Contrasting Effects of Singlet Oxygen and Hydrogen Peroxide on Bacterial Community Composition in a Humic Lake
腐殖质的轻微激发在水生生态系统的表层水中产生活性氧（ROS）。富含腐殖质的湖泊中产生的丰富ROS包括单线态氧（1O2）和过氧化氢（H2O2）。
因为这些ROS的半衰期和毒性不同，我们比较了它们对腐殖质湖Grosse Fuchskuhle（德国东北部）地表水中微生物活性（14C-亮氨酸掺入）和细菌群落组成（BCC）的影响。为此，在2006年7月，2008年9月和2009年8月进行了从湖中采集的水样的实验。人工增加的1O2和H2O2浓度在相似程度上抑制了水样中的微生物活动，但是相应的ROS对BCC的影响。变化很大。通过16S rRNA基因克隆文库和RT-PCR DGGE进行的BCC分析揭示了主要细菌群的相对丰度和活性以及主要系统发育型的组成的ROS特异性变化。
这些变化在不同年份进行的三次实验中是一致的。多核杆菌必需品，Limnohabitans相关的种系型（Betaproteobacteria）和Novosphingobium acidiphilum（Alphaproteobacteria）的相对丰度增加或不受光敏1O2暴露的影响，但在H2O2暴露后降低。对于淡水AcI-B簇的放线菌，发现了相反的模式，其对1O2高度敏感但对H2O2暴露不敏感。此外，组特异性RT-PCR DGGE分析显示，与自由生活群体相比，附着颗粒的P. essentialarius和Limnohabitans相关系统型对1O2暴露表现出更高的抗性。这些结果意味着1O2作为紧密附属的多核杆菌和Limnohabitans相关系统型的生态位分离的一个因素。因此，光化学ROS产生引起的氧化应激应被视为决定环境相关细菌群的丰度，活性和系统型组成的环境变量，特别是在光照和腐殖质丰富的水域中。
单线态氧（1O2）具有高活性，在水中的半衰期约为3.5μs[17]，并且通过氧化脂质，核酸和蛋白质导致细胞损伤[18][19]。相比之下，H2O2在淡水中的半衰期长达8小时[20]。此外，H2O2通过生物膜扩散，主要与铁 - 硫簇反应，导致随后的细胞内羟基自由基形成和生物分子的损伤[21]。因此，由1O2和H2O2引起的细胞损伤的潜力显著不同。
类胡萝卜素在光合细菌和植物叶绿体中是不可避免的，以防止光系统产生1O2 [42]，[43]。非光合细菌也表现出类胡萝卜素，其可能作为由细胞光敏剂如黄素（flavin)[42]或各种细胞外来源产生的1O2的猝灭剂。细胞清除剂，包括氨基酸如L-组氨酸和trytophan，还原硫醇（谷胱甘肽，硫氧还蛋白），肌氨酸赖氨酸和多胺也使细胞损伤最小化1O2。这些清除剂在与1O2反应后需要再生，因此需要激活参与调节细胞氧化还原稳态的酶（综述见[43]）。
过氧化氢被细胞酶如过氧化氢酶(catalase)和过氧化物酶（谷胱甘肽过氧化物酶和过氧化物酶）降解[21]。增加的H2O2浓度导致铁 - 硫簇的氧化和分解导致细胞快速死亡，这在电子传递链组分中是常见的。过氧化氢与游离铁（II）一起通过芬顿反应导致形成高毒性的羟基径向，其迅速与大多数细胞组分反应并促进细胞死亡。
因此，H2O2的细胞水平紧密平衡，细胞反应受到例如OxyR或PerR的良好调节，OxyR或PerR协调H2O2降解基因，谷胱甘肽转换，产生氧化还原缓冲液作为谷氧还蛋白和硫氧还蛋白以及参与控制铁的基因代谢。因此，所有具有有氧代谢的细菌都需要防御H2O2暴露的防御系统。这可以解释为什么H2O2对BCC的影响比环境中的1O2小得多。
Abstract
Light excitation of humic matter generates reactive oxygen species (ROS) in surface waters of aquatic ecosystems. Abundant ROS generated in humic matter rich lakes include singlet oxygen (1O2) and hydrogen peroxide (H2O2).
