AvrHah1 is both necessary and sufficient for X. gardneri-dependent effects on S. enterica persistence
Previously, we had shown that infection with X. gardneri promotes survival of S. enterica on the leaves of tomato plants19. To determine the importance of AvrHah1 for X. gardneri-dependent effects on S. enterica, we examined the effects of co-inoculation with the X. gardneri AvrHah1 DNA-binding domain mutant avrHah1ΔDBD 30. Bacterial populations were monitored on tomato plants dip-inoculated with S. enterica in the presence or absence of wildtype X. gardneri or the X. gardneri avrHah1ΔDBD mutant. The presence of wildtype X. gardneri led to approximately ten-fold higher S. enterica populations than the S. enterica alone treatment by 6 dpi (Fig. 1a; P=0.0058). Conversely, co-inoculation of S. enterica with the X. gardneri avrHah1ΔDBD mutant had no significant impact on S. enterica (Fig. 1a; P=0.889). Xanthomonas populations were monitored over time, and there were no significant differences between treatments (Fig. 1b; P>0.556 for all comparisons).
To determine if avrHah1 is sufficient for Xanthomonas-mediated effects on S. enterica persistence, X. vesicatoria, which does not impact S. enterica persistence, was transformed with a plasmid carrying the avrHah1 gene (pUFR034 (avrHah1)) and tested for its effects on S. enterica. Bacterial populations were determined in tomato plants dip-inoculated with S. enterica in the presence or absence of X. gardneri, X. vesicatoria + pUFR034 (vector alone control), or X. vesicatoria + pUFR034 (avrHah1). Co-inoculation of X. gardneri or X. vesicatoria + pUFR034 (avrHah1) with S. enterica resulted in approximately ten-fold higher S. enterica populations at 6 dpi than in plants co-inoculated with X. vesicatoria + pURF034 and S. enterica or inoculated with S. enterica alone (Fig. 2a; X. gardneri P=6.40 x 10−7 and 1.74x10−6, respectively; X. vesicatoria + pUFR034 (avrHah1) P=8.17x10−5 and 0.00019, respectively). There was no significant difference between the X. vesicatoria + pURF034 and S. enterica alone treatments (P=0.8203). Xanthomonas populations were monitored over time, and there were no significant differences between treatments (Fig. 2b; P>0.97). These data demonstrate that the presence of AvrHah1 in Xanthomonas spp. can increase the persistence of S. enterica in co-inoculated leaves.
The X. gardneri avrHah1ΔDBD mutant induces transcription of previously identified AvrHah1 targets.
Previous work has shown that AvrHah1 activates expression of multiple tomato genes, including bHLH3, bHLH6, PL, and PE29. These targets provide the foundation for the current model of AvrHah1 water soaking. To examine whether these targets play a role in the mechanism by which X. gardneri enhances S. enterica persistence, we monitored transcription of the genes in leaf samples after inoculation with S. enterica in the presence or absence of wildtype X. gardneri or the X. gardneri avrHah1ΔDBD mutant. Under our experimental conditions, plants infected with the X. gardneri avrHah1ΔDBD mutant had the same or higher levels of the four targets compared to plants infected with wildtype X. gardneri at 1, 3, and 6 dpi (Table 1). In contrast, published results show that wildtype X. gardneri Xg153 significantly induces transcription of these genes (~90-1,300-fold) compared to a corresponding avrHah1ΔDBD mutant when infiltrated into tomato Heinz 1706 leaves for 48 hours29. To determine the effect of inoculation method (dip-inoculation vs infiltration) on our differing results, we infiltrated MoneyMaker tomato leaves with X. gardneri 444 wildtype and avrHah1ΔDBD mutant following published protocols29. Leaf samples were collected 48 hours post-infiltration, and plant gene expression was measured using quantitative PCR (qRTPCR). As with the dip-inoculation experiments, there were no significant differences between leaves inoculated with wildtype X. gardneri or the X. gardneri avrHah1ΔDBD mutant for bHLH3, PL, or PE (Table 1). Plants inoculated with wildtype X. gardneri had ~30-fold more bHLH6 transcription than plants inoculated with the X. gardneri avrHah1ΔDBD mutant (Table 1). These data indicate that AvrHah1 may target other host genes in the mechanism that leads to increased S. enterica persistence in this system.
