Differential expression of circulating miRNAs after alemtuzumab induction therapy in lung transplantation

Alemtuzumab is a monoclonal antibody targeting CD52, used as induction therapy after lung transplantation (LTx). Its engagement produces a long-lasting immunodepletion; however, the mechanisms driving cell reconstitution are poorly defined. We hypothesized that miRNAs are involved in this process. The expression of a set of miRNAs, cytokines and co-signaling molecules was measured with RT-qPCR and flow cytometry in prospectively collected serum samples of LTx recipients, after alemtuzumab or no induction therapy. Twenty-six LTx recipients who received alemtuzumab and twenty-seven matched LTx recipients without induction therapy were included in the analysis. One year after transplantation four miRNAs were differentially regulated: miR-23b (p = 0.05) miR-146 (p = 0.04), miR-155 (p < 0.001) and miR-486 (p < 0.001). Expression of 3 miRNAs changed within the alemtuzumab group: miR-146 (p < 0.001), miR-155 (p < 0.001) and miR-31 (p < 0.001). Levels of IL-13, IL-4, IFN-γ, BAFF, IL-5, IL-9, IL-17F, IL-17A and IL-22 were different one year after transplantation compared to baseline. In no-induction group, concentration of sCD27, sB7.2 and sPD-L1 increased overtime. Expression of miR-23b, miR-146, miR-486, miR-155 and miR-31 was different in LTx recipients who received alemtuzumab compared to recipients without induction therapy. The observed cytokine pattern suggested proliferation of specific B cell subsets in alemtuzumab group and co-stimulation of T-cells in no-induction group.

The no-induction group did not show any change in expression of miRNAs overtime. Instead, patients who received alemtuzumab had significantly higher expression of miR-146a and miR-155 one year after transplantation (p < 0.001 and p < 0.001, respectively) ( Table 2, Fig. 1). MiR-31 was found downregulated one year after transplantation in the alemtuzumab group (eightfold change, p < 0.001, Fig. 1).

Comparison of miRNA expression between groups.
One year after transplantation, five miRNAs showed different expression levels among the groups: miR-23b, miR-486, miR-146a and miR-155 and miR-31 ( Fig. 2, Table 3). MiR-23b was upregulated with a fourfold change in the control group and an eightfold change in the alemtuzumab group (p = 0.046). Both miR-146 and miR-155 were upregulated in alemtuzumab groups and downregulated in the no-induction group (p < 0.001 and p < 0.001, respectively). MiR-486 was dysregulated MiRNA gene target prediction and pathway enrichment analysis. After identification of dysregulated miRNAs, gene target prediction analysis was performed using miR-155, miR-146a, miR-23b and miR-31.   www.nature.com/scientificreports/ The total number of identified gene targets after accounting for overlap was 10,461. Among the identified genes, IL-17A and TNFSF13B genes, which encode IL-17A and BAFF respectively, were found to be targets of the dysregulated miRNAs. Using the identified gene targets, we performed enrichment analysis to identify significantly enriched pathways using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. MultiMir revealed 39 significantly enriched pathways with a p-value < 0.05 and a false discovery rate (FDR) < 0.10 (Benjamini-Hochberg procedure) in alemtuzumab group compared to controls at one year after transplantation. Similarly, 20 significantly enriched pathways were identified at one year after transplantation compared to baseline in alemtuzumab patients. The pathway names, odds ratios and levels of significance are listed in Tables S4 and S5. Analysis of the identified gene targets showed that signaling pathways involved in immune cell proliferation were enriched: Toll-like receptor (TLRs) signaling pathway, cytokine downstream signaling pathways, mammalian target of rapamycin (mTOR) pathway, Janus kinase (JAK)-Signal Transducers and Activators of Transcription (STAT) signaling pathway, mitogen-activated protein kinase (MAPK) pathway and PI3K-Akt signaling pathway (Tables S3 and S4). Moreover, transcription factors fundamental in immune cell activation such as forkhead box protein O1 (FOXO1), signal transducer and activator of transcription 1 and 5 (STAT1 and STAT 5), suppressor of cytokine signaling 1 (SOCS 1), nuclear factor kappa B (NF-kB), TNF receptor associated factor 6 (TRAF 6), interleukin-1 receptor-associated kinase 1 (IRAK 1), bHLH transcription factor (c-Myc) and SMAD family members were found as main targets of the differentially regulated miRNAs.

