5.3. Cautionary Tales from Evolution: Trade-Offs Are Favorable, Trade-Ups Are Not
5.3.进化的警示故事：权衡利弊是有利的，取舍是不利的
As emphasized many times in this review, caution is warranted when choosing a phage or phage combination for use in therapy. For example, some phage(s) may drive a deleterious trade-up that increases bacterial resistance to one or more antibiotics, as opposed to steering a desired trade-off, which is obviously not a favorable outcome for the patient (114). For instance, coliphages T6 and U115 bind to Tsx porins, which also permit cell entry of the antibiotic albicidin. Bacteria that evolved resistance against each of the two phages or to the antibiotic harbored mutations in Tsx, resulting in increased cross-resistance to all three antimicrobial agents (37). This example highlights that careful characterization of phages is needed before deploying them for therapeutic use. Accurately predicting phage-driven trade-offs requires determination of phage receptor binding and would benefit from high-throughput screening of phage candidates to identify those that select for trade-offs useful in the clinic.
正如本综述多次强调的那样，在选择噬菌体或噬菌体组合用于治疗时必须谨慎。例如，一些噬菌体可能会导致有害的权衡，增加细菌对一种或多种抗生素的耐药性，而不是引导理想的权衡，这对病人来说显然不是一个有利的结果（114）。例如，大肠杆菌 T6 和 U115 与 Tsx 门蛋白结合，而 Tsx 门蛋白也允许抗生素 albicidin 进入细胞。对这两种噬菌体或抗生素都产生抗药性的细菌在 Tsx 中携带突变，导致对这三种抗菌剂的交叉抗药性增加 ( 37)。这个例子突出表明，在将噬菌体用于治疗之前，需要仔细鉴定噬菌体的特性。准确预测噬菌体驱动的权衡需要确定噬菌体受体的结合情况，而对候选噬菌体进行高通量筛选以确定那些能选择在临床上有用的权衡的噬菌体将使我们受益匪浅。
Phages are biological entities, with the expected potential to evolve greater phenotypic diversity over time, if they undergo genetic changes during patient treatment. Mounting evidence suggests that nongenetic phenotypic variation (physiological differences) can accumulate in phage populations, as recently reviewed by Igler (115). Such phenotypic variation in phages may complicate the interactions between bacteria and phages. Furthermore, the evolution of bacterial resistance in response to differing phages may take myriad evolutionary paths, and changes in bacterial diversity are difficult to predict (114). Variation in target bacteria populations raises similar concerns. For example, differing expression or underlying mutations in E. coli efflux pump gene tolC can drive pleiotropy between phage resistance and antibiotic resistance for phages that use TolC as a cellular receptor. In an example we discussed above, while a majority of the mutants resistant to the TolC-targeting phage U136B became more susceptible to antibiotics, a few variants were observed to become more drug resistant, likely due to synergistic pleiotropy (39). Thus, complexities of genotypic and phenotypic variation in both phage and bacteria—arising spontaneously as well as via their interactions—deserve much greater attention. This is important in order to refine our predictions for phage therapies designed to steer bacterial evolution and to minimize surprising pleiotropic effects and outcomes.