Table 1
– Transcription of AvrHah1 targets in tomato leaves after infection with the X. gardneri avrHah1ΔDBD mutant.
| Dip-inoculation | Infiltration |
Gene | 1 dpia | 3 dpi | 6 dpi | 48h |
bHLH3 | 3.28 ± 2.60 | 1.10 ± 0.21 | 1.21 ± 0.82 | 1.71 ± 0.14 |
bHLH6 | 1.88 ± 0.70 | 0.76 ± 0.17 | 1.52 ± 1.26 | 0.04 ± 0.01 |
PL | 2.00 ± 1.42 | 1.20 ± 0.39 | 4.38 ± 4.75 | 9.99 ± 2.17 |
PE | 1.63 ± 1.02 | 0.85 ± 0.26 | 2.49 ± 2.71 | 0.90 ± 0.19 |
a. Gene expression is displayed as the relative expression ratio of transcription levels after infection with the X. gardneri avrHah1ΔDBD mutant using the plant response to wildtype X. gardneri as the calibrator (i.e., the response to wildtype X. gardneri is set to 1.0). |
X. gardneri and X. vesicatoria elicit different immune responses in tomato leaves, which are further affected by co-inoculation with S. enterica.
To test the hypothesis that infection with X. gardneri alters the plant immune response to the benefit of S. enterica, plant defense gene expression was monitored over time in tomato leaves inoculated with S. enterica, X. gardneri, X. vesicatoria, or a combination of each xanthomonad with S. enterica. The SA-inducible pathogenesis related protein gene pr1a1 and the JA-inducible proteinase inhibitor gene pin1 were used as established markers31,32 to indirectly monitor these two plant defense pathways with qRTPCR.
At 1 dpi, there were few differences in pr1a1 expression between treatments (Fig. 3a). The X. gardneri with S. enterica co-inoculation treatment was significantly different from the negative control (plants treated with water) (P= 0.0018), but all other treatments were statistically the same as the negative control at this time point (Fig. 3a). Compared to the negative control, tomatoes that were inoculated with X. gardneri, X. gardneri and S. enterica, or X. vesicatoria showed significant increases in pr1a1 expression by 3 dpi (Fig. 3a; P= 0.0205, 0.0005, and 0.0032, respectively). Contrastingly, tomato plants treated with X. vesicatoria and S. enterica or S. enterica alone had no change in pr1a1 expression compared to the water control at 3 dpi (Fig. 3a; P= 0.901 and 0.889, respectively). By 6 dpi, pr1a1 levels had increased in plants inoculated with X. gardneri, X. gardneri and S. enterica, or X. vesicatoria with changes reaching approximately 100-10,000-fold compared to water controls (Fig. 3a; P= 5x10−7, <1x10−7, and 7.46x10−5, respectively). Plants treated with X. vesicatoria and S. enterica or S. enterica alone had no change in pr1a1 expression compared to the water control at 6 dpi (Fig. 3a; P= 0.244 and 1.00, respectively). Plants inoculated with X. vesicatoria and S. enterica had pr1a1 levels that were statistically the same as both the water control and the X. vesicatoria alone treatment (Fig. 3a; P= 0.244 and 0.084, respectively).
At 1 dpi, pin1 expression levels showed similar trends as pr1a1. There were few differences from the water control except for the X. vesicatoria with S. enterica treatment which showed 10-fold higher levels of pin1 expression (Fig. 3b; P<1x10−7). By 3 dpi, treatment with X. gardneri led to reduced levels of pin1 while treatment with X. vesicatoria and S. enterica resulted in increased pin1 expression compared to the water control (Fig. 3b; P= 0.0011 and 0.00097, respectively). Although inoculation with X. vesicatoria or S. enterica alone had no change in pin1 expression compared to the water-inoculated plants at 6 dpi (Fig. 3b; P= 0.071 and 0.939, respectively), inoculation with X. gardneri either with or without S. enterica led to a 1-3 log decrease in pin1 levels (Fig. 3b; P<1x10−7 and 3x10−7, respectively). Co-inoculation with X. vesicatoria and S. enterica led to a 10-fold increase in pin1 expression at 6 dpi (Fig. 3b, P= 0.00048). Taken together, these data show that the two Xanthomonas spp. induce different plant immune responses.
