Bioprospecting Soil Bacteria from Arid Zones to Increase Plant Tolerance to Drought: Growth and Biochemical Status of Maize Inoculated with Plant Growth-Promoting Bacteria Isolated from Sal Island, Cape Verde

Climate change and anthropogenic activities are responsible for extensive crop yield losses, with negative impact on global agricultural production. The occurrence of extreme weather events such as drought is a big challenge for agriculture, negatively impacting crops. Thus, methodologies reducing crop dependence on water will be a great advantage. Plant roots are colonized by soil bacteria, that can establish beneficial associations with plants, increasing crop productivity and plant tolerance to abiotic stresses. The aim of this study was to promote plant growth and to increase crop tolerance to drought by inoculation with osmotolerant bacterial strains. For that, bacteria were isolated from plants growing in Sal Island (Cape Verde) and identified. The osmotolerance and plant-growth promotion (PGP) abilities of the strains were determined. A maize seed cultivar tolerant to drought was inoculated with the strains evidencing best PGP capacity and osmo-tolerance. Results evidenced the ability of some bacterial strains increasing the development and inducing osmotolerance in plants. These results evidence the potential of osmotolerant bacteria to further increase the level of tolerance of maize varieties tolerant to drought, decreasing the dependence of this crop on irrigation, and open new perspectives to growth maize in drought affected areas and to use water more efficiently.


Introduction
According to the United Nations [1] by 2050 the world's population is expected to increase from 7.87 billion (in 2021) to 9.7 billion, about 23% higher than today and 6% higher than the predicted by FAO [2]. Nowadays, the prevalence of malnutrition worldwide is about 11% and expected to be higher in developing regions in the years to come [1]. Therefore, food production must increase by 70% in order to feed the growing human population, and the production of cereals such as maize, wheat and rice will need to rise to about 3 billion tones [2][3][4]. Maize is a crucial crop to ensure global food security, since it is highly used as biofuel, food source and feed crop globally. In fact, maize is the third most cultivated crop worldwide, and contributes about 62% to the total cereal production, thus any factor impacting maize production will have a serious impact on food security [5,6].
Environmental changes negatively impacts global food production. Climate change has caused inter-seasonal yield variability and negative impact on crop growth, development and production. The major challenges to agriculture are related to climate change severity and occurrence of extreme weather events, such as floods, drought and heat haves but also to agricultural management (application of fertilizers, tillage, agrochemicals) [5,7,8]. maize variety to osmotic stress was assessed, by evaluating germination, growth and the physiological and biochemical status of plants. The comparison between plants exposed to watered and drought conditions allowed to elucidate the bacterial influence on the biochemistry of plants that contributed to stress alleviation.

Bacterial Plant Growth-Promoting (PGP) Traits
All strains have the ability to produce IAA, siderophores and alginate (Table 1). Three strains are not able to solubilize phosphate, but the others evidence a moderate solubilization ability (1.04−1.97) with strains H, J, O and U having the higher ratios (1.97, 1.88, 1.94 and 1.88, respectively). Strain S stands out by exhibiting the highest production of IAA, siderophores and alginate, although it is not able to solubilize phosphate. On the other hand, strain K is the highest producer of alginate and IAA. Table 1. Identification and plant growth-promoting abilities of bacteria isolated from Cape Verde. Determination of the percentage of polyethylene glycol that inhibits 50% of bacterial growth (IC50 PEG). Indole acetic acid (IAA) and alginate production were expressed in µg/mL/DO. Siderophore production was expressed in percent siderophore units per optical density (PSU/DO). Phosphate solubilization was expressed by the ratio between halo and colony diameter. The plant emergence rate was determined in the 4th day of growth, in comparison with not-inoculated control (↑-increase, ↓-decrease, ↔-no change). Exposure to osmotic stress significantly increased lipid peroxidation in most bacterial strains ( Figure 1A), with G and S showing the highest increases (500% and 800%, respec-tively). Only 4 strains significantly showed lower LPO levels when exposed to osmotic stress (K, L, M and N). pared to control (non-inoculated seeds) and the remaining strains had a marginal effect ( Exposure to osmotic stress significantly increased lipid peroxidation in most bacterial strains ( Figure 1A), with G and S showing the highest increases (500% and 800%, respectively). Only 4 strains significantly showed lower LPO levels when exposed to osmotic stress (K, L, M and N).  . Values are means of three replicates + standard deviation. Significant differences compared to non-inoculated watered plants (control) were marked with single asterisks (p < 0.05) or double asterisks (p < 0.01). For additional information (means, standard deviation and statistical significance) see supplementary Table S1. Osmotic stress also increased the damage in proteins ( Figure 1B) in most strains, being the damage significant in 46% of them.

