrow cells and 70.106 splenocytes from B10.D2 donor mice after 7 Gy irradiation. Mice were then given sterile water or imatinib from day +7 after transplantation to the end of the experiment (day +52). Results Imatinib decreased the proliferation of total T cells (P = 0.02), CD8+ T cells (P = 0.01), and of regulatory T cells (Tregs) (P = 0.02) in the spleen. In the severe scl-cGVHD model, imatinib-treated mice had significantly lower levels of PDGF-r phosphorylation than control mice on day 29 after transplantation (P = 0.008). However, scl-cGVHD scores were similar between vehicle- and imatinib-treated mice during the whole experiment, while there was a suggestion for less weight loss in imatinib-treated mice that reached statistical significance at day +52 following transplantation (P = 0.02). Conclusions Imatinib had a limited impact in murine scl-cGVHD despite significant inhibition of PDGF-r.


Introduction
Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the main curative treatment for many hematological malignancies [1]. Its anti-tumor activity relies in large part on immune-mediated graft-versus-tumor effects (GvT effects) [2,3]. However, donor immune cells contained in the graft can also attack healthy host tissues causing graft-versus-host disease (GVHD) [4][5][6][7]. GVHD can be divided into two syndromes, acute GVHD, historically defined as a GVHD reaction occurring within the first 100 days after allo-SCT and chronic GVHD (cGVHD), that generally occurs beyond day 100 [8,9]. While cGVHD has been associated with graft-versus-tumor effects [3,10], it is also a major cause of morbidity/mortality in longterm transplant recipients [11].
Sclerodermatous cGVHD (scl-cGVHD) is one of the most severe form of cGVHD and develops in approximately 20% of cGVHD patients [12]. Although scl-cGVHD shares common features with systemic fibrosis, the two syndromes differ both in terms of pathology (scl-cGVHD usually begins in the superficial layer of the skin and then extents to deeper layers of the skin while the opposite is generally true in systemic sclerosis), and in terms of clinical symptoms, with clinical features such as Raynaud's syndrome, pulmonary hypertension and cardiac dysfunction being frequently observed in patients with systemic sclerosis but infrequently in scl-cGVHD patients [13,14].
The pathogenesis of cGVHD remains not fully understood. It is generally accepted that donor T cells are largely involved [4]. Specifically, data from murine models of cGVHD suggest that donor T cells involved in cGVHD are mainly CD4 + T helper 2 (Th2) cells [15]. These Th2 cells secrete IL-4, IL-5, IL-10, IL-11 and IL-13 that stimulate other cells to release fibrosing factors such as IL-13, PDGF and TGF-β.These ones then induce fibrosis in the skin and other affected organs. Histocompatibility antigenic disparities between donor and recipient are also a risk factor for cGVHD (although to a lesser extent than for acute GVHD [16]), suggesting that cGVHD manisfestations are due to recognition of allogeneic antigens, such as major or minor histocompatibility antigens by donor T cell. Host thymus integrity could also play a role, as suggested by the lower incidence of chronic GVHD in younger recipients [16], although some studies failed to observe an association between thymic function and subsequent occurrence of cGVHD [17,18]. Finally, emerging data have also demonstrated an important role for B cells in cGVHD pathogenesis [19][20][21]. Imatinib (Glivec 1 ; Novartis Pharmaceuticals) is a tyrosine kinase inhibitor developed as a competitive inhibitor of ATP for binding to BCR-ABL inducing apoptosis of BCR-ABL dependent leukemic cells [22]. However, imatinib is not specific towards BCR-ABL and also targets other tyrosine kinases such as the stem cell factor c-kit, c-Abl (involved in transforming growth factor (TGF)-β signaling pathway), and platelet-derived growth factor receptor (PDGF-r) [22]. Given that the TGF-β and PDGF signaling pathways are largely involved in the fibrogenesis process in scl-cGVHD [15,23], and given the ability of imatinib to inhibit T-cell proliferation in vitro [24], some clinical studies have assessed the impact of imatinib in patients with steroidrefractory cGVHD [25][26][27][28][29]. Unfortunately, these studies yielded conflicting results underlying the importance of re-assessing the impact of imatinib in scl-cGVHD in pre-clinical models.

