Complications from sepsis are largely driven by maladaptive immune responses. Platelets impact innate and adaptive immune responses, but whether platelets regulate adaptive immune responses in sepsis, including CD8+ T cell responses, remains unknown. To explore a possible contribution of platelets to the altered immune response in sepsis, we used deep sequencing of ribosome-protected RNA (RPR) fragments (also called ribosome footprint profiling) to acquire a global picture of translation in platelets from healthy and septic individuals. Using this method, we identified altered translation of transcripts related to adaptive CD8+T cell responses. We further confirmed these changes using megakaryocyte cultures, and an in vivo model of murine sepsis. Finally, we employed in vitro models to examine their functional relevance.
Genome-wide sequencing of ribosome-protected mRNA fragments of human platelets isolated from septic patients and healthy donors revealed more than 1,000 genes under active translation in platelets. Sepsis altered the total RNA expression and translation of more than 300 genes by at least 2 fold compared to healthy individuals. Among these, classic MHC class I genes including HLA-A, B, and C were actively translated, and increased in sepsis. For example the expression of HLA-A increased almost 3 fold (RPR of healthy 62±77 vs sepsis 166±340; N=4-5). HLA-A mRNA expression changes were confirmed by real-time qPCR. We further demonstrated consistent changes in platelet surface expression of MHC class I proteins by flow cytometry.
Since platelets are anucleated and don't actively transcribe RNAs, we hypothesized that the increased mRNA of HLA-A in platelets initiated within megakaryocytes. To mimic the acute inflammation in vitro, human CD34+ derived megakaryocytes were treated with either interferon gamma at 250U/mL or LPS that was supplemented with lipid binding protein and CD14 at 100ng/mL for 16 hours. Treatments significantly increased HLA-A RNA expression (IFN-gamma 443±163% and LPS 161±41% as compared to untreated, N=5, P<0.01,) and cell surface MHC class I protein expression (MFI: IFN-gamma 220±60% and LPS 155±28% as compared to untreated, N=6, P<0.01).
To examine whether sepsis directly induced MHC class I in mice, we used a murine sepsis model induced by cecal ligation and puncture (CLP). CLP induced approximately 40% mortality and a 50% decrease of platelet count. Deep sequencing of ribosome protected RNAs from control and CLP mice platelets showed a 6 fold increase in murine MHC class I expression in platelets (H2D and H2K) after CLP, (RPR counts: H-2D 36±29, 227±103; H-2K 2.5±1.3 vs 16±2.6; N=3). H2D protein increased on the platelet surface by 1.5 fold 48 h after CLP. This was associated with a significantly decreased CD8+ T cell population in the spleen.
Activated platelets have been shown able to prime CD8+ T cells through MHC class I. To measure antigen cross-presentation by MHC class I, isolated platelets from CLP or sham control mice were pulsed with ovalbumin at 0.5mg/mL for 2 hours, and antigen cross-presentation was detected by an antibody specifically targeting the ovalbumin peptide SIINFEKL presented by H-2K using flow cytometry. Our studies indicated antigen cross-presentation by platelets significantly increased in the CLP group (MFI: Sham 21±2, CLP 42±24, N=3-8, P<0.05) compared to controls.
Together, our studies demonstrate dynamic regulation of MHC class I on platelets. Expression of MHC class I mRNAs and active translation increase during sepsis in both human and mouse platelets. Furthermore, antigen cross-presentation by platelet MHC class I could be upregulated in sepsis and its effect on CD8+ T cell responses is under investigation.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.