摘要
出身背景
随着抗微生物耐药性（AMR）风险的不断上升，能够全面评估临床/环境样本携带的抗微生物耐药性负担的分析选择有限。食物可能是人类AMR细菌的潜在来源，但其在推动AMR临床传播方面的意义尚不清楚，这主要是由于缺乏全面而敏感的监测和评估工具。宏基因组学是一种独立于培养的方法，非常适合揭示未知细菌群落中存在的特定微生物性状的遗传决定因素，如AMR。尽管它很受欢迎，但非选择性测序样本宏基因组的传统方法（即霰弹枪宏基因组学）有几个技术缺陷，导致其对AMR评估的有效性不确定；例如，抗性相关基因的低发现率是由于它们在庞大的宏基因组中的自然小基因组足迹。在这里，我们描述了一种靶向耐药性测序方法的开发，并展示了其在与几种零售食品相关的细菌AMR基因图谱表征中的应用。
后果
使用定制的诱饵捕获系统靶向4000多个参考AMR基因和263个质粒复制子序列的靶向宏基因组测序工作流程针对模拟和样品衍生的细菌群落制剂进行了验证。与霰弹枪宏基因组学相比，靶向方法始终提高了抗性基因靶标的回收率，并大大提高了靶标检测效率（> 300倍）。对36份零售获得的食品样本（新鲜芽苗菜 = 10; 绞肉，n = 26）及其相应的细菌富集培养物（n = 36）揭示了AMR基因的身份和多样性的深入特征，其中大多数在其他方面未被全宏基因组鸟枪测序方法检测到。此外，我们的研究结果表明，食源性伽马射线菌可能是食品相关AMR遗传决定因素的主要来源，并且所选高风险食品的抗性结构在很大程度上由微生物组组成决定。
结论
对于基于宏基因组测序的AMR监测，本文提出的靶标捕获方法代表了一种更灵敏和有效的方法来评估复杂食物或环境样本的耐药性。这项研究还进一步表明，零售食品是多种耐药基因的载体，表明其对AMR的传播有潜在影响。
Abstract
Background
With the escalating risk of antimicrobial resistance (AMR), there are limited analytical options available that can comprehensively assess the burden of AMR carried by clinical/environmental samples. Food can be a potential source of AMR bacteria for humans, but its significance in driving the clinical spread of AMR remains unclear, largely due to the lack of holistic-yet-sensitive tools for surveillance and evaluation. Metagenomics is a culture-independent approach well suited for uncovering genetic determinants of defined microbial traits, such as AMR, present within unknown bacterial communities. Despite its popularity, the conventional approach of non-selectively sequencing a sample’s metagenome (namely, shotgun-metagenomics) has several technical drawbacks that lead to uncertainty about its effectiveness for AMR assessment; for instance, the low discovery rate of resistance-associated genes due to their naturally small genomic footprint within the vast metagenome. Here, we describe the development of a targeted resistome sequencing method and demonstrate its application in the characterization of the AMR gene profile of bacteria associated with several retail foods.

Result
A targeted-metagenomic sequencing workflow using a customized bait-capture system targeting over 4,000 referenced AMR genes and 263 plasmid replicon sequences was validated against both mock and sample-derived bacterial community preparations. Compared to shotgun-metagenomics, the targeted method consistently provided for improved recovery of resistance gene targets with a much-improved target detection efficiency (> 300-fold). Targeted resistome analyses conducted on 36 retail-acquired food samples (fresh sprouts, n = 10; ground meat, n = 26) and their corresponding bacterial enrichment cultures (n = 36) reveals in-depth features regarding the identity and diversity of AMR genes, most of which were otherwise undetected by the whole-metagenome shotgun sequencing method. Furthermore, our findings suggest that foodborne Gammaproteobacteria could be the major reservoir of food-associated AMR genetic determinants, and that the resistome structure of the selected high-risk food commodities are, to a large extent, dictated by microbiome composition.

Conclusions
For metagenomic sequencing-based surveillance of AMR, the target-capture method presented herein represents a more sensitive and efficient approach to evaluate the resistome profile of complex food or environmental samples. This study also further implicates retail foods as carriers of diverse resistance-conferring genes indicating a potential impact on the dissemination of AMR.

https://link.springer.com/article/10.1186/s40793-023-00482-0