The Plant Journal (2006)

doi: 10.1111/j.1365-313X.2006.02758.x

TECHNICAL ADVANCE

Quantitative in situ assay of salicylic acid in tobacco leaves using a genetically modified biosensor strain of Acinetobacter sp. ADP1 Wei E. Huang1, Linfeng Huang1,†, Gail M. Preston2, Martin Naylor1, John P. Carr3, Yanhong Li1,4, Andrew C. Singer1, Andrew S. Whiteley1 and Hui Wang1,* 1 NERC/Centre for Ecology and Hydrology-Oxford, Mansfield Road, Oxford OX1 3SR, UK, 2 Department of Plant Sciences, University of Oxford, South Parks Road, Oxford, OX1 3RB, UK, 3 Department of Plant Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EA, UK, and 4 College of Biology, Capital Normal University, Xisanhuanbeilu 105, Beijing 100037, China Received 19 January 2006; revised 3 March 2006; accepted 8 March 2006. * For correspondence (fax þ44 1865 281 696; e-mail [email protected]). † Present address: Sainsbury Laboratory, John Innes Centre, Colney Lane, Norwich NR4 7UH, UK.

Summary Salicylic acid (SA) plays important roles in plants, most notably in the induction of systemic acquired resistance (SAR) against pathogens. A non-destructive in situ assay for SA would provide new insights into the functions of SA in SAR and other SA-regulated phenomena. We assessed a genetically engineered strain of Acinetobacter sp. ADP1, which proportionally produces bioluminescence in response to salicylates including SA and methylsalicylate, as a reporter for salicylate accumulation in the apoplast of plant leaves. SA was measured quantitatively in situ in NN genotype tobacco (Nicotiana tabacum L. cv Xanthi-nc) leaves inoculated with tobacco mosaic virus (TMV). The biosensor revealed accumulation of apoplastic SA before the visible appearance of hypersensitive response (HR) lesions. When the biosensor was infiltrated into TMV-inoculated leaves displaying HR lesions at 90 and 168 h post-inoculation, salicylate accumulation was detected predominantly in tissues surrounding the lesions and in veins adjacent to HR lesions. These images are consistent with previous data demonstrating that SA accumulation occurs prior to and following the onset of visible HR lesions. We also used the biosensor to observe apoplastic SA accumulation in tobacco leaves inoculated with virulent and HR-eliciting strains of the bacterial plant pathogen Pseudomonas syringae. The work demonstrates that the Acinetobacter sp. ADP1 biosensor is a useful new tool to non-destructively assay salicylates in situ and to map their spatial distribution in plant tissues. Keywords: salicylic acid, biosensor, in situ assay, hypersensitive response, systemic acquired resistance, Acinetobacter sp. ADP1.

Introduction Biosynthesis of the important plant signal chemical salicylic acid (SA; 2-hydroxybenzoic acid) is increased by multiple stimuli including certain abiotic stresses (Surplus et al., 1998), pathogen infection (Malamy et al., 1990; Me´traux et al., 1990; Wildermuth et al., 2001) and developmental triggers (Morris et al., 2000; Raskin et al., 1987). SA has been most thoroughly studied with respect to its function in ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd

defensive signalling (see reviews by Alvarez, 2000; Durrant and Dong, 2004; Shah, 2003). In plant–pathogen interactions that conform to the gene-for-gene model, plants possessing a dominant resistance (R) gene against a specific (‘avirulent’) pathogen respond to inoculation by exhibiting a hypersensitive response (HR). The HR limits infection to a small number of cells, which may undergo programmed cell 1

2 Wei E. Huang et al. death. SA is required for successful pathogen localization, so that, for example, inhibiting SA accumulation in tobacco plants possessing the N-resistance gene compromises resistance to tobacco mosaic virus (TMV; Mur et al., 1997). Local and systemic acquired resistance (SAR), enhanced resistance states that can occur following infection of a resistant host with an avirulent pathogen (Ross, 1961a,b), also require SA (Gaffney et al., 1993). When levels of SA in plant tissues are elevated as a result of SAR induction, or through overexpression of SA biosynthetic transgenes, or simply by treating plants with SA or its derivatives, this decreases plant susceptibility to broad range of microbes, including virulent pathogens to which the host would normally be fully susceptible (Naylor et al., 1998; Verberne et al., 2000; White, 1979; White et al., 1983). SA levels also increase over the course of plant development, and it has been suggested that they may reach sufficiently high levels to have a direct inhibitory effect on the growth of certain microbes (Cameron and Zaton, 2004). Many biologically active low-molecular-weight molecules in plants are glycosylated to yield a biologically inactive form. The predominant glycosylated metabolite of SA is SAglucoside (SAG; salicylic acid 2-O-b-D-glucoside), which is stored exclusively in the vacuole (Dean et al., 2005; Enyedi et al., 1992; Hennig et al., 1993; Lee et al., 1995). In leaves undergoing the HR, the levels of SAG exceed the levels of free SA by approximately 10-fold, and in non-inoculated leaves exhibiting SAR, only 5–6% of the SA content is in the free form (Verberne et al., 2000). It has been suggested that SAG contributes to SAR by acting as an inactive storage form of SA that can be rapidly reconverted to the active form, free SA (Hennig et al., 1993). SA can also be converted into a volatile methyl ester (methylsalicylate; MeSA), which is active as an inducer of resistance within the plant producing MeSA and neighbouring plants (Huang et al., 2003a; Lee et al., 1995; Shulaev et al., 1997). Like SA, MeSA can be converted into an inactive glucose conjugate (Dean et al., 2005). To date, quantitative measurements of SA in plant tissues have most typically involved extraction of phenolic compounds from tissue fragments by homogenization in 90% methanol and partitioning into a non-aqueous solvent. This is followed by chromatographic analysis, usually highperformance liquid chromatography, and in some studies the presence of SA has been further authenticated by mass spectroscopy (Malamy et al., 1990; Me´traux et al., 1990; Raskin et al., 1987). This type of methodology can be adapted for high-throughput analysis of many samples at once (Muller et al., 2002). Although not as widely adopted, some workers have taken the approach of analysing plant extracts for the presence of SA using immunological methods (Wang et al., 2002). These methods are highly accurate and quantitative. However, they are destructive and can provide at best only limited information on the distribution of SA within living plant tissues.