Because these ROS differ in half-life time and toxicity, we compared their effects on microbial activity (14C-Leucine incorporation) and bacterial community composition (BCC) in surface waters of humic Lake Grosse Fuchskuhle (North-eastern Germany). For this purpose, experiments with water samples collected from the lake were conducted in July 2006, September 2008 and August 2009. Artificially increased 1O2 and H2O2 concentrations inhibited microbial activity in water samples to a similar extent, but the effect of the respective ROS on BCC varied strongly. BCC analysis by 16S rRNA gene clone libraries and RT-PCR DGGE revealed ROS specific changes in relative abundance and activity of major bacterial groups and composition of dominating phylotypes.
These changes were consistent in the three experiments performed in different years. The relative abundance of Polynucleobacter necessarius, Limnohabitans-related phylotypes (Betaproteobacteria), and Novosphingobium acidiphilum (Alphaproteobacteria) increased or was not affected by photo-sensitized 1O2 exposure, but decreased after H2O2 exposure. The opposite pattern was found for Actinobacteria of the freshwater AcI-B cluster which were highly sensitive to 1O2 but not to H2O2 exposure. Furthermore, group-specific RT-PCR DGGE analysis revealed that particle-attached P. necessarius and Limnohabitans-related phylotypes exhibit higher resistance to 1O2 exposure compared to free-living populations. These results imply that 1O2 acts as a factor in niche separation of closely affiliated Polynucleobacter and Limnohabitans-related phylotypes. Consequently, oxidative stress caused by photochemical ROS generation should be regarded as an environmental variable determining abundance, activity, and phylotype composition of environmentally relevant bacterial groups, in particular in illuminated and humic matter rich waters.
Singlet oxygen is highly reactive, exhibits a half-life time in water of ∼3.5 μs [17], and causes cell damage by oxidation of lipids, nucleic acids, and proteins [18], [19]. In contrast, H2O2 has a half-life time of up to 8 hours in freshwater [20]. Moreover, H2O2 diffuses through biological membranes and mainly reacts with iron-sulphur clusters leading to subsequent intracellular hydroxyl radical formation and damage of biomolecules [21]. Hence, potentials for cell damage caused by 1O2 and H2O2 differ substantially.
Comparison of 1O2 and H2O2 Toxicity
Moderately increased 1O2 and highly increased H2O2 concentrations caused similar inhibition of 14C-leucine incorporation suggesting different toxic potentials of 1O2 and H2O2. This finding also indicates that small changes of 1O2 generation (frequent during diurnal changes in sunlight intensity) may hamper microbial activity in surface waters of humic lakes. In contrast, only large changes in H2O2 concentrations may affect the activity of dominant bacterial species. However, the H2O2 concentrations applied in our experiments were not exaggerated and the natural potential of H2O2 formation in 0.22 μm filtered lake water of the SW basin was high (Fig. S7). In H2O2 depleted water samples, H2O2 concentrations in the μM range can be reached rapidly after irradiation with sunlight or UV-A/B which has been frequently observed for boreal lakes [25], [26]. Microorganisms strongly contribute to the decay of H2O2 [27]. This is indicated by 2.4-fold higher H2O2 decay rates in our unfiltered water samples compared to those filtered through 0.22 μm (Materials S1). Obviously, the bacterial community or at least some phylotypes can detoxify H2O2 and therefore balances H2O2 levels in their environment. This notion is in line with earlier findings that bacteria are involved in H2O2 degradation in marine surface waters [27] and that H2O2 degradation by some bacterial populations is important for growth of other bacteria in aquatic environments [28]. Hence, bacteria thriving in surface waters of humic lakes are well adapted to H2O2 exposure and may prevent accumulation of toxic H2O2 concentrations.
Defence Mechanisms Against Environmental ROS Exposure
Details on the presence of molecular response mechanisms against environmental ROS exposure in typical freshwater bacteria are elusive. Recently, molecular defence systems against 1O2 exposure were found in bacteria [42], [43] and defence strategies against H2O2 generated in aerobic metabolism are known in detail for several bacterial model systems [21].
