Root inoculation with Azotobacter chroococcum 76A enhances tomato plants adaptation to salt stress under low N conditions

Plant growth conditions

A greenhouse experiment was carried at the experimental station of the University of Naples Federico II, Southern Italy (lat. 43°31’N, long. 14°58′E; alt. 60 m above sea level) with Micro Tom tomato plants. Tomato seeds were germinated in peat on May 2015 and grown until the 3rd-4th true leaf. Plants were transplanted in 10 cm Ø plastic pots at 30 Days-After-Sowing (DAS) containing pure peat moss (100%) and drip irrigated with nutrient solutions from 35 DAS. At transplanting, the growth substrate was inoculated with Azotobacter chrococcum 76A strain as described in section “Bacterial strain and inoculum preparation”. The irrigation water was characterized by a high bicarbonate concentration, presence of Na⁺ and Cl⁻ (8.63 mg l^− 1 Na⁺ and 10.3 mg l^− 1 Cl⁻) and with values of pH and electrical conductivity (EC) of 7.3 and 0.58 dS m^− 1, respectively. Plants were fertigated daily using six nutrient compositions, calculated from a standard nutrient solution (SNS) previously used for Micro Tom [29, 44]. The SNS composition was 1.93 mM NO₃, 2.53 mM P₂O₅, 7.64 mM K₂O, 1.48 mM MgO, 0.84 μM CuEDTA, 10 μM Fe DTPA, 3.45 μM Mn EDTA, 2.08 μM Mo, 0.83 μM Zn EDTA.

The SNS was distributed at two different concentrations, optimal (Opt - 100% SNS augmented with 3.25 mM NH NO₃) and sub-optimal (Sub - 50% of SNS without NH NO₃). Opt and Sub solutions were salinized by adding NaCl at concentrations of 50 or 100 mM. Non-salinized controls were provided for both Opt and Sub nutrient solutions. Overall, 6 nutrient solutions were compared: Sub-0NaCl, Sub-50 NaCl, Sub-100 NaCl, Opt-0 NaCl, Opt-50 NaCl, Opt-100 NaCl. The six nutrient solutions were pumped from independent 100 L tanks through a drip-irrigation system, with one emitter per plant (2 l h^− 1). Two fertilization treatments were applied per day, each of 1–3 min duration. The A. chroococcum inoculum [54] was given at 30 and 71 DAS. Non-inoculated plants were included as controls for all treatments.

Bacterial strain and inoculum preparation

The strain A. chroococcum 76A, previously selected for its multiple plant growth promotion activities as well as antimicrobial activity and tolerance to salt and drought stress [2, 54], was used. For inoculum preparation, the strain was grown in Yeast Mannitol (YM) liquid medium at 28 °C for 24 h in a rotary shaker (150 rpm). The culture was harvested at the late exponential phase of growth, centrifuged at 3293 x g, and the bacterial cells were suspended in a 5% sucrose solution at the ratio 1:5 (w:v). The strain was freeze-dried and added to quartz sand to reach a microbial concentration of approximately 1 × 10⁷ CFU g^− 1. The inoculum was applied in the central part of the pot in the peat substrate before tomato transplanting. For non-inoculated controls, a 5% sucrose solution lacking bacterial cells was applied to substrate similar to inoculated treatment. Viable microbial counts were performed immediately after transplanting and at the end of the experiment (90 DAS). Rhizosphere samples (10 g) were suspended in 90 mL of quarter strength Ringer’s solution (Oxoid, Milan, Italy). After shaking, suitable dilutions (1:10) were performed and used to inoculate nitrogen-free medium LG agar [40] by using the Surface Spread Plate Count Method. The plates were incubated for 48–72 h at 28 °C.

Biometric and physiological measurements

At the end of the experiment (90 DAS), plants were separated in leaves, stems, roots and fruits for fresh biomass determination and their tissues were dried in a forced-air oven at 80C for 72 h for the dry biomass determination. The final plant height, the number of leaves, fruits, leaf area and number were also recorded. The leaf area was measured with a Li-Cor 3000 area meter (Li-Cor, Lincoln, NE-USA).

