Bioimaging, GSK, Collegeville, PA, United States.
Non Clinical Safety, GSK, Collegeville, PA, United States.
Front Immunol. 2023 Jan 11;13:1081156. doi: 10.3389/fimmu.2022.1081156. eCollection 2022.
The goal of this study was to utilize a multimodal magnetic resonance imaging (MRI) and positron emission tomography (PET) imaging approach to assess the local innate immune response in skeletal muscle and draining lymph node following vaccination in rats using two different vaccine platforms (AS01 adjuvanted protein and lipid nanoparticle (LNP) encapsulated Self-Amplifying mRNA (SAM)). MRI and FDG PET imaging were performed temporally at baseline, 4, 24, 48, and 72 hr post Prime and Prime-Boost vaccination in hindlimb with Cytomegalovirus (CMV) gB and pentamer proteins formulated with AS01, LNP encapsulated CMV gB protein-encoding SAM (CMV SAM), AS01 or with LNP carrier controls. Both CMV AS01 and CMV SAM resulted in a rapid MRI and PET signal enhancement in hindlimb muscles and draining popliteal lymph node reflecting innate and possibly adaptive immune response. MRI signal enhancement and total FDG uptake observed in the hindlimb was greater in the CMV SAM vs CMV AS01 group (↑2.3 - 4.3-fold in AUC) and the MRI signal enhancement peak and duration were temporally shifted right in the CMV SAM group following both Prime and Prime-Boost administration. While cytokine profiles were similar among groups, there was good temporal correlation only between IL-6, IL-13, and MRI/PET endpoints. Imaging mass cytometry was performed on lymph node sections at 72 hr post Prime and Prime-Boost vaccination to characterize the innate and adaptive immune cell signatures. Cell proximity analysis indicated that each follicular dendritic cell interacted with more follicular B cells in the CMV AS01 than in the CMV SAM group, supporting the stronger humoral immune response observed in the CMV AS01 group. A strong correlation between lymph node MRI T2 value and nearest-neighbor analysis of follicular dendritic cell and follicular B cells was observed (r=0.808, P<0.01). These data suggest that spatiotemporal imaging data together with AI/ML approaches may help establish whether imaging biomarkers can predict local and systemic immune responses following vaccination.
本研究旨在利用多模态磁共振成像（MRI）和正电子发射断层扫描（PET）成像方法，评估大鼠接种两种不同疫苗平台（AS01 佐剂蛋白和脂质纳米颗粒（LNP）包裹的自我扩增信使 RNA（SAM））后骨骼肌和引流淋巴结的局部固有免疫反应。在基线、第 4、24、48 和 72 小时进行 MRI 和 FDG PET 成像，以评估后腿的巨细胞病毒（CMV）gB 和五聚体蛋白的疫苗接种情况，这些蛋白与 AS01、LNP 包裹的 CMV gB 蛋白编码 SAM（CMV SAM）、AS01 或 LNP 载体对照联合使用。CMV AS01 和 CMV SAM 均导致后腿肌肉和引流的腘淋巴结的 MRI 和 PET 信号迅速增强，反映了固有和可能的适应性免疫反应。与 CMV AS01 组相比，CMV SAM 组后腿的 MRI 信号增强和总 FDG 摄取更高（AUC 增加 2.3-4.3 倍），并且在 Prime 和 Prime-Boost 给药后，CMV SAM 组的 MRI 信号增强峰值和持续时间向右移动。虽然各组的细胞因子谱相似，但只有 IL-6、IL-13 和 MRI/PET 终点之间存在良好的时间相关性。在 Prime 和 Prime-Boost 疫苗接种后 72 小时对淋巴结切片进行成像质谱细胞术，以表征固有和适应性免疫细胞特征。细胞邻近分析表明，在 CMV AS01 组中，每个滤泡树突状细胞与滤泡 B 细胞的相互作用多于 CMV SAM 组，这支持了在 CMV AS01 组中观察到的更强的体液免疫反应。淋巴结 MRI T2 值与滤泡树突状细胞和滤泡 B 细胞的最近邻分析之间存在强烈相关性（r=0.808，P<0.01）。这些数据表明，时空成像数据结合人工智能/机器学习方法可能有助于确定成像生物标志物是否可以预测疫苗接种后局部和全身免疫反应。