The non-pathogenic soil bacterium Acinetobacter sp. ADP1 is versatile in its use of carbon sources, is amenable to genetic modification and its genome has been fully sequenced (Barbe et al., 2004; Juni and Janik, 1969; Young et al., 2005). Acinetobacter sp. ADP1 naturally contains the salicylate-responsive sal operon that allows the bacteria to use salicylate as a sole carbon source (Jones et al., 2000). We recently reported construction of a Acinetobacter sp. ADP1-derived salicylate biosensor, ADPWH_lux (Huang et al., 2005b). This biosensor strain contains a chromosomally located fusion of the salA (SA hydroxylase) gene of the Acinetobacter sal operon (Jones et al., 2000) with the promoterless bioluminescence gene cassette luxCDABE from Photorhabdus luminescens (Winson et al., 1998a,b). The five lux genes encode the components of a selfcontained light-emitting system that does not require exogenous substrate. Luciferase encoded by luxA and B oxidizes reduced flavin mononucleotide and a long-chain aliphatic aldehyde in the presence of molecular oxygen to generate green–blue light at 490 nm. The synthesis and recycling of the aldehyde substrate for bioluminescence are catalyzed by a multi-enzyme reductase complex encoded by the luxC, luxD and luxE genes (Chatterjee and Meighen, 1995; Meighen, 1991). The polycistronic luxCDABE from P. luminescens was selected because its product is thermally stable in comparison with the functionally equivalent lux gene products from Vibrio harveyi (Chatterjee and Meighen, 1995; Winson et al., 1998a,b). The sal operon in Acinetobacter sp. ADP1 is controlled by LysR-type regulators (Jones et al., 2000). LysR-type regulators are usually tetramer proteins. In an non-induced condition, two subunits of regulator bind to repressor binding sites (RBS) and two other subunits bind to activator binding sites (ABS) present on the )35 promoter region. In the presence of inducer, the regulator protein changes its conformation and shifts to the )42 region of the ABS. This shift helps interaction with the Cterminal domain of the a-subunit (a-CTD) domain of the RNA polymerase (RNAP), directing RNAP to bind the UP-DNA sequence motif, therefore initiating salA transcription (Jones et al., 2000; Tropel and van der Meer, 2004). In Acinetobacter sp. ADPWH_lux, SA-induced bioluminescence is proportional to SA over a wide concentration range and can be activated by concentrations as low as 5 nM in 15 min (W.E. Huang, unpublished data). From a screen of 26 plant secondary metabolites and derivates, it was found that the biosensor is highly specific to SA, MeSA and the synthetic SA derivative acetylsalicylic acid (aspirin; Huang et al., 2005b; Y. Li and W.E. Huang, unpublished data). Significantly, Acinetobacter sp. ADPWH_lux does not respond to the SA isomers 4-hydroxybenzoic acid and 3hydroxybenzoic acid, or to the related phenolic compounds benzoate and catechol (Huang et al., 2005b). It also does not respond to cis-3-hexen-1-ol, linalool, farnesol, geraniol, citral, ())-b-pinene, (1R)-(þ)-a-pinene, (1S)-())-a-pinene, coumarin,

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

Quantitative salicylic acid in situ assay using biosensor 3 menthone, pulegone, (R)-(þ)-limonene, (R)-())-carvone, (S)-(þ)-carvone, cineole, p-cymene, cumene, 2-phenylpropionaldehyde, 4-isopropylphenol, 2-phenylpropionic acid, methyljasmonate and salicin (Y. Li and W.E. Huang, unpublished data). Importantly, the Luria–Bertani (LB) medium used for growth of Acinetobacter sp. ADPWH_lux does not induce bioluminescence unless SA is added. In LB medium, the biosensor strain does not display any detectable consumption of SA for least 10 h of growth, by which time its preferred carbon sources in the medium have, presumably, become limiting (Huang et al., 2005b). In this study, Acinetobacter sp. ADPWH_lux cells were infiltrated into leaves to report the spatial and temporal pattern of increased SA accumulation before and after the appearance of the HR elicited by TMV in NN genotype tobacco. We also used the biosensor to detect SA accumulation during compatible and incompatible interactions between tobacco and Pseudomonas syringae. The results obtained indicate that Acinetobacter sp. ADPWH_lux is highly effective as a biosensor to quantitatively assess and visualize changes in SA accumulation in plant tissue nondestructively and in situ. Results Comparison of SA measurement by bacterial biosensor to that using GC/MS To confirm that measurements of SA obtained using the Acinetobacter sp. ADPWH_lux biosensor are as accurate as conventional methods, the SA concentration in an LB extract of a TMV-inoculated NN genotype tobacco leaf showing HR lesions (9 days post-inoculation) was measured. This was carried out on the same extract using the biosensor as described previously (Huang et al., 2005b) and by gas chromatography/mass spectrometry (GC/MS). The two methods gave almost identical results: 840 ng g)1 (SA/fresh leaf) and 897 ng g)1 by the biosensor and GC/MS methods, respectively. This result also indicates that the biosensor was exclusively responding to SA and was not sensitive to the SAG that would have been present in the leaf and released from vacuoles during extraction into LB medium. Detection of SA standard solutions infiltrated into tobacco leaves Before attempting to use the biosensor strain to detect increases in endogenous SA levels in leaves, it was necessary to determine whether or not the Acinetobacter sp. ADPWH_lux cells could survive and remain capable of reporting the presence of SA following infiltration into plant leaves. In these experiments, the biosensor was introduced into the apoplast (extracellular space) of leaves by infiltration through the lower epidermis using a syringe with no

