This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by DeAngelis, K. M.
Right arrow Articles by Lindow, S. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by DeAngelis, K. M.
Right arrow Articles by Lindow, S. E.
Agricola
Right arrow Articles by DeAngelis, K. M.
Right arrow Articles by Lindow, S. E.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, December 2005, p. 8537-8547, Vol. 71, No. 12
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.12.8537-8547.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Two Novel Bacterial Biosensors for Detection of Nitrate Availability in the Rhizosphere

Kristen M. DeAngelis,{dagger} Pingsheng Ji,{dagger},{ddagger} Mary K. Firestone, and Steven E. Lindow*

Department of Plant and Microbial Biology and Department of Environmental Science Policy and Management, University of California, Berkeley, California 94720

Received 29 September 2004/ Accepted 12 September 2005


arrow
ABSTRACT
 
The nitrate-regulated promoter of narG in Escherichia coli was fused to promoterless ice nucleation (inaZ) and green fluorescent protein (GFP) reporter genes to yield the nitrate-responsive gene fusions in plasmids pNice and pNgfp, respectively. While the promoter of narG is normally nitrate responsive only under anaerobic conditions, the L28H-fnr gene was provided in trans to enable nitrate-dependent expression of these reporter gene fusions even under aerobic conditions in both E. coli DH5{alpha} and Enterobacter cloacae EcCT501R. E. cloacae and E. coli cells containing the fusion plasmid pNice exhibited more than 100-fold-higher ice nucleation activity in cultures amended with 10 mM sodium nitrate than in nitrate-free media. The GFP fluorescence of E. cloacae cells harboring pNgfp was uniform at a given concentration of nitrate and increased about 1,000-fold when nitrate increased from 0 to 1 mM. Measurable induction of ice nucleation in E. cloacae EcCT501R harboring pNice occurred at nitrate concentrations of as low as 0.1 µM, while GFP fluorescence was detected in cells harboring pNgfp at about 10 µM. In the rhizosphere of wild oat (Avena fatua), the whole-cell bioreporter E.cloacae(pNgfp) or E. cloacae(pNice) expressed significantly higher GFP fluorescence or ice nucleation activity when the plants were grown in natural soils amended with nitrate than in unamended natural soils. Significantly lower nitrate abundance was detected by the E. cloacae(pNgfp) reporter in the A. fatua rhizosphere compared to in bulk soil, indicating plant competition for nitrate. Ice- and GFP-based bacterial sensors thus are useful for estimating nitrate availability in relevant microbial niches in natural environments.


arrow
INTRODUCTION
 
A major part of the nitrogen cycle results from interactions between plants and soil microorganisms in the rhizosphere. As a critical compound in many transfers and transformations of nitrogen, nitrate is a labile, dynamic pool in rhizosphere N cycling. Nitrate is a significant source of plant N, and it can also be incorporated by a variety of microorganisms through assimilatory nitrate reduction (15) and is subject to denitrification in the absence of oxygen (13). Roots are successful competitors with microorganisms for limited NO3–-N (40). However, depletion of nitrate in various root zones may be different, as in the case of rice and maize, where uptake of nitrate reaches a maximum in the zone where root hairs emerge (9). Since the spatial patterns of nitrate concentration in the rhizosphere may correspond to patterns of root ion uptake as well as reflect microbial N transformations, a clear understanding of nitrate availability in the root zone is essential to understanding the mechanisms of rhizosphere N cycling.

It has been difficult to characterize nitrogen cycling processes in the rhizosphere because of the dynamic and heterogeneous nature of the intact root-soil-microbe system. A fine-scale study of nitrate is even more complicated because nitrate pools are quite transient due to the ion's great mobility; it is rapidly produced through nitrification and utilized by plants and soil microorganisms. Previous techniques developed to study soil NO3– have relied on ion-specific electrodes (21), chemical extraction of bulk soil samples coupled with ion chromatography (12, 39), or colorimetric techniques (21) or have used the stable isotope 15N in tracer studies (18) or 15N pool dilution studies (10, 11, 18) to measure gross rates of nitrate production and consumption. Although some of these methods can provide considerable information on nitrogen content and transformation in soil, they require a relatively large sample mass and have limited fine-scale spatial resolution. Nitrate availability in the microsites colonized by bacteria may be different from the nitrate content of bulk soil samples; hence, the amount of nitrate actually available to rhizosphere bacteria cannot be accurately estimated by conventional means. Therefore, complementary methods with improved spatial resolution for the analysis of nitrate ion are desirable.

We report here the development of two whole-cell bacterial reporters to detect the availability of nitrate in soil around roots that is based on Enterobacter cloacae harboring nitrate-responsive reporter genes. These biosensors differ in their sensitivity and ease of application to soil systems. An ice nucleation reporter gene, inaZ, from the bacterium Pseudomonas syringae (17, 29) and a stable green fluorescent protein (GFP) reporter gene (36) were placed under the regulatory control of the promoter of Escherichia coli nitrate reductase gene (narG), which confers the reduction of the electron acceptor nitrate to nitrite. Since expression of nitrate reductase in E. coli occurs only under anaerobic conditions due to its secondary regulation by the transcriptional regulator Fnr, we provided cells with an altered fnr gene (22, 23) that enabled nitrate-inducible transcription by the narG promoter in the presence of oxygen. We provide evidence that both GFP and inaZ whole-cell biosensors are highly responsive to nitrate in culture and in the rhizosphere and that the population-level estimates of nitrate availability provided by the ice nucleation-based biosensors are complementary to the estimates of nitrate availability at the single-cell level provided by the GFP-based biosensor.


arrow
MATERIALS AND METHODS
 
Bacterial strains, plasmids, and culture conditions.
The bacterial strains and plasmids used are described in Table 1. Erwinia herbicola (Pantoea agglomerans) strain 299R is a spontaneous rifampin-resistant mutant of a culture originally isolated from pear leaves (3, 4). Enterobacter cloacae EcCT501R is a spontaneous rifampin-resistant mutant of E. cloacae EcCT501, a plant-beneficial bacterium isolated from cotton (38).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Bacterial strains and plasmids used in this study

For routine studies, bacterial cells were grown in Luria-Bertani (LB) medium (44) at 37°C for E. coli or 28°C for Erwinia and Enterobacter strains except as otherwise stated. When cells were prepared for fluorescence measurements in culture or in the rhizosphere, cultures were grown aerobically in M9 minimal medium (without thiamine) (44) containing 0.4% glucose. To achieve anaerobic growth, media were inoculated with about 106 bacterial cells ml–1 and filled to the tops of 250-ml bottles. Bacterial cultures were incubated for 6 h at 26°C without shaking to deplete oxygen. Determination of ice nucleation activity of bacterial strains was performed in LB or M9 minimal medium containing 0.4% glucose as a carbon source. In all growth conditions, the antibiotic levels used to select strains or plasmids were as follows: rifampin, 100 µg ml–1; streptomycin, 20 µg ml–1; kanamycin, 50 µg ml–1; ampicillin, 150 µg ml–1; and chloramphenicol,20 µg ml–1 (Sigma-Aldrich). Cycloheximide (actidione) at 100 µg ml–1 and benomyl at 25 µg ml–1 (Sigma-Aldrich) were added to media to inhibit fungi when bacteria were recovered from soils and roots.

