The future is now: single-cell genomics of bacteria and archaea

Abstract

Interest in the expanding catalog of uncultivated microorganisms, increasing recognition of heterogeneity among seemingly similar cells, and technological advances in whole-genome amplification and single-cell manipulation are driving considerable progress in single-cell genomics. Here, the spectrum of applications for single-cell genomics, key advances in the development of the field, and emerging methodology for single-cell genome sequencing are reviewed by example with attention to the diversity of approaches and their unique characteristics. Experimental strategies transcending specific methodologies are identified and organized as a road map for future studies in single-cell genomics of environmental microorganisms. Over the next decade, increasingly powerful tools for single-cell genome sequencing and analysis will play key roles in accessing the genomes of uncultivated organisms, determining the basis of microbial community functions, and fundamental aspects of microbial population biology.

Fig. 1

Methods for microbial genomics. (a) Standard metagenomics and sequence 'binning' to produce composite microbial genomes. (b) Targeted metagenomics and sequence 'binning' to produce composite microbial genomes. (c) Targeted enrichment of an organism to produce a single composite microbial genome. (d) Culture-based isolation for production of an axenic culture and a clonal microbial genome. (e) Multiplex PCR-based single-cell gene sequencing to obtain the sequence of multiple loci in single cells. (f) Single-cell genome sequencing utilizes cell isolation and single-cell WGA to produce single-cell microbial genomes. (g) Table summarizing characteristics of the genomic methods indicated in parts (a–f).

Fig. 2

Sources of contamination and the effect of reaction volume. (a) Three major sources of contamination: sample, laboratory, and reagent. (b) Schematic showing cross section of fluid volumes at the microliter, nanoliter, and picoliter scales. Fixed-concentration contaminant molecules are indicated. (c) Features of WGA reactions at the three volume scales.

Fig. 3

Two classes of cell isolation methods, random encapsulation and micromanipulation. Four methods for random encapsulation: (a) manual dilution, (b) microfluidic array, (c) flow cytometry, (d) microdroplet emulsion. Four methods for micromanipulation: (e) micropipetting, (f) microfluidic flow, (g) laser tweezers, (h) optoelectronic tweezers. (i) Table summarizing characteristics of the cell isolation methods indicated in parts (a–h).

Fig. 4

Methods for WGA. (a) tinker-adapter PCR. (b) Interspersed repetitive sequence PCR. (c) Primer extension preamplification. (d) Degenerate oligonucleotide-primed PCR. (e) Displacement degenerate oligonucleotide-primed PCR. (f) MDA. (g) Single primer isothermal amplification. (h) MALBAC. (i) Table summarizing characteristics of the WGA methods indicated in parts (a–h).

Fig. 5

Amplification bias and chimerism in five single-cell data sets. (a) Coverage of the segmented filamentous bacteria (SFB) genome by nonchimeric reads based on independent WGA and sequencing of five SFB filaments from mouse gut (Pamp et al., 2012). The inset shows Pearson correlation coefficients for the coverage profile of all pairs of cells revealing very weak correlation in the high-quality read bias from cell to cell. (b) Coverage of the SFB genome by MDA-induced chimeric reads from the same five SFB filaments introduced in part (a). The inset shows Pearson correlation coefficients for the coverage profile of all pairs of cells revealing very weak correlation in the distribution of artificial chimeras from cell to cell.