needle fitted, which is the same method typically used for agroinfiltration. We found that when the cells were mixed with known concentrations of SA and infiltrated into lamina panels (relatively flat areas of leaf tissue separated from each other by the secondary veins) of healthy leaves, bioluminescence was readily detectable (Figure 1a,b). In fact, the bioluminescence induced by many of the SA concentrations used was visible to the naked eye. Infiltration of biosensor cells in the absence of SA did not result in bioluminescence (Figure 1a,b), demonstrating that the lux reporter sequence was induced specifically by SA and not by any stress imposed upon the bacterial cells resulting from their introduction into the plant tissue. These infiltrations also demonstrated that Acinetobacter sp. ADPWH_lux did not elicit sizeable increases in SA biosynthesis by the plant. As expected, Acinetobacter sp. ADPWH_lux had no pathogenic effect when infiltrated into leaves (Figure 2a). When recording images of bioluminescence in infiltrated leaves, we always included a freshly prepared in vitro concentration ‘ladder’ consisting of tubes containing a standard amount of biosensor cells mixed with a known concentration of SA (Figure 1a,b). Leaf-to-leaf variation was low as shown in Figure 1. Variation due to the biosensor condition occurred between experiments, but could be levelled out by in silico calibration using the in vitro SA standard ladder. Using image-processing techniques previously reported, the images of bioluminescence from these SA ladders were used as calibration standards to allow comparison between images of different biosensor-infiltrated leaves (Huang et al., 2002, 2003b). Thus, when the leaf images shown in Figure 1(a,b) were calibrated using the in vitro SA ladders, it was possible to combine the data and plot the values for the bioluminescence induced in leaves against the corresponding concentration of SA that had been co-infiltrated with the biosensor cells (Figure 1c). Bioluminescence from the biosensor was proportional to the concentration of SA up to 400 lM, and an equation for SA concentration was derived: SA (lM) ¼ 3.1382 · intensity ) 17.208. This was used to estimate leaf SA concentrations in subsequent experiments. SA concentrations could not be measured when the bioluminescence intensity was below 5.48. SA distribution in TMV-inoculated leaves Leaves of NN genotype tobacco were inoculated with TMV, or mock-inoculated, and at various times before and after HR lesion formation were infiltrated with either Acinetobacter sp. ADPWH_lux cells in LB medium, or with sterile LB medium. At each time point, four TMV-inoculated leaves, each from an individual plant, were excised and examined. Two mock leaves were also examined each time. Two TMVinoculated leaves and two mock leaves were infiltrated with the biosensor culture, while the other two TMV-inoculated

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

4 Wei E. Huang et al. Figure 1. Calibration of the in planta biosensor assay by comparison of in vitro and in vivo salicylic acid dilution series. The in vitro concentration ‘ladders’ were prepared in PCR tubes by dilution of a standard salicylic acid (SA) solution in the presence of Acinetobacter sp. ADP1 SA biosensor (50 ll) to yield final SA concentrations of 0, 0.1, 0.2, 0.3, 0.4, 0.5, 1.0 and 2.0 lM in a final volume of 100 ll. For imaging, tubes containing the SA dilutions plus biosensor were arranged serially in rows next to plant leaves (a, b). Tobacco leaves from healthy, unstressed plants were used to produce in vivo SA concentration ladders by mixing various concentrations of SA with the biosensor prior to infiltration into separate leaf panels. The final concentrations of SA (lM) infiltrated into each leaf panel are indicated on the images in (a) and (b). Leaves and in vitro concentration ladders were imaged together in the dark to reveal bioluminescence. Raw pictures of (a) a low- and (b) a high-concentration in vivo ladder are presented. Inclusion of the in vitro SA concentration ladder as an image calibration standard allows separate images of in vivo SA concentration ladders to be directly compared. Image calibration of images (a) and (b) using the in vitro SA concentration ladders was used to construct a dose–response curve for the infiltrated biosensor (c). Using the linear range of the dose–response curve, a correlation equation of SA (lM) ¼ 3.1382 · pixel intensity ) 17.208 (R2 ¼ 99.5%, P < 0.001, range: 0–400 lM) was derived based on the average (circles) of the low- and high-concentration curves (standard deviations represented by error bars). Data points from the low- and high-concentration ladders are indicated by triangles and squares, respectively.

leaves were infiltrated with LB medium only. Using a needleless syringe, the biosensor cells were infiltrated uniformly into the leaf apoplast of the entire leaf. As expected, no bioluminescence was detected in TMV-inoculated leaves when medium alone was injected (Figure 2). However, bioluminescence was clearly visible in TMV-inoculated leaves (NN genotype) by 40 h post-inoculation (hpi), which is prior to the appearance of visible HR lesions (leaf 4 in Figure 2a,b). Bioluminescence at this stage was visible with the naked eye. The increased accumulation of SA in pre-necrotic tissue is consistent with previous studies showing, for example, that a pre-necrosis increase in SA corresponds to other preHR physiological effects such as local increases in tem-

perature, transpiration rate and alterations in chlorophyll fluorescence (Chaerle et al., 1999, 2004). The appearance of increased levels of SA at 40 hpi is also consistent with work showing that SA is required for successful localization of the virus (Mur et al., 1997), and suggests that virus spread may already be subject to SA-mediated resistance mechanisms prior to the occurrence of cell death. Observations of bioluminescence generation following infiltration of leaves with biosensor at time points after the formation of visible lesions, at 90 and 168 hpi, showed that bioluminescence was strongest in areas with high HR lesion density (Figure 2a,b, leaves 5 and 6). The data indicate that SA accumulation is strongest in tissues closest to the primary inoculation sites and falls off with increasing distance from the lesion. No bioluminescence was detected in directly inoculated nn genotype (TMV-susceptible) tobacco leaves (observed at 90 hpi; data not shown), consistent with earlier studies on the induction of SA biosynthesis by TMV (Malamy et al., 1990). After image calibration using the in vitro SA concentration ladders and the bioluminescence data obtained by injecting biosensor plus known amounts of SA into healthy leaves (Figure 1), the data from the leaf images shown in Figure 2 were processed and the concentrations of SA of each pixel were estimated and plotted over their locations in the leaf images to produce a false colour-coded map of SA concentration (Figure 3). SA concentration mapping revealed bioluminescence signals that were difficult to visualize directly from unprocessed images. For example, from the falsecolour SA concentration mapping (Figure 3), it is apparent that in the mock-inoculated leaf infiltrated with biosensor cells (leaf 2) there is a low level of diffuse bioluminescence (not apparent in Figure 2), although this is below the level of bioluminescence induced in cells infiltrated into