Construction of reporter gene fusions. (i) pNarG-Ice.
Preparation of genomic DNA from E. coli strain MG1655 was done by the method of Saiki (43). The narG promoter-regulatory region was obtained as a 592-bp HindIII-EcoRI fragment by using the primers 5'-TTAAGCTTCTATATCGCCTGCGTAGTG-3' and 5'-GGTCCAGGAATTCACTCATC-3' (Integrated DNA Technologies, Coralville IA), based on the published sequence (28, 42). PCR mixtures included 25 mM MgCl2 (4 µl), 10x buffer (5 µl), 5 U Taq polymerase (0.2 µl; Promega), 2.5 mM deoxynucleoside triphosphates (1 µl each), H2O (26.8 µl), 1 µl of each primer, and 5 µl of E. coli MG1655 DNA. PCR was performed with a Perkin-Elmer Cetus DNA thermal cycler. Following initial denaturation (94°C, 2 min), 30 cycles were performed at 94°C for 1 min, 55°C for 1 min, and 72°C for 1.5 min, and there was a final extension cycle at 72°C for 10 min. The amplified products were analyzed by electrophoresis through a 2% agarose gel (Gibco BRL) and visualized after staining with ethidium bromide on a UV transilluminator.

The amplified fragments were cloned into the PCR cloning vector pCR2.1-TOPO by using the TOPO TA cloning kit (Invitrogen, Carlsbad, CA). Plasmid DNAs from selected white colonies were prepared using the QIAprep Spin miniprep kit (QIAGEN Inc., Valencia, CA), digested with EcoRI (New England BioLabs, Beverly, MA), and analyzed by gel electrophoresis as described above. The PCR products were confirmed by sequencing using M13 forward (–20) and M13 reverse primers (Invitrogen, Carlsbad, CA) on an ABI model 377 DNA sequencer (ABI Prism). The plasmid harboring the PCR-amplified fragment was designated pNARG-TOPO.

To construct the reporter gene fusion pNarG-Ice, plasmid DNA of pNARG-TOPO was digested with HindIII (New England BioLabs) and EcoRI, and a 592-bp HindIII-EcoRI fragment was extracted using a QIAEX II gel extraction kit (QIAGEN Inc., Valencia, CA) and cloned immediately 5' to the promoterless inaZ gene in the broad-host-range vector pPROBE-NI (36) (Fig. 1). The inaZ reporter gene encodes an ice nucleation protein that is incorporated into the outer membranes of bacterial cells and catalyzes ice formation (16, 31). Plasmid pPROBE-NI contains a transcriptional fusion cassette bounded by terminators, in which the promoterless inaZ reporter gene is located downstream of a pUC18 polylinker (36). The resulting plasmid, designated pNarG-Ice, was transformed into competent cells of E. coli DH5{alpha} (44) and introduced into E. herbicola 299R by triparental mating using pRK2073 as a mobilizing plasmid.



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 1. Schematic diagram of the construction of the pNice fusion plasmid containing the L28H-fnr gene. Cmr, Kmr, and Apr, resistance to chloramphenicol, kanamycin, and ampicillin, respectively. narGp indicates a 592-bp HindIII-EcoRI fragment of the narG promoter-regulatory region from the genome of E. coli K-12 strain MG1655.

For assessment of nitrate-regulated ice nucleation activity of E. herbicola 299R and E. coli DH5{alpha} containing pNarG-Ice, bacterial cultures grown overnight in LB broth amended with 50 µg/ml of kanamycin (LB-km) were used to inoculate 250-ml flasks containing 50 ml of LB-km or 250-ml bottles filled with LB-km at a final density of approximately 106 CFU ml–1. The LB medium was amended with or without 10 mM sodium nitrate or ammonium sulfate, and the flasks or bottles were incubated at 26°C with or without shaking (250 rpm). Samples of bacterial cultures were taken at different time intervals, and ice nucleation activity was measured by a droplet freezing assay of serial dilutions of the bacterial cultures as described elsewhere (34). Bacterial population sizes were determined by plating appropriate dilutions of the cultures on LB agar (LA) amended with kanamycin, using a spiral plater (Spiral Systems, Inc., Cincinnati, OH). After 30 h of incubation at 28°C, colonies on the plates were counted, and the population size was used to yield cell-normalized estimates of ice nucleation activity.

(ii) pNice.
A 1.6-kb BamHI-XbaI fragment containing the L28H-fnr structural gene in plasmid pPK434 was inserted into plasmid pNARG-TOPO to replace the BamHI-XbaI fragment containing the narG promoter (Fig. 1). The L28H-fnr gene, previously described as LH27-fnr, encodes Fnr with a single amino acid substitution that allows Fnr-dependent gene expression in the presence of oxygen (1, 23). The resulting plasmid contains a HindIII fragment harboring the L28H-fnr gene and was designated pFNR-TOPO. The HindIII fragment in pFNR-TOPO was then cloned immediately 5' to the narG promoter in plasmid pNarG-Ice. The resulting plasmid was designated pNice (Fig. 1). pNice was introduced into E. herbicola 299R and E. cloacae EcCT501R by triparental mating as described above. Nitrate-regulated ice nucleation activity of DH5{alpha}, 299R, and EcCT501R containing pNice was assessed in M9 with various concentrations of sodium nitrate by using methods mentioned above.

(iii) pNgfp.
The inaZ gene was removed from E. cloacae(pNice) by digestion with EcoRI and NotI (New England Biolabs) and replaced with a GFP reporter gene isolated as an EcoRI and NotI restriction digest fragment from pPROBE-NT (43) to produce pNgfp. This plasmid was transformed into E. coli strain DH5{alpha} by electroporation under standard conditions (44). The proper orientation of the cloned fragment was confirmed by restriction mapping of purified plasmids from individual colonies by using enzymes that cut asymmetrically in the plasmid.

(iv) pNgfp and pVS1-dsRed.
In order to account for all biosensor bacteria applied to the rhizosphere, we introduced a compatible plasmid harboring a constitutively expressed red fluorescent protein (RFP) marker gene into the E. cloacae(pNgfp) reporter strain. E. cloacae(pNgfp) was electroporated with pVS1-dsRed, which contains the dsRed gene (M. Marco, unpublished data) downstream of a kanamycin promoter from E. coli conferring constitutive transcription in plasmid pVSP61 (33) harboring a spectinomycin resistance gene. After electroporation into E. cloacae(pNgfp) by standard methods (44), the transformants were recovered on LA containing spectinomycin. Transformants containing pVS1-dsRed were identified by their pink colony color and were verified to have constitutive RFP expression and nitrate-dependent GFP expression by plating cells onto LA with or without 10 mM NaNO3, fixing cells, and visualizing them by microscopy as described below.

To quantify nitrate-dependent GFP expression, single colonies of E. cloacae(pNgfp)(pVS1-dsRed) were inoculated into LB medium containing kanamycin and spectinomycin and incubated overnight. This overnight culture was used to seed a 100-ml culture of M9 minimal medium with kanamycin and spectinomycin. Cells were allowed to reach a density of 109 cells ml–1 before sodium nitrate was added, and GFP fluorescence of cells was monitored 2 and 4 h later, as described below.

Preparation of soil and plants.
Soils were collected from the Red Hill, under the growing zones of the annual graminoid Avena fatua, at the University of California Hopland Research and Extension Center, Hopland, California. This soil is a medium-texture loam derived from hard sandstone and shale, classified as an ultic haploxeralf (7). Soils were collected to a depth of 10 cm and immediately processed for planting. Soil nitrate concentrations were reported to be 0.47 mg N kg–1 soil at 0- to 10-cm depth during the spring growing season (7).

For testing of in situ performance of the biosensor cells containing pNice, soils were immediately sieved to 2 mm, homogenized, and mixed with fine quartz sand (1:1 by volume) before planting to improve drainage (19). Avena fatua (wild oat) and Phaseolus vulgaris (snap bean, cv. Bush Blue Lake 274) (Valley Seed Service, Fresno, CA) were grown in conical pots (Cone-Tainer Nursery, Canby, OR) that were cut vertically into two parts and reannealed with tape before planting to facilitate access to developing roots. The conical pots with seeds were incubated at room temperature until germination and then transferred to a greenhouse and watered daily with tap water. After about 10 days following germination, plants were moved to a growth chamber, maintaining temperatures of 22 to 25°C with a 12-h photoperiod and approximately 450 µmol photons/m2 s–1, and treated with 20 ml of either 10 mM or 0.1 mM sodium nitrate 12 h before inoculation with bacterial sensor strains; control plants received 20 ml of tap water.