Antioxidant Response
Exposure to PEG increased catalase activity in most strains, significantly in B, D, G, H, K, L, M, Q, R, S, T and U strains ( Figure 1C).
Relatively to SOD activity, exposure to osmotic stress only significantly increased SOD in strains A and D and decreased it in L, R and T. In fact, 28% of the bacterial strains had almost no change in SOD activity ( Figure 1D). Glutathione S-transferases activity was poorly influenced by the presence of PEG, with 67% of the strains not evidencing changes in the enzyme activity. K and L significantly increased GST activity (>50%) ( Figure 1E).

Osmolyte Production
Out of the 18 strains, 13 were able to increase proline production between 20% and 80% ( Figure 1F), significantly in D, L and U.

Protein Content
Half of the bacterial strains were able to increase protein content between 30% and 300% relatively to control ( Figure 1G), significantly higher in B, F, G, L, Q and S. On the other hand, M and T showed significant reduction in protein levels, 44% and 63% respectively.

Multivariate Analysis
The PCO was performed with the difference between drought and control, showing the effect of drought in bacteria. In the multivariate analysis ( Figure 1H) it is observed that PCO1 explains most of the drought effects on bacterial biochemistry.
Alterations induced by exposure to PEG are highly correlated to the activity of antioxidant enzymes (GSTs and CAT), osmolyte production (proline) and protein carbonylation, with Pearson correlations higher than 0.85. In the negative axes of PCO1 and PCO2 it is possible to observe that strain K is highly influenced by PC, GST and SOD activity. On the other hand, A, F and N evidenced similar responses to drought, with increases of protein carbonylation (protein damage), but were able to trigger the antioxidant response (SOD activity). Being near the origin, M strain displayed low influence of drought on its biochemistry. In the positive axis of PCO2 it is noticeable the high correlation of strains S, L, Q and U to catalase activity (S, Q and U), LPO (S), proline (L and U) and protein content (Q). The positive axis of PCO1 reveals the majority of the strains that increased lipid peroxidation when exposed to drought stress (B, D, E, G, H, J, O and T) and the ones that decreased PC (B and O), GST activity (D, E, G, H, O and R) and proline content (H and R).

Biometric Parameters
It is observed that the inoculation with bacterial strains did not have a significant effect on shoot growth and biomass both in watered and drought plants (Figure 2A,B).  Values are means of three replicates + standard deviation. Significant differences compared to non-inoculated watered plants (control) were marked with single asterisks (p < 0.05) or double asterisks (p < 0.01). For additional information (means, standard deviation and statistical significance) see supplementary Table S2.
The influence of bacterial inoculation is more noticeable in plant roots. Under drought, some bacterial strains significantly mitigated the negative impact of drought on root weight ( Figure 2C). When inoculated with R and T under stress it is noticed a significant promotion of 24% and 41% (respectively) in comparison with drought control. Root length significantly increased when plants were inoculated with A and S, under drought stress ( Figure 2D). Values are means of three replicates + standard deviation. Significant differences compared to non-inoculated watered plants (control) were marked with single asterisks (p < 0.05) or double asterisks (p < 0.01). For additional information (means, standard deviation and statistical significance) see supplementary Table S2.
The influence of bacterial inoculation is more noticeable in plant roots. Under drought, some bacterial strains significantly mitigated the negative impact of drought on root weight ( Figure 2C). When inoculated with R and T under stress it is noticed a significant promotion of 24% and 41% (respectively) in comparison with drought control. Root length significantly increased when plants were inoculated with A and S, under drought stress ( Figure 2D).