Discussion
Over the last 20 years, constant refinements of the immunosuppression protocols contributed to the improvement of long-term outcome after lung transplantation. Alemtuzumab is a monoclonal antibody, recently used as induction agent in solid organ transplantation. Based on the published experience, this agent shows excellent results in terms of rejection rates and survival [4][5][6]23 . MiRNAs are post-transcriptional regulators of complex biological and pathological processes and have been identified to exert a key function in hematopoiesis and immune cell homeostasis and activation. In the current study, we identified a set of differentially regulated miRNAs (miR-155, www.nature.com/scientificreports/ miR-146a, miR-23b, miR-31 and miR-486) one year after alemtuzumab induction therapy. Moreover, a panel of cytokines were found to be differently in both no-induction and alemtuzumab. Only in the alemtuzumab group, BAFF and IL-17A showed a positive and negative correlations with miR-146 and miR-155, respectively. Finally, enrichment pathway analysis supported the hypothesis that these five miRNAs are key regulators of immune cell proliferation, lineage commitment and inflammatory response. Alemtuzumab is a monoclonal antibody against CD52, which is a glycoprotein found on the surface of lymphocytes, monocytes and dendritic cells. The engagement of alemtuzumab on CD52 causes a profound longlasting immunodepletion 2,3 . B cell counts recover within 3-6 months and T cell counts within 12-24 months after treatment 2,24 . After reconstitution, phenotypic changes in the immune cell subsets seem to play a crucial role in allograft tolerance mechanisms 2,24-30 .
MicroRNAs are known key players in the pathogenesis of autoimmune diseases, fibrosis and cancer but their role in the context of immunosuppression after lung transplantation has not been studied yet. For the first time, could show herein that miR-155 and miR-146a, two key orchestrators of hematopoiesis and immune cell function are upregulated one year after alemtuzumab treatment. MiR-155 controls T-cell lineage commitment, by promoting T helper type 1 (Th1) over T helper type 2 (Th2) cells 14 . It has been previously described that deficiency of miR-155 leads to decreased germinal center (GC) size and response, lower levels of plasma and memory cells and impaired production of high-affinity class-switched IgGs 14,31,32 . MiR-146a is dynamically regulated during CD4 + cell differentiation, with an upregulation in differentiated Th1 and a downregulation in Th2 cells 33 . Thus, it is reasonable to hypothesize that both miRNAs may actively shape the reconstitution of immune cell subsets observed after alemtuzumab treatment 24,[27][28][29] .
B-cell activating factor is a cytokine belonging to TNF family, emerged as a critical factor for B cell survival and maturation 34 . Three main receptors of BAFF are known: B-cell maturation antigen (BCMA), Transmembrane activator and CAML interactor (TACI) and BAFF-R. BAFF binds strongly to TACI and BAFF-R but only weakly with BCMA 35 . BAFF-R is critical for survival and maturation of B cell precursors, it is expressed at late transitional stage and is present on all mature B cells 36 . TACI is present on late transitional stage and BCMA is present on plasma cells and is responsible for their long-lived survival [37][38][39] . The cytokine A proliferation-inducing ligand (APRIL) shares the receptor BCMA and TACI with BAFF, however, contrarily to BAFF, it strongly binds to BCMA and weakly to TACI. This slightly different expression pattern independently regulates homeostatic proliferation of different B-cell subsets 40 . In our cohort, we observed a strong increase in BAFF expression but www.nature.com/scientificreports/ not of APRIL. Concomitantly, serum levels of sCD40L and IL-6, which are both required for B cell maturation in plasma cells and IL-6 production 41,42 , did not increase overtime. Thus, the isolated increased expression of BAFF may indicate the proliferation of memory B cells, a cell subset which is spared by alemtuzumab 2,24 . Similar to our findings, serum BAFF levels but not APRIL were found to be increased in a cohort of kidney recipients after alemtuzumab induction 43 . We found a positive correlation between miR-155 and BAFF. A similar finding was observed in rheumatoid arthritis patients, whose CD19 + B cells expressed miR-155 under stimulation by BAFF 44 . As miR-155 plays a critical role in early B cell commitment and proliferation, it is reasonable to hypothesize that, after alemtuzumab immunodepletion, the memory B cells compartment undergoes a homeostatic proliferation driven by miR-155 through the mediation of BAFF. However, this mechanism needs to be further clarified in future experiments.
