Production of the Phytohormone Indole-3-Acetic Acid by Estuarine Species of the Genus Vibrio

Abstract

Strains of Vibrio spp. isolated from roots of the estuarine grasses Spartina alterniflora and Juncus roemerianus produce the phytohormone indole-3-acetic acid (IAA). The colorimetric Salkowski assay was used for initial screening of IAA production. Gas chromatography-mass spectroscopy (GC-MS) was then employed to confirm and quantify IAA production. The accuracy of IAA quantification by the Salkowski assay was examined by comparison to GC-MS assay values. Indole-3-acetamide, an intermediate in IAA biosynthesis by the indole-3-acetamide pathway, was also identified by GC-MS. Multilocus sequence typing of concatenated 16S rRNA, recA, and rpoA genes was used for phylogenetic analysis of environmental isolates within the genus Vibrio. Eight Vibrio type strains and five additional species-level clades containing a total of 16 environmental isolates and representing five presumptive new species were identified as IAA-producing Vibrio species. Six additional environmental isolates similar to four of the Vibrio type strains were also IAA producers. To our knowledge, this is the first report of IAA production by species of the genus Vibrio or by bacteria isolated from an estuarine environment.

Estuaries along the east coast of temperate North America are ecologically valuable, productive systems dominated by only a few species of plants. Spartina alterniflora (smooth cord grass; hereinafter referred to as Spartina) is a keystone species responsible for very high rates of primary production in Atlantic coast marshes and is a major contributor to the global cycling of several elements (10, 14, 15, 35, 38, 39, 45). Juncus roemerianus (black needle rush; hereinafter referred to as Juncus) is a common subdominant species (28) residing in areas of higher elevation, lower salinity, and less frequent tidal inundation. The roots of these macrophytes are associated with a diverse assemblage of microorganisms, including N₂-fixing and sulfate-reducing bacteria, which greatly contribute to their productivity (30, 31).

The phytohormone indole-3-acetic acid (IAA) is the most commonly occurring naturally produced auxin and the most thoroughly studied plant growth regulator. IAA directs several aspects of plant growth and development (37), including the induction and regulation of a variety of processes: e.g., cell division, root extension, vascularization, apical dominance, and tropisms (6, 32). The effects of IAA on plant root tissue are concentration dependent and can be species specific. Responses to increasing IAA concentrations advance from the stimulation of primary root tissue to the development of lateral and adventitious roots and finally to the complete cessation of root growth (1, 6, 16, 29, 32, 37, 44).

Many microorganisms interact with and affect their environment through the production and transudation of signal compounds (17). The findings of numerous studies (see, e.g., references 8, 23, 25, and 37) demonstrate that a variety of plant-associated terrestrial bacteria produce and exude IAA. Auxin synthesis by cyanobacteria has also been reported previously (40). IAA is thought to reduce the integrity of plant cell walls by upregulating the production of cellulases and hemicelluloses, resulting in the leakage of some simple sugars and other nutrients that would benefit root-associated microorganisms (17). Likewise, root growth would be an advantage to resident bacteria due to the increased availability of root exudates and root surface for growth. Microorganisms that produce IAA can influence the host plant and function as pathogens, symbionts, or growth regulators, depending on how their IAA production influences the concentration of the plant's endogenous IAA pool and on the sensitivity of the plant to auxin. Organisms such as Erwinia chrysanthemi, Pseudomonas savastanoi, and Agrobacterium tumefaciens are phytopathogens of many host plants (11, 21, 23, 46). Other organisms, including Azospirillum brasilense and Pseudomonas putida GR12-2, have proven beneficial to plants, and many IAA producers have been shown to stimulate increases in root mass and/or length (20, 37, 44).