Photosynthetic Pigments
In watered plants, there is a trend to decrease photosynthetic pigments in inoculated plants, significantly reduce when inoculated with strain T ( Figure 2E−G). Under drought, inoculation with D significantly promoted chlorophyll a synthesis by 58% ( Figure 2E). The effect of drought on chlorophyll b and carotenoids content was mostly negative or had no impact, but inoculation with some strains (A, D, S and T) alleviated the significant negative impact of drought observed in non-inoculated plants ( Figure 2F,G).

Cellular Damage
Most LPO changes due to drought and inoculation were not significant. Inoculation with T increased significantly LPO levels ( Figure 3A).  Values are means of three replicates + standard deviation. Significant differences compared to non-inoculated watered plants (control) were marked with single asterisks (p < 0.05) or double asterisks (p < 0.01). For additional information (means, standard deviation and statistical significance) see supplementary Table S3.
Overall, drought and inoculation decreased PC levels, significantly for Q, R and S in control conditions and for D, F, G and Q under osmotic stress ( Figure 3B).

Antioxidant Response
Under control conditions, inoculation led to reduced CAT activity ( Figure 3C), significantly for Q and R (by 37% and 47% respectively), exception being made for strain A that significantly increased it. Under drought stress, inoculation with A and D significantly reduced CAT activity.
Under control conditions, inoculation with G and R strains reduced significantly SOD activity ( Figure 3D). Exposure to drought decreases significantly SOD activity in Values are means of three replicates + standard deviation. Significant differences compared to non-inoculated watered plants (control) were marked with single asterisks (p < 0.05) or double asterisks (p < 0.01). For additional information (means, standard deviation and statistical significance) see supplementary Table S3.
Overall, drought and inoculation decreased PC levels, significantly for Q, R and S in control conditions and for D, F, G and Q under osmotic stress ( Figure 3B).

Antioxidant Response
Under control conditions, inoculation led to reduced CAT activity ( Figure 3C), significantly for Q and R (by 37% and 47% respectively), exception being made for strain A that significantly increased it. Under drought stress, inoculation with A and D significantly reduced CAT activity.
Under control conditions, inoculation with G and R strains reduced significantly SOD activity ( Figure 3D). Exposure to drought decreases significantly SOD activity in non-inoculated plants and plants inoculated with D and S.

Metabolism
Under stress, strains D and G significantly reduced ETS, but overall ETS activity was not significantly affected by drought or inoculation.
Relatively to protein content ( Figure 3H), both drought and inoculation led to higher protein, in most plants. Significant increases in protein production are observed in plants inoculated with Q, R, S and T, with increases over 30%.

Osmolyte Production
Overall, the accumulation of soluble sugars (SS) was not significantly affected by drought or inoculation ( Figure 3F). Exception being made for Q, R and T that significantly decreased SS content under control conditions (Q) and increased under drought stress (R and T).

Multivariate Analysis
The PCO shows the effect of bacterial inoculation on the response of plants to drought stress. In the PCO ( Figure 3H) it is observed that all parameters have high correlations, ranging from 0.88 to 0.98, with LPO and CAT having the highest values (0.97 and 0.98, respectively). The drought effects on non-inoculated plants (control), and plants inoculated with A and F positioned them on the negative side of PCO1 and PCO2 axes, showing low influence of bacterial inoculation on the biochemical changes plant underwent when exposed to drought. On the positive side of PCO1 and PCO2 axes it is observed the strain that influenced positively most parameters (R). Strains T and Q are close positively influencing mostly SOD and LPO. Strains F and G are positioned in the positive side of PCO2 and on the negative side of PCO1 axes, showing the ability to reduce protein carbonylation and to increase antioxidant response (SOD activity) under stress. S strain is positioned in the positive PCO1 and negative PCO2 axes evidencing the influence of this strain on the increase of PC on drought compared to watered plants. Strain R stands out by being positioned in the most positive value of PCO1 increasing CAT activity, soluble sugars accumulation and protein and proline content in plants as a response to water shortage.

