Suppression of different classes of somatic mutations in Arabidopsis by vir gene-expressing Agrobacterium strains

Abstract

Background: Agrobacterium infection, which is widely used to generate transgenic plants, is often accompanied by T-DNA-linked mutations and transpositions in flowering plants. It is not known if Agrobacterium infection also affects the rates of point mutations, somatic homologous recombinations (SHR) and frame-shift mutations (FSM). We examined the effects of Agrobacterium infection on five types of somatic mutations using a set of mutation detector lines of Arabidopsis thaliana. To verify the effect of secreted factors, we exposed the plants to different Agrobacterium strains, including wild type (Ach5), its derivatives lacking vir genes, oncogenes or T-DNA, and the heat-killed form for 48 h post-infection; also, for a smaller set of strains, we examined the rates of three types of mutations at multiple time-points. The mutation detector lines carried a non-functional β-glucuronidase gene (GUS) and a reversion of mutated GUS to its functional form resulted in blue spots. Based on the number of blue spots visible in plants grown for a further two weeks, we estimated the mutation frequencies.

Results: For plants co-cultivated for 48 h with Agrobacterium, if the strain contained vir genes, then the rates of transversions, SHRs and FSMs (measured 2 weeks later) were lower than those of uninfected controls. In contrast, co-cultivation for 48 h with any of the Agrobacterium strains raised the transposition rates above control levels. The multiple time-point study showed that in seedlings co-cultivated with wild type Ach5, the reduced rates of transversions and SHRs after 48 h co-cultivation represent an apparent suppression of an earlier short-lived increase in mutation rates (peaking for plants co-cultivated for 3 h). An increase after 3 h co-cultivation was also seen for rates of transversions (but not SHR) in seedlings exposed to the strain lacking vir genes, oncogenes and T-DNA. However, the mutation rates in plants co-cultivated for longer times with this strain subsequently dropped below levels seen in uninfected controls, consistent with the results of the single time-point study.

Conclusions: The rates of various classes of mutations that result from Agrobacterium infection depend upon the duration of infection and the type of pathogen derived factors (such as Vir proteins, oncoproteins or T-DNA) possessed by the strain. Strains with vir genes, including the type used for plant transformation, suppressed selected classes of somatic mutations. Our study also provides evidence of a pathogen that can at least partly counter the induction of mutations in an infected plant.

Fig. 1

Constructs for scoring somatic mutation rates. a. Construct for scoring point mutations, showing the positions and types of mutated bases in the open reading frame of the GUS gene, [32, 33]. b. Gene cassette with an inverted repeat of the truncated GUS gene [34]. c. Construct to detect frame-shift mutations [36]. d. Construct to score transposition frequency; the Tag1 element is engineered between the GUS and a CaMV 35S promoter [37]. e. A blue spot (arrow), showing functional GUS reversions after histochemical staining

Fig. 2

Point mutation rates in different Arabidopsis detector lines, after infection with various Agrobacterium strains, E. coli and heat-killed bacteria for 48 h. a. T → G transversion rates in line 693. b. T → G transversion rates in line 747. c. T → A transversion rates in line 699. d. C → T transition rates in line M4. Bars indicate the standard error of the mean of three biological repeats, each consisting of about 140–160 plants. Numbers on top of the bars show the respective mean values. Numbers at the bottom show the respective number of seedlings analysed. Vertical axis shows the mutation rates. ^{a, b, c, d, e}Means not followed by the same letter are significantly different at 5 % level as determined by Duncan’s multiple range test. C, control; HK-A, heat-killed-Ach5; E, E. coli; HK-E, heat-killed E. coli

Fig. 3

Intrachromosomal somatic homologous recombination rates in Arabidopsis detector lines 651 and R2L1, after infection with different Agrobacterium strains, E. coli and heat-killed bacteria for 48 h. a. Line 651 b. Line R2L1. Bars indicate the standard error of the mean of three biological repeats, each consisting of about 140–160 plants. Numbers on top of the bars show the respective mean values. Numbers at the bottom show the respective number of seedlings analysed. Vertical axis shows the mutation rates.^{a, b}Means not followed by the same letter are significantly different at 5 % level as determined by Duncan’s multiple range test. C, control; HK-A, heat-killed-Ach5; E, E. coli; HK-E, heat-killed E. coli

Fig. 4

Frame-shift mutation and transposition rates in transgenic Arabidopsis detector lines after infection with different Agrobacterium strains, E. coli and heat-killed bacteria for 48 h. a. Frame-shift mutations rates in line G10. b. Tag1 transposition rates. Bars indicate the standard error of the mean of three biological repeats, each consisting of about 60–80 plants. Numbers on top of the bars show the respective mean values. Numbers at the bottom show the respective number of seedlings analysed. Vertical axis shows the mutation rates. ^{a, b, c, d, e, f}Means not followed by the same letter are significantly different at 5 % level as determined by Duncan’s multiple range test. C, control; HK-A, heat-killed-Ach5; E, E. coli; HK-E, heat-killed E. coli

Fig. 5

Transversion rates in Arabidopsis, after infecting with Agrobacterium strain VOT or XXX for various time periods. a. T → G transversion rates in line 747 after VOT infection. b. T → A transversion rates in line 699 after VOT infection. c. T → G transversion rates in line 747 after XXX infection. d. T → A transversion rates in line 699 after XXX infection. Bars indicate the standard error of the mean of three biological repeats, each consisting of about 140–160 plants. Numbers on top of the bars show the respective mean values. Vertical axis shows the mutation rates. ^{a, b, c, d}Means not followed by the same letter are significantly different at 5 % level as determined by Duncan’s multiple range test

Fig. 6

Point mutation and intrachromosomal homologous recombination rates in plants after infection with E. coli and VOT respectively, for various time periods. a. T → G transversion rates in line 747 after infection with E. coli. b. C → T transition rates in line M4 after subjecting to E. coli treatment c. Intrachromosomal homologous recombination rates in line 651 after VOT infection. d. Intrachromosomal homologous recombination rates in line R2L1 after VOT infection. Bars indicate the standard error of the mean of three biological repeats, each consisting of about 140–160 plants. Numbers on top of the bars show the respective mean values. Vertical axis shows the mutation rates.^{a, b, c, d, e}Means not followed by the same letter are significantly different at 5 % level as determined by Duncan’s multiple range test