Activation of Fms-Like Tyrosine Kinase 3 Signaling Enhances Survivin Expression in a Mouse Model of Rheumatoid Arthritis

Survivin is known as an inhibitor of apoptosis and a positive regulator of cell division. We have recently identified survivin as a predictor of joint destruction in patients with rheumatoid arthritis (RA). Flt3 ligand (Flt3L) is expressed in the inflamed joints and has adjuvant properties in arthritis. Studies on 90 RA patients (median age 60.5 years [range, 24–87], disease duration 10.5 years [range, 0–35]) show a strong positive association between the levels of survivin and Flt3L in blood. Here, we present experimental evidence connecting survivin and Flt3L signaling. Treatment of BALB/c mice with Flt3L led to an increase of survivin in the bone marrow and in splenic dendritic cells. Flt3L changed the profile of survivin splice variants, increasing transcription of the short survivin40 in the bone marrow. Treatment with an Flt3 inhibitor reduced total survivin expression in bone marrow and in the dendritic cell population in spleen. Inhibition of survivin transcription in mice, by shRNA lentiviral constructs, reduced the gene expression of Flt3L. We conclude that expression of survivin is a downstream event of Flt3 signaling, which serves as an essential mechanism supporting survival of leukocytes during their differentiation, and maturation of dendritic cells, in RA.


Introduction
Survivin is an intracellular protein and a member of the inhibitor of apoptosis proteins (IAPs), with many functions in cytoprotection, cell division and cellular adaptation. It is encoded by the BIRC5 gene on the human chromosome 17q25, into a 142 amino acid, 16.5 kDa protein but can be extensively alternatively spliced into several variants and form homodimers and heterodimers with two different splice variants, or an IAP-IAP complex by pairing with X-linked IAP (XIAP) [1,2]. Transcription of survivin is negatively regulated by p53 [3] and positively regulated by STAT3 [4] and TCF-4 [5].
Similar to human protein, full-length murine survivin (survi-vin140) contains a single BIR-domain, which is critical for its antiapoptotic function [6], and a carboxy-terminal coiled-coil domain that links its function to the cell cycle [7]. Murine survivin121 contains a BIR-domain that makes it able to inhibit caspase activity but lacks the coiled-coil structure. Additionally, there is a splice variant that predicts a 40 aa residue protein (survivin40), which lacks both the BIR-and a coiled-coil regions. The differential expression of these forms of survivin is believed to affect the balance between cell proliferation and programmed cell death [8].
Survivin has only limited expression in adult tissues, but is overexpressed in tumors and is therefore regarded as a cancer gene. Survivin is expressed in a cell-cycle dependent manner with a peak at G2/M. Together with Aurora B, borealin and INCENP, survivin forms the chromosomal passenger complex, which is recruited to chromosomes by the phosphorylation by histones which is recognized by the BIR domain of survivin. In anaphase, the chromosomal passenger complex relocalizes to form the mitotic spindle and stimulate cytokinesis. The requirement of survivin during fetal development has been demonstrated by lethality of knockout embryos [1].
Survivin is also expressed independently of cell cycle progression and is linked to the inhibition of apoptosis [3]. The pool of survivin with apoptotic inhibiting properties seems to be localized to the mitochondria [1,9], and is released into the cytoplasm in response to death stimuli. Survivin inhibits apoptosis in complex with hepatitis B X interacting protein [10] or XIAP, perhaps by separating Smac/Diablo from XIAP, thus enabling caspase degradation [1]. There is an emerging role for survivin in normal adult CD34+ hematopoetic stem cells [11,12] and in the development, maturation and survival of immune cells, for example in T cells [13][14][15][16] and neutrophils [17]. It is upregulated in response to stimulation with hematopoetic cytokines and growth factors [18][19][20].
Expression of survivin in malignancies is associated with unfavorable outcome [21,22] and resistance to cytotoxic treatment [23][24][25]. In the context of rheumatoid arthritis (RA), extracellular survivin is a marker of poor prognosis. A prospective study on 651 patients at the early stage of RA with the mean disease duration of 6 month showed that high serum levels of survivin were predictive for severe cause of RA, characterized by persistent joint inflammation and progressive joint destructions [26]. The proportion of survivin-positive patients may vary from 20-30% in the group of established and treated RA patients [27] to 60% in the population of early RA patients [26]. Successful antirheumatic treatment may reduce serum levels of survivin, while survivin-positive patients accumulate among the patients who do not respond to anti-rheumatic treatment [26][27][28]. A growing number of publications [26,27,[29][30][31] support the idea that survivin has a role in progression of RA. We have previously shown that survivin has a key function in the regulation of invasive properties of fibroblasts in the inflamed rheumatic joint, and that intracellular survivin is essential for urokinase expression and for the up-regulation of urokinase receptor [32]. Also, serum survivin modifies surface pattern of leukocyte adhesion molecules [33].
Fms-like tyrosine kinase 3 (Flt3) is a receptor tyrosine kinase class III, a class that also includes c-KIT, PDGFRa/b and c-FMS receptors. It is expressed on hematopoietic stem cells and promotes differentiation, maturation and survival of lymphoid progenitors in the bone marrow [34]. Flt3 is essential for the development of antigen presenting cells, such as B cells and dendritic cells (DC) [35][36][37]. Mutations in the Flt3 receptor, leading to its constant activation, are frequently found in acute myeloid leukemia (AML) [38]. It has been shown that bone marrow cells expressing Flt3 with internal tandem duplication mutations which keeps Flt3 signaling constantly active, have an increased survivin expression [39,40] and this is associated with development of resistance to Flt3 inhibition [39].
Flt3 ligand (Flt3L) is expressed by many cell types, and is present in a soluble intracellular form, and in a membrane bound form [41]. Flt3L has recently been outlined within a panel of preclinical biomarkers of predictive value for the development of RA [42]. We have recently shown that Flt3L is elevated in RA patients and that Flt3L has adjuvant properties when injected into the joint, facilitating development of arthritis in experimental settings [43]. Blockade of Flt3 signaling using a synthetic Flt3 inhibitor alleviates signs of synovitis and cartilage destruction in antigen-induced model of arthritis [44].
Since expression of the Flt3 receptor and survivin in adults is present in hematopoietic stem cells, we address the question whether Flt3L signaling could actually be linked to survivin in rheumatoid arthritis. We further tested the pathological significance of this relation using in vivo treatment with Flt3L and evaluated the possibility of targeting this pathway with an Flt3 inhibitor and lentiviral shRNA knockdown of survivin.