Carotenoids are inevitable in photosynthetic bacteria and in the chloroplasts of plants to prevent photosystem based generation of 1O2 [42], [43]. Non-photosynthetic bacteria also exhibit carotenoids, which likely serve as quenchers of 1O2 generated by cellular photosensitizers such as flavins [42] or by various extracellular sources. Cellular scavengers, which include amino acids such as L-histidine and trypotphan, reduced thiols (glutathione, thioredoxin), mycosoprine lysine and polyamines also minimize cellular damages by 1O2. Such scavengers need to be regenerated after their reaction with 1O2, and therefore enzymes involved in adjusting the cellular redox homeostasis need to be activated (reviewed in [43]).
In photosynthetic Alphaproteobacteria, response mechanisms to 1O2 exposure are controlled by the alternative sigmafactor RpoE, which is bound to the anti-sigmafactor ChrR under non-stress conditions. The release of RpoE from ChrR after 1O2 exposure triggers the induction of genes encoding stress response mechanisms and further regulatory factors, including RpoHII and several small regulatory RNAs [42]. Homologs of these sigmafactors are conserved in photosynthetic Alphaproteobacteria and have been found in several Beta- and Gammaproteobacteria lineages [45]. Genomes of species representing abundant freshwater bacterial clades did not harbour homologous genes. Hence, defence systems and their control in abundant freshwater bacteria may substantially differ from established bacterial model systems.
Very likely, individual bacterial lineages use different strategies to overcome natural 1O2 exposure, which could explain very well the species specific sensitivity to 1O2 exposure in our study.
Hydrogen peroxide is detoxified by cellular enzymes such as catalases and peroxidases (glutathione peroxidases and peroxiredoxin) [21]. Increased H2O2 concentrations lead to rapid cell death by the oxidation and disassembly of iron-sulphur clusters, which are common in electron transport chain components. Hydrogen peroxide together with free iron(II) leads to the formation of highly toxic hydroxyl radials by the Fenton reaction, which rapidly react with most cellular components and facilitate cell mortality. Therefore, cellular levels of H2O2 are tightly balanced and the cellular response is well regulated by, for example, OxyR or PerR which coordinate genes for H2O2 degradation, glutathione turnover, production of redox buffers as glutaredoxin and thioredoxin as well as genes involved in controlling iron metabolism. All bacteria with an aerobic metabolism, therefore, require defence systems against H2O2 exposure. This may explain, why H2O2 had a much smaller effect on BCC compared to 1O2 in the environment.
CONCLUSION
From our data we conclude that differences in sensitivity to 1O2 and H2O2 may explain the negative correlation in abundance of Actinobacteria and Betaproteobacteria in the surface waters of Lake Grosse Fuchskuhle and elsewhere. The exclusion of specific bacterial lineages from humic matter rich particles and the presence of species-like taxa due to ROS specific separation of ecological niches should be regarded as an ecological factor shaping natural microbial communities. Hence, temporal and spatial differences in ROS generation, particularly in humic matter rich aquatic ecosystems, have the potential to affect major microbial processes and their rates. For example, niche separation by ROS has strong implications for bacterial adaptation and evolution in natural ecosystems. We propose that changes in 1O2 exposure have a larger impact on BCC than H2O2, because 1O2 is i) more toxic compared to H2O2 and ii) defence mechanisms against H2O2 are present in all aerobic organisms, whereas putative defences against singlet oxygen exposure may only occur in bacteria specifically adapted to cellular or environmental 1O2 formation. Further, insights into the molecular mechanisms of cellular defences against environmental ROS in general and singlet oxygen in particular are necessary to understand in detail the role of 1O2 and H2O2 for controlling activity and composition of aquatic microbial communities.
SOURCE:
Stefanie P. Glaeser, 1 , 2 Bork A. Berghoff, 1 , 3 Verena Stratmann, 1 Hans-Peter Grossart, 4 , 5 , * and Jens Glaeser 1 , *
PLoS One. 2014; 9(3): e92518.
Published online 2014 Mar 25. doi: [10.1371/journal.pone.0092518]
PMCID: PMC3965437
PMID: 24667441
Contrasting Effects of Singlet Oxygen and Hydrogen Peroxide on Bacterial Community Composition in a Humic Lake https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3965437/