Rhizosphere nitrate availability was also tested on plants of Avena fatua by using the E. cloacae pNgfp biosensor. Seeds were rinsed in tap water and pregerminated prior to being planted under a slow drip of tap water for 4 days in darkness. Once germinated, seeds were planted in Hopland field soil as described above. To prepare them for planting, soils were hydrated to 75% water holding capacity, corresponding to 15 ml sterile double-distilled water per 100 g field moist soil. Additional small microcosms were designed to create a rhizosphere environment that could be placed directly under a microscope, measuring 100 by 30 by 4 mm with one face made of a double-wide no. 1 coverslip (Proscitech, Australia). Soils were packed into microcosms and pregerminated seeds placed at the top and covered with a thin layer of soil. The microcosms were then inclined to a 45-degree angle relative to the horizontal, so that the roots would grow along the coverslip face of the microcosm; this set-up is analogous to that used by Jaeger et al. (19). After 4 to 5 days of growth in growth chambers (16 h light days, approximately 450 µmol photons/m2 s–1, 20% humidity), seedlings were inoculated with bacterial biosensors.

Root inoculation and sampling of whole-cell sensor bacteria.
E. cloacae sensor strain EcCT501R(pNice) was grown in M9 medium amended with rifampin and kanamycin and incubated overnight at 26°C with shaking. Cells were collected by centrifugation, washed twice with 10 mM sterile potassium phosphate buffer (PPB) (pH 7), and resuspended in PPB at a concentration of about 107 CFU ml–1. Bacterial suspensions were spray inoculated onto plant roots and surrounding soil at 3 ml per pot by using an artist's airbrush sprayer (Paasche Airbrush Company, Harwood Heights, IL). The inoculated roots were again enclosed within the conical pots, and the plants were returned to the growth chamber. The pots were placed in three replicate racks; five pots for each treatment were randomly arranged in each rack.

E. cloacae sensor strain EcCT501R(pNgfp)(pVS1-dsRed) was prepared for introduction into the rhizosphere by growing a 48-h culture from a single colony in M9 minimal medium with kanamycin and spectinomycin. Cells from the 48-h culture were washed twice in PPB, resuspended in 1x M9 salts to a concentration of 108 cells ml–1, and applied to A. fatua roots grown in rhizotrons designed for microscopy, using an artist's airbrush sprayer, as described above, and applying a total of 108 cells ml–1 in 0.5 ml of liquid per microcosm. A new coverslip was then placed over the microcosm, and cells were incubated in the rhizosphere for 24 h before viewing. Two hours prior to application of the biosensor strain to the rhizosphere, three replicate rhizotrons were amended with 10 mM NaNO3 only, 10 mM NaNO3 plus 1.6% glucose, 1.6% glucose only, or water. After application of amendments and biosensor suspensions, the soil water content was near water holding capacity but was not measured explicitly.

Measurement of ice nucleation activity.
For determination of ice nucleation activity of EcCT501R(pNice) in the rhizosphere, roots (about 100 mg) from three replicate pots were excised at various times after inoculation with bacteria and placed into 2.5 ml of sterile PPB (10 mM) in 5-ml tubes. The tubes were vortexed and sonicated for 5 min in an ultrasonic cleaning bath to dislodge the bacterial cells. Ice nucleation activity was measured by a droplet freezing assay of serial dilutions of rhizosphere soil washings as described elsewhere (34), and bacterial population sizes were determined by plating appropriate dilutions of root washings on LA amended with rifampin, kanamycin, and cycloheximide or benomyl, using a spiral plater (Spiral Systems, Inc., Cincinnati, OH). After 30 h of incubation at 28°C, colonies on the plates were counted and the population size was used to yield cell-normalized estimates of ice nucleation activity in the rhizosphere. Analysis of variance (ANOVA) was performed using the ANOVA or GLM procedures of the Statistical Analysis System (SAS Institute Inc., Cary, NC). Means were compared using Duncan's multiple-range test at a P value of 0.05. The experiment was repeated under the growth chamber conditions. The results from the duplicate experiments were similar, and those from a representative experiment are presented.

Measurement of GFP fluorescence.
The GFP fluorescence of bacterial cultures was determined using a TD-700 fluorometer (Turner Designs, Sunnyvale, CA) with a 486-nm bandpass excitation filter and a 510- to 700-nm combination emission filter. Cells were washed with 10 mM PPB and diluted to a concentration of about 106 cells ml–1. The GFP fluorescence of individual cells was determined by microscopy. Washed cells were fixed in a solution of three parts 4% paraformaldehyde and one part phosphate-buffered saline. Cells were incubated at 4°C for 2 hours, washed to remove paraformaldehyde, and resuspended in 50% phosphate-buffered saline-50% ethanol. Fixed cells were then mounted on slides and examined by epifluorescence microscopy for GFP and RFP fluorescence. To capture GFP fluorescence, an Endow filter with an excitation wavelength of 450 to 490 nm, a dichroic mirror wavelength of 495 nm, and a barrier wavelength of 500 to 550 nm was used. For dsRed fluorescence, a dsRed filter with an excitation wavelength of 541 to 551 nm, a dichroic mirror wavelength of 575 nm, and a barrier wavelength of greater than 580 nm was used. All images were captured using a Zeiss Axiophot epifluorescence microscope equipped with an Optronics DEI 750 low-light, cooled charge-coupled-device color video camera.

Nitrate availability was estimated in the three different rhizosphere zones of wild oat, root tip, root hairs, mature root, and bulk soil. Digital images of 10 to 20 fields of view were captured for each sample. For each field of view, two images were captured; GFP-fluorescent cells were visualized through an Endow filter, and subsequently dsRed-fluorescent cells were visualized with a dsRed filter. Image processing was done using NIH Image, IPLab, and Photoshop 7.0, as in another study (27). For rhizotron samples, both the total number of red cells and the number of bright GFP cells (those that could be detected without adjusting contrast levels) were enumerated. In a second analysis of the same images, the total number of GFP cells observable after enhancing the contrast to reveal dimly green fluorescent cells was counted. One-way ANOVA was performed using JMP (SAS Institute, Inc., Pacific Grove, PA).


arrow
RESULTS
 
Nitrate response of pNarG-Ice fusions.
Nitrate-dependent ice nucleation was observed under anaerobic conditions in both E. coli DH5{alpha} and E. herbicola 299R harboring a transcriptional fusion of a 592-bp HindIII-EcoRI fragment containing the narG promoter-regulatory region to an inaZ reporter gene on plasmid pNarG-Ice. In LB medium amended with 10 mM sodium nitrate, the ice nucleation activities of E. coli(pNarG-Ice) and E. herbicola(pNarG-Ice) were 85- and 300-fold greater than those in LB medium without nitrate amendment under anaerobic conditions, respectively (Table 2). Both strains expressed ice nucleation activity that increased with increasing nitrate concentration in the range of 0.1 µM to 10 mM under anaerobic conditions (data not shown). Neither E. herbicola 299R nor E. coli DH5{alpha} expressed detectable icenucleation activity with or without parental plasmid pPROBE-NI under either anaerobic or aerobic incubation conditions (data not shown). The ice nucleation activity of E. herbicola(pNarG-Ice) or E. coli(pNarG-Ice) was not enhanced by nitrate addition under aerobic incubation conditions, nor was E. herbicola(pNarG-Ice) influenced by ammonium under either anaerobic or aerobic conditions (Table 2). In the presence of nitrate under aerobic incubation, both E. herbicola(pNarG-Ice) and E. coli(pNarG-Ice) expressed only low levels of ice nucleation activity (about 10–6 to 10–7 ice nuclei/CFU), which were about 100,000- to 1,000,000-fold lower than those in strains grown anaerobically (Table 2).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Ice nucleation activity of Erwinia herbicola 299R(pNarG-Ice) and Escherichia coli DH5{alpha}(pNarG-Ice) in the presence of various nitrogen-containing compounds under anaerobic and aerobic conditions

