Microbial scout hypothesis and microbial discovery

Abstract

In this study, we examine the temporal pattern of colony appearance during cultivation experiments, and whether this pattern could inform on optimizing the process of microbial discovery. In a series of long-term cultivation experiments, we observed an expected gradual increase over time of the total number of microbial isolates, culminating in a 700-fold colony count increase at 18 months. Conventional thought suggests that long-term incubations result in a culture collection enriched with species that are slow growing or rare, may be unavailable from short-term experiments, and likely are novel. However, after we examined the phylogenetic novelty of the isolates as a function of the time of their isolation, we found no correlation between the two. The probability of discovering either a new or rare species late in the incubation matched that of species isolated earlier. These outcomes are especially notable because of their generality: observations were essentially identical for marine and soil bacteria as well as for spore formers and non-spore formers. These findings are consistent with the idea of the stochastic awakening of dormant cells, thus lending support to the scout model. The process of microbial discovery is central to the study of environmental microorganisms and the human microbiome. While long-term incubation does not appear to increase the probability of discovering novel species, the technology enabling such incubations, i.e., single-cell cultivation, may still be the method of choice. While it does not necessarily allow more species to grow from a given inoculum, it minimizes the overall isolation effort and supplies needed.

Fig 1

Accumulation of growth events over time in long-term cultivation experiments using microorganisms from marine (○) and soil (●) samples.

Fig 2

Proportion of singletons (species met once study-wide) from marine microorganisms (○) and soil microorganisms (●) at each time point.

Fig 3

Proportion of novel isolates among all isolates, separately for each time point for marine microorganisms (○), non-spore-forming soil microorganisms (■), and spore-forming soil microorganisms (□). In all cases, the isolate was considered novel if its 16S rRNA gene sequence shared less than 99% identity with previously reported isolates. Note the low value of R², indicating that the percentage of novel isolates does not depend on the time required to form a visible biomass.

Fig 4

Proportion of soil isolates (●) representing novel genera among all soil isolates. Data on weekly observations were pooled by month to minimize scatter. Scatter was due to the relatively low number of substantially novel isolates observed on a weekly basis. Regarding the low R² value, see the legend to Fig. 3.

Fig 5

Average phylogenetic novelty of isolates depending on the time of the appearance of marine microorganisms (○) and soil microorganisms (●). The low value of R² indicates that the novelty of an isolate does not depend on the time of its isolation.

Fig 6

Proportion of sporulating and nonsporulating soil microorganisms isolated in long-term cultivation experiments.