噬菌体是生物实体，如果在患者治疗过程中发生基因变化，随着时间的推移，噬菌体可能会进化出更多的表型多样性。越来越多的证据表明，非遗传表型变异（生理差异）可能会在噬菌体群体中累积，伊格勒（Igler）最近对此进行了综述（115）。噬菌体的这种表型变异可能会使细菌与噬菌体之间的相互作用复杂化。此外，细菌对不同噬菌体的抗性进化可能有无数种进化途径，细菌多样性的变化难以预测 ( 114)。目标细菌种群的变化也会引起类似的问题。例如，对于使用 TolC 作为细胞受体的噬菌体来说，大肠杆菌外排泵基因 tolC 的不同表达或潜在突变可能会导致噬菌体抗药性和抗生素抗药性之间的多义性。在我们上面讨论的一个例子中，虽然大多数对 TolC 靶向噬菌体 U136B 具有抗药性的突变体变得对抗生素更易感，但也观察到少数变体变得更耐药，这很可能是由于协同多效性所致 ( 39)。因此，噬菌体和细菌中基因型和表型变异的复杂性–自发产生的以及通过相互作用产生的–值得更多关注。这对于完善我们对噬菌体疗法的预测非常重要，噬菌体疗法旨在引导细菌进化，并最大限度地减少令人惊讶的多生物效应和结果。
6. CONCLUDING REMARKS AND FUTURE DIRECTIONS
6.结束语和未来方向
The nearly inevitable evolution of bacterial resistance to phage attack during patient treatment will continue to pose challenges for widespread development of therapeutic phages. It is remarkable that phage therapeutic approaches have historically worked safely and effectively despite abundant knowledge gaps regarding the evolution of phage resistance in clinical settings. The rules of engagement between lytic phages and their bacterial hosts, studied using in vitro and in vivo models in the laboratory, can sometimes translate to accurate predictions of phage utility in human patients, as observed during some recent cases of emergency personalized treatment (96, 108, 116). These observations suggest that phage-bacteria interactions may obey (at least some) general rules that are robust across disparate environments, ranging from well-oxygenated liquid growth medium housed in shaking culture flasks to the relative black box of oxygen-limited biofilm-structured infections in the human body. However, these supposed rules may reflect constraints such as the biophysics of phage attachment to binding receptors on host-cell surfaces or ability of phages to hijack key nodes in cellular metabolic networks to achieve intracellular replication despite environmental differences. We should thus avoid naïve assumptions regarding consistencies of phage-bacteria interactions and acknowledge that crucial details of bacterial resistance to phages in clinical settings remain largely unknown.
在患者治疗过程中，细菌对噬菌体攻击几乎不可避免地会产生抗药性，这将继续给治疗性噬菌体的广泛开发带来挑战。令人瞩目的是，尽管在临床环境中噬菌体抗药性的演变方面存在大量知识空白，但噬菌体治疗方法历来都能安全有效地发挥作用。在实验室中利用体外和体内模型研究的溶菌噬菌体与其细菌宿主之间的接触规则，有时能准确预测噬菌体在人类患者中的效用，正如在最近的一些紧急个性化治疗病例中观察到的那样（96、108、116）。这些观察结果表明，噬菌体与细菌之间的相互作用可能遵守（至少是某些）一般规则，这些规则在不同的环境中都能保持稳定，这些环境既包括摇瓶中氧气充足的液体生长培养基，也包括人体内氧气有限的生物膜结构感染的相对黑箱环境。然而，这些假定的规则可能反映了一些限制因素，如噬菌体附着到宿主细胞表面结合受体的生物物理学原理，或噬菌体劫持细胞代谢网络中关键节点的能力，以实现细胞内复制（尽管存在环境差异）。因此，我们应该避免天真地假设噬菌体与细菌相互作用的一致性，并承认临床环境中细菌对噬菌体产生耐药性的关键细节在很大程度上仍然未知。
To fill this knowledge gap, we recommend that attempts to isolate and characterize phage-resistant bacteria should be an increasing focus of clinical trials testing safety and efficacy of phage treatments (52). For example, bacteria harbor a wide variety of phage defense mechanisms (18, 19, 27, 30–34, 117, 118), but it is unclear which of these should be most impactful in contributing to treatment failures in the clinic. Moreover, if phage-bacteria coevolution occurs in the treated patient, it is uncertain how the dynamics and type of interaction could alter the course of treatment success. The spatial distribution of infecting bacteria (e.g., biofilm structures that may protect susceptible cells against collisions with phages) should influence the relative importance of certain defenses. Genotypes of target bacteria may also play crucial roles. For instance, bacteria may harbor prophages that prevent lytic phages from completing their intracellular