Host population diversity as a driver of viral infection cycle in wild populations of green sulfur bacteria with long standing virus-host interactions

Abstract

Viral infections of bacterial hosts range from highly lytic to lysogenic, where highly lytic viruses undergo viral replication and immediately lyse their hosts, and lysogenic viruses have a latency period before replication and host lysis. While both types of infections are routinely observed in the environment, the ecological and evolutionary processes that regulate these different viral dynamics are still not well understood. In this study, we identify and characterize the long-term dynamics of uncultivated viruses infecting green sulfur bacteria (GSB) in a model freshwater lake sampled from 2005-2018. Because of the additional requirements for the laboratory cultivation of anaerobes like GSB, viruses infecting GSB have yet to be formally identified, leaving their diversity and impact on natural populations of GSB virtually unknown. In this study, we used two approaches to identify viruses infecting GSB; one in vitro based on flow cytometry cell sorting, the other in silico based on CRISPR spacer sequences. We then took advantage of existing bulk metagenomes derived from Trout Bog Lake covering the 2005-2018 period to examine the interactions between GSB hosts and their viruses across multiple years and seasons. From our data, GSB populations in Trout Bog Lake were found to be concurrently infected with at least 2-8 viruses each, many of which were lysogenic viruses; one GSB host population in particular was consistently associated with two lysogens with a nearly 100% infection rate for over 10 years. We illustrate with a theoretical infection model that such an interaction can be stable over multiple years given a low, but persistent level of lysogen induction in host populations with already high infection rates. Overall, our data suggest that single GSB populations are typically infected by multiple viruses at the same time, that lytic and lysogenic viruses can readily co-infect the same host population in the same ecosystem, and that host strain-level diversity might be an important factor controlling the lytic/lysogeny switch.