After flowering, the following physiological measurements were performed: leaf relative water content (RWC), photosynthetic rate, leaf stomatal conductance, water potential and SPAD index. The RWC was measured on the fully expanded fourth or fifth leaf from the top of the plant. RWC was calculated according to Jones and Turner [22]. The photosynthetic rate (P) and stomatal conductance (gs) were determined with a gas exchange ADS LCA-4 infrared gas analyzer (Analytical Development Company, Hoddesdon, UK). The SPAD index was measured with a SPAD 502-Plus chlorophyll meter (Konica Minolta, New Jersey, USA).

The water potential was measured psychrometrically using a dew-point psychrometer (WP4, Decagon Devices, Pullman, Washington, USA). The osmotic potential (Ψπ) was measured on frozen leaf samples and the pressure potential (Ψp) was estimated as the difference between Ψw and Ψπ, assuming a matric potential equal to zero.

Mineral analysis

Mineral composition was determined using dried, finely ground (mesh 0.5 mm) samples of leaves and roots. Anions and cations were extracted in Milli-Q water (Merck Millipore, Darmstadt, Germany) in a thermostatic bath at 80 °C for 10 min (ShakeTemp SW22, Julabo, Seelbach, Germany). After centrifuging at 6000 g for 10 min, the supernatant was filtered (0.2 μm) and analyzed by ion chromatography with suppressed conductivity detection using a Dionex ICS-3000 system (Sunnyvale, CA, USA). Cations analysis was carried out with isocratic method (20 mM; flow rate 1 ml/min) using an IonPac CS12A column with a CG12A guard column and methanesulfonic acid as eluent. Anions analysis was performed with NaOH gradient (1 mM – 50 mM; flow rate 1.5 ml/min) using an IonPac AS11HC column with an AG11HC guard column.

RNA extraction and qRT-PCR

Leaves of 12-week-old plants (90 DAS, 19 DAST) from all treatments were harvested and immediately frozen in liquid nitrogen and stored at − 80 °C. Leaves from the same treatment were mixed and three replications per bulk were analyzed. 100 mg of fresh leaf tissue per sample was homogenized with liquid nitrogen and extracted with 1 ml of TRIzol (Life Technologies, Carlsbad, CA, USA). First-strand synthesis was performed with a QuantiTect Reverse Transcription Kit (QIAGEN, Valencia, CA, USA) using 1 μg of RNA. Real-time qPCR reactions, using 10 ng of cDNA per reaction, two experiments, four replicates per experiment, were carried out on an ABI Instruments 7900HT qPCR detection system (Applied Biosystems, Foster City, CA, USA) using Platinum SYBR Green qPCR SuperMix-UDG with ROX (Life Technologies, Carlsbad, CA, USA). All qRT-PCR primers were determined to be within 3% efficiency of each other. Relative expression levels were calculated using Actin2 as an internal standard and the ΔΔCt method for relative quantification. The primers used in this study are listed in Additional file 1: Tables S1.

Statistical analysis

Data were analyzed with a two way ANOVA (Nutrient solution x Inoculum). Least Significant Difference (LSD) multiple range comparison tests were used to determine differences between means (P ≤ 0.05). For gene expression analysis, the ΔCt values of each gene of interest and the actin reference gene were analyzed using Student’s T-test. Single asterisks denote significant differences according to Student (P < 0.1) between untreated controls and inoculated, double asterisks denote (P < 0.05) and triple asterisks denote (P < 0.01) between untreated controls and inoculated plants.