Bacterial inaZ-based sensors regulated by nitrate under both aerobic and anaerobic incubation.
We expected that transcription of NarG should no longer be suppressed in the presence of oxygen in strains containing the mutant L28H-fnr gene. Under aerobic conditions, the ice nucleation activity of E. coli(pNice) was more than 100-fold higher in media amended with 10 mM sodium nitrate than in media lacking nitrate (about 10–1 ice nuclei/CFU and 10–3 ice nuclei/CFU, respectively) (Fig. 2A). The ice nucleation activity of E. coli(pNice) increased with increasing concentrations of nitrate in the medium in the range of about 10–3 to 10–7 M (Fig. 2B). These nitrate-mediated changes in ice nucleation activity were detectable within 2 hours after inoculation of the bacterium into nitrate-containing media, and little further increase in ice nucleation occurred during an additional 6-hour exposure (Fig. 2A). Since ice nucleation proteins are stable, continued cell growth was balanced by continued ice protein production in these cultures.



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 2. Ice nucleation activity of E. coli DH5{alpha}(pNice) (A and B) and E. cloacae EcCT501R(pNice) (C and D) as a function of incubation time (A and C) and nitrate concentration (B and D). Values are the means for three replicate cultures, and error bars indicate standard errors of the means. (A) E. coli DH5{alpha}(pNice) grown aerobically in M9 medium with 10 mM sodium nitrate (open circles) or without nitrate amendment (closed circles) over time. (B) E. coli DH5{alpha}(pNice) 6 h after incubation in M9 amended with different concentrations of NaNO3 under aerobic conditions. (C) E. cloacae EcCT501R(pNice) grown in M9 under aerobic conditions with 10 mM sodium nitrate (open circles) or without nitrate amendment (closed circles) over time. (D) E. cloacae EcCT501R(pNice) 6 h after incubation in M9 amended with different concentrations of NaNO3 under aerobic (open circles) or anaerobic (closed circles) conditions. The regression lines were based on the mean ice nucleation activity at each concentration of nitrate: y = 1.184x + 2.568 and r2 = 0.987 for "aerobic" regression line (thin line); y = 1.233x + 2.759 and r2 = 0.988 for "anaerobic" regression line (thick line).

While pNarG-Ice conferred nitrate-dependent ice nucleation under anaerobic conditions in E. herbicola 299R, pNice did not confer expression of greater ice nucleation activity in E. herbicola 299R in either LB or M9 medium amended with 10 mM sodium nitrate, compared with media without added nitrate, under aerobic conditions (data not shown). While higher than that exhibited by E. herbicola(pPROBE-NI), the activity of E. herbicola(pNice) was nearly as low as that exhibited by E.herbicola(pNarG-Ice) under aerobic conditions, suggesting either that the modified Fnr did not function in E. herbicola 299R or that its effect was overcome by its native oxygen-dependent Fnr.

In order to develop a nitrate reporting strain that would be more environmentally competent than E. coli derivatives, pNice was introduced into the soil bacterium E. cloacae strain EcCT501R. Under aerobic conditions, EcCT501R(pNice) expressed much higher ice nucleation activity in M9 medium amended with 10 mM sodium nitrate (Fig. 2C) than did E. coli DH5{alpha}(pNice) (about 10–1 ice nuclei/CFU and 10–3 ice nuclei/CFU, respectively) (Fig. 2A). Strain EcCT501R(pNice) exhibited increased ice nucleation within 1 hour after exposure to nitrate under aerobic conditions, and 10,000-fold greater ice nucleation activity was expressed in M9 medium containing 10 mM sodium nitrate compared to that in medium without added nitrate (Fig. 2C). Under aerobic conditions, the ice nucleation activity of EcCT501R(pNice) increased with increasing concentrations of nitrate in M9 medium in the range of 10–7 to 10–3 M (Fig. 2D). A similar nitrate-responsive ice nucleation activity was seen under anaerobic conditions (Fig. 2D).

Assessment of nitrate availability in the rhizospheres of wild oat and snap bean by using an inaZ-based biosensor.
In contrast to that of E. cloacae EcCT501R, the population size of E. coli DH5{alpha} declined rapidly after inoculation into the rhizosphere of snap bean (Fig. 3C). The population size of the E. cloacae biosensor strain remained at the relatively high numbers at which it was inoculated onto roots for over a day (Fig. 3A and 3B). For this reason, E. cloacae EcCT501R was used as a host strain for reporter gene fusions to evaluate nitrate availability in the rhizosphere.



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 3. Population dynamics of Enterobacter cloacae EcCT501R(pNice) and Escherichia coli DH5{alpha}R(pNice) in the rhizosphere of wild oat or snap bean. (A and B) E. cloacae EcCT501R(pNice) in the rhizospheres of wild oat (A) and snap bean (B), which were grown in soil amended with 10 mM or0.1mM of sodium nitrate or without nitrate amendment. (C) E. coli DH5{alpha}R(pNice) in the rhizosphere of snap bean in unamended soil. Error bars indicate the standard errors of the mean log-transformed population sizes obtained for three replicates.

Population sizes of EcCT501R(pNice) in the rhizospheres of wild oat (Fig. 3A) and snap bean (Fig. 3B) amended with 10 mM nitrate were only slightly higher than in those amended with 0.1 mM nitrate or without nitrate amendment. On snap bean roots, about 106.5 CFU were recovered per gram of rhizosphere soil amended with 10 mM sodium nitrate, compared to 106.3 CFU per gram of rhizosphere soil with lower nitrate concentrations (Fig. 3B), and this difference was not significant. Population sizes of EcCT501R(pNice) in the rhizo- spheres of wild oat and bean plants both remained relatively constant. Thus, we found no evidence for nitrate-mediated increases in populations of the biosensor strains over this short time period.

The ice nucleation activity of EcCT501R(pNice) was about 10–2 ice nuclei/CFU at 3 hours after inoculation onto the rhizosphere of wild oat amended with 10 mM sodium nitrate, which was about 100-fold higher than that for roots grown in soil without nitrate amendment (Fig. 4), at both 3 and 6 h after inoculation of the biosensor bacteria. Since the ice nucleation activity of the cells used as inoculum was less than 10–5 ice nuclei/CFU, it is clear that nitrate-dependent expression of ice nucleation occurred in both soils. The ice nucleation activity in the rhizosphere of wild oat amended with 0.1 mM of sodium nitrate was also significantly higher than that with no nitrate amendment at 3 and 6 h after inoculation of the bacterium (Fig. 4). The ice nucleation activity of the nitrate biosensor in the rhizosphere of bean plants was similar at a given nitrate amendment to that for wild oat (data not shown).



View larger version (14K):
[in this window]
[in a new window]
  		FIG. 4. Ice nucleation activity of Enterobacter cloacae EcCT501R(pNice) in the rhizosphere of wild oat. Wild oat was grown in soil amended with 10 mM, 0.1 mM, or no sodium nitrate, and bacterial cells were spray inoculated onto the roots 12 h after soil nitrate amendment. Error bars indicate the standard errors of the mean cell-normalized ice nucleation activities for three replicates.