Higher Levels of Flt3 Ligand in Survivin Positive Rheumatoid Arthritis Patients
Survivin levels in the blood of 104 healthy controls (male n = 16, female n = 88, mean age 52.4 years [range 18-67]) were measured by ELISA. The evaluation of the results showed a non-Gaussian distribution of survivin levels in the studied group. The 95% confidence interval was calculated which indicated that the levels of survivin above 450 pg/ml was present in less than 5% of healthy individuals. Thus, the patients with serum or synovial fluid levels of survivin above 450 pg/ml represented the survivinpositive group (n = 29) and the remaining were survivin-negative (n = 61, Table 1). In the present material, 32% of the RA patients were survivin-positive. This is in agreement with our previous reports [27]. As expected, survivin-positive RA patients were significantly more often RF-positive (p = 0.0006) and a majority of the survivin positive patients had erosive joint disease (p = 0.003). The comparison between the groups revealed significantly higher levels of Flt3L in the blood of the survivin-positive RA patients compared to survivin-negative (pg/ml: 110 [range  vs 70 [range 10-230], p = 0.003, Fig. 1A). This was also the case for synovial fluid (pg/ml: 200 [range  vs 150 [range 30-930], p = 0.055), although not statistically significant (Fig. 1A). There was no significant difference in the levels of acute-phase reactants (e.g., C-reactive protein, serum amyloid A protein levels, IL-6 levels) between the survivin-positive and survivin-negative groups ( Table 1).