Bacterial growth

Plant growth responses and fruit yield

Inoculation (I) with A. chroococcum 76A and nutrient solutions (N) had both a remarkable effect on plant growth and fruit yield, with a significant interaction (IxN) for some of the biometric parameters considered (Table 2). Shoot dry weight (SDW), fruit fresh weight (FFW) and fruit number per plant (FN) all increased, compared to untreated control plants, upon microbial inoculation (+ 45% SDW, + 39% FFW, + 49% FN). These parameters were also similarly affected by different nutrient solutions (Table 2). The addition of NaCl caused a decline in terms of plant growth and yield regardless of the nutritional level. At 50 mM and 100 mM NaCl, the average SDW, FFW and FN reduction vs. non-salinized solutions was − 15% and − 44% (for SDW), − 37% and 64% (for FFW), − 31% and − 53% (for FN), respectively (means of Sub + Opt nutritional regimens). It should be noted that in the absence of NaCl, the fruit number (9.3 vs. 8.3) at sub-optimal vs. optimal nutritional levels was similar, whereas SDW and FFW were reduced by 21% and 25% respectively in optimal conditions. These results indicate that higher nutrient concentration and additional nitrogen provided as NH NO with the optimal nutritional regimen did not further improve the number of fruit per plant and it was indeed inhibitory for shoot and fruit growth. A significant interaction between microbial inoculation (I) and nutrient solutions (N) was found for shoot fresh weight (SFW), root dry weight (RDW) and Fruit Dry Weight (FDW) (Fig. 1). For these parameters, although it was observed an overall proportional decline with salinity under both sub-optimal and optimal nutritional regimen, there were substantial differences between inoculated plants and non-inoculated controls. Specifically, at moderate (50 mM NaCl) and severe (100 mM NaCl) salinity, A. chroococcum 76A treated plants performed always better than uninoculated plants under sub-optimal nutritional regimen in terms of SFW (+ 44% as average of the two salinity treatments), RDW (+ 45%) and FDW (+ 56%). Higher nutrients concentration and addition of NH NO (optimal regimen) did not further improve the overall plant performance and it indeed flattened down the differences between A. chroococcum 76A and uninoculated plants. The optimal nutritional level and most likely the extra nitrogen caused a partial inhibition of Azotobacter growth (Table 1) and, consequently, its effects on some growth parameters (Fig. 1).

	Shoots FW	Shoots DW	Roots DW	Fruits FW	Fruits number	Fruits DW	Leaves FW
	g	g	g	g		g	g
Inoculum (I)
Control	15.0 b	1.1 b	0.1 b	9.6 b	5.1 b	0.9 b	0.7 b
76 A	19.7 a	1.6 a	0.2 a	13.3 a	7.6 a	1.4 a	1.0 a
Nutrent Solution (N)
Sub - 0 mM NaCl	27.7 a	1.9 a	0.2 a	19.8 a	9.3 a	1.8 a	1.1 a
Sub - 50 mM NaCl	18.2 b	1.7 ab	0.2 a	11.3 bc	6.4 b	1.3 b	0.9 ab
Sub - 100 mM NaCl	10.7 c	1.0 c	0.1 b	6.3 d	4.2 c	0.7 c	0.6 b
Opt - 0 mM NaCl	21.4 ab	1.5 b	0.2 a	14.8 b	8.3 a	1.6 ab	1.0 a
Opt - 50 mM NaCl	16.6 b	1.2 bc	0.1 b	10.5 c	5.8 b	1.0 b	0.7 b
Opt - 100 mM NaCl	9.4 c	0.9 c	0.1 c	6.0 d	4.0 c	0.6 c	0.5 b
Significance
I	***	***	***	***	***	***	*
N	***	***	***	***	***	***	***
IxN	*	ns	*	ns	ns	*	ns

ns, *; **, ***Non significant or significant at P ≤ 0.05, 0.01, and 0.001, respectively

Gas exchange analysis, water relations and SPAD

With respect to leaf water potential and gas exchanges (Table 4), there were no differences between control and A. chroococcum 76A treated plants, whereas salinity reduced photosynthesis under sub-optimal nutritional regimen (− 50% at 100 mM NaCl). The photosynthetic levels under optimal nutritional regimen were generally lower than those at suboptimal nutrition and they did not differ at varying salinity. RWC and SPAD values were 5% and 15% higher in A. chroococcum 76A treated than uninoculated control plants, respectively. These parameters also decreased with salinity under both suboptimal and optimal nutritional levels. In the absence of NaCl, the RWC under sub-optimal nutritional level was 7% higher than under optimal nutritional regimen (compare Sub-Opt 0 mM NaCl vs. Opt 0 mM NaCl in Table 4).