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

Quantitative salicylic acid in situ assay using biosensor 5 Figure 2. Detection of endogenous salicylic acid in tobacco leaves. Leaves of TMV-resistant (i.e. hypersensitively responding) tobacco were inoculated with the virus or mock-inoculated, and at various times later were uniformly infiltrated with either the Acinetobacter sp. ADP1 SA biosensor in LB medium or with sterile LB medium. Leaves were photographed under room light (a) immediately before being imaged in the dark to reveal SAinduced bioluminescence from the infiltrated bacteria (b). In vitro SA concentration ladders were included with each leaf imaged to allow comparison of in vivo bioluminescence between separate images. Leaf 1 was inoculated with TMV and infiltrated at 90 h post-inoculation (hpi) with LB medium and leaf 2 was infiltrated with the biosensor 90 h after mock-inoculation. Leaves 3– 6 were inoculated with TMV and infiltrated with the biosensor and imaged at 16, 40, 90 and 168 hpi, respectively.

TMV-challenged leaves at 16 hpi (leaf 4: Figures 2, 3). Such a bioluminescence background as shown in Figure 3 (leaf 2) did not correlate to incubation time and occurred regardless of conditions (data not shown). It is probably due to very weak non-specific leakage of the bioluminescence operon resulting in a trace level of luminescence expression. Although the background was very low, unnoticeable in raw pictures, it could be detected after data processing. However, due to the application of the in vivo ladder

(Figure 1), the bioluminescence background signal was eliminated from SA measurements. TMV inoculation induced an accumulation of SA at 6.5
0.5 lM (mean
SE) at 16 hpi based on an average of two leaves examined. No SA accumulation was detected (bioluminescence intensity <5.48) in mock-inoculated leaves at 16 hpi. The presence of increased SA levels in TMV-inoculated tissues at 16 hpi suggests that increased biosynthesis of SA commences very early during the defence reaction, around 30 h before the

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

6 Wei E. Huang et al.

Figure 3. Quantification of endogenous SA accumulation in hypersensitively responding tobacco leaves. False-colour-coded SA concentration map of the leaves shown in Figure 2. Using the signals from the in vitro concentration ladders in Figure 2 to normalize the leaf bioluminescence images to the in vivo concentration ladders in Figure 1, it was possible to determine the concentration of endogenous SA accumulating in response to TMV infection (Figure 2).

appearance of visible lesion formation and thus around 8 h earlier than has been documented using conventional analytical techniques (Malamy et al., 1990). At 40 hpi, which was before HR lesions became visible, false-colour SA concentration mapping revealed localized areas of extremely high levels of SA that often exceeded 200 lM (Figure 3, leaf 4), twice the concentration at which SA starts to have phytotoxic effects (100 lM; Lee et al., 1995). In the ‘hottest’ spot, SA accumulated to 380 lM (Figure 3). These areas of accumulation of phytotoxic levels of SA were less apparent at later time points. By 90 and 168 hpi, local SA concentrations were highest around HR lesions (Figure 3) but did not exceed 81.3
1.3 and 82.9
1.2 lM (mean
SE), respectively. Presumably, by these time points, the infiltrated biosensor cells could no longer penetrate to the centres of the lesions, where SA concentrations had reached their highest levels, because of desiccation and tissue collapse. Not surprisingly, infiltration of Acinetobacter sp. ADPWH_lux biosensor cells into the leaf apoplast did not result in observable bioluminescence signals within the veins (Figures 1, 2), although signals adjacent to the veins were

seen in images of leaves taken at 90 and 168 hpi. In order to determine whether the biosensor could be used to detect SA in vein tissue, cells were introduced, using a syringe fitted with a needle, into the main (primary) vein of a detached tobacco leaf exhibiting HR lesions (90 hpi; Figure 4a). Bioluminescence was apparent in the primary vein and some of the secondary veins, showing that salicylate is present in this tissue at this stage of the response to the virus and indicating the extent to which it is possible to introduce biosensor cells into the vasculature (Figure 4b). However, bioluminescence was not distributed continuously along the veins, particularly in the secondary veins (Figure 4b). The areas of higher luminosity appear to correspond to points where HR lesions are adjacent to the vein tissue. As in the leaf apoplast, no bioluminescence could be detected in veins of mock-inoculated leaves (data not shown). SA accumulation in P. syringae-infected leaves Having confirmed the utility of the biosensor for studying plant–virus interactions, we tested whether it could be

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

Quantitative salicylic acid in situ assay using biosensor 7

Figure 4. Detection of endogenous salicylate in veins. Using a syringe fitted with a needle, a TMV-inoculated leaf was injected with biosensor via the petiole. The leaf was injected and imaged at 90 h postinoculation, by which time the hypersensitive response (HR) lesions were easily apparent under normal illumination (a). When imaged in the dark (b), bioluminescence was apparent in the primary and secondary veins, with some areas showing particularly strong signals possibly corresponding to points at which vein tissue was undergoing HR or came close to necrotic lamina tissue (arrows).

used to map SA accumulation in situ following interactions with other plant pathogens. The accumulation of SA during compatible and incompatible P. syringae–plant interactions has been reported in a number of previous studies (Huang et al., 2003a, 2005a; Mur et al., 2000; Smith-Becker et al., 1998; Zeier et al., 2004; Zhou et al., 1998). Therefore, we infiltrated the tobacco pathogen P. syringae pv. tabaci 11528 into tobacco cv. SR-1 leaves at concentrations ranging from 5 · 104 to 5 · 107 colonyforming units (cfu) ml)1. After 23 h, necrosis was clearly visible in panels inoculated with P. syringae pv. tabaci at 5 · 107 cfu ml)1. We infiltrated the inoculated leaves with the biosensor and observed bioluminescence 1 h after infiltration (Figure 5). The tobacco pathogen P. s. pv. tabaci elicited detectable SA accumulation, with the strongest signal occurring at 5 · 105 cfu ml)1. The confluent necrosis elicited at higher inoculum densities prevented biosensor infiltration, although dark necrotic areas were surrounded by halos of