Whole-cell GFP-based biosensors regulated by nitrate in culture.
Since E. cloacae EcCT501R(pNice) exhibited a population-level nitrate-responsive ice nucleation activity and environmental competence in the rhizosphere, we developed a GFP-based variant to enable observations of nitrate availability to single cells as well as population-based estimates. The ice nucleation reporter gene was replaced in pNice to produce pNgfp. As expected, cells of EcCT501R harboring pNgfp exhibited nitrate-dependent GFP fluorescence under both aerobic and anaerobic conditions. The expression of GFP in the presence of nitrate did not alter the growth of this strain (data not shown). Little GFP fluorescence was observed in cells of this strain in media lacking nitrate, but strong green fluorescence was observed in nitrate-containing media (Fig. 5). The GFP fluorescences of single cells grown in broth cultures containing a given concentration of nitrate were similar (Fig. 5). Because of the narrow range of GFP fluorescence of cells exposed to a given nitrate concentration, the GFP fluorescence of individual cells provides a relatively unambiguous estimate of its nitrate environment.



View larger version (12K):
[in this window]
[in a new window]
  		FIG. 5. Histogram showing the fluorescence of individual cells of EcCT501R(pNgfp) grown in M9 minimal medium containing no nitrate (A), 0.1 µM nitrate (B), 10 µM nitrate (C), or 1 mM nitrate (D). After 4 h of incubation, cells were fixed, placed on slides, and immediately viewed with an epifluorescence microscope. The means (variances) of cells exposed to 0, 10–7, 10–5, and 10–3 M NaNO3 were 7.171 (3.519), 27.25 (32.00), 277.4 (204.0), and 457.8 (362.3), respectively.

To enable the visualization of all cells of the nitrate biosensor and only cells of this strain, this strain was also transformed with pVS1-dsRed, a plasmid that is compatible with pNgfp and which harbors a constitutively expressed dsRed marker gene. This red-marked nitrate biosensor grew slightly slower than the strain without pVS1-dsRed but exhibited a nitrate-induced GFP fluorescence similar to that of EcCT501R(pNgfp) (Fig. 6). The measurement of GFP fluorescence in response to nitrate by using a fluorometer (Fig. 6) gives slightly different results than single-cell measurements using microscopy (Fig. 5); this could be due to the different methods used for detection. Little increase in GFP fluorescence was observed in cells grown in less than about 10–5 M nitrate, but fluorescence increased rapidly thereafter with increasing nitrate concentrations to about 10–2 M nitrate. The GFP fluorescence of cells exposed to 10–3 M nitrate was about 1,000-fold greater than that of control cells without nitrate exposure.



View larger version (10K):
[in this window]
[in a new window]
  		FIG. 6. GFP fluorescence, determined using a fluorometer, of cell cultures of E. cloacae EcCT501R(pNgfp)(pVS1-dsRed) in minimal medium containing different concentrations of nitrate. There was no detected GFP fluorescence without added nitrate (data not shown). OD, optical density unit.

Assessment of nitrate availability in the rhizosphere of wild oat by using a GFP-based NO3– sensor.
The GFP-based nitrate biosensor was applied to roots to determine if these whole-cell biosensors would yield sufficient GFP fluorescence to detect nitrate in natural environments. Microcosms were treated with one of four treatments: nitrate alone, glucose alone, nitrate plus glucose, or water. The effect of each variable (nitrate, glucose, and root zone) on the incidence of induction of nitrate biosensor cells was calculated based on measured fluorescence per cell. Glucose-only and nitrate-only treatments did not have a significant effect and are omitted in Fig. 7. Two replicate experiments were conducted, and a Student t test on the results (P< 0.05) determined that the two experiments' results were not significantly different and could be combined into one large data set.



View larger version (21K):
[in this window]
[in a new window]
  		FIG. 7. GFP expression by E. cloacae EcCT501R(pNgfp)(pVS1-dsRed) in the rhizosphere of Avena fatua; error bars indicate one standard error. Two hours prior to application of sensors to the rhizosphere, microcosms were amended with the equivalent of 6.6 mM NaNO3 plus 1.6% glucose (dark bars) or water (light bars). (A) Cells exhibiting any GFP fluorescence expressed as the total number of green fluorescing cells of any brightness divided by total number of red fluorescing cells (x100). (B) Cells exhibiting bright GFP fluorescence expressed as total number of very bright green fluorescing cells divided by total number of red fluorescing cells (x100). Statistically significant differences in either percent GFP induction or percent bright GFP induction are represented by different letters (for total percent induction, n = 140; for percent bright cells, n = 84).

While we did not attempt to quantify GFP fluorescence in this experiment, it was obvious that there were three recognizably distinct populations of cells in the rhizosphere: those that were very strongly induced, exhibiting "very bright" GFP fluorescence; those that were only slightly induced, exhibiting "dim" but detectable GFP fluorescence; and those that were uninduced, having undetectable fluorescence. The fluorescence of very bright cells was apparent without any image adjustment, while the majority of cells were dim cells, unambiguously discernible only after image contrast adjustment. The bright cells had presumably sensed high local concentrations of nitrate; these were analyzed separately, either as a fraction of total cells with RFP fluorescence (Fig. 7B) or as a fraction of total cells with GFP fluorescence (Table 3), from the total dim plus bright fluorescent cells (Table 3 and Fig. 7A).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Model of contributing effects of nitrate addition, root zone location, and glucose addition on E. cloacae(pNgfp) in soil microcosmsa

By this approach we found that nitrate has a strong effect on both the incidence of total cells responding detectably to nitrate and the percentage of highly induced cells (Fig. 7; Table 3). Location along the root significantly affected the proportion of highly induced cells but not that of cells as a whole that responded to nitrate by GFP fluorescence (Fig. 7A). The nitrate biosensor E. cloacae EcCT501R(pNgfp)(pVS1-dsRed) sensed 10 times as many regions of apparent high nitrate availability (P = 0.0073) in the bulk soil compared to the root hair and root tip regions of the rhizosphere in the absence of added nitrate (Fig. 7B). One-way ANOVA established significant differences (F = 2.0582; P = 0.0524) in overall nitrate availability between the root hairs without added nitrate and root hairs with added nitrate (Fig. 7A). One-way ANOVA also indicated significant differences (F = 2.930; P = 0.0091) in high nitrate availability between the bulk soil and rhizosphere, particularly root hairs and mature roots, when nitrate was added (Fig. 7B). Glucose treatment did not affect overall GFP fluorescence or induction of bright fluorescence.

Multiple regression analysis was used to build a model that would reflect the contributions of different factors to the incidence of detectable nitrate responsiveness of the nitrate biosensor and that proportion of highly responsive cells (Table 3). The variable that best fits the data is reported first, and the variable that fits the residuals of the data from the first variable is reported second, etc. For each iteration of the model, the F statistic (as part of the whole-model F test), P value, and r2 are reported (Table 3). This analysis shows that the only factors that contribute to GFP induction of the nitrate biosensor are nitrate and root zone, while glucose has a negligible effect on bacterial GFP expression of cells as a whole, and there is no interaction between added nitrate and glucose. Root zone contributed to induction of cells with very bright GFP fluorescence (P = 0.0073), as did nitrate and the cross product of nitrate and glucose (P = 0.070 and P = 0.0743), but not glucose alone (Table 3).


arrow
DISCUSSION
 
Transcription of a nitrate-inducible promoter by the root-colonizing bacterium E. cloacae, both in culture media and in the rhizospheres of wild oat and snap bean, was easily detected by an ice nucleation reporter gene system as well as a GFP reporter gene system coupled with a dsRed marker gene. Using the two different reporter systems, we were able to estimate the amount of nitrate available to rhizosphere bacteria. The inaZ-based nitrate biosensor expressed significantly higher levels of ice nucleation activity in the rhizospheres of wild oat and snap bean grown in soils amended with nitrate than in the unamended control (Fig. 4). These findings were in agreement with nitrate levels that were detected using the GFP marker reporter construct, which reported much higher levels of available nitrate in the bulk soil than in the rhizosphere (Fig. 7B).