The aim of the present study was to assess IAA synthesis by Vibrio strains isolated from the roots of highly productive salt marsh grasses. The Salkowski assay was used to perform an initial screening for the presence of IAA, gas chromatography-mass spectroscopy (GC-MS) verified and quantified IAA production, and multilocus sequence typing (MLST) analysis classified all isolates within the genus Vibrio.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

All environmental strains used in this study were isolated from the rhizoplanes of Spartina and Juncus specimens collected from the North Inlet salt marsh near Georgetown, SC (33°20′N, 79°12′W) by Bagwell et al. (3). Briefly, short-form Spartina was collected in January 1994, and tall-form Spartina and Juncus were collected in June 1995. Roots measuring ∼1.25 mm or less in diameter from several plants were rinsed with deionized water and then stab inoculated into nitrogen-free semisolid media having a pH of either 7.0 or 7.5 and containing various carbon sources (3). These root cultures were incubated in the dark at 30°C until bacterial outgrowths from the root surfaces were apparent, and then the outgrowths were transferred by stab inoculation into fresh semisolid media. Strains were isolated by being streaked onto nitrogen-free medium plates and maintained on Bacto marine agar (Difco, Sparks, MD). Strain designations indicate the source plant (J for Juncus, S for short-form Spartina, and T for tall-form Spartina), the carbon source used in the enrichment medium (C for citrate, G for glucose, M for malate, and S for sucrose), and the pH of the medium (1 for pH 7.0 and 2 for pH 7.5). For example, strain J-C2-35 was isolated from Juncus by using citrate as a carbon source in a medium of pH 7.5 and was the 35th strain isolated (35). These rhizoplane isolates were evaluated by Bagwell et al. (3) for the Gram reaction; cell size, shape, arrangement, and motility; the production of endospores; and the production of cytochrome oxidase, peroxidase, and several extracellular enzymes and were then further examined for a variety of physiological properties and the utilization of various substrates by using API test strips (Fisher, Pittsburg, PA) and plates from Biolog (Hayward, CA). The 25 strains employed in this study were identified as Vibrio species on the basis of results from the above-mentioned tests and 16S rRNA sequencing (see below). Eight Vibrio type strains (Table 1) were included in the Salkowski and GC-MS analyses, and sequences from 39 type strains were included in the multilocus sequence analysis (Fig. 1).

TABLE 1.

Comparison of results of IAA quantification by GC-MS and the Salkowski assay^a

Type strain or environmental isolate	IAA concn (μg ml−1) determined by:		No. of cells ml−1	IAA production (fg cell−1)
	GC-MS	Salkowski assay
S-C1-5	0.00	4.75 ± 0.33	1.80 × 109	0.00
S-C1-13	0.00	1.46 ± 1.57	1.70 × 109	0.00
T-G2-10-B2	0.00	0.92 ± 0.03	5.70 × 109	0.00
J-C1-1aGR2	0.12 ± 0.04	10.00 ± 1.15	4.00 × 108	0.30
V. hispanicus LMG 13240	0.13 ± 0.02	1.09 ± 0.28	4.28 × 108	0.30
J-C1-25	0.14 ± 0.01	8.98 ± 0.53	1.32 × 109	0.11
V. fischeri ATCC 700601	0.17 ± 0.03	0.00	5.10 × 109	0.033
V. alginolyticus ATCC 17749	0.22 ± 0.06	4.06 ± 1.21	1.25 × 109	0.18
T-C2-11	0.28 ± 0.05	8.00 ± 0.87	2.48 × 109	0.11
S-M2-14-B1	0.28 ± 0.12	3.96 ± 0.15	7.33 × 108	0.38
V. diazotrophicus ATCC 33466	0.42 ± 0.05	1.47 ± 0.97	1.22 × 109	0.34
V. parahaemolyticus ATCC 17802	0.42 ± 0.06	8.91 ± 0.57	1.81 × 109	0.23
T-S2-8	0.46 ± 0.05	5.67 ± 0.11	4.16 × 109	0.11
S-G1-1-B1	0.58 ± 0.18	4.83 ± 0.74	1.59 × 109	0.36
J-S2-6	0.45 ± 0.19	4.10 ± 0.72	1.58 × 109	0.29
T-S2-7	0.77 ± 0.33	4.21 ± 0.17	1.92 × 109	0.40
V. fluvialis ATCC 33809	0.88 ± 0.14	12.03 ± 0.58	2.05 × 109	0.43
J-C2-38	0.92 ± 0.26	7.74 ± 1.61	2.79 × 109	0.33
J-C2-40	1.26 ± 0.43	9.83 ± 0.83	2.09 × 109	0.60
J-S2-26	2.01 ± 0.16	5.52 ± 0.36	3.42 × 109	0.59
V. natriegens ATCC 14048	2.45 ± 0.42	2.69 ± 0.12	4.85 × 108	5.1
J-S2-17	2.46 ± 0.13	4.75 ± 1.09	1.03 × 109	2.38
J-S2-12	3.08 ± 0.91	5.53 ± 0.60	2.73 × 109	1.13
T-G2-12w-B2	3.40 ± 0.38	5.63 ± 1.02	5.68 × 109	0.60
J-S2-25	4.09 ± 0.80	4.85 ± 0.19	2.49 × 109	1.64
J-S2-8	4.80 ± 1.15	21.43 ± 3.58	3.64 × 109	1.32
J-C2-35	4.83 ± 0.74	12.57 ± 0.84	2.36 × 109	2.05
T-C2-8	4.90 ± 2.27	20.15 ± 1.99	3.07 × 109	1.60
V. pacinii LMG 19999	5.20 ± 0.64	10.39 ± 0.77	6.80 × 108	7.65
J-C2-20op	5.22 ± 1.09	8.23 ± 0.91	2.12 × 109	2.46
J-M2-6	5.71 ± 1.07	6.69 ± 1.02	7.90 × 108	7.23
T-C2-3	8.04 ± 3.12	16.69 ± 1.52	3.20 × 109	2.51
T-S2-9	12.78 ± 1.45	18.60 ± 2.77	3.00 × 109	4.27