Figure 5. Detection of salicylate in tobacco infected by pathogen P. syringae pv. tabaci 11528. The pathogen was infiltrated to one side of leaf at concentrations ranging from 5 · 104 to 5 · 107 cfu ml)1, and the other half of the leaf was infiltrated with buffer (mock). At 23 h post-inoculation, necrosis was clearly apparent in the panel inoculated with bacteria at 5 · 107 cfu ml)1. The biosensor was injected into the petiole and infiltrated into leaf panels and the leaf was photographed 1 h later (a). When imaged in the dark (b), bioluminescence was apparent in both the infected areas and the veins.

SA-induced bioluminescence. Figure 5(b) also shows that SA could be detected in the veins of leaves inoculated with P. s. pv. tabaci, as observed for TMV. We also confirmed that the biosensor could be used to detect a bacteria-induced HR. We infiltrated P. syringae pv. tomato DC3000 and the type III secretion mutant P. syringae pv. tomato DC3000 hrcC) into tobacco cv. SR-1 leaves at concentrations ranging from 5 · 104 to 5 · 107 cfu ml)1. At 23 hpi, leaf panels inoculated with P. syringae pv. tomato at 5 · 106 and 5 · 107 had collapsed as a result of the HR, while the hrcC mutant of P. syringae failed to elicit the HR. The biosensor was then applied as for P. s. pv. tabaci. Bioluminescence was strongest in leaf panels inoculated with P. s. pv. tomato at 5 · 105 cfu ml)1 and no signal was detected in panels inoculated with P. syringae pv. tomato DC3000 hrcC) (Figure 6). This is consistent with previous observations by Mur et al. (2000) and Huang et al. (2003a), who recorded SA levels of at least 400 ng g)1 fresh weight 24 h after inoculating HR-eliciting P. syringae pv. tomato into tobacco, but observed no SA accumulation in response to a P. syringae pv. tomato DC3000 hrcC) mutant.

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

8 Wei E. Huang et al.

Figure 6. Detection of salicylate in tobacco infected by the non-host pathogen P. syringae pv. tomato and a type III secretion mutant. P. syringae pv. tomato DC3000 and the type III secretion-deficient mutant P. syringae pv. tomato DC3000 hrcC) were infiltrated into tobacco leaf panels at concentrations ranging from 5 · 104 to 5 · 107 cfu ml)1. At 23 h postinoculation, the biosensor was infiltrated into leaf panels and the leaf was photographed 1 h later (a). When imaged in the dark (b), strong bioluminescence was apparent in the panel infected with 5 · 105 cfu ml)1, the lowest concentration at which there was no visible cell death and at which the biosensor could be infiltrated effectively. No signal was detected in panels treated with the type III secretion mutant.

Discussion We have shown that Acinetobacter sp. ADPWH_lux is usable as a biosensor for the non-destructive detection and measurement of increased SA accumulation in planta following infection with viral and bacterial pathogens. This nonpathogenic bacterial biosensor can be used to make similar observations in any plant species that are amenable to infiltration, including the model plant Arabidopsis (Huang et al., unpublished preliminary data). However, because the method that we have employed introduces the biosensor cells into the apoplast of the leaf or into the xylem, this means that the salicylate signal we have detected will be predominantly due to free, unconjugated SA. Thus, the method as we have employed it does not monitor salicylate in the form of SAG, the inactive SA store, because this is located within the cell vacuole and will not be accessible to the biosensor cells (Dean et al., 2005; Hennig

et al., 1993). In any case, our experiment in which tissue was homogenized before assay using GC/MS and the biosensor in parallel indicates that the biosensor does not respond to SAG. Presumably Acinetobacter lacks a suitable glucosidase capable of releasing the SA. The method can detect SA in veins (most likely the xylem), but due to the non-uniformity in the distribution of the infiltrated biosensor it is not possible to quantify this as can be done for intercellular SA in the leaf laminae. Therefore, the total amount of free SA has not been measured in vivo in intact leaves, but such information can be obtained by in vitro measurements by using either the biosensor or biochemical methods for leaf extracts. As with other existing methods, the biosensor cannot distinguish between SA that is synthesized by isochorismate synthase and isochorismate pyruvate lyase, or via the phenylpropanoid pathway. Although SA is the major soluble active salicylate in plants, the biosensor is also sensitive to MeSA, which is also biologically active but volatile and highly insoluble in water and therefore unlikely to accumulate in the intercellular fluid or xylem sap. Thus, it is unlikely that a significant portion of the bioluminescence signal is a response to MeSA. In future work, it may be possible to adapt the biosensor method to detect emanation of MeSA from plants by passing air from an enclosed growth chamber over a culture of the biosensor. In conclusion, we believe that the approach of using Acinetobacter-based biosensors offers a useful and adaptable technique for the detection of SA and possibly other biologically active molecules in plants. The method would complement other non-destructive monitoring techniques, such as infra-red and fluorescence imaging (Chaerle et al., 1999, 2004) or biophoton detection (Bennett et al., 2005), that have recently been applied to the investigation of plant– pathogen interactions. The SA biosensor permits detailed observations in real-time of spatially resolved changes in SA concentration, for example the occurrence of transient, localized high concentrations of SA that are undetectable by conventional analytical techniques. Experimental procedures Biosensor measurement for SA in leaf extracts A TMV-inoculated leaf (with HR lesions) and a mock-inoculated leaf were excised at 9 days post-inoculation. Leaves were extracted in fresh LB liquid medium (2.5 ml LB per 1 g of leaf) by grinding then vortexing for 30 sec. After sonicating for another 5 min on ice, the homogenates were centrifuged at 12 000 g for 15 min. The supernatants were used for SA measurement using the biosensor as described previously (Huang et al., 2005b), except that the mock extract was used to make an SA standard ladder (SA final concentrations of 0, 0.08, 0.8, 3.8 and 7.7 lM). In brief, 60 ll of LB medium and 50 ll of salicylate biosensor culture were mixed with 20 ll of the SA ladder or the TMV extract, respectively. The 130 ll mixtures were