Mice and drugs
Twelve to 14 week-old B10.D2 (H-2 d , Jackson Laboratories, Bar Harbor, USA) and Balb/cJ (H-2 d , Jackson Laboratories) mice were used as donors and recipients, respectively, in a MHC-matched minor antigens disparate scl-cGvHD model [30,31]. All mice were maintained in top-filtered cages in a standard animal facility and provided with sterilized food. Sterilized water supplemented with Baytril 1 1% (Bayer HealthCare, Diegem, Belgium) was given from 3 days before transplantation until the end of the experiment (day +52). Water was changed every 2-3 days. All animal experiments were approved by the animal ethic committee of the University of Liege (local file number: 1438).
Imatinib (Glivec 1 ) was kindly provided by Novartis Pharmaceuticals (Basel, Switzerland). Imatinib was dissolved in sterile water and given by oral gavage at a dose of 150 mg/kg/day (50 mg/kg every morning + 100 mg/kg every evening).

Bone marrow transplantation and scl-cGVHD
Balb/cJ recipient mice were transplanted as previously described [31]. Briefly, recipient Balb/cJ mice (H-2 d ) were lethally irradiated with 7 Gy total body irradiation using a 137 Cs irradiator (GammaCell 40, Nordion, Ontario, Canada). Six hours later, recipients were injected with 200 μL of the prepared cell suspension (2x10 6 spleen cells + 1x10 6 bone marrow cells for the moderate scl-cGVHD model [32] or 70x10 6 spleen cells + 10x10 6 bone marrow cells for the severe scl-cGVHD model [31, 33]) from B10.D2 (H-2 d ) mice. Recipient mice were monitored daily, while recipient body weights and GVHD scores (see next section) were recorded every 3 days. Mice were observed for a maximum of 52 days after transplantation. Imatinib treatment was started on day +7 post-transplant and was prepared in sterile water. A dose of 150 mg/kg/ day (50 mg/kg every morning + 100 mg/kg every evening, as previously reported [34]) was given orally, while control mice received the same volume of sterile water. This high dose of imatinib has been used in several previous in vivo studies [32,35,36]. At the end of the experiment (day +52), mice were sacrificed and splenocytes and bone marrow cells were collected in sterile RPMI + FBS 10% + P/S 1% while blood was collected in Microtainer 1 K2E tubes (BD Biosciences, San Diego, California, USA).

Assessment of scl-cGVHD
Scl-cGvHD severity was assessed using a clinical scoring system as previously reported [31,33]. Briefly, ear-tagged animals were individually scored 3 times/week for five parameters: weight loss (grade 1, loss between 10-20%; grade 2, loss > 20%), posture (1, kyphosis only at rest; 2, severe kyphosis when the animal moves), activity (1, moderate activity impairment; 2, no move unless stimulated), skin (1, erythema or scaling tail; 2, lesions on the body surface), and hair loss (1, loss > 1 cm 2 ; 2 loss > 2 cm 2 ). The clinical score was generated by summation of the five criteria scores (0 to 10/10). Animals reaching a score of 8/10 were sacrificed to avoid unnecessary pain according to our local ethic committee. Final scores for dead animals reaching the ethical limit score were carried forward for the remaining time points.

