Rising Through the Ranks: Seasonal and Diel Patterns of Marine Viruses
Total Page:16
File Type:pdf, Size:1020Kb
bioRxiv preprint doi: https://doi.org/10.1101/2020.05.07.082883. this version posted May 7, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. Rising through the Ranks: Seasonal and Diel Patterns of Marine Viruses Gur Hevroni1,2 , José Flores-Uribe1,3, Oded Béjà1, and Alon Philosof1,4, 1Faculty of Biology, Technion - Israel Institute of Technology, Israel. 2Current address: Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel. 3Department of Plant Microbe Interactions, Max Planck Institute for Plant Breeding Research, Cologne, 50829, Germany. 4Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, United States. Virus-microbe interactions have been studied in great molec- infecting marine bacteria to be expressing AMGs in diur- ular details for many years in cultured model systems, yield- nal patterns, coupled with their host metabolism and repro- ing a plethora of knowledge on how viruses use and manipu- duction cycle (18–22). Relying on their host for propaga- late host machinery. Since the advent of molecular techniques tion, virus abundance is predicted to follow that of their and high-throughput sequencing, viruses have been deemed the hosts ("Kill-the-Winner" model (23), and the "Bank" model most abundant organisms on earth and methods such as co- (24)). These models predict that in any given environment, occurrence, nucleotide composition and other statistical frame- a small fraction of viruses is highly abundant while the rest works have been widely used to infer virus-microbe interactions, overcoming the limitations of culturing methods. However, their are rarer, waiting for the right conditions (i.e. host) to infect accuracy and relevance is still debatable, as co-occurrence does their hosts. Such abundance patterns are typical of micro- not necessarily mean interaction. Here, we introduce an eco- bial and viral communities (16, 24, 25), and their rank abun- logical perspective of marine viral communities and potential dance distribution often fits a log-normal curve (26). While interaction with their hosts, using analyses that make no prior these concepts have been useful for describing the distribu- assumptions on specific virus-host pairs. By size fractioning wa- tion of viruses in a given sample, our perception of the re- ter samples into "free viruses" and "microbes" (i.e. also viruses lationship between temporal abundance variation of marine inside or attached to their hosts) and looking at how viral groups viruses and host interaction remains largely obscure. Cur- abundance changes over time along both fractions, we show that rently only a few environmental studies of diel patterns in the viral community is undergoing a change in rank abundance marine viruses (18, 19, 22) and seasonality effects on viral across seasons, suggesting a seasonal succession of viruses in the communities (21, 27–32) have been reported, and none of Red Sea. We use abundance patterns in the different size frac- tions to classify viral populations, indicating potential diverse these incorporate both diel and seasonal time-scales. interactions with their hosts and potential differences in life his- tory traits between major viral groups. Finally, we show hourly To resolve whether seasonal and diel patterns exist in the resolved variations of intracellular abundance of similar viral viral community (specifically, double-strand DNA viruses of groups, which might indicate differences in their infection cy- marine prokaryotes) and to study the link between viral abun- cles or metabolic capacities. dance and the potential interaction with their host, we col- lected two 24-hour time-series of coastal water samples dur- marine viruses | seasonal succession | virus-host interaction | diel oscillation ing two seasons. We separated the cellular fraction (gDNA Correspondence: [email protected], [email protected] or metagenome), free virus fraction (vDNA or virome), and RNA fraction (RNA or metatranscriptome. Fig. S1b), al- lowing us to look at these fractions in both seasonal and diel Introduction time-scales. We used the Bank model (24) as a framework to Viruses of marine microorganisms outnumber their hosts classify viral active or inactive state based on their seasonal and are considered the most abundant biological entities in abundance across sample types (i.e. RNA, gDNA or vDNA), the ocean (1). They comprise the largest reservoir of ge- that is, classifying the abundant viruses as the "Active" group, netic diversity in the oceans, and they are major partici- and the rest as the "Bank" group ("inactive") (Fig. 1a). By ap- pants in oceanic biotic and abiotic processes (2–7). It is plying this framework, we show that most of the seasonally estimated that ~50% of marine microbial production is me- abundant viruses (Active) go through a seasonal change in diated by virus-induced release of dissolved