^aIAA concentrations are means ± standard deviations (n = 3). Calculations of cell-specific IAA production employed results from the GC-MS method.

FIG. 1.

Neighbor-joining phylogenetic tree of concatenated 16S rRNA, recA, and rpoA gene sequences constructed using the Jukes-Cantor method of correction. Strains represented in bold were tested for IAA production. Bootstrap values represent 1,000 replications; values of less than 50 are not shown.

The minimal medium employed by Bagwell et al. (3) was used for experiments examining IAA production. The basal medium contained the following (in grams per liter): NaCl, 28; Na₂MoO₄, 0.01; Tris HCl, 6.0; and NH₄Cl, 5.0 (pH 7.0). After being autoclaved, this medium was amended with the following (final concentrations): 2 mM MgSO₄, 400 μM CaCl₂, 11 mM glucose, 30 mM K₂HPO₄, 50 μM FeCl₃, and 0.98 mM (200 μg ml⁻¹) tryptophan (Trp). IAA production was not detectable in media lacking Trp.

Quantification of IAA by the Salkowski assay.

All strains were incubated in triplicate in 2 ml of minimal medium at 30°C in the dark with shaking for 72 h. Culture supernatants were recovered after centrifugation at 6,000 × g for 10 min. One milliliter of supernatant was mixed with 1 ml of Salkowski's reagent R1 (12 g liter⁻¹ FeCl₃ in 429 ml liter⁻¹ H₂SO₄). After room temperature incubation in the dark for 20 min, absorbance at 535 nm was determined (18). IAA concentrations were determined using triplicate standard curves for authentic IAA (Sigma-Aldrich, St. Louis, MO) prepared in basal medium.

Quantification of IAA by GC-MS.