Leaf ion contents

The leaf ion profile was significantly altered in response to both microbial inoculation (I) and nutritional regimen (N) (Table 3). Leaf ammonia concentration was higher under optimal fertilization (3.1 mg/kg dry matter) compared to the sub-optimal regimen (0.8 mg/kg dry matter). Potassium decreased with respect to both the microbial treatment (− 15%) and nutritional regimens with a 36% and 45% lower K tissue concentrations at 50 and 100 mM NaCl, respectively, compared to non-salinized solutions as average of sub-optimal and optimal nutritional regimen, (Table 3). For all other ions, a significant interaction was found between microbial treatment and nutritional regimen (Fig. 2). Na⁺ and Cl⁻ levels in leaves increased upon addition of NaCl to the nutrient solution under both sub-optimal and optimal nutritional regimens. However, the concentrations of these ions were remarkably higher in A. chroococcum 76A treated plants vs. uninoculated controls in the absence of NaCl stress with + 91% for Na⁺, under optimal nutrient regimen (Fig. 2a) and + 56% and + 76% for Cl⁻ under sub-optimal and optimal nutritional regimens, respectively (Fig. 2b). Ca²⁺ was always higher in A. chroococcum 76A plants under sub-optimal nutritional levels (Fig. 2c). The differences between inoculated plants and controls were generally attenuated under optimal nutritional regimen, although in the absence of NaCl the leaf Ca²⁺ level of inoculated plants was still twice that of uninoculated controls. Mg²⁺ levels were rather stable in non-inoculated plants under all nutritional regimens, with a moderate decline under optimal nutritional conditions, in the absence of NaCl. 76A treated plants again performed slightly better under sub-optimal nutritional regimen with highest values at 0 and 50 mM NaCl compared to all other treatments (Fig. 2d). NO₃⁻ levels generally declined with salinization and were lower in 76A plants vs. uninoculated controls at 0 and 50 NaCl (− 63% and − 59%, respectively) (Fig. 2e). This pattern was somehow reverted under optimal fertilization, at least in the absence of NaCl (+ 32% in 76A plants). For PO₃ ions, we found similar levels in inoculated vs. non-inoculated plants at all nutrient solutions tested, with the exception of significant 47% and 46% lower levels at sub-optimal nutritional level + 100 mM NaCl and at optimal nutritional level in the absence of salt (Fig. 2f).

	Na	NH4	K	Mg	Ca	Cl	NO3	PO4
	mg/kg d.m.	mg/kg d.m.	mg/kg d.m.	mg/kg d.m.	mg/kg d.m.	mg/kg d.m.	mg/kg d.m.	mg/kg d.m.
Inoculum (I)
Control	19.6	2.6	29.1 a	3.7 b	10.7 b	42.5 b	14.3 a	18.6 a
76 A	22.2	2.3	24.7 b	4.1 a	14.6 a	57.1 a	11.2 b	13.7 b
Nutrient Solution (N)
Sub - 0 mM NaCl	1.8 c	0.8 c	42.8 a	4.7 a	13.7 a	25.9 d	19.0 a	16.3 ab
Sub - 50 mM NaCl	22.8 b	1.2 c	23.9 c	4.4 a	14.8 a	55.6 bc	10.3 bc	13.8 c
Sub - 100 mM NaCl	33.0 a	1.4 c	21.2 cd	4.2 a	10.3 b	61.6 b	7.9 c	15.9 b
Opt - 0 mM NaCl	7.8 c	3.1 b	30.9 b	3.2 b	13.4 a	31.8 d	19.6 a	18.3 a
Opt - 50 mM NaCl	24.5 b	3.0 b	23.6 c	3.3 b	14.9 a	50.5 c	11.8 b	14.6 bc
Opt - 100 mM NaCl	35.5 a	5.2 a	19.1 d	3.5 b	9.0 b	73.5 a	7.8 c	18.2 a
Significance
I	ns	ns	**	***	***	***	***	***
N	***	***	***	***	**	***	***	**
IxN	*	ns	ns	***	**	***	***	***

ns, *; **, ***Non significant or significant at P ≤ 0.05, 0.01, and 0.001, respectively