replication (119). The relative frequencies of these prophage-carrying cells in a bacterial population are difficult to gauge accurately. These and many other factors can determine whether phage particles can gain access to host cells and successfully infect cell variants that may spatiotemporally differ in genotype, metabolism, and/or physiology. Therefore, characterizations of evolved phage-resistant bacteria in the clinic present vital topics for developing trusted strategies in phage therapy. Mitigating bacterial resistance during phage therapy is an exemplar of evolutionary medicine goals that seek to harness evolution thinking to improve the understanding and treatment of infectious diseases (120).为了填补这一知识空白，我们建议在评估噬菌体治疗的安全性和有效性的临床试验中，越来越多地关注分离和表征耐噬菌体细菌的尝试 (52)。例如，细菌拥有各种各样的噬菌体防御机制 (18, 19, 27, 30-34, 117, 118)，但目前尚不清楚哪些机制会对治疗失败产生最大影响。此外，如果噬菌体-细菌共进化发生在接受治疗的患者身上，那么这种相互作用的动态和类型将如何改变治疗成功的进程尚不清楚。感染细菌的空间分布（例如，可能保护易感细胞免于与噬菌体碰撞的生物膜结构）会影响某些防御机制的相对重要性。目标细菌的基因型也可能发挥关键作用。例如，细菌可能携带阻止裂解噬菌体完成细胞内复制的前噬菌体 (119)。这些携带前噬菌体的细胞在细菌群体中的相对频率很难准确评估。这些以及许多其他因素可以决定噬菌体颗粒是否能进入宿主细胞并成功感染可能在时空上基因型、代谢和/或生理上存在差异的细胞变异体。因此，对临床进化耐噬菌体细菌的表征，是为噬菌体治疗开发可靠策略的重要课题。减轻噬菌体治疗过程中的细菌耐药性，是进化医学目标的一个例子，该目标旨在利用进化思维来改善对传染病的理解和治疗 (120)。
The tools used in microbiology research have advanced tremendously since the discovery of phages and the early work that first used these viruses in attempts to cure infections in humans and other animals. These pioneering efforts occurred when laboratory techniques were crude relative to modern advances in culture technologies, molecular biology, high-throughput genetic sequencing, and microscopy. A goal for phage therapy is to characterize details of phage traits and interactions with host bacteria using modern methods that recapitulate the environments of clinical settings. Examples include laboratory microcosms such as pathogenicity models in tissue culture, cell systems with tissue differentiation (e.g., air-liquid interface culture), and animal models with relevant biotic and abiotic features (e.g., robust innate and adaptive immunity that mimic potential synergies between host defenses and therapy strategies) (121). Without these efforts, we face perpetual mismatches between presumed microbial fitness and (co)evolution in culture flasks versus actual outcomes in relevant circumstances. Importantly, these data should bring greater reality to studies of evolved phage resistance, with the possibility of accumulating knowledge that hastens our abilities to make clinical treatment decisions to benefit patient health and therapy outcomes. Given the power for target bacteria to leverage existing cellular defenses against virus attack and to rapidly evolve increased phage resistance, it is crucial to utilize experiments and analyses that address these perpetual challenges, in order to develop therapies that minimize the potency for bacteria to evolve and/or steer evolved phage resistance along favorable paths for treatment success.自噬菌体被发现以及早期尝试利用这些病毒治疗人类和动物感染以来，微生物研究的工具取得了巨大进步。这些开创性的工作发生在实验室技术还比较粗糙的年代，远远比不上现代培养技术、分子生物学、高通量基因测序和显微镜技术的进步。噬菌体治疗的一大目标是利用现代方法表征噬菌体特性以及它们与宿主细菌的相互作用细节，这些方法需要能够模拟临床环境。例如，实验室微观模型，例如组织培养中的致病性模型、具有组织分化的细胞系统（例如气-液界面培养）以及具有相关生物和非生物特征的动物模型（例如，强大的先天性和适应性免疫，模拟宿主防御和治疗策略之间潜在的协同作用） (121)。如果不进行这些尝试，我们就会永远面临培养瓶中推定的微生物适应性和（共）进化与实际相关环境下结果之间的不匹配。重要的是，这些数据应该为研究噬菌体抗性进化带来更大的真实性，有可能积累知识，从而加快我们做出临床治疗决策的能力，以改善患者健康和治疗效果。鉴于靶向细菌利用现有细胞防御系统抵抗病毒攻击并快速进化出更强噬菌体抗性的能力，因此至关重要的是利用实验和分析来应对这些持续存在的挑战，以开发能够最小化细菌进化潜力的治疗方法，并/或引导噬菌体抗性进化到有利于治疗成功的方向。