In the current study, miR-31 was found to be significantly down-regulated in alemtuzumab patients. It is well described that miR-31 positively correlates with T cell activation and induces the transcription of IL-2 in T cells by suppression of RhoA and activation of nuclear factor of activated T-cells (NF-AT) 45,46 . Moreover, miR-31 directly targets IL-25 transcription, with the consequent production of IL-12 and IL-23 by antigen-presenting cells and an increased Th1/Th17 response 47 . Thus, downregulation of miR-31 may significantly contribute to www.nature.com/scientificreports/ contain T cell activation. An analogous effect can be ascribed to the upregulation of miR-146a, an anti-inflammatory microRNA, which we found in the alemtuzumab group. Engagement of T-cell receptor (TCR) triggers expression of miR-146a which negatively targets NF-kB and inflammatory response 48 . Taken together, it is reasonable to hypothesize that the downregulation of miR-31 and upregulation of miR-146 might participate in the regulation of T cell responses and the Tregs/Th17 balance, thus favoring graft tolerance. Alemtuzumab, in fact, has been associated with an increase in peripheral levels of anti-inflammatory cytokines and with a suppression of Th1 and Th17 transcription factors 25,26 . This is further supported in our findings by the negative correlation found between miR-155 and IL-17A. The frequency of Treg cells seems to increase after alemtuzumab immunodepletion 24,27,28 . MicroRNAs have been found to be fundamental regulators of Tregs commitment and proliferation 49 . MiR-155 binds to the promoter region of FoxP3 50 . Increased FoxP3 transcription induces miR-155 expression, which further stimulates Tregs proliferation in a feedback loop 51 . Rudensky's group also showed that miR-146 is upregulated in Tregs 52 . MiR-31 is known to target FoxP3 mRNA and negatively influence its transcription 53 , on the contrary miR-23 is a FoxP3 transcriptional target and its expression is involved in Tregs commitment 54 . This is in line with our As ancillary finding, serum levels of three soluble co-signaling molecules, namely sCD27, B7.2 and sPD-L1, were increased one year after transplantation in no-induction group while remained unchanged in alemtuzumab group. Soluble CD27 is secreted by activated T-cells and plays a role in their maturation and proliferation. Increased levels have been detected in inflammatory and autoimmune diseases, viral infections and lymphoid malignancies, making it a potential marker for T-cell-mediated inflammation 55 . Similarly, soluble B7.2 is a costimulatory molecule of human T cells 56 and it has been found increased in systemic lupus erythematosus, asthma and leukemia [57][58][59] . It is reasonable to hypothesize that soluble B7.2 could have a similar stimulatory effect in no-induction group. Finally, higher levels of sPD-L1 were associated with poorer long-term graft outcome in kidney recipients 60 . The soluble form of this molecule may play as antagonist of PD-1/PD-L1 engagement, thereby stimulating proliferation and cytokine production of T cells 61 . Serum levels of a set of cytokines have been found significantly different one year after transplantation compared to baseline. IL-4, IL-13, IL-5, IL-9 and IL-22 are all reduced one year after transplantation. This finding is most likely due to the suppressing effects of the maintenance immunosuppression protocol (mainly CNI and steroids) on T cell activation. On the contrary, IFN-γ, a critical cytokine for the innate and adaptive immunity, is increased in both groups. This cytokine is produced mainly by NK and NKT cells and by CD4 + and CD8 + , it is, therefore, reasonable to hypothesize that the increase of IFN-γ is associated to the antigen-specific immune response.
Our study is not free of limitations. First, quantification of immune cell subsets by FACS is lacking. This could have strengthened our results, by correlating the presented molecular signature with the cellular findings 2,24 . Thus, we can only speculate on which immune cells are responsible for the observed miRNA dysregulation. Further, serum samples beyond one year after transplantation were not available for analysis. This would have deepened our knowledge about the repopulation of other immune cell subsets as well as the possible long-term effects of induction therapy. Furthermore, the explanatory and pilot nature of the study may solely suggest possible underlying processes, which have to be further confirmed by mechanistic studies. On the other side, it should be acknowledged that the current analysis represents the first study describing different miRNAs expression after LTx in a prospectively collected cohort after alemtuzumab induction.
In conclusion, this report provides the first evidence of miRNAs and cytokine dysregulation following alemtuzumab induction therapy in lung transplantation. Since these dysregulated mediators play a role in the homeostatic reconstitution of different B cell subsets and in the regulation of T cell response, their role in regulation of graft tolerance and in humoral complications can be envisaged. Further research, however, is necessary to correlate their expression with the activation of specific molecular pathways in different T and B cell subsets in order to increase our knowledge and ultimately improve the efficacy of our immunosuppressive strategies.