Gene expression

Four genes were chosen to correlate the rates of nitrogen content to molecular mechanisms that play a role in nitrogen uptake: the NH₄⁺ transporter (AMT2), the NO₃ transporter (NRT2.1), nitrite reductase (Nii), and nitrate reductase (NR) The plasma membrane sodium antiporter, SALT OVERLY SENSITIVE 1 gene (SOS1), two members of the Na⁺-K⁺/H⁺ antiporter family (NHX1 and NHX2), and the Na⁺-selective class I-HKT transporter (HKT1;1) were used to assess ion transport. Optimal N levels led to a general decreased expression of AMT2, NRT2.1 and NR in leaves (Fig. 4). For these genes, inoculation with A. chroococcum 76A did not induce notable changes to expression under sub-optimal N, with the exception of NRT2.1 whose relative expression was less than half of the untreated control at moderate and severe salinity (Fig. 4). The LATE EMBRYOGENESIS ABUNDANT (LEA) gene was chosen for its high inducibility under both salt and drought stress as previously reported [19]. LEA expression in inoculated plants under stress treatment was higher than in uninoculated, implying a level of crosstalk between signaling and nutrition. LEA serves as an indicator of stress response and the increased expression levels seen at optimal N conditions indicate that additional nitrogen exacerbated the stress phenotype. The expression of the plasma membrane sodium antiporter, SOS1 was also evaluated. SOS1 is the primary transporter responsible for exclusion of sodium ions from the cytoplasm and required for salt tolerance [17]. Expression in sub-optimal conditions was slightly lower in inoculated plants (Fig. 4). Conversely, in the optimal nutrient regimen, expression of SOS1 was higher in inoculated plants. A second sodium transporter, the Na⁺-selective class I-HKT transporter (HKT1;1), was selected to evaluate the ion homeostasis (Ansins, et al., 2012). HKT1;1 expression was not induced upon inoculation with A. chroococcum 76A or by salt stress in sub-optimal nutrient condition in leaves of MicroTom plants. Expression of HKT1;1 was slightly increased in optimal nutrients and upon Azotobacter inoculation (Additional file 1: Figure S1). Finally, expression of NHX1 and NHX2 showed increases in all salt stress conditions, with no remarkable differences between inoculated and uninoculated plants (Additional file 1: Figure S1).

A. chroococcum 76A is an effective salt stress protectant

Ionic profile reveals a stress pre-adaptation state of A. chroococcum 76A treated plants