Survivin Expression is Upregulated in Mouse Bone Marrow Following Treatment with Flt3 Ligand
To elucidate a causative relation behind the observed association between survivin and Flt3L levels in RA patients, arthritic mice were challenged with Flt3L. Two independent experiments were performed containing 15 Flt3L-treated mice and 19 PBStreated controls. After 14 days of Flt3L treatment, survivin levels in bone marrow cell lysates were increased compared to the PBS treated group (pg/ml: 472.56193.4 vs 286.5648.9, p = 0.01, Fig. 1B). At day 28, the number of survivin+ cells in bone marrow of Flt3L treated mice was higher compared to PBS treated controls (28.166.3% vs 21.964.3%, p = 0.02, Fig. 1B). Additionally, Flt3Ltreated mice tended to have somewhat higher survivin concentration per cell, measured as the intensity of intracellular staining (gMFI: 336654 vs 288629, n.s.). The intensity of survivin staining was higher in the Flt3+ cells compared to the Flt32 cell population (p = 0.003). Flt3L treatment tended to upregulate survivin expression in both Flt3+ and Flt32 cell populations (Fig. 1C).
Despite the similar size of survivin+ population in bone marrow and spleen, survivin gene expression in spleen was 4 times lower compared to bone marrow (p = 0.002, Fig. 2A, PBS). Flt3L treatment of arthritic mice did not affect total survivin levels in the spleen, since the percentage of survivin+ spleen cells was similar in the Flt3L treated group and the control group (20.365.3% vs. 19.566.3%). High intensity of survivin expression was attributed to the MHCII + CD11c hi dendritic cell population in the spleen (Fig. 1D). Flt3L treatment increased the proportion of survivin+ cells within the MHCII + CD11c hi population (73.966.2% vs 66.064.2%, p = 0.005, Fig. 1E). When cultured in vitro, splenocytes of Flt3L-treated mice (n = 7) secreted significantly higher levels of survivin compared to splenocytes of PBS-treated controls (n = 7) (pg/ml: 32.667.7 vs 24.366.2, p = 0.04), but Flt3L treatment did not increase serum levels of survivin (pg/ml: 177262503 vs 119061335).
in spleen was lower compared to the expression in the bone marrow (p = 0.002, Fig. 2A). Further analysis of different splice variants of survivin ( Fig. 2C) was done on spleen and bone marrow samples. Flt3L treatment increased transcription of the short survivin40 variant in bone marrow, while no increase in the transcription of the full-length variant survivin140 could be detected (Fig. 2D). Since it was hard to design specific primers for the survivin121 transcript, a primer pair localized in the exon 1 and exon 2 was used. These primers were able to detect survivin140 and survivin121 transcripts, but not survivin40 transcript. The transcription of survivin140 and survivin121 variants was similar in Flt3L treated and in PBS treated mice. Western blot showed the dominating expression of survivin140 in spleen and survivin140 and survivin121 in bone marrow samples (Fig. 2B).

Survivin Expression in the Bone Marrow is Dependent on Flt3 Receptor Activity
We hypothesized that blockade of Flt3 receptor would affect survivin production in the bone marrow. Survivin was measured in arthritic mice treated with the Flt3 inhibitor SU11248 (sunitinib) in 3 independent experiments containing totally 24 sunitinib-treated mice and 25 control mice treated with citric acid. On day 28, a reduction of total survivin levels was observed in bone marrow lysates of sunitinib treated mice compared to the citric acid treated controls (pg/ml: 17966388 vs. 259061107, p = 0.049, Fig. 3A). Sunitinib significantly reduced the number of survivin+ cells in mouse bone marrow (46.864.2% vs 59.366.8%, p = 0.0007), and the intensity of intracellular survivin staining (p = 0.002, Fig. 3C). This reduction of survivin expression could be seen mainly on Flt32 cells (p = 0.003, Fig. 3B).
To assess whether sunitinib suppressed the expression of survivin in secondary lymphoid organs, we analyzed survivin levels in the spleen. The total survivin levels in the spleen were not reduced following sunitinib treatment (Fig. 3C), however a significant reduction of survivin expression in the MHCII + CD11c hi dendritic cell population was identified (gMFI: 26846607 vs 21796207, p = 0.02). The number of survivin+ dendritic cells was also reduced with sunitinib treatment (58.165.3 vs 67.6611.5%, p = 0.05, Fig. 1E). In spleen, Flt3 inhibition resulted in a reduction of CD11c+ cell population (p = 0.02, Table 2). The latter observation was in consistence with our previous findings on the inhibitory effect of sunitinib on the development of dendritic cells [44].