To determine if there was a connection between AvrHah1 and the observed induction of pr1a1 transcription, we monitored pr1a1 levels at 6 dpi in the X. gardneri avrHah1ΔDBD mutant. Wildtype X. gardneri infection led to increased pr1a1expression (Fig. 3c; P=1x10−7 and 5x10−7). Inoculation with the X. gardneri avrHah1ΔDBD mutant also induced pr1a1 expression compared to the water control (Fig. 3c; P< 1x10−7). Further, we measured both free and conjugated forms of SA in leaf tissue at 1 and 3 dpi after dip-inoculation with water, S. enterica, X. gardneri, X. gardneri avrHah1ΔDBD, X. vesicatoria, or S. enterica with each xanthomonad. SA levels between the treatments were not significantly different (P>0.05; Fig. S1). These data demonstrate that avrHah1 is not required for the induction of pr1a1 during X. gardneri infection.
To examine the impact of AvrHah1 on plant gene expression in response to X. vesicatoria, leaf samples were taken at 6 dpi and examined for pr1a1 expression. Compared to the negative control, tomatoes that were inoculated with X. gardneri, X. gardneri and S. enterica, X. vesicatoria + pURF034, X. vesicatoria + pUFR034 (avrHah1), X. vesicatoria + pUFR034 (avrHah1) and S. enterica showed significant increases in pr1a1 expression by 6 dpi (Fig. 3d; P= 1x10−7, 8x10−7, 0.00048, 1.1x10−5, and 7x10−7, respectively). Contrastingly, tomato plants treated with X. vesicatoria + pURF034 and S. enterica or S. enterica alone had no change in pr1a1 expression compared to the water control at 6 dpi (Fig. 3d; P= 0.165 and 0.174, respectively). Thus, addition of avrHah1 to X. vesicatoria alters the immune response to resemble the host response to X. gardneri.
AvrHah1 is not required for electrolyte leakage in X. gardneri-infected tomato leaves.
Previous work has demonstrated that X. gardneri infection leads to more cellular damage in tomato leaves, as measured through electrolyte leakage, than X. vesicatoria infection17. From those data, it was hypothesized that this increase in cellular damage led to a resulting increase in S. enterica persistence in the phyllosphere17. Separately, it was shown that AvrHah1 is required for induction of electrolyte leakage in pepper leaves that have been infiltrated with X. gardneri30. To determine if AvrHah1 is linked to cellular damage and the resulting increases in S. enterica populations in tomato, plants were dip-inoculated with water, S. enterica, X. gardneri, X. gardneri avrHah1ΔDBD, X. vesicatoria, or S. enterica with each xanthomonad. Leaf samples were collected at 6 dpi. Plants treated with wildtype X. gardneri and S. enterica or X. vesicatoria with or without S. enterica had higher levels of electrolyte leakage than the water or S. enterica controls (P<0.01, Fig. S2). The X. gardneri avrHah1ΔDBD treatment resulted in intermediate electrolyte leakage levels that were statistically similar to the water and S. enterica controls and both the wildtype X. gardneri and X. vesicatoria treatments (P>0.01; Fig. S2). The X. gardneri avrHah1ΔDBD mutant and S. enterica treatment gave similar results as the avrHah1ΔDBD mutant treatment except that it had statistically lower levels of electrolyte leakage than the two X. vesicatoria treatments (P<0.01; Fig. S2). These results indicate that specific Xanthomonads cause different levels of electrolyte leakage, but these differences do not correlate with increased S. enterica persistence.