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

Quantitative salicylic acid in situ assay using biosensor 9 added into a black clear-bottom 96-well microplate (Fisher Scientific, Loughborough, Leicestershire, UK). The microplate was then incubated at 37
C for 1 h before bioluminescence and OD600 were measured using a SynergyTM HT Multi-Detection Microplate Reader (Bio-Tek Instruments Inc., Winooski, VT, USA) with the plate reading and analysis software of Bio-Tek KC4 (version 3.1). Relative bioluminescence was obtained by dividing bioluminescence by OD600 to estimate the SA concentration as described previously (Huang et al., 2005b). This SA biosensor Acinetobacter sp. ADPWH_lux will be made readily available to academic researchers through H.W. and W.E.H. ([email protected]).

GC/MS measurement of SA in the TMV-infected leaf extract To confirm SA estimation made by the biosensor, LB extracts of TMV-infected leaves were analysed by GC/MS (gas chromatography/mass spectrometry). A 0.7 ll aliquot of 20-fold concentrated extract was removed from the supernatant after centrifuging at 18 400 g for 20 min at 4
C, and 10 ll of 6 N NaOH was added to the sample, immediately followed by extraction with 1.4 ml dichloromethane (DCM) in a 4 ml borosilicate vial with a PTFE-lined cap. The organic layer was removed, after which 10 ll of 6 N HCl was added followed by another 1.4 ml DCM extraction. The DCM extract was subsequently removed and evaporated to dryness. Anhydrous sodium sulphate was added to remove any water. Samples were resuspended in approximately 0.2 ml of pyridine and transferred to a gas chromatography vial where the samples were evaporated to dryness and resuspended in 50 ll pyridine and 50 ll N-trimethylsilyl-N-methyl-trilfluoroacetamide (MSTFA). The sample was incubated at 50
C on a heating block for 2 h, after which it was analysed by GC/MS (Perkin-Elmer Autosystem XL TurboMass, Seer Green, Bucks, UK) equipped with a fused silica column, Agilent DB5ms (Stockport, Cheshire, UK), 30 m · 0.25 mm, 25 lm) under temperature programmed conditions (starting at 75
C hold for 1 min, ramp 15
C min)1 to 200
C, ramp 25
C min)1 to 280
C, hold 2.5 min). Quantification of trimethylsilyl (TMS) salicylate film thickness was based on the major ion fragment m/z 73.

Plant growth conditions and inoculation with TMV and P. syringae Tobacco (Nicotiana tabacum L.) cultivars Xanthi-nc (NN genotype), Xanthi (nn genotype) and SR-1 were grown in a glasshouse before use when they were approximately 6 weeks old. For TMV experiments, tobacco leaves were inoculated evenly over the upper surface with TMV strain U1 (11 lg ml)1 in water) or water (mock inoculation) using carborundum as an abrasive. For P. syringae experiments, P. syringae pv. tomato DC3000 (D. Cuppels, ATTC No. BAA-871), P. syringae pv. tomato DC3000 hrcC) (Yuan and He, 1996), and P. syringae pv. tabaci 11528 (ATCC No. 11528) were grown overnight on LB agar (Sambrook et al., 1989). Bacteria were resuspended in 10 mM MgCl2, diluted to concentrations ranging from 5 · 104 to 5 · 107 cfu ml)1 and inoculated into tobacco leaves using a blunt 1 ml syringe. Inoculated and mock-inoculated plants were maintained in a Gallenkamp Plus incubator (Sanyo Ltd, Loughborough, UK) at 23
C under a 12 h light/dark regime.

Handling of the salicylate biosensor bacteria Acinetobacter sp. ADP1 has been completely sequenced (NCBI accession number NC005966; Young et al., 2005) and its modifica-

tion to yield the SA biosensor strain Acinetobacter sp. ADPWH_lux has been described previously (Huang et al., 2005b). To prepare bacteria for infiltration, 10 ml of LB medium was inoculated with a single colony of Acinetobacter sp. ADPWH_lux and incubated at 37
C with shaking at a rate of 150 rpm. The overnight culture was diluted to 200 ml with fresh LB medium, and incubated with shaking at 37
C for approximately 4 h to reach OD600 ¼ 0.4 (equivalent to 109 cfu ml)1). The bacterial suspension was maintained at 20
C and was usable for infiltration or in preparation of in vitro or in vivo SA concentration standards for up to 8 h. Infiltration of the biosensor cell suspensions into the leaf apoplast was achieved using a 5 ml needle-less syringe to introduce the bacteria through the lower epidermis. To detect salicylate within vein tissues, a syringe fitted with a needle was used to introduce biosensor cell suspensions into the petioles of detached leaves. Following infiltration of the biosensor cells, plant leaves were incubated for 1 h (P. syringae) or 2 h (TMV) prior to imaging.

Imaging and quantification of bacterial luminescence SA-induced bioluminescence was imaged in the dark using a VersaDoc Imaging System (BioRad Laboratories, Hemel Hempstead, Hertfordshire, UK) using the ‘high sensitive chemiluminescent’ setting for 150 sec with a Nikon 50 mm lens at f 1.4. Images of leaves were also taken under the normal lab light ‘photo’ setting for 3 sec with the same lens at f 11. Image data were converted to bmp files and imported into Matlab version 6.1 (The MathWorks Inc., Natick, MA, USA). After calibration to a common level based on in vitro SA concentration ladders (included in all images), the intensity of each pixel of a image was obtained using the UTHSCSA image tool (version 2.0 for Windows, University of Texas Health Science Center in San Antonio) as described previously (Huang et al., 2002, 2003b). For the in planta SA dilution series (Figure 1), the bioluminescence intensities were averaged and plotted against SA concentration. Colour contour maps of SA distribution were generated using a home-made Matlab code (described in Huang et al., 2002, 2003b). Regression analysis was performed in using Minitab-14 (Minitab Inc., State College, PA, USA).