Skin biopsies and immunohistochemistry
On day 29 after transplantation, mice were anesthetized with an i.p. injection of a solution of xylasin/ketamin (10 mg/kg and 75 mg/kg, respectively, given in i.p.). Skin tissue samples of 0.5 cm 2 from the upper back were collected and directly fixed in paraformaldehyde 10% before being paraffin-embedded and sectioned at 4-5 μm thickness. Parrafin-embedded sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Antigen was retrieved with citrate buffer combined with heating in a microwave oven. Slides were peroxidase blocked with a commercial protein block serum-free solution (Dako, Glostrup, Danemark) for 10 minutes and were then incubated with one of the following antibodies at 4˚C overnight: polyclonal rabbit anti-mouse phospho-c-Abl or polyclonal rabbit anti-mouse phospho-PDGFR (Abcam PLC, Cambridge, United Kingdom) and detected with the Envision system (Dako) for 30 minutes. Colorimetric detection was completed with diaminobenzidine (Dako) and slides were then counterstained with hematoxylin. Phospho-c-Abl and phospho-PDGF-r was evaluated in the dermis by a semi-quantitative score of the intensity and extent of staining according to an arbitrary scale

In vivo proliferation experiment
Balb/cJ recipient mice were lethally (7 Gy Total body irradiation) irradiated and then injected with 10x10 6 B10.D2 bone marrow cells on day 0. Fourteen days later, spleens from B10.D2 donor mice were harvested and cells were isolated as previously described. Splenocytes were labeled with carboxyfluorescein succinimidyl ester (CFSE, Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Cells were then resuspended in PBS at a concentration of 70x10 6 cells/200 μL. Recipient mice were injected i.v. with 70x10 6 cells and received imatinib or sterile water treatment for 5 days as described above. Mice were then sacrificed and blood, lymph nodes and spleen were collected. Harvested cells were then stained for CD3, CD4, CD8 and FoxP3 expression.

Flow cytometry
The reconstitution of the different T-cell sub-populations was evaluated by flow cytometry at the indicated time-points. Specifically, blood from tail vein was collected in a BD Microtainer 1 K2E Tube (BD Biosciences, San Diego, USA). Blood volumes were measured and transferred into a FACS tube. Blood cells were washed with PBS + FBS 3% + P/S 1% (staining buffer) and supernatants were then removed. Biotinilated antibody was added and incubated with cells for 30 minutes in the dark at 4˚C. Cells were washed twice with staining buffer. Extracellular antibodies, including the streptavidin antibody, were added as a mix directly in tubes and incubated for 30 minutes at 4˚C in the dark. Cells were washed twice with staining buffer. Cells were then fixed and permeabilized using the FoxP3 staining buffer set (eBioscience, San Diego, CA, USA) according to the manufacturer's instructions. Cells were then incubated for 30 minutes at 4˚C in the dark with intracellular antibodies and finally washed twice and resuspended with pure PBS in BD Trucount Tubes 1 to determine the absolute number of each lymphocyte population per μL of blood.
Data were acquired on a FACS Canto II (Becton Dickinson) and were analyzed with FlowJo 10.0.7 (Treestar Inc., San Carlos, CA).

Statistics
Statistical analyses were performed with the Graphpad1 Prism 5.00 software (La Jolla, CA, US). The Mann-Whitney U test was used to compare GVHD or histological scores and flow cytometry data in the different experimental groups. Results were expressed as median ± interquartile range or mean ± SEM with P-values considered as significant at 0.05. All P-values were 2-sided.

Imatinib decreases in vivo proliferation of T-cell subsets
Given the primordial role of donor T cells in the pathogenesis of cGVHD, we first assessed the impact of imatinib on T-cell proliferation in vivo. Specifically, after 7 Gy TBI, Balb/cJ mice were transplanted with 10x10 6 bone marrow cells from B10.D2 mice on day 0, followed by the injection of 70x10 6 CFSE-labeled spleen cells on day + 14. Treatment with sterile water or imatinib was started immediately after splenocyte injection. Mice were sacrificed 5 days later and spleen, lymph nodes, blood and bone marrow were collected and stained for flow cytometry experiments. Imatinib decreased the proliferation of total T cells (defined as CD3 + cells) in the spleen (P = 0.018), the blood (P = 0.0271), and the lymph nodes (P = 0.015), the proliferation of CD4 + T cells in the lymph nodes (P = 0.005), the proliferation of CD8 + T cells in the spleen (P = 0.007), the blood (P = 0.05), and the lymph nodes (P = 0.044), and the proliferation of Tregs in the spleen (P = 0.023) and the lymph nodes (P = 0.002) (Fig 1). scl-cGVHD model. Imatinib did not impact cGVHD severity since a comparable evolution of cGVHD scores was observed in the two groups of mice although there was a suggestion for less weight loss in imatinib-treated mice that reached statistical significance at day +52 following transplantation (P = 0.0226) (Fig 2B). Moreover, imatinib did not reduce scl-cGVHD scores in the moderate model of scl-cGVHD (S1 Fig).