Survivin shRNA Reduces Flt3 Ligand Expression in the Spleen
Finally, we studied if Flt3/Flt3L system was changed by survivin inhibition. BALB/c mice received lentiviral constructs, containing a sequence coding for shRNA targeting survivin gene transcripts (n = 8), or the same lentiviral construct containing a non-targeting shRNA sequence (n = 9), by a single intra-articular injection. On day 28, the efficacy of survivin inhibition was confirmed on the

Discussion
High levels of survivin and Flt3L in blood and synovial fluid of patients with rheumatoid arthritis are implied in the pathogenesis of joint inflammation [26,27,43]. Additionally, a tight clinical association observed between these parameters with respect to joint destruction and effect of anti-rheumatic treatment [28] prompted us to study common biological processes linking high expression of survivin and Flt3L in RA patients. The analysis of a cohort of RA patients with the established joint disease showed a clear association of high Flt3L levels with high levels of survivin in blood. Neither Flt3L nor survivin was a consequence of inflammation, measured by CRP, WBC count, serum amyloid A protein and IL6. Due to the cross-sectional nature of the study, one can see a clear difference in the frequency of radiologic erosions, which is a sign of severe disease, in the survivin positive group compared to the survivin negative group. The survivinpositive group contains higher proportion of patients treated with immunosuppressive anti-rheumatic drugs compared to the survivin-negative group. High levels of survivin characterize patients with therapy resistant cause of RA, suggesting that many patients responding to anti-rheumatic treatment convert from survivinpositive to survivin-negative. The results from the patient material show an association between Flt3L and survivin in RA, suggesting the existence of a mutual mechanism regulating levels of these proteins. Similarity of the groups with respect to disease duration, age, levels of C-reactive protein and white blood cell count further support this argument.
Our clinical observations were further extended in the experimental setting, where the requirement of Flt3/Flt3L signaling for expression of survivin in vivo was assessed. Survivin levels were increased following treatment with recombinant Flt3L and they were decreased with Flt3 receptor blockade. This proves that survivin production in the bone marrow is dependent on Flt3 signaling and makes Flt3L a major trigger of survivin transcription and expression in leukocytes. Interestingly, we showed the existence of a negative regulation loop where the low expression of survivin in spleen (induced by shRNA) resulted in the low levels of Flt3L in spleen. In the joints of RA patients, the elevated levels of Flt3L could originate from immune cells or fibroblast-like stromal cells in the synovium, by an inflammation caused shedding of the membrane bound form [45] or as a response to lack of antigen presenting cells. It is for example known that the human syndrome of dendritic cell, monocyte, B and NK lymphoid deficiency is associated with elevated levels of Flt3L in humans [37] and the same is true in mice following Flt3 inhibition [44,46].
Flt3/Flt3L signaling is important for the maturation of hematopoietic cells [34]. Here, we show that Flt3L induces survivin production in the Flt3+ progenitors in bone marrow, and in dendritic MHCII + CD11c hi cells in spleen. Survivin is increased after Flt3L treatment also in the Flt32 population in the bone marrow, suggesting that survivin is induced in Flt3 positive hematopoietic progenitors and remains upregulated after internalization of activated Flt3. Thus, survivin expression induced by Flt3L is associated with differentiation of cells rather than cell proliferation. Our findings are in agreement with previous reports on the induction of survivin production in CD34+ stem cells [20] prior to their differentiation.
Spleen is a maturation site for the peripheral dendritic cells and the process has recently been shown to be dependent on Flt3 expression [47,48]. In our study, a reduction of the intracellular pool of survivin by inhibition of Flt3 led to a pronounced reduction of CD11c+ population of dendritic cells in the spleen. The obvious effect of Flt3L treatment/Flt3 inhibition on survivin levels in the MHCII + CD11c hi population in the spleen further supports a strong relationship between the induction of survivin and DC development. We have previously shown that inhibition of Flt3 receptor reduces the DC population in spleen [44]. In RA patients, we have seen that high levels of extracellular survivin are  associated with an increased number of circulating CD11c+ survivin expressing cells in blood [33]. Our present results suggest that modulation of survivin levels is an essential mechanism regulating homeostasis of peripheral dendritic cells. Interestingly, studies on the Flt3L-induced expressions of different survivin-transcripts revealed an effect on the short survivin40-transcript in bone marrow. This variant lacks both the BIR-domain, essential for the interaction with caspases and for the anti-apoptotic effects of survivin [6], as well as the coiled coildomain required for the formation of chromosomal passenger complex [7]. As survivin40 retains the N-terminal domain, it has been believed to bind other survivin variants [8], and to modulate their function by preventing homodimerization and facilitating interaction through the BIR-and coiled-coil domains. However, we could not detect survivin40 on protein level with western blot, as this has been shown earlier by another group [8] and might either indicate that this transcript is not translated or that the antibodies used are not able to detect the survivin40. The increase in survivin protein after Flt3L treatment can either be a result of an increased survivin40-expression or of a decrease in survivin140or survivin121-turnover. In spleen, high survivin expression is seen within the MHCII + CD11c hi population. This population represents a small part of the total amount of spleen cells and verification of survivin changes requires cell sorting prior to the RNA analyses.
In conclusion, our findings indicate that expression of survivin is a downstream event of Flt3L signaling, which serves as an essential mechanism supporting survival of hematopoietic cells during their differentiation and maturation of antigen presenting cells in bone marrow and spleen during arthritis.

