The bacterium Staphylococcus aureus is a major human pathogen for which the emergence of antibiotic resistance is a global public health concern. Infection severity, and in particular bacteraemia-associated mortality, has been attributed to several host-related factors, such as age and the presence of comorbidities. The role of the bacterium in infection severity is less well understood, as it is complicated by the multifaceted nature of bacterial virulence, which has so far prevented a robust mapping between genotype, phenotype and infection outcome. To investigate the role of bacterial factors in contributing to bacteraemia-associated mortality, we phenotyped a collection of sequenced clinical S. aureus isolates from patients with bloodstream infections, representing two globally important clonal types, CC22 and CC30. By adopting a genome-wide association study approach we identified and functionally verified several genetic loci that affect the expression of cytolytic toxicity and biofilm formation. By analysing the pooled data comprising bacterial genotype and phenotype together with clinical metadata within a machine-learning framework, we found significant clonal differences in the determinants most predictive of poor infection outcome. Whereas elevated cytolytic toxicity in combination with low levels of biofilm formation was predictive of an increased risk of mortality in infections by strains of a CC22 background, these virulence-specific factors had little influence on mortality rates associated with CC30 infections. Our results therefore suggest that different clones may have adopted different strategies to overcome host responses and cause severe pathology. Our study further demonstrates the use of a combined genomics and data analytic approach to enhance our understanding of bacterial pathogenesis at the individual level, which will be an important step towards personalized medicine and infectious disease management.

中文翻译：

---

金黄色葡萄球菌菌血症相关死亡率的克隆差异。

金黄色葡萄球菌是一种主要的人类病原体，其耐药性的出现引起了全球公共卫生的关注。感染的严重程度，尤其是与菌血症相关的死亡率，已归因于几种与宿主相关的因素，例如年龄和合并症。由于细菌毒力的多方面性质使细菌复杂化，迄今为止，细菌在感染严重性中的作用还不太清楚，迄今为止，细菌毒力已阻止了基因型，表型和感染结果之间的稳固映射。为了调查细菌因素在与菌血症相关的死亡率中所起的作用，我们对来自血液感染患者的一系列临床金黄色葡萄球菌分离株进行了表型鉴定，代表了两种全球重要的克隆类型，CC22和CC30。通过采用全基因组关联研究方法，我们鉴定并在功能上验证了几个影响溶细胞毒性表达和生物膜形成的遗传基因座。通过在机器学习框架内分析包含细菌基因型和表型以及临床元数据的汇总数据，我们发现在最能预测不良感染结果的决定因素中存在明显的克隆差异。尽管溶细胞毒性增加和生物膜形成水平降低预示着CC22背景菌株的感染致死风险增加，但这些毒力特异性因素对与CC30感染相关的死亡率几乎没有影响。因此，我们的结果表明，不同的克隆可能采取了不同的策略来克服宿主反应并引起严重的病理。我们的研究进一步证明了结合使用基因组学和数据分析方法来增强我们对个体细菌发病机制的了解，这将是朝着个性化医学和传染病管理迈出的重要一步。