Acknowledgements We are grateful to Mrs Delia McCall for maintenance of glasshouse plants. H.W. is supported by the CEH Biodiversity Program (Project C02875).

References Alvarez, M.E. (2000) Salicylic acid in the machinery of hypersensitive cell death and disease resistance. Plant Mol. Biol. 44, 429–442. Barbe, V., Vallenet, D., Fonknechten, N. et al. (2004) Unique features revealed by the genome sequence of Acinetobacter sp ADP1, a versatile and naturally transformation competent bacterium. Nucleic Acids Res. 32, 5766–5779. Bennett, M., Mehta, M. and Grant, M. (2005) Biophoton imaging: a nondestructive method for assaying R gene responses. Mol. Plant Microbe. Interact. 18, 95–102. Cameron, R.K. and Zaton, K. (2004) Intercellular salicylic acid accumulation is important for age-related resistance in Arabidopsis to Pseudomonas syringae. Physiol. Mol. Plant Pathol. 65, 197–209.

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

10 Wei E. Huang et al. Chaerle, L., van Caeneghem, W., Messens, E., Lambers, H., van Montagu, M. and van der Straeten, D. (1999) Presymptomatic visualization of plant–virus interactions by thermography. Nat. Biotechnol. 17, 813–816. Chaerle, L., Hagenbeek, D., De Bruyne, E., Valcke, R. and van der Straeten, D. (2004) Thermal and chlorophyll-fluorescence imaging distinguish plant–pathogen interactions at an early stage. Plant Cell Physiol. 45, 887–896. Chatterjee, J. and Meighen, E.A. (1995) Biotechnological applications of bacterial bioluminescence (Lux) genes. Photochem. Photobiol. 62, 641–650. Dean, J.V., Mohammed, L.A. and Fitzpatrick, T. (2005) The formation, vacuolar localization, and tonoplast transport of salicylic acid glucose conjugates in tobacco cell suspension cultures. Planta, 221, 287–296. Durrant, W.E. and Dong, X. (2004) Systemic acquired resistance. Annu. Rev. Phytopathol. 42, 185–209. Enyedi, A.J., Yalpani, N., Silverman, P. and Raskin, I. (1992) Localization, conjugation, and function of salicylic acid in tobacco during the hypersensitive reaction to tobacco mosaic virus. Proc. Natl Acad. Sci. USA 89, 2480–2484. Gaffney, T., Friedrich, L., Vernooij, B., Negrotto, D., Nye, G., Uknes, S., Ward, E., Kessmann, H. and Ryals, J. (1993) Requirement of salicylic acid for the induction of systemic acquired resistance. Science, 261, 754–756. Hennig, J., Malamy, J., Grynkiewicz, G., Indulski, J. and Klessig, D.F. (1993) Interconversion of the salicylic acid signal and its glucoside in tobacco. Plant J. 4, 593–600. Huang, W.E., Smith, C.C., Lerner, D.N., Thornton, S.F. and Oram, A. (2002) Physical modelling of solute transport in porous media: evaluation of an imaging technique using UV excited fluorescent dye. Water Res. 36, 1843–1853. Huang, J., Cardoza, Y.J., Schmelz, E.A., Raina, R., Engelberth, J. and Tumlinson, J.H. (2003a) Differential volatile emissions and salicylic acid levels from tobacco plants in response to different strains of Pseudomonas syringae. Planta, 217, 767– 775. Huang, W.E., Oswald, S.E., Lerner, D.N., Smith, C.C. and Zheng, C.M. (2003b) Dissolved oxygen imaging in a porous medium to investigate biodegradation in a plume with limited electron acceptor supply. Environ. Sci. Technol. 37, 1905–1911. Huang, J., Schmelz, E.A., Alborn, H., Engelberth, J. and Tumlinson, J.H. (2005a) Phytohormones mediate volatile emissions during the interaction of compatible and incompatible pathogens: the role of ethylene in Pseudomonas syringae infected tobacco. J. Chem. Ecol. 31, 439–459. Huang, W.E., Wang, H., Huang, L.F., Zheng, H.J., Singer, A.C., Thompson, I.P. and Whiteley, A.S. (2005b) Chromosomally located gene fusions constructed in Acinetobacter sp. ADP1 for the environmental detection of salicylate. Environ. Microbiol. 7, 1339–1348. Jones, R.M., Pagmantidis, V. and Williams, P.A. (2000) sal genes determining the catabolism of salicylate esters are part of a supraoperonic cluster of catabolic genes in Acinetobacter sp. strain ADP1. J. Bacteriol. 182, 2018–2025. Juni, E. and Janik, A. (1969) Transformation of Acinetobacter calcoaceticus (Bacterium anitratum). J. Bacteriol. 98, 281–288. Lee, H.I., Leon, J. and Raskin, I. (1995) Biosynthesis and metabolism of salicylic acid. Proc. Natl Acad. Sci. USA 92, 4076–4079. Malamy, J., Carr, J.P., Klessig, D.F. and Raskin, I. (1990) Salicylic acid – a likely endogenous signal in the resistance response of tobacco to viral infection. Science, 250, 1002–1004. Meighen, E.A. (1991) Molecular biology of bacterial bioluminescence. Microbiol. Rev. 55, 123–142.