Host strain survival in the rhizosphere was an important determinant of the success of our whole-cell biosensors. E. coli DH5{alpha} exhibited significant declines in population sizes in the rhizospheres of plants (Fig. 3C) and could not be used in this study. We directly applied the E. cloacae EcCT501R biosensor to roots and soils and incubated for 24 h, a period sufficient for narG induction. In Loper and Henkels' study (32) of iron availability using a pvd-inaZ ice nucleation reporter gene fusion, lack of growth in the rhizosphere did not hinder the usefulness of the inaZ reporter construct. Leveau and Lindow (27) pointed out that GFP fluorescence per cell is not only a function of the activity of the promoter but also a function of the rate at which the cell is reproducing; fast growth of the GFP biosensor population will result in dilution of the GFP signal upon cell division, causing an underestimate of the queried environmental signal. In our study, the population size of E. cloacae strain EcCT501R did not change significantly during the 24-hour course of our experiments (Fig. 3). Thus, such growth rate effects were probably not a contributing factor to the interpretation of the biosensors' reports of bioavailable nitrate.

Because of the ability of the L28H-fnr gene product to overcome oxygen repression in E. cloacae EcCT501R, we used this strain for all in situ experiments (Fig. 3) (23, 42). E. coli DH5{alpha} exhibited narG promoter transcription and oxygen-dependent repression of NarG similar to those of EcCT501R (1, 23). Kiley and Reznikoff demonstrated that an Fnr mutant strain of E. coli containing the L28H-fnr gene product exhibited a 47-fold increase in beta-galactosidase with nitrate addition under aerobic conditions (23). E. herbicola 299R exhibited good nitrate induction of NarG under anaerobic conditions and has performed well in the rhizosphere as a whole-cell biosensor (3, 4, 18). However, the introduction of the L28H-fnr gene product did not overcome oxygen repression as did the E. coli DH5{alpha} and E. cloacae EcCT501R strains in this study. That the reporter strains are nitrate inducible aerobically is crucial to their application in soils, since in this application the microcosm soil was well aerated, and most applications would be under aerobic conditions (6, 18, 42, 46).

The inaZ reporter gene system is a sensitive and convenient tool for studies of nutrient conditions in natural environments such as the rhizosphere, partly because ice nucleation activity increases with roughly the square of ice protein content (29). This exponential increase in ice nucleation activity with increased transcription results in very high induction levels compared to other reporter genes. Out of a set of three sucrose-regulated reporter gene fusions made with GFP, inaZ, or lacZ, the inaZ reporter had the widest range of activity and had much higher sensitivity than GFP or lacZ (35, 41). The extremely low background level of ice nucleation activity in soil also makes inaZ a good reporter gene for use in the rhizosphere. Transcription of narG is induced about 18-fold (45) to 118-fold (2) by addition of 40 mM nitrate in culture media under anaerobic conditions. Given the exponential relationship between ice protein content and ice nucleation activity, we would expect about a 13,000-fold increase in ice nucleation activity in cells harboring the narG-inaZ gene fusions. Indeed, we observed about a 1,000- to 60,000-fold increase in ice nucleation activity upon addition of different concentrations of nitrate to E. coli and E. cloacae biosensor strains in culture (Fig. 2). Since even very low levels of gene expression are measurable using inaZ reporter gene fusions (34), this system often provides evidence of gene expression that is not detectable with other reporter genes (35). The inaZ-based nitrate biosensor was responsive to nitrate concentrations of as low as 10–7 M under both aerobic and anaerobic conditions, about 10-fold lower than that detected using the GFP-based biosensor (Fig. 2 and 6), making it particularly useful for evaluation of nitrate availability in complex natural environments such as the rhizosphere.

The GFP-based nitrate biosensor provides small-scale spatial information about nitrate availability near roots that is not possible to obtain with the ice nucleation biosensor (6, 27, 30). The use of dsRed as a marker gene enables recognition of all cells in a population of bioreporter cells, so that GFP fluorescence of individual cells can be resolved. However, high and variable background autofluorescence in the soil can make quantification of GFP fluorescence intensity difficult for some soils. Furthermore, subtracting high background levels of fluorescence from images of bioreporter strains in situ can artificially lower GFP fluorescence measurements and underestimate effector molecule concentrations. Relatively low background fluorescence in our soil allowed us to evaluate the availability of nitrate at the single-cell scale by using microscopy. We have also pooled the results and done statistical analyses to examine the availability of nitrate and the impact of root uptake on residual soil nitrate in soil near roots. This type of analysis eliminates some of the heterogeneity observed at the single-cell level, and it allows us to quantify larger spatial trends in nitrate availability at meso-spatial scales. Using the GFP reporter at the single-cell scale allows assessment of nitrate availability in soil microsites (niches) with dimensions of micrometers. Such resolution of soil bacterial environments at the niche scale is critical to understanding the in situ physiological ecology of soil bacteria.

Other nitrate or nitrogen biosensors have been constructed using different reporter genes, but each has their drawbacks to applications in the rhizosphere. An E. coli nitrate biosensor harboring a fusion of the narG promoter region of E. coli to the lux genes of Photorhabdus luminescens (42) could detect nitrate at as low as 0.5 x 10–5 M in culture media. While this is on par with the detection limit of the pNice sensor (Fig. 2) and the pNgfp sensor (Fig. 6), this lux-based sensor functions only under anaerobic conditions. While reporter systems employing the bacterial luciferase genes luxAB have been used for in situ detection of other nutritional responses of bacteria in the rhizosphere (20, 26), the use of lux-based reporter gene systems in soil environments is confounded by soluble pigments and soil particles which can attenuate light emitted by bioluminescent strains, as well as a low level of metabolic energy of cells that may limit light production independently of transcription of narG (24, 30). In studies using a lux reporter gene for the detection of nitrogen starvation in the barley rhizosphere, there was much less bioluminescence of the reporter bacteria in unamended soil than in nutrient-amended soil (20). Although the reporter was designed to detect N limitation, it was unclear whether the lack of signal was due to general insufficient metabolic energy or specific N limitation (20).

In our study, nitrate concentrations in unamended soils were barely detectable using either the GFP- or the ice nucleation-based nitrate sensor, confirming that our soil-plant mesocosms initially provided a relatively nitrogen-limited environment (Fig. 4 and 7). Dim cells were similar in intensity to cultured cells seeing 10–7 M or less nitrate (Fig. 5). Low apparent nitrate availability, even after nitrate addition (Fig. 7), suggested that during the 12-h preincubation prior to sensor addition, nitrate was rapidly consumed by roots and rhizosphere microorganisms. Independent estimates of bulk field nitrate availability in this California grassland soil done by colorimetric analysis of KCl extracts of soil solutions found that concentrations of nitrate, while variable, were about 0.5 µg NO3– N g–1 soil (about 100 µM) during periods of active plant uptake (7). The amount of nitrate available to the bioreporter in the rhizosphere in these studies was estimated to be between 0.1 and 1 µM in nonamended soils from measurements of nitrate-dependent ice nucleation activity in defined culture conditions (compare Fig. 2 and 4). We estimate the nitrate available to the E. cloacae strain in the rhizospheres of soils amended with 10 mM nitrate to be about 0.1 mM (Fig. 2 and 4). These biosensors thus report the biological availability of extremely dynamic nitrate pools. Based on the concentration of the nitrate solution sprayed on these root-soil systems, the GFP reports should have been dominated by brightly fluorescent cells. However, GFP reports from the rhizosphere indicted that plant/microbial uptake had rapidly depleted the added nitrate.