Methods
Study design. This is a comparative study, including serum samples of lung transplant recipients prospectively collected between June 2008 and December 2013 at two timepoints: time of transplantation and 12 months post-transplant. Within this period a total of 710 lung transplant recipients consented for prospective storage of biological samples including serum, plasma or bronchoalveolar lavage (BAL) for scientific purposes. Written informed consent was obtained from all recipients. Among them, 252 received alemtuzumab as induction therapy and for 30 patients, serum samples for the two timepoints defined by the study protocol were available. The remaining 222 patients could not be included in the analysis for the following reasons: (1) they did not have serum samples at both timepoints, (2) the serum samples were obtained at day of listing, (3) the collected biological samples did not include serum. The 30 control patients were chosen among the 233 patients who did not receive any induction therapy and had serum samples at both timepoints. Control patients were matched based on the following variables: gender, age, underlying diagnosis, CMV risk, type of transplantation and primary graft dysfunction (PGD) grade (Table 4).
Inclusion criteria were primary lung transplantation, adult age and availability of serum samples both at time of transplantation and 1 year after transplantation. Exclusion criteria were multi-organ transplantation, retransplantations and pediatric age. No patient had signs or diagnosis of acute cellular rejection, antibodymediated rejection, infection or CLAD at time of blood sampling. All experiments and methods were performed in accordance with relevant guidelines and regulations. The study has been approved by the Institutional Ethical Committee of the Medical University of Vienna [ECS 1729/2020] and was conducted according the declaration of Helsinki. Sanofi, Cambridge, USA) as a single intravenous dose of 30 mg after LTx at arrival at the intensive care unit (ICU) or no induction therapy. Recipients after alemtuzumab induction received a low-dose CNI-based maintenance immunosuppression, while recipients without induction therapy received a triple-drug protocol (IS protocols and target blood levels are shown in Table 1). During follow-up, in case of deterioration of kidney function, infections requiring hospital admission or rejection, the target blood levels were adapted to the low or high end of the respective target range (Table 1). Perioperative infectious prophylaxis was based on broadspectrum antibiotics or adapted to resistance testing. All patients received a lifelong Pneumocystis prophylaxis with trimethoprim-sulfamethoxazole or atovaquon. Prophylactic inhalation therapy with amphotericin B and gentamycin or according to pretransplant airway colonization was provided for 1-3 months. CMV prophylaxis included CMV hyperimmunoglobulines (POD 1,7,14 and 21) together with valganciclovir for a minimum of 3 months. In high-risk patients (donor + /recipient −) a 12 months prophylaxis was performed. Surveillance bronchoscopy with transbronchial biopsy (TBB) and BAL were performed 2 weeks and 1, 2, 3, 6, 12 months after transplantation and whenever clinically indicated. All patients received a chest computed tomography (CT) once a year. Biopsies were classified according to ISHLT criteria 62 . ACR grade A2 and LB grade B2 or higher were treated with a pulse of steroids for 3 days with consecutive dose tapering. In case of inadequate clinical response, ATG (2 mg/kg) was administered for 5 days.
Cytokine expression by flow cytometry. A total of 17 cytokines were isolated from 1:1 diluted serum samples using the LEGENDplexTM Human Th Cytokine Panel (12-plex), the LEGENDplexTM Human B Cell Panel (13-plex) and Human Immune Checkpoint Panel 1 (12-plex) as per the manufacturer's instruction. These panels are bead-based multiplex assays, which use fluorescence-encoded beads. Quantification was performed using BD LSRFortessa™ Cell Analyzer. Each patients' sample was measured in duplicates. Bead fluorescence was analyzed using LEGENDplexTM Software, which allows quantification of cytokines in pg/ml.
Genes targeted by miRNAs were identified using the multiMiR package. First, the "get_multimir" function was used to retrieve all targeted genes. The analysis was limited to experimentally validated miRNA-target interactions. MiRNA-target pairs with support type "negative" were not considered for further analysis. Next, the remaining target genes were mapped to KEGG pathways using the R package "org.Hs.eg.db". Finally, pathways significantly enriched in genes targeted by miRNAs were identified using Fisher's exact test. Across all performed tests, p-values were corrected for multiple testing using the "fdr" approach. For visualization purposes, significantly enriched pathways sharing at least 10 genes were connected, and the resulting graphs were plotted using the R package "ggraph".
Statistical analysis. Categorical variables were reported as absolute and relative frequencies (%), continuous variables as median (interquartile range, IQR) or mean (± standard deviation). Relative frequencies were calculated based on the number of patients alive in follow-up at the respective timepoint. Chi-square tests, Fisher exact tests, Mann-Whitney U-tests, or ANOVA were used to compare variables as applicable. Correlations were calculated using Spearman's correlation test. Data was analyzed using SPSS version 26.0 software or R 4.0.5 and graphics were designed with GraphPad Prism 6.