organic matter abundance rank and that the high similarity in viral richness or ’viral shunt’ (8, 9). Viruses are suggested to alter mi- across seasons is largely derived from low abundance viruses crobial primary and secondary production (10–12), impact (Bank). Furthermore, we assigned a taxonomic and ecolog- the population dynamics and diversity of microbial commu- ical classification to hundreds of viral contigs with different nities (13), and play an indispensable role in marine bio- patterns of abundance in the viral and cellular fractions. This geochemical fluxes (14). Virus proliferation strongly de- classification indicates that similar seasonal abundance pat- pends on host metabolism (15), often by manipulating the terns of free viruses do not directly translate to similar abun- host’s metabolic and transcriptional machinery using aux- dance of these viruses inside their hosts (cellular samples), in iliary metabolic genes (AMGs) (5, 16, 17). Additionally, contrast to what is often implied in viral metagenomic (i.e. recent environmental studies have reported several viruses viromes) studies. Finally, we observed many virus-host in- Hevroni et al. | bioRχiv | May 7, 2020 | 1–18 bioRxiv preprint doi: https://doi.org/10.1101/2020.05.07.082883. this version posted May 7, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. teractions (predicted by virus abundance in the cellular and winter viral fraction Fig. S3). A similar mechanism, where RNA samples) having a differential signature between light high seasonal host diversity generates high viral diversity has and dark hours of the day. Such light-dependent interac- been recently hypothesized to be in effect in the Arctic ma- tions are prevalent in cyanophage populations (viruses infect- rine environment (35). Our observations suggest that sea- ing Cyanobacteria), but also include several viral populations sonal variations in host diversity could drive an increase in predicted to infect a heterotrophic host. viral diversity in a coastal and more temperate marine envi- ronment, and are not confined to the Arctic. Results and discussion We classified the Active viral group into viral clusters (VC) Two sets of 24-hour time-series samples with two hour in- based on gene-sharing network (36) (Fig. 1b and Extended tervals between samples were collected for this study. The Information) and a taxonomic annotation was assigned based first set was collected during the summertime (11-12 August, on the identity of Refseq viral genomes found in that clus- 2015) when the water column was stratified and the second ter, or according to the closest Refseq genomes that repre- was collected during late winter (7-8 February, 2016) when sented a distinct hub. We kept the annotation at the viral the upper water column was mixed. In addition, we used genus level (38) (e.g. cyanopodovirus, SAR11 virus, etc.) metagenomic data from a diel sample of Red Sea pelagic and considered highly connected VCs with the same taxo- water collected during fall time (October, 2012) (33) (Fig. nomic annotation as "viral populations". Most of the VCs S1a). Assembled contigs (length ≥ 5kb) from all 48 DNA did not cluster with a Refseq representative and were classi- samples (cellular and viral fractions) were filtered for pre- fied as "Uncultured virus" in our data (Table S2), accentuat- dicted viral contigs using two virus classifiers (see Extended ing the high proportion of yet uncharacterized viruses in the Information). A total of 32,496 contigs were assigned as environment (2, 18, 25, 35). A closer look at the abundance viral by both algorithms (minimum probability cutoff 0.75) of different viral populations across seasons and sample types and were mostly enriched in the viromes compared to the reveals different patterns for distinct populations. For exam- metagenomes (Fig. S2a). These contigs were considered of ple, cyanophages, one of the most abundant viral groups in viral origin for downstream analysis. The rank abundance our data, was grouped into four populations according to their distribution of the viral contigs best fitted a log-normal dis- viral families, Cyanomyoviridae, Cyanopodoviridae, and two tribution in all the datasets (34) (i.e. different sample types Cyanosiphoviridae populations (Fig.2). The cyanomyovirus and seasons. (Fig. S2b) and Table S1), indicative of a rel- population (300 contigs) was abundant in both viral and cel- atively small number of viruses with high abundance, while lular samples in all seasons. The cyanopodovirus population the majority have a very low abundance (in agreement with (220 contigs) was abundant in the viral samples across sea- the Bank model and other reports (18, 21, 24, 25)). We des- sons, while in the cellular samples it was highly abundant ignated the most abundant viral contigs as the Active group in the summer, but mostly absent from the autumn and win- and the rest as the Bank group (abundance threshold was set ter samples.