Ethics Statement
The study involving human participants (S441-01) was approved by the Ethics Committee of Sahlgrenska University Hospital. All studies were conducted in compliance with the Declaration of Helsinki, and all patients gave written informed consent to participate in the study. All animal experiments (Protocol Numbers: 176-2008, 328-2008) were approved by the Ethical Committee on Animal Experiments in Gothenburg.

Patients
Synovial fluid and serum samples were obtained from 90 patients with RA attending the Rheumatology Clinic at Sahlgrenska University Hospital, Göteborg, Sweden, for acute joint effusion. All the patients met the diagnostic criteria for RA suggested by the American College of Rheumatology (ACR) [49]. The presence of bone erosions, defined as the loss of cortical definition at the joint, was recorded in proximal interphalangeal, metacarpophalangeal, carpus, wrist and metatarsophalangeal joints using recent radiographs of hands and feet. The presence of one erosion was sufficient to fulfill the requirement of an erosive disease. The presence of rheumatoid factor of any of the immunoglobulin isotypes was considered positive. Clinical characteristics of the patients are presented in Table 1. Blood from 104 healthy blood donors (male n = 16, female n = 88, mean age 52,4 years [range 18-67]) was used to calculate normal levels of survivin.

Collection and Preparation of Samples
Synovial fluid was obtained by arthrocentesis, aseptically aspirated and transmitted into sodium citrate (0.129 mol/l; pH 7.4) containing tubes. In most cases synovial fluid was obtained from knee joints. Synovial fluid samples were aliquoted and frozen in 270uC immediately after joint injections. Blood samples were centrifuged at 1000 rpm and serum was aliquoted and stored at -70uC.

Laboratory Measures of Disease Activity
Serum levels of C-reactive protein (CRP) were measured by standard nephelometry, with established normal range 0-5 mg/ ml. Serum amyloid A protein (SAA) levels were detected by an ELISA (Biosource, Camarillo, CA). White blood cell (WBC) counts in blood and synovial fluid samples were obtained with a microcell counter F300 (Sysmex, Toa, Japan). Levels of IL-6 were measured as described previously [43], and rheumatoid factor (RF) were measured using standard laboratory techniques at Sahlgrenska University Hospital, Göteborg, Sweden.