Triplicate 10-ml cultures of each isolate were grown, with shaking, at 30°C in the dark. After incubation for 72 h, the internal standard 5-methoxy-indole-3-acetic acid (5-Me-IAA; Sigma-Aldrich) was added to a final concentration of 5 μg ml⁻¹. Supernatants were collected immediately after centrifugation at 4°C (6,000 × g) for 10 min. IAA and similar compounds were extracted from the supernatants as described by Minamisawa et al. (34) with minor modifications. Briefly, 5 ml of supernatant was adjusted to pH 2.5 to 3.0 with HCl and partitioned three times against a 1/3 volume of spectranalyzed-grade diethyl ether (Fisher). Ether phases were combined and evaporated to dryness in the dark under a stream of N₂ gas. Samples were then reconstituted in 5 ml of Optima liquid chromatography-mass spectrometry-grade acetonitrile (Fisher) and stored overnight in the dark at 4°C. A 1-ml sample was removed from each extract and evaporated to dryness in a rotary evaporator. Samples were reconstituted in 50 μl of acetonitrile, mixed 1:1 with bis(trimethylsilyl)trifluoroacetamide (BSTFA; Sigma-Aldrich), and trimethylsilyated for 15 min at 70°C. Samples were analyzed with a Hewlett Packard 5890 series II gas chromatograph-5971 series mass selective detector system with an autosampler. Electron impact ionization at 70 eV was used. GC-MS conditions were as follows: column, DB-5 (30 m by 0.25 mm by 0.025-mm film thickness); carrier gas, He; injection temperature, 280°C; initial temperature, 100°C for 3 min, increasing by 5°C min⁻¹ to a final temperature of 230°C; and detector temperature, 280°C. Mass-to-charge ratios (m/z values) from 50 to 335 were monitored using the scan mode. Integrated IAA and 5-Me-IAA peak areas were compared to triplicate standard curves for authentic IAA and 5-Me-IAA and used to calculate IAA quantities. Cell counts were performed using a hemocytometer (Hausser Scientific, Horsham, PA) with phase-contrast microscopy at a total magnification of ×1,000.

MLST.

Bacterial genomic DNA was extracted using the Wizard genomic DNA purification kit (Promega, Madison, WI). PCR was performed to amplify near-full-length 16S rRNA gene sequences by using primers 27F and 1492R (26) and to amplify partial recA and rpoA sequences by using recA1F, recA2R, rpoAF1, and rpoA3R (43) as listed in Table 2. Taq DNA polymerase (Qiagen, Valencia, CA) was used for all PCRs. The PCR program for the amplification of 16S rRNA gene sequences was as follows: initial denaturation at 94°C for 1 min; 3 cycles of 95°C for 1 min, 40°C for 1 min, and 72°C for 1 min 30 s; 30 cycles of 95°C for 1 min, 43°C for 1 min, and 72°C for 1 min 30 s; and final elongation at 72°C for 5 min. The PCR program used for the amplification of the recA and rpoA sequences consisted of initial denaturation at 95°C for 5 min and then 3 cycles of 95°C for 1 min, 55°C for 2 min 15 s, and 72°C for 1 min 15 s, followed by 30 cycles of 95°C for 35 s, 55°C for 1 min 15 s, and 72°C for 1 min 15 s and final elongation at 72°C for 7 min (43). PCR products were sequenced using primers 519F, 529R, 907R, 1099F, and 1240R for 16S rRNA genes (26, 47) and primers recA1F, recA2R, recA3F, recA4R, rpoA1F, rpoA3R, rpoA5F, and rpoA6R for recA and rpoA (43) (Table 2). The sequencing of amplicons was performed using a BigDye Terminator cycle sequencing kit, version 3.1 (Applied Biosystems, Foster City, CA), and an ABI Prism 3730 DNA analyzer. Sequences were edited using BioEdit version 7.0.9 (http://www.mbio.ncsu.edu/BioEdit/BioEdit.html) and ClustalX2 (27). Neighbor-joining trees were constructed from concatenated 16S rRNA, recA, and rpoA gene sequences with MEGA version 4.0 (42) using the Jukes-Cantor nucleotide substitution model. Sequence data for reference Vibrio species were obtained from GenBank.

TABLE 2.

Primers used in this study for the amplification and sequencing of recA, rpoA, and 16S rRNA genes

Primer	Sequence (5′-3′)	Reference
recA1F	TGARAARCARTTYGGTAAAGG	43
recA2R	TCRCCNTTRTAGCTRTACC	43
recA3F	TYGGBGTGATGTTYGGTAACC	43
recA4R	GGGTTACCRAACATCACVCC	43
rpoA1F	ATGCAGGGTTCTGTDACAG	43
rpoA3R	GHGGCCARTTTTCHARRCGC	43
rpoA5F	GCAGCDCGTGTWGARCARCG	43
rpoA6R	CGYTGYTCWACACGHGCTGC	43
27F	AGAGTTTGATCMTGGCTCAG	26
1492R	GCYTACCTTGTTACGACTT	26
519F	CAGCAGCCGCGGTAA	12
529R	CGCGGCTGCTGGCAC	12
907R	CCCCGTCAATTCCTTTGAGTTT	12
1099F	GCAACGAGCGCAACCC	12
1240R	CCATTGTAGCACGTGT	12

Nucleotide sequence accession numbers.

