aGVHD is one of the most frequent and lethal complications after allo HSCT, underscoring the need to develop novel therapies. To achieve this goal, aGVHD mechanisms needs to be further elucidated. Recently it was reported that miRNAs are modulating aGVHD. In addition miRNAs are also present in the human serum and regulate immune responses. Here, we hypothesize that serum miRNAs expression is deregulated in aGVHD and could play a role in aGVHD pathogenesis.
To identify miRNAs associated with aGVHD we performed serum miRNA expression analysis using deep-sequencing from allo HSCT recipients samples at the time of clinical suspicion of aGVHD. Peripheral blood (PB) samples were collected weekly until day 100+ and at the time of clinical diagnosis of aGVHD from allo HSCT patients enrolled into OSU11002. After serum separation, total RNA was extracted using Trizol. Libraries were constructed using the small RNA profiling kit and sequenced on the Solid analyzer. A mouse model of aGVHD (B6 mice donor splenocytes and BM cells transplanted to lethally irradiated F1 recipients) was used to assess serum miRNA expression in animals with aGVHD.
In this study we included 10 patients with aGVHD (bowel n=2; skin (n=5) and both skin and bowel aGVHD (n=3). Median age was 51.9, conditioning regimens were mainly non-myeloablative (n=9), with unrelated donors (n=9). PB samples from allo HSCT patients with no aGVHD and matched for age, disease, conditioning regimen, donor and timing of sample collection were obtained and used as controls. Sequence alignment was performed using miRBase. Normalization as reads per million was followed by quantiles. We compared miRNA expression between all patients with aGVHD (n=10) and controls (n=7) using class comparison (BRB). We found 7 miRNAs up-regulated (miR-146a, miR-323-b, miR-34c, miR-363, miR-4245, miR-29a, miR-181a* ) and 3 miRNAs down-regulated (miR-3168, miR-662, miR-550a) (Fold change (FC) >2, p<0.01). Since miR-146 and miR-29a were both involved in immune regulation we further validated these miRNAs by RT-PCR in the B6-F1 model of murine aGVHD. We found up-regulation of miR-146 FC 2, p<0.01 and miR-29a FC 4.9 p<0.01) in mice with aGVHD (n=6) with respect to controls (n=4). Next, we focus on miR-29a since our group found that this miR binds as ligands to TLR8. We hypothesized that serum miR-29a could bind to TLR8 of APCs activating NFkB and enhancing alloreactive responses during aGVHD. First, we examined whether extracellular miR-29a could activate dendritic cells (DCs). B6 splenocytes were stimulated with Dotap formulations (mimicking exosomes) of miR-29a. Negative controls included Dotap alone or Dotap-miR16 formulation. We found that CD69 expression measured by FACS is significantly elevated in CD11c+ DCs (34%), and CD8+ T cells (56%) populations treated with miR-29a compared to controls (p<0.01). CD86, a co-stimulatory molecule on DCs, was also significantly up-regulated after miR-29a stimulation (33%, p<0.01). To investigate whether T cells could be activated by the miR alone, independently of APCs, we isolated untouched resting T cells from mouse splenocytes suspension using the Pan-T cell isolation kit and stimulated them with Dotap-miR-29a, Dotap-miR-16 or Dotap alone. CD69 was not up-regulated under these culture conditions indicating that the activation of T cells was dependent on APCs activation. To further confirm that miR-29a could activate DCs, we isolated DCs from B6 mice using the pan DC isolation kit and repeated the above experiment. We found that miR-29a stimulation of DCs but not controls induced the up-regulation of both CD69 and CD86 (20%). Furthermore, miR-29a Dotap treatment of isolated DCs stimulated the release of TNFα in the supernatant (114.2±14.3 pg/ml vs. controls 26.98±2.09 pg/ml, p<0.01). We also performed coimmunoprecipitation assays for TLR8 in HEK-293 cells expressing GFP-TLR8 and treated with Dotap-miR-16, Dotap-miR-29a, or Dotap alone and determined miRNA levels by qRT-PCR. Only miR-29a expression was highly enriched (>50-fold). This binding leads to the activation of NFkB as measured by a NF-κB assay in TLR8–HEK-293 cells treated only with Dotap-miR-29a. Validation of these results using murine and human DCs are undergoing.
Altogether, our results indicate that serum miR-29a is up-regulated during aGVHD and activates DCs, likely by direct binding to TLR8 and inducing NFkB activation
No relevant conflicts of interest to declare.