Antigen Induced Arthritis using Methylated Bovine Serum Albumin (mBSA)
Female Balb/c mice were purchased from the Charles River Laboratory (Germany) and housed at the animal facility at the Department of Rheumatology & Inflammation Research, under

In Vivo Treatment with Flt3 Ligand and Flt3 Inhibitor
Arthritic mice were treated with a daily intraperitoneal injection of recombinant Flt3L, in the dose of 1.5mg/mouse, starting 4 days prior to first immunization. Flt3L was purchased from Creative BioMart (Shirley, NY) or purified from Sp2.0 hybridoma transfected with Flt3L gene [43]. Control mice were treated with PBS injection or a control cell line supernatant purified with the same procedure. Flt3 receptor was inhibited using sunitinib (SU11248, Pfizer). Treatment was given by gavage (day 7-28, 40 mg/kg mouse), and the control group was treated with the same volume of citrate buffer as previously described [44]. Specificity of sunitinib for Flt3 inhibition in our model has been previously shown [44].

Knockdown of Survivin using Lentiviral Vector
Lentiviral construct, encoding short hairpin (sh)RNA targeting survivin with the sequence CCGGCAAAGACTACCCGT-CAGTCAACTCGAGTTGACTGACGGGTAGTCTTTGTT-TTTG (TRCN0000054616, Sigma-Aldrich), were selected after in vitro evaluation. Lentiviral particles were given by a single intra-articular injection at immunization day 0 (10 7 transduction particles/knee, n = 8). The control group received the same amount of lentiviral particles with non-targeting scrambled sequence (n = 9). Downregulation of survivin was confirmed in bone marrow and spleen.

Protein Preparation and Western Blotting
Total protein was prepared from cells and tissue by homogenization and sonication in the presence of protease inhibitors (Complete Mini, Roche Diagnostics GmbH, Mannheim, Germany). Protein concentrations were measured using the BCA Protein Assay kit (Pierce, Rockford, IL) according to the manufacturer's protocol. Proteins were separated on SDS-PAGE 4-12% Bis-Tris gels (NuPAGE, Invitrogen), and transferred to PVDF membranes (NuPAGE, Invitrogen). Membranes were blocked with 5% non-fat milk and incubated with rabbit-antisurvivin antibodies (840471, R&D systems) at 4uC over night. Detection was performed with peroxidase conjugated anti-rabbit secondary antibody (NA934VS, GE Healthcare) and ECL Prime TM Western Blotting Reagents (GE Healthcare). Chemilu-

Gene Expression Analysis
Total RNA from spleen tissue and bone marrow cells was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. Concentration and quality of the RNA was evaluated with NanoDrop spectrophotometer (Thermo Scientific, USA) and Experion (BioRad Laboratories). 400 ng RNA was used for cDNA synthesis using RT 2 First Strand Kit (SABiosciences, Qiagen). Real-time amplification was performed with RT 2 SYBRH Green qPCR Mastermix (SABioscences, Qiagen) using a ViiA TM 7 Real-Time PCR System (Applied Biosystems) according to the manufacturer's instructions. A negative control (no template) reaction was also performed for each primer pair tested. A melting curve for each PCR was performed between 60 and 95uC to ensure that only a single product had been amplified. For Flt3 and Flt3L PCR Assays from SABiosciences (Qiagen) were used (sequences available upon request). Primers used for quantification of survivin expression were designed using Primer3 [50], uniqueness checked using the In silico PCR feature of the UCSC Genome Browser, and NetPrimer (PremierBiosoft) was used to check for secondary structures. Primers were ordered from Sigma-Aldrich, and are noted in table 3. Expression levels of the genes were normalized to the two reference genes, Gapdh and Ppia (TATAA, Sweden, sequences available upon request). Primer concentration used was 0.5 mM. The results were expressed as the fold change compared with the expression level in the control cells with the ddCq-method.

Statistical Analysis
Patient data was expressed as mean (6SD) and analyzed with GraphPad Prism, version 5.0 for Mac (GraphPad, San Diego, CA). The patient material was stratified by the level of survivin, and the difference between the groups regarding Flt3L levels, age, rheumatoid factor, disease duration, presence of erosions, CRP, and WBC count in blood and synovial fluid, was calculated using the Mann-Whitney test or Chi-Square test for categorical data. Experimental data was also analyzed using GraphPad Prism, expressed as mean with SD and significance regarding differences between groups was calculated using the Mann-Whitney test. The two-tailed tests were used and for the statistical evaluation of the results p values ,0.05 was considered significant.