Nucleotide sequences determined in this study were submitted to GenBank under the accession numbers FJ176405 to FJ176467 and FJ464357 to FJ464371.

RESULTS AND DISCUSSION

Eight Vibrio type strains and 25 environmental strains of Vibrio species isolated from Spartina and Juncus rhizoplanes were screened for IAA production by using the colorimetric Salkowski assay and GC-MS (Table 1). Uninoculated media contained no detectable IAA. The Salkowski assay indicated a range of IAA concentrations from 0.00 μg ml⁻¹ for Vibrio fischeri (ATCC 700601) to 21.43 ± 3.58 μg ml⁻¹ (mean ± standard deviation) for strain J-S2-8 (Table 1). Of the eight type strains, only V. fischeri was negative for IAA production by the Salkowski assay. All of the Vibrio type strains and all 25 environmental isolates listed in Table 1 were analyzed using GC-MS for the verification and quantification of auxin production. IAA production by 22 of the 25 environmental isolates and all 8 type strains was quantified. The other three environmental strains had no detectable IAA, with a detection sensitivity limit of 0.01 μg ml⁻¹. IAA from Vibrio culture supernatants had an observed retention time of 25.4 min and produced a spectrum identical to that of authentic IAA, with a parent ion exhibiting an m/z of 319 and daughter ions exhibiting m/z values of 304, 276, 202, 186, and 147 (Fig. 2A and B). Strain T-S2-9 produced a considerably higher concentration of IAA (13 μg ml⁻¹) than the others, and strain J-C1-1a-GR2 produced the lowest concentration (0.12 μg ml⁻¹). GC-MS results were substantially lower than those of the Salkowski assay for 32 of the 33 IAA-producing strains. This finding is likely due to the reaction of the Salkowski reagent with indole derivatives other than IAA (13, 18). Given this potential for cross-reaction with other compounds, the Salkowski assay would be considered less accurate than the GC-MS assay, though still useful for screening large numbers of strains due to its relative ease of use and low cost. As data from V. fischeri indicate (Table 1), the relative insensitivity of the Salkowski assay can lead to false-negative results for strains producing low levels of IAA. Confirmation of the Salkowski assay results by GC-MS is important for the assurance of accurate IAA detection and quantification.

FIG. 2.

Mass spectra of authentic IAA, with a parent ion exhibiting an m/z of 319 (peak 6) and daughter ions exhibiting m/z values of 147 (peak 1), 186 (peak 2), 202 (peak 3), 276 (peak 4), and 304 (peak 5) (A); IAA extracted from strain J-C1-25 culture supernatant, with parent and daughter ions present (B); and IAM extracted from J-C1-25 culture supernatant, with a parent ion exhibiting an m/z of 318 (peak 4) and daughter ions exhibiting m/z values of 130 (peak 1), 202 (peak 2), and 303 (peak 3) (C). Abundance is an expression of ion intensity with arbitrary units.

Indole-3-acetamide (IAM) was detected in the culture supernatants of 21 of the 22 IAA-producing environmental isolates and in those of 6 of the 8 type strains analyzed. The observed spectrum matched that of authentic IAM (33), with a retention time of approximately 28.2 min, a parent ion with an m/z of 318, and daughter ions with m/z values of 303, 202, and 130 (Fig. 2C). V. diazotrophicus (ATCC 33466), V. hispanicus (LMG 13240), and J-C2-38 supernatants did not contain IAM. The three known pathways for the microbial synthesis of IAA are (i) the indole-3-acetonitrile (IAN) pathway (Trp → indole-3-acetaldoxime → IAN → IAA), (ii) the IAM pathway (Trp → IAM → IAA), and (iii) the indole-3-pyruvic acid (IPA) pathway (Trp → IPA → indole-3-acetaldehyde [IAAld] → IAA). The IAM and IPA pathways appear to be utilized by bacteria more commonly than the IAN pathway, although there have also been reports of auxin production via the IAN route (22, 23). While these pathways are all Trp dependent, Trp-independent pathways, as yet uncharacterized, have also been proposed (6).

