2025 年在世界卫生组织已选定流行株后，H3N2 又出现多重突变，最终在美洲与欧洲引发了更早且更重的流感季，说明按上一季流行谱“猜测”配方的年度策略天然有滞后风险。常规疫苗主要依赖适应性免疫识别特定抗原位点，而这些位点恰是病毒最易变异的区域，因此当演化速度超过制备节奏时，保护缺口会扩大。
加州理工学院 Pamela Björkman 团队提出 60 面“马赛克”冠状病毒纳米颗粒：每颗含 8 个刺突蛋白片段（1 个来自 SARS‑CoV‑2，其余来自其他 sarbecovirus），在小鼠和食蟾猴中更容易诱导对保守区域的 B 细胞反应，并对未纳入设计的 SARS 仍有保护。为提升可生产性与可监管性，他们正在推进 mRNA 版本，由宿主细胞膜组装颗粒；与此同时，杜克大学用 80,000 个血凝素变体的混合物把免疫重点从可变“头部”转向更稳定的“茎部”，但政治上对 mRNA 的质疑仍可能影响推进。
斯坦福团队转向“先天免疫预激活”路线：用鼻喷疫苗在数周内每只小鼠接种 4 次，肺部 SARS‑CoV‑2 相关冠状病毒载量约下降 700 倍，且至少维持 3 个月，同时对金黄色葡萄球菌（Staphylococcus aureus）、鲍曼不动杆菌（Acinetobacter baumannii）及过敏反应也出现抑制。传统抗原特异性疫苗已有约 200 年应用历史，因此这类广谱方向更可能成为补充，而不是立即替代。
In 2025, after the WHO had selected strains for the flu vaccine, H3N2 accumulated multiple mutations, leading to an earlier and more severe season in America and Europe, showing the inherent lag risk in annual strain-guessing strategies. Conventional vaccines rely on adaptive immunity against specific antigenic sites, which are also among the fastest-changing viral regions, so protection gaps widen whenever viral evolution outruns formulation cycles.
The Caltech group led by Pamela Björkman built a 60-surface coronavirus mosaic nanoparticle with 8 spike fragments (1 from SARS‑CoV‑2 and the rest from other sarbecoviruses), and in mice and macaques it preferentially amplified B-cell responses to conserved regions while still protecting against a SARS virus not included in the design. To improve manufacturability, a parallel mRNA approach now programs host cells to assemble such mosaic particles, while a Duke strategy using 80,000 haemagglutinin variants shifts immunity toward the stable stalk instead of the variable head in influenza models.
At Stanford, researchers pursued an innate-readiness approach with a nasal vaccine given four times over several weeks, producing about a 700-fold lower lung viral burden in mice against SARS‑CoV‑2–related coronaviruses for at least three months and also reducing responses to Staphylococcus aureus, Acinetobacter baumannii, and allergens. The traditional vaccine platform has roughly a 200-year empirical record, so broad-spectrum methods are likely to remain complementary, but mouse-to-human translation and human immune-genetic diversity remain major uncertainties.
Source: Scientists are working on “everything vaccines”
Subtitle: A single jab could protect against a wide range of pathogens
Dateline: 4月 01, 2026 06:04 上午