Me´traux, J.-P., Signer, H., Ryals, J., Ward, E., Wyss-Benz, M., Gaudin, J., Raschdorf, K., Schmid, E., Blum, W. and Inverardi, B. (1990) Increase in salicylic acid at the onset of systemic acquired resistance in cucumber. Science, 250, 1004–1006. Morris, K., A.-H. Mackerness, S., Page, T., John, C.F., Murphy, A.M., Carr, J.P. and Buchanan-Wollaston, V. (2000) Salicylic acid has a role in regulating gene expression during leaf senescence. Plant J. 23, 677–686. Muller, A., Duchting, P. and Weiler, E.W. (2002) A multiplex GC-MS/ MS technique for the sensitive and quantitative single-run analysis of acidic phytohormones and related compounds, and its application to Arabidopsis thaliana. Planta, 216, 44–56. Mur, L.A., Bi, Y.M., Darby, R.M., Firek, S. and Draper, J. (1997) Compromising early salicylic acid accumulation delays the hypersensitive response and increases viral dispersal during lesion establishment in TMV-infected tobacco. Plant J. 12, 1113–1126. Mur, L.A., Brown, I.R., Darby, R.M., Bestwick, C.S., Bi, Y.-M., Mansfield, J. and Draper, J. (2000) A loss of resistance to avirulent bacterial pathogens in tobacco is associated with the attenuation of a salicylic acid-potentiated oxidative burst. Plant J. 23, 609–621 Naylor, M., Murphy, A.M., Berry, J.O. and Carr, J.P. (1998) Salicylic acid can induce resistance to plant virus movement. Mol. Plant Microbe. Interact. 11, 860–868. Raskin, I., Ehman, A., Melander, W.R. and Meeuse, B.J.D. (1987) Salicylic acid: a natural inducer of heat production in Arum lilies. Science, 237, 1601–1602. Ross, A.F. (1961a) Localized acquired resistance to plant virus infection in hypersensitive hosts. Virology, 14, 329–339. Ross, A.F. (1961b) Systemic acquired resistance induced by localized virus infection in plants. Virology, 14, 340–358. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboartory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Shah, J. (2003) The salicylic acid loop in plant defense. Curr. Opin. Plant Biol. 6, 365–371. Shulaev, V., Silverman, P. and Raskin, I. (1997) Airborne signalling by methyl salicylate in plant pathogen resistance. Nature, 385, 718–721. Smith-Becker, J., Marois, E., Huguet, E.J., Midland, S.L., Sims, J.J. and Keen, N.T. (1998) Accumulation of salicylic acid and 4-hydroxybenzoic acid in phloem fluids of cucumber during systemic acquired resistance is preceded by a transient increase in phenylalanine ammonia-lyase activity in petioles and stems. Plant Physiol. 116, 231–238. Surplus, S.L., Jordan, B.R., Murphy, A.M., Carr, J.P., Thomas, B. and A.-H. Mackerness, S. (1998) Ultraviolet-B-induced responses in Arabidopsis thaliana: role of salicylic acid and reactive oxygen species in the regulation of transcripts encoding photosynthetic and acidic pathogenesis-related proteins. Plant Cell Environ. 21, 685–694. Tropel, D and van der Meer, JR (2004) Bacterial transcriptional regulators for degradation pathways of aromatic compounds. Microbiol. Mol. Biol. Rev. 68, 474–500. Verberne, M.C., Verpoorte, R., Bol, J.F., Mercado-Blanco, J. and Linthorst, H.J. (2000) Overproduction of salicylic acid in plants by bacterial transgenes enhances pathogen resistance. Nat. Biotechnol. 18, 779–783. Wang, S.C., Xu, L.L., Li, G.J., Chen, P.Y., Xia, K. and Zhou, X. (2002) An ELISA for the determination of salicylic acid in plants using a monoclonal antibody. Plant Sci. 162, 529–535. White, R.F. (1979) Acetylsalicylic acid (aspirin) induces resistance to tobacco mosaic virus in tobacco. Virology, 99, 410–412. White, R. F., Antoniw, J. F., Carr, J. P. and Woods, R. F. (1983) The effects of aspirin and polyacrylic acid on the multiplication and

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x

Quantitative salicylic acid in situ assay using biosensor 11 spread of TMV in different cultivars of tobacco with and without the N-gene. Phytopathol. Z. (J. Phytopathol.) 107, 224–232. Wildermuth, M.C., Dewdney, J., Wu, G. and Ausubel, F.M. (2001) Isochorismate synthase is required to synthesize salicylic acid for plant defense. Nature, 414, 562–565. Winson, M.K., Swift, S., Hill, P.J., Sims, C.M., Griesmayr, G., Bycroft, B.W., Williams, P. and Stewart, G. (1998a) Engineering the luxCDABE genes from Photorhabdus luminescens to provide a bioluminescent reporter for constitutive and promoter probe plasmids and mini-Tn5 constructs. FEMS Microbiol. Lett. 163, 193–202. Winson, M.K., Swift, S., Fish, L., Throup, J.P., Jorgensen, F., Chhabra, S.R., Bycroft, B.W., Williams, P. and Stewart, G. (1998b) Construction and analysis of luxCDABE-based plasmid sensors for investigating N-acyl homoserine lactone-mediated quorum sensing. FEMS Microbiol. Lett. 163, 185–192.

Young, D.M., Parke, D. and Ornston, L.N. (2005) Opportunities for genetic investigation afforded by Acinetobacter baylyi, a nutritionally versatile bacterial species that is highly competent for natural transformation. Annu. Rev. Microbiol. 59, 519–551. Yuan, J. and He, S.Y. (1996) The Pseudomonas syringae Hrp regulation and secretion system controls the production and secretion of multiple extracellular proteins. J. Bacteriol. 178, 6399–6402. Zeier, J., Pink, B., Mueller, M.J. and Berger, S. (2004) Light conditions influence specific defence responses in incompatible plant– pathogen interactions: uncoupling systemic resistance from salicylic acid and PR-1 accumulation. Planta, 219, 673–683. Zhou, N., Tootle, T.L., Tsui, F., Klessig, D.F. and Glazebrook, J. (1998) PAD4 functions upstream from salicylic acid to control defense responses in Arabidopsis. Plant Cell, 10, 1021–1030.

ª 2006 The Authors Journal compilation ª 2006 Blackwell Publishing Ltd, The Plant Journal, (2006), doi: 10.1111/j.1365-313X.2006.02758.x