Imatinib decreases phosphorylation of PDGF-r and C-Abl
We next assessed on skin biopsy harvested on day +29 whether PDGF-r and c-Abl signaling pathways were impacted by imatinib. Indeed, it has been previously reported that prevention of fibrosis by tyrosine kinase inhibitors could be attributed at least in part to a direct inhibition of c-Abl and PDGF-r signaling [32]. As shown in Fig 3, there were comparable levels of phosphorylated c-Abl on day 29 in imatinib-treated and control mice (P = 0.6), while imatinibtreated mice had significantly lower levels of PDGF-r phosphorylation than control ones (P = 0.0079) (Fig 3). These data demonstrate that imatinib inhibits PDGF-r signaling.

Imatinib does not affect T-cell counts on day +21
We next assessed the impact of imatinib on T-cell subsets in scl-cGVHD mice 14 days after treatment initiation (D+21 post-transplantation).
We also compared the proliferation of the different T-cell subsets on day 21 by measuring their expression of KI67. As shown in Fig 5, compared to severe scl-cGVHD control mice, mice given syngeneic grafts had higher proliferation of CD4 + T cells (P = 0.0072), naïve CD4 + T cells (P = 0.0019) and Tregs (P = 0.0061) but lower proliferation of CD8 + T cells (P = NS).
-Comparison between control and imatinib-treated cGVHD mice. As shown in Fig 4, absolute counts of the various T-cell subsets were comparable in imatinib-treated and control mice in the severe scl-cGVHD model. In the moderate scl-cGVHD model, in comparison to control mice, mice given imatinib had slightly lower counts of total (P = 0.0101) and of effector Finally, compared to control mice, imatinib-treated mice displayed higher proliferation of CD4 + T cells (P = 0.0464), CD8 + T cells (P = 0.0496), and Tregs (P = 0.035) in the severe scl-cGVHD model (Fig 5). This was not observed in the moderate scl-cGVHD model (S3 Fig).  Imatinib does not affect T-cell counts on day +35 As for day +21, we also assessed the impact of imatinib on T-cell subsets in blood and on their proliferation on day +35 after transplantation (28 days after starting treatments).
-Comparison between severe scl-cGVHD and syngeneic transplants. Fig 6 shows that, compared to severe scl-cGVHD control mice, mice receiving syngeneic grafts had higher absolute counts of all T-cell subsets (P values ranged from 0.0004 to 0.0167) except for effector memory CD8 + T-cells that, on the contrary, were higher in severe scl-cGVHD than in syngeneic mice (P = 0.0167) and counts of central memory CD8 + T cells were comparable in the two groups.
Regarding analyses of proliferation assessed by the expression of KI67, proliferation of CD8 + T-cells was lower in syngeneic mice than in severe scl-cGVHD control mice (P = 0.0007) while the opposite was observed for naïve (P = 0.033) and central memory (P = 0.003) CD4 + T cells (Fig 7).
-Comparison between control and imatinib-treated cGVHD mice. As observed on day +21, absolute counts of T-cell subsets were also comparable whether animals received or not imatinib on day +35 in the severe scl-cGVHD model (Fig 6). In the moderate scl-cGVHD model, in comparison to control mice, mice given imatinib had slightly lower counts of total (P = 0.008) and naive (P = 0.03) CD4 + T cells, and of total (P = 0.03) and naive (P = 0.008) CD8 + T cells (S4 Fig). Interestingly, proliferation of naive CD4 + T cells was higher in imatinib-treated than in control mice both in the severe scl-cGVHD (P = 0.03) (Fig 7) and in the moderate scl-cGVHD (P = 0.016) model (S5 Fig). First, in vivo administration of imatinib in the severe scl-cGVHD model was able to inhibit its targets. Indeed, phosphorylation levels of PDGF-r were significantly decreased in imatinibtreated compared to control mice, while perhaps a similar trend was observed for c-Abl. This is in concordance with the observations