Based on measurement of fluorescence intensities of individual cells in the rhizospheres of nitrate-amended soils, we estimate that the weakly fluorescent cells in the rhizosphere sensed at most 1 µM nitrate. Strongly fluorescent cells were documented to occur at a concentration of 1 mM nitrate (Fig. 5). Assuming a soil water content of about 15%, the amount of nitrate added to the mesocosms would yield a soil solution nitrate concentration of about 6.6 mM. In cultures this concentration of nitrate induces high levels of GFP fluorescence that are nearly insensitive to further increases in nitrate concentrations. However, in soil receiving added nitrate, only a small subset of the bacterial population reported such high concentrations (ranging from 0.5% of the cells near the root tip to 3% of cells in the bulk soil). Thus, nitrate was not depleted from a small subset of soil microsites, microsites that were apparently isolated from active root uptake. Documentation by GFP reporters of the number and distribution of soil microsites diffusionally isolated from plant root uptake is in itself an interesting result. Such heterogeneity may be very important for explaining the ability of microbes and plants to efficiently compete with one another for inorganic nitrogen (14, 25).

Examination of different root zones by using GFP biosensors for nitrate availability reveals that there can be higher nitrate available to soil bacteria in the bulk soil compared to those near roots. When nitrate concentrations were very low (no nitrate added), the root hairs were most active in uptake (Fig. 7A). After nitrate addition, soil near roots was rapidly depleted (Fig. 7B). More conventional methods for nitrate detection in the soil are unable to detect small-scale heterogeneity of nitrate availability in the rhizosphere. In other work, we have successfully documented the mesoscale patterns of mineralization and nitrification in the A. barbata rhizosphere (D. J. Herman et al., submitted for publication) by using 15N-based methods. Patterns of N mineralization and nitrification were found to vary substantially along the root, in part due to plant-microbe competition for inorganic N (5, 8). In the current study, we also observe mesoscale differences in nitrate uptake in different root zones, but in addition, we discerned microscale patterns of nitrate access and utilization by roots and microbes. Bacterial cell biosensors allow us to map the spatial patterns of interaction between bacteria and plant roots at a scale that reveals important understanding of the interactions occurring.

This is the first application of a whole-cell biosensor for the detection of nitrate availability in the rhizosphere. The use of a reporter system linked to nitrate reductase genes achieves high specificity and sensitivity while providing either mesoscale (millimeters) or microscale (micrometers) spatial resolution. Further studies that use nitrate biosensors to monitor the spatial and temporal patterns of nitrate availability in the root systems under natural environmental conditions will likely provide new insight into nitrogen dynamics at scales that will enable better understanding of root-microbe interactions.


arrow
ACKNOWLEDGMENTS
 
We thank P. J. Kiley and W. S. Reznikoff for providing plasmid pPK434, Joyce Loper (USDA-ARS, Corvallis, OR) for providing Enterobacter cloacae strain EcCT501, and Beatriz Quiñones for providing the constitutively expressing dsRed plasmid. We also thank Daniel Gage for helpful comments regarding the microcosms.

This work was supported by grant 98-35107-6369 from the National Research Institute of the U.S. Department of Agriculture.


arrow
FOOTNOTES
 
* Corresponding author. Present address: Department of Plant and Microbial Biology, 111 Koshland Hall, University of California, Berkeley, CA 94720-3102. Phone: (510) 642-4174. Fax: (510) 642-4995. E-mail: icelab{at}socrates.berkeley.edu. Back

{dagger} Kristen DeAngelis and Pingsheng Ji contributed equally to this work. Back

{ddagger} Present address: North Florida Research and Education Center, University of Florida, Quincy, FL 32351. Back