Tomato plants treated with A. chroococcum 76A had higher levels of Na⁺ and Cl⁻ ions compared to their relative untreated controls (Table 3). The bacterial inoculum seemed to have a tissue concentration effect on the low levels of Na⁺ and Cl⁻ ions dissolved in non-salinized irrigation water (Table 3; Fig. 2a, b). This increase mirrored the increase of Ca²⁺ (Fig. 2c) that typically is associated with protection from osmotic and ionic stress [56]. While Na⁺ ions may serve as cheap osmoticum in stress adaptation [30] and may have stimulated root growth (Fig. 1b), Ca²⁺ is a fundamental component of the regulatory machinery for cellular ion homeostasis during salinity stress but also hormonal regulation and growth control [4]. The increase of cytosolic Ca²⁺ (which was 30% higher in 76A plants vs. control plants) under salt stress triggers the SOS pathway, through the activation of Ca²⁺-binding protein SOS3 and the kinase SOS2, which, upon SOS3 activation, form a complex that stimulates activity of the Na⁺/H⁺ exchanger SOS1. Activation of SOS1 in turn will remove Na⁺ from the cytoplasm relieving its toxic effects [8]. High apoplastic Na⁺ would serve as osmoticum without impairing cellular functions [17]. Consistent with this mechanism of Na⁺ detoxification, expression of SOS1 increased in all salt stress treatments. However, under sub-optimal nutrient regimen, plants inoculated with 76A had lower levels of SOS1 expression than uninoculated controls (Fig. 4) and accumulated more sodium in salt stress conditions. Lower levels of SOS1 expression have been shown to increase root and shoot accumulation of sodium in tomato [37]. While salt stress reduced the K⁺/Na⁺ ratio under both nutritional regimens (Fig. 3), the only notable differences in this ratio between controls and inoculated plants were observed in non-stress conditions. In the absence of stress, plants inoculated with A. chroococcum 76A had much lower K⁺/Na⁺ ratios than controls. The strain A. chroococcum 76A appears to affect the K⁺/Na⁺ ratio by increasing the sodium content. However, this effect was not observed when salt stress was imposed. The mechanism responsible for increased Na⁺ and Cl⁻ is unclear, but it does not appear to inhibit growth in control conditions and interestingly does not come into play under moderate or severe salt stress. A side effect of high leaf Cl⁻ was a reduced PO₄⁻concentration (Fig. 2f), possibly due to competition between these two anions [9, 41]. Consistent with the PO₄⁻ pattern, NO₃⁻ also decreased in sub-optimal nutritional conditions (Figs. 2e). Low NO₃⁻ levels have been associated to high Na⁺ and Cl⁻ availability/uptake in/from the root zone [10, 49]. However nitrate increased in response to optimal N with inoculated plants containing more nitrate than controls in unstressed conditions. A. chroococcum 76A may have likely facilitated assimilation of nitrogen although high salinity was observed to inhibit increased nitrate content. Synergistic effects when nitrogen supplementation is used in combination with Azotobacter strains have been reported [47].

Molecular mechanisms underlying ion partitioning and A. chroococcum 76A induced stress protection

AMT2 expression was barely detectable under optimal nutritional regimen and moderate salt stress (Fig. 4). Nitrogen transporters such as, NRT2.1 and AMT2 as well as Nii and NR are known to be transcriptionally down regulated by high levels of N and metabolites such as amino acids [32]. In sub-optimal nutritional conditions, inoculation did not significantly increase plant NH₄⁺ content. Under carbon limiting conditions, uptake of both amino acids and ammonium may be down regulated. The primary uptake of nitrogen into the root cells may be in the form of amino acids instead of NO₃⁻ and NH₄⁺ [6]. The presence of A. chroococcum 76A appears to synergistically enhance the expression of LEA under salt stress, with higher expression levels observed in inoculated plants. Interestingly, under optimal nutritional conditions, expression was elevated even in unstressed plants, with drastically higher levels of expression under salt stress conditions. Possibly the optimal nutrient solution we used, which was adapted from standard solution for soil-less production of cherry tomato plants which are larger and typically have an indeterminate growth, was not ideal for the smaller determinate MicroTom tomato in our specific experimental conditions.

The presence of A. chroococcum 76A appears to prime plant responses to stress. Instances of this PGPR increasing stress tolerance have been documented, but the molecular mechanisms remain unknown [48, 50, 58]. Stress priming for salt tolerance has been observed in wheat seed treated with Nitragin biofertilizer, a mixture of Azotobacter spp., Azospirillum spp. and Pseudomonas spp. Inoculation increased also germination under saline conditions [15]. Our results indicate that inoculation increases expression of at least one key gene (LEA) involved in salt and drought responses. Inoculation with 76A did not appear to affect the expression of the two evaluated NHX genes (Additional file 1: Figure S1) and we observed no significant differences in K⁺ accumulation in inoculated plants. Sodium accumulation (Fig. 2) corresponded with the expression patterns of the sodium extruder, SOS1, with lower levels of expression in sub-optimal nutrients in inoculated plants and higher levels of expression in optimal nutrients. The sodium transporter HKT1;1 was not significantly affected by inoculation but has been previously reported to not be highly inducible in leaves of tomato plants [52]. Inoculation with 76A appears to alter sodium transport while not affecting potassium uptake. Overall, ion compartmentation could be one mechanism that has contributed in 76A plants to better perform under salt stress (see moderately higher Na⁺ and Cl⁻ associated with better growth), whereas tissue Ca²⁺ accumulation may have played a predominant signaling role in growth control.