Rapid functional activation of a horizontally transferred eukaryotic gene in a bacterial genome in the absence of selection

Abstract

Horizontal gene transfer has occurred between organisms of all domains of life and contributed substantially to genome evolution in both prokaryotes and eukaryotes. Phylogenetic evidence suggests that eukaryotic genes horizontally transferred to bacteria provided useful new gene functions that improved metabolic plasticity and facilitated adaptation to new environments. How these eukaryotic genes evolved into functional bacterial genes is not known. Here, we have conducted a genetic screen to identify the mechanisms involved in functional activation of a eukaryotic gene after its transfer into a bacterial genome. We integrated a eukaryotic selectable marker gene cassette driven by expression elements from the red alga Porphyridium purpureum into the genome of Escherichia coli. Following growth under non-selective conditions, gene activation events were indentified by antibiotic selection. We show that gene activation in the bacterial recipient occurs at high frequency and involves two major types of spontaneous mutations: deletion and gene amplification. We further show that both mechanisms result in promoter capture and are frequently triggered by microhomology-mediated recombination. Our data suggest that horizontally transferred genes have a high probability of acquiring functionality, resulting in their maintenance if they confer a selective advantage.

Figure 1.

Genetic screen for functional activation of a eukaryotic gene following horizontal transfer into a prokaryote (Escherichia coli). (A) Schematic map of the eukaryotic ble-GFP gene transferred to the lacZ locus in the E. coli genome by selection for kanamycin resistance (conferred by the nptII gene). The ble-GFP fusion gene is driven by the tubulin gene promoter (P_Tub) and terminator (T_Tub) from the red alga Porphyridium purpureum (35) and confers resistance to the antibiotic zeocin. Ec-ZL26 denotes the E. coli strain harboring the construct in its lacZ locus. P_lac, lac operon promoter; P_rrn, rRNA operon promoter from the tobacco plastid genome; TrbcL, 3′ UTR from the tobacco rbcL gene. Endogenous bacterial genes are represented as black boxes, coding regions of transgenes as white boxes and promoters and terminators (driving transgenes and the lac operon) as gray boxes. (B) Selection for functional activation of the ble-GFP gene. A large-scale genetic screen for functional activation of the ble gene in bacteria was conducted by exposing cultures of Ec-ZL26 to stringent selection for ble-mediated resistance to zeocin. The 30 ml LB medium inoculated with a single colony of Ec-ZL26 was cultivated for 1 day, followed by plating of equal aliquots onto 10 LB plates with zeocin for overnight selection of resistant colonies. The selection process was repeated 30 times (i.e. with 30 independent cultures of 30 ml each). In total, approximate 10¹²E. coli cells were subjected to selection on 300 plates. (C) Zeocin-resistant colonies appeared on selection plates from 8 of the 30 cultures. P, number of the selection plate; A, gene activation event; 1 (in bold), the colony representing the initially characterized activation event from cultures 13 and 16. (D) Analysis of GFP accumulation in a dividing zeocin-resistant cell from activation event Ec-ZL26-A1 by confocal laser-scanning microscopy. Scale bars: 1 μm.

Figure 2.

Expression levels of the ble-GFP gene differ between the activation events. (A) Analysis of GFP fluorescence of zeocin-resistant colonies. The left plate was photographed under daylight, the right plate was scanned with laser light to visualize the green fluorescence of GFP (see ‘Materials and Methods’ section). (B) Zeocin tolerance tests to determine the antibiotic resistance level by assaying growth on LB plates with zeocin concentrations between 100 and 700 μg/mL. Question marks indicate that the resistance level may be even higher, but concentrations higher than 700 μg/mL were not tested. (C) Quantification of GFP mRNA accumulation in zeocin-resistant clones by qRT-PCR. Relative expression levels (rel. expr.; normalized to Ec-ZL26-A8, the strain with the lowest GFP expression levels) are shown. Error bars represent the standard deviation from three biological replicates.

Figure 3.

Identification of genomic rearrangements leading to functional activation of the eukaryotic ble-GFP gene in Escherichia coli. (A) Map of the region transferred into the E. coli genome. Restriction sites used in RFLP analyses and primers used for PCR amplification and DNA sequencing are indicated, and expected fragment sizes are given in kb. (B) Southern blot analyses of zeocin-resistant clones. Total bacterial DNA was digested with two different restriction enzymes. While ClaI does not have a recognition site in the transferred DNA, NdeI cuts once in the transgenic region (at the start codon of the nptII coding region). The blots were hybridized to a radiolabeled probe covering the ble coding region. The sizes of the restriction fragments recognized in the Ec-ZL26 strain are 9023 bp for ClaI and 8191 bp for NdeI. (C) Quantification of the bands in the Southern blot assays (with the Image Lab 6.0 software) to estimate the copy number of the ble gene in the bacterial strains with activated ble expression. The cross-hybridizing band of ∼5.1 kb (in the ClaI blot) that is also visible in the wild type was used as reference band for quantitation. (D) Gel electrophoretic analysis of PCR products of the region harboring the activated ble (obtained with the primers indicated in panel A). The magnified area shows a stronger exposure of the amplified bands corresponding to the tandem duplications in strains A4 and A5 (cf. Figure 4).

Figure 4.

Overview of the molecular rearrangements leading to functional activation of the eukaryotic ble gene in the Escherichia coli genome. Nucleotide positions at the break points of deletions (left) or duplications (right) are given and refer to the start of the coding region in which they occur (lacZ, ble or nptII), with negative numbers indicating positions upstream and positive numbers indicating positions downstream of the first nucleotide of the respective coding region. The three transgenic coding regions (ble, GFP, nptII) are distinguished by three different striping patterns. The green Δ and the green dotted lines denote deletions, the red solid lines indicate tandem duplications. The multi-copy duplications in strain A4 are schematically shown by thick red lines (representing copies 2–8). Short directly repeated sequences found at the break points are represented by their nucleotide sequences.

Figure 5.

Gene activation events occur in the absence of selection. The cultivation and selection process can be divided into three phases: the early stage (lag and log phases of growth in the absence of selection), the late stage (stationary growth phase in the absence of selection) and the selection phase (growth on solid medium in the presence of the antibiotic). Only at the early stage, a single activation event can lead to more than one zeocin-resistant colony. Six and three resistant colonies were obtained from cultures 13 (A5) and 16 (A6), respectively, and all colonies from one culture showed the same molecular rearrangement, strongly suggesting that the colonies go back to one and the same gene activation event (and that this event occurred early in the absence of antibiotic selection).