arrow
REFERENCES
 
    1
  1. Bates, D. M., C. V. Popescu, N. Khoroshilova, K. Vogt, H. Beinert, E. Munck, and P. J. Kiley. 2000. Substitution of leucine 28 with histidine in the Escherichia coli transcription factor FNR results in increased stability of the 4Fe-4S (2+) cluster to oxygen. J. Biol. Chem. 275:6234-6240.[Abstract/Free Full Text]
    2. 2
  2. Berg, B. L., and V. Stewart. 1990. Structural genes for nitrate-inducible formate dehydrogenase in Escherichia coli K-12. Genetics 125:691-702.[Abstract]
    3. 3
  3. Brandl, M., E. M. Clark, and S. E. Lindow. 1996. Characterization of the indole-3 acetic acid (IAA) biosynthetic pathway in an epiphytic strain of Erwinia herbicola and IAA production in vitro. Can. J. Microbiol. 42:586-592.
    4. 4
  4. Brandl, M. T., and S. E. Lindow. 1996. Cloning and characterization of a locus encoding an indolepyruvate decarboxylase involved in indole-3-acetic acid synthesis in Erwinia herbicola. Appl. Environ. Microbiol. 62:4121-4128.[Abstract]
    5. 5
  5. Bremer, E., and P. Kuikman. 1997. Influence of competition for nitrogen in soil on net mineralization of nitrogen. Plant Soil 190:119-126.[CrossRef]
    6. 6
  6. Bringhurst, R. M., Z. G. Cardon, and D. J. Gage. 2001. Galactosides in the rhizosphere: utilization by Sinorhizobium meliloti and development of a biosensor. Proc. Natl. Acad. Sci. USA 98:4540-4545.[Abstract/Free Full Text]
    7. 7
  7. Canals, R. M., D. J. Herman, and M. K. Firestone. 2003. How disturbance by fossorial mammals alters N cycling in a California annual grassland. Ecology 84:875-881.
    8. 8
  8. Christie, P., and E. A. Wasson. 2001. Short-term immobilization of ammonium and nitrate added to a grassland soil. Soil Biol. Biochem. 33:1277-1278.[CrossRef]
    9. 9
  9. Colmer, T. D., and A. J. Bloom. 1998. A comparison of NH4+ and NO3– net fluxes along roots of rice and maize. Plant Cell Environ. 21:240-246.[CrossRef]
    10. 10
  10. Davidson, E. A., S. C. Hart, C. A. Shanks, and M. K. Firestone. 1991. Measuring gross nitrogen mineralization, immobilization, and nitrification by nitrogen-15 isotopic pool dilution in intact soil cores. J. Soil Sci. 42:335-350.[CrossRef]
    11. 11
  11. Davidson, E. A., J. M. Stark, and M. K. Firestone. 1990. Microbial production and consumption of nitrate in an annual grassland. Ecology 71:1968-1975.[CrossRef]
    12. 12
  12. Dick, W. A., and M. A. Tabatabai. 1979. Ion chromatographic determination of sulfate and nitrate in soils. Soil Sci. Soc. Am. J. 43:899-904.[Abstract/Free Full Text]
    13. 13
  13. Firestone, M. K. 1982. Nitrogen in agricultural soils, p. 289-326. Agronomy monograph no. 22. ASA-CSSA-SSSA, Madison, Wis.
    14. 14
  14. Gallardo, A., J. J. Rodriguez-Saucedo, F. Covelo, and R. Fernandez-Ales. 2000. Soil nitrogen heterogeneity in a Dehesa ecosystem. Plant Soil 222:71-82.[CrossRef]
    15. 15
  15. Gottschalk, G. 1986. Bacterial metabolism. Springer-Verlag, New York, N.Y.
    16. 16
  16. Govindarajan, A. G., and S. E. Lindow. 1988. Size of bacterial ice-nucleation sites measured in situ by radiation inactivation analysis. Proc. Natl. Acad. Sci. USA 85:1334-1338.[Abstract/Free Full Text]
    17. 17
  17. Green, R. L., and G. J. Warren. 1985. Physical and functional repetition in a bacterial ice nucleation gene. Nature 317:645-648.[CrossRef]
    18. 18
  18. Herman, D. J., L. J. Halverson, and M. K. Firestone. 2003. Nitrogen dynamics in an annual grassland: oak canopy, climate and microbial population effects. Ecol. Appl. 13:593-604.
    19. 19
  19. Jaeger, C. H., S. E. Lindow, W. Miller, E. Clark, and M. K. Firestone. 1999. Mapping of sugar and amino acid availability in soil around roots with bacterial sensors of sucrose and tryptophan. Appl. Environ. Microbiol. 65:2685-2690.[Abstract/Free Full Text]
    20. 20
  20. Jensen, L. E., and O. Nybroe. 1999. Nitrogen availability to Pseudomonas fluorescens DF57 is limited during decomposition of barley straw in bulk soil and in the barley rhizosphere. Appl. Environ. Microbiol. 65:4320-4328.[Abstract/Free Full Text]
    21. 21
  21. Keeney, D. R., and E. B. Nelson. 1982. Nitrogen-inorganic form, p. 643-698. In A. L. Page et al. (ed.), Methods of soil analysis, part 2, 2nd ed. Agronomy monograph. ASA and SSSA, Madison Wis.
    22. 22
  22. Kiley, P. J., and W. S. Reznikoff. Personal communication.
    23. 23
  23. Kiley, P. J., and W. S. Reznikoff. 1991. Fnr mutants that activate gene expression in the presence of oxygen. J. Bacteriol. 173:16-22.[Abstract/Free Full Text]
    24. 24
  24. King, J. M. H., P. M. Digrazia, B. Applegate, R. Burlage, J. Sanseverino, P. Dunbar, F. Larimer, and G. S. Sayler. 1990. Rapid, sensitive bioluminescent reporter technology for naphthalene exposure and biodegradation. Science 249:778-781.[Abstract/Free Full Text]
    25. 25
  25. Korsaeth, A., L. Molstad, and L. R. Bakken. 2001. Modeling the competition for nitrogen between plants and microflora as a function of soil heterogeneity. Soil Biol. Biochem. 33:215-226.[CrossRef]
    26. 26
  26. Kragelund, L., C. Hosbond, and O. Nybroe. 1997. Distribution of metabolic activity and phosphate starvation response of lux-tagged Pseudomonas fluorescens reporter bacteria in the barley rhizosphere. Appl. Environ. Microbiol. 63:4920-4928.[Abstract]
    27. 27
  27. Leveau, J. H. J., and S. E. Lindow. 2002. Bioreporters in microbial ecology. Curr. Opin. Microbiol. 5:259-265.[CrossRef][Medline]
    28. 28
  28. Li, S. F., and J. A. Demoss. 1987. Promoter region of the Nar operon of Escherichia coli: nucleotide sequence and transcription initiation signals. J. Bacteriol. 169:4614-4620.[Abstract/Free Full Text]
    29. 29
  29. Lindgren, P. B., R. Frederick, A. G. Govindarajan, N. J. Panopoulos, B. J. Staskawicz, and S. E. Lindow. 1989. An ice nucleation reporter gene system—identification of inducible pathogenicity genes in Pseudomonas syringae pv. phaseolicola. EMBO J. 8:1291-1301.[Medline]
    30. 30
  30. Lindow, S. E. 1995. The use of reporter genes in the study of microbial ecology. Mol. Ecol. 4:555-566.
    31. 31
  31. Lindow, S. E., E. Lahue, A. G. Govindarajan, N. J. Panopoulos, and D. Gies. 1989. Localization of ice nucleation activity and the IceC gene product in Pseudomonas syringae and Escherichia coli. Mol. Plant-Microbe Interact. 2:262-272.[Medline]
    32. 32
  32. Loper, J. E., and M. D. Henkels. 1997. Availability of iron to Pseudomonas fluorescens in rhizosphere and bulk soil evaluated with an ice nucleation reporter gene. Appl. Environ. Microbiol. 63:99-105.[Abstract]
    33. 33
  33. Loper, J. E., and S. E. Lindow. 1994. A biological sensor for iron available to bacteria in their habitats on plant surfaces. Appl. Environ. Microbiol. 60:1934-1941.[Abstract/Free Full Text]
    34. 34
  34. Loper, J. E., and S. E. Lindow. 1996. Reporter gene systems useful in evaluating in situ gene expression by soil- and plant-associated bacteria, p. 482-491. In C. J. Hurst (ed.), Manual of environmental microbiology. American Society for Microbiology, Washington, D.C.
    35. 35
  35. Miller, W. G., M. T. Brandl, B. Quinones, and S. E. Lindow. 2001. Biological sensor for sucrose availability: relative sensitivities of various reporter genes. Appl. Environ. Microbiol. 67:1308-1317.[Abstract/Free Full Text]
    36. 36
  36. Miller, W. G., J. H. J. Leveau, and S. E. Lindow. 2000. Improved gfp and inaZ broad-host-range promoter-probe vectors. Mol. Plant-Microbe Interact. 13:1243-1250.[Medline]
    37. 37
  37. National Research Council. 1989. Alternative agriculture. Board on Agriculture. National Academy Press, Washington, D.C.
    38. 38
  38. Nelson, E. B. 1988. Biological control of Pythium seed rot and preemergence damping-off of cotton with Enterobacter cloacae and Erwinia herbicola applied as seed treatments. Plant Dis. 72:140-142.[CrossRef]
    39. 39
  39. Nieto, K. F., and W. T. Frankenberger. 1985. Single column ion chromatography. 1. Analysis of inorganic anions in soils. Soil Sci. Soc. Am. J. 49:587-592.[Abstract/Free Full Text]
    40. 40
  40. Norton, J. M., and M. K. Firestone. 1996. N dynamics in the rhizosphere of Pinus ponderosa seedlings. Soil Biol. Biochem. 28:351-362.[CrossRef]
    41. 41
  41. Prest, A. G., M. K. Winson, J. R. M. Hammond, and G. Stewart. 1997. The construction and application of a lux-based nitrate biosensor. Lett. Appl. Microbiol. 24:355-360.[CrossRef][Medline]
    42. 42
  42. Rattray, E. A. S., J. I. Prosser, L. A. Glover, and K. Killham. 1995. Characterization of rhizosphere colonization by luminescent Enterobacter cloacae at the population and single-cell levels. Appl. Environ. Microbiol. 61:2950-2957.[Abstract]
    43. 43
  43. Saiki, R. K. 1990. Amplification of genomic DNA, p. 13-20. In M. A. Innis (ed.), PCR protocols, a guide to methods and applications. Academic Press Inc., San Diego, Calif.
    44. 44
  44. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
    45. 45
  45. Schroder, I., S. Darie, and R. P. Gunsalus. 1993. Activation of the Escherichia coli nitrate reductase (NarGHJI) operon by NarL and Fnr requires integration host factor. J. Biol. Chem. 268:771-774.[Abstract/Free Full Text]
    46. 46
  46. Sorensen, J., L. E. Jensen, and O. Nybroe. 2001. Soil and rhizosphere as habitats for Pseudomonas inoculants: new knowledge on distribution, activity and physiological state derived from microscale and single-cell studies. Plant Soil 232:97-108.[CrossRef]

Applied and Environmental Microbiology, December 2005, p. 8537-8547, Vol. 71, No. 12
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.12.8537-8547.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • de Werra, P., Baehler, E., Huser, A., Keel, C., Maurhofer, M. (2008). Detection of Plant-Modulated Alterations in Antifungal Gene Expression in Pseudomonas fluorescens CHA0 on Roots by Flow Cytometry. Appl. Environ. Microbiol. 74: 1339-1349 [Abstract] [Full Text]  
  • DeAngelis, K. M., Firestone, M. K., Lindow, S. E. (2007). Sensitive Whole-Cell Biosensor Suitable for Detecting a Variety of N-Acyl Homoserine Lactones in Intact Rhizosphere Microbial Communities. Appl. Environ. Microbiol. 73: 3724-3727 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by DeAngelis, K. M.
Right arrow Articles by Lindow, S. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by DeAngelis, K. M.
Right arrow Articles by Lindow, S. E.
Agricola
Right arrow Articles by DeAngelis, K. M.
Right arrow Articles by Lindow, S. E.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»