reported by Zerr et al. in a similar murine model of scl-cGVHD than the moderate scl-cGVHD model used in the current study [32]. Secondly, imatinib administration failed to significantly ameliorate severe scl-cGVHD, although there was a trend for less weight losses in imatinib-treated mice. This is in discordance to what has been reported by Zerr et al. who observed that imatinib significantly decreased scl-cGVHD scores [32]. The reason for this apparent discrepancy is unclear and is not related to the lower numbers of splenocytes injected (35 times less) by Zerr et al. since in our hands, imatinib administration did not decrease scl-cGCHD either when assessed in the moderate scl-cGVHD model (in which similar number of donor cells were infused than in the study by Zerr et al.). Further, the total dose of imatinib administered was similar in both studies (150 mg/kg), although it was administered once daily in the Zerr et al. study and twice daily in the current study. It has been previously reported that a twice-daily schedule of administration of imatinib was more efficient than a one daily schedule [35]. Taken together our data suggest that a partial inhibition of PDGF-r is insufficient to significantly ameliorate scl- cGVHD. Based on this hypothesis, a more potent PDGF-r and c-abl blockade by combining imatinib with nilotinib could be assessed in this mouse model [37]. On the other hand, a possible negative impact of imatinib on scl-cGVHD is its ability to decrease Treg [38]. Indeed, previous studies have demonstrated that imatinib impaired Treg proliferation and function on one hand [38,39], and, on the other hand, potentiated antitumor T cell responses in an gastrointestinal stromal tumor model through CD8 + T cell activation inducing Treg apoptosis within the tumor by depleting indoleamine 2,3-dioxygenase [40]. In agreement with these data, we demonstrated lower Treg proliferation in the spleen and the lymph nodes by imatinib. Previous studies have demonstrated the ability of Tregs to prevent acute [41], xenogeneic [42,43], and chronic [40,44] GVHD, while we have recently observed that the hypomethylating agent azacytidine was able to significantly ameliorate severe scl-cGVHD at least in part by promoting Treg [31]. These observations might justify further studies in murine models of GVHD combining imatinib and azacytidine for severe scl-CGVHD prevention.
Thirdly, compared to mice transplanted with syngeneic grafts, we observed that allogeneic scl-cGVHD mice had higher counts of effector memory CD8 + T cells but significantly lower counts of all other T-cell subsets. These data suggest that the scl-cGVHD process induces the expansion of effector memory CD8+ T cells while impacting immune reconstitution probably in part by decreasing thymic neo-generation of T cells and increasing T-cell apoptosis as previously reported [45,46]. Further, these data are consistent with recent observations by Binsfeld et al. in a murine model of graft-versus-myeloma effect consisting in the transplantation of BALB/cJ mice, previously injected with luciferase-transfected MOPC315.BM myeloma cells, with bone marrow cells and splenocytes from B10.D2 mice [33].
Fourthly, we observed that imatinib had no impact on T cell reconstitution in severe scl-cGVHD mice, while it slightly impaired it in the moderate scl-cGVHD model. Previous studies have evidenced the in vitro inhibitory effects of imatinib on T-cell activation and T-cell proliferation (through direct inhibition of LCK by imatinib) [24,47]. In accordance with these in vitro published data, we observed that imatinib slightly decreased T-cell proliferation (including Treg proliferation) in vivo in a lymphopenic host.
In conclusion, imatinib failed to significantly alleviate scl-cGVHD in a severe murine model of scl-cGVHD despite it partially inhibited PDGF-r.