The presence of IAM in the culture supernatants of these Vibrio species may indicate the use of the IAM pathway for the synthesis of IAA, but microorganisms have been shown previously to use multiple IAA synthesis pathways (8), and this may be the case with some of these strains. At present, our investigations of these strains have not yielded any evidence of the IPA pathway, as we did not detect the IPA pathway intermediate IAAld. IAAld may not be produced by these vibrios, or it may have spontaneously oxidized to IAA or indole-3-ethanol (6, 13, 24). The detection of IAM in culture supernatants provides a useful starting point for investigations of the pathway(s) employed by vibrios for auxin production, which will be a subject of future studies.

MLST, employing concatenated near-full-length 16S rRNA gene sequences and >80% complete recA and rpoA gene sequences (43), was used to determine phylogenetic affiliations of the 25 environmental isolates analyzed for IAA production (Fig. 1). We included 39 reference species (with sequence data obtained from GenBank) in our analysis. Estuarine isolates were found to form five clades. Clade 1 includes J-C1-1a-GR2 and J-C1-25; clade 2 includes J-M2-6; clade 3 includes J-S2-12, J-S2-17, J-S2-25, J-S2-26, and T-G2-12w-B2; clade 4 includes J-S2-6, T-G2-10-B2, T-S2-7, T-S2-8, and S-G1-1-B1; and finally, clade 5 is composed of J-S2-8, T-C2-3, T-C2-8, and T-S2-9. Isolates J-C2-35, J-C2-40, and T-C2-11 grouped closely to V. natriegens, while J-C2-20op, S-C1-5, and S-C1-13 grouped with V. parahaemolyticus. Isolates S-M2-14-B1 and J-C2-38 appear to be strains of V. pacinii and V. fluvialis, respectively (Fig. 1). The production of IAA was broadly distributed among members of the genus Vibrio, and species recovered from quite different niches produced auxin in culture. This finding may imply a broader association of Vibrio species with marine and estuarine plants than previously thought, which is consistent with the frequent recovery of these organisms (see, e.g., references 3, 4, 7, and 28) and their signature sequences (see, e.g., references 5, 9, 30, and 31) from such sources.

The synthesis of IAA by many terrestrial bacteria has been well documented (for a recent review, see reference 41). IAA production by freshwater wetland rhizosphere bacteria (19) and an ascosporogenous yeast (Pichia spartinae) associated with Spartina in Louisiana marshes (36) has also been reported, but prior to this study, auxin production by marine or estuarine bacteria had not been examined. Auxin production by estuarine Vibrio presents the potential for dynamic interspecies communication. To date, the highly diverse assemblage of microorganisms in the salt marsh remains largely unexplored. In these systems, primary production and decomposition are limited by combined nitrogen, which is obtained primarily through diazotrophy (2, 35). Most of the Vibrio strains analyzed in this study are diazotrophic and were recovered from the rhizoplane (C. R. Lovell, unpublished data). The N₂ fixation activity of salt marsh diazotrophs has been shown to increase in response to treatments that encourage plant growth and development, indicating a tight coupling between plant productivity and closely associated diazotrophs (2). An increase in root surface area, due to stimulation by IAA, would provide enhanced opportunity for the synthesis of usable nitrogen.

The Vibrio strains examined in this study have the capacity to interact with their host plants through molecular signaling pathways, possibly contributing to cycles of growth and senescence. Though we presently know very little of these interactions or their consequences, they may play a role in shaping the estuarine landscape, contributing to the accretion of plant biomass and the cycling of carbon. While physical factors such as drought, season, and semidiurnal tidal flushing of the rhizosphere are likely to affect the exogenous IAA concentration and therefore its effects on host plants, the finding of IAA production by plant-associated Vibrio species is certainly of interest and is being explored through ongoing plant inoculation studies.