Discussion
The present study aimed to investigate the effects of osmotic stress on bacteria and maize, the biochemical mechanisms associated, and the contribution of PGPB tolerant to osmotic stress, isolated from an arid environment, the Sal Island in Cape Verde, on plant tolerance to drought. The use of a drought tolerant variety of Z. mays allowed to understand the biochemical mechanisms underlying the plant tolerance, and to evaluate if Cape Verde strains can further increase the plant tolerance to drought. Several studies already described the effects of osmotic stress on bacteria [36][37][38], and the contribution of plant growth-promoting bacteria on plant tolerance to drought [39][40][41][42][43][44][45]. Some studies reported the use of drought tolerant bacterial strains to enhance plant growth [33][34][35], showing that osmotolerant strains can be more efficient in drought stress mitigation in inoculated plants.
Following the classification of Sá et al. (2019), six strains were moderately tolerant to osmotic stress (IC 50 between 7.5% and 10% PEG, that is between −0.13 and −0.19 MPa, according to Guo et al. [46]). The plant growth-promoting traits evaluated showed that all strains are capable of siderophore and IAA production. Siderophores are proven to be one of the most crucial survival strategies for microorganisms, for they form complexes with iron (Fe) and other metal ions, improving their solubility and uptake, which can enhance plant growth and development [33,47,48]. On the other hand, was already reported that higher production of siderophores by bacterial strains is associated with increased drought resistance in inoculated plants [49].
The phytohormone IAA, the most common plant hormone of the auxin class, plays a crucial role in cell division, fruit development and cell elongation [50]. When inoculated with bacteria able to produce IAA, plants benefit from increased root length and density (primary and lateral), having a well-developed root system, imperative for efficient water and nutrient uptake and to anchor plants in the soil [50][51][52]. Besides increasing plant length and biomass, IAA producing bacteria are involved in the induction of tolerance to drought stress, and are reported to boost chlorophyll contents and nutrient uptake [52,53].
Out of the 18 strains, 15 are able to moderately solubilize phosphate. Phosphorus represents a crucial role in vital plant functions, such as growth, sustenance and development. Phosphate solubilization is a plant growth-promoting trait described as the ability to convert unavailable forms of phosphorus into forms easily absorbed by plants, promoting plant growth and yield [54]. Additionally, this PGPB trait was associated with increased drought tolerance in plants [41,[55][56][57].
Osmotic stress can cause oxidative damage in cells, resulting from the production of reactive oxygen species (ROS) [58]. ROS can react with metabolic and physiologic important biomolecules causing irreversible damage in lipids, proteins and DNA [59], or even cell necrosis and death [27]. Organisms deal with oxidative stress through different mechanisms that regulate ROS concentrations, for example, by producing ROS scavenging enzymes, such as superoxide dismutase and catalase [27].
When exposed to osmotic stress, a response pattern was observed in the different strains. Overall, osmotic stress led to high protein carbonylation and to high lipid peroxidation, evidencing that both membranes and proteins were negatively affected by osmotic stress, as already reported in other studies (e.g., [36]). As a mechanism to cope with the oxidative stress generated, most strains showed higher SOD and CAT activity. A significant increase in proline content was also observed and helped in the osmoregulation of plant cells and protecting proteins from oxidative damage. Indeed, several studies showed that proline is essential to cellular osmoregulation, and its synthesis under drought conditions is associated with stress tolerance both in bacteria and plants [36,39,[59][60][61][62]. The increase in GST activity induced by osmotic stress was also observed in the present study, and evidences that the biotransformation activity was triggered to minimize the toxicity cause by endoxenobiotics produced during stress, further protecting cells from damage [63].
Drought in one of the major causes of agricultural loss. Presently, almost 40% of arable land is in arid and semi-arid regions, being drought a real concern. PGPB are being recognized to play a crucial role in mitigation of drought effects on plants, and the use of bacteria isolated from semi-arid soils is being regard as a promising strategy for increasing plant tolerance to stress, and to reduce the amount of water necessary for irrigation without compromising yields. The results of the present study evidenced that some strains (R and T) induced higher biomass in roots, privileging branching, while others (A and S) induced higher length in drought stressed plants, evidencing the potential increase in water uptake efficiency of these plants, as reported in other studies [64,65]. Given the fact that strains R and T are moderately tolerant to osmotic stress, it corroborates the fact that osmotolerant bacteria are able to increase plant tolerance to drought, by facilitating the absorption of water in conditions of water shortage scarcity, and enabling a significant increasing of plant biomass compared to non-inoculated plants.
As observed in the greenhouse experiment, drought negatively affected the photosynthetic pigments, which may lead to lower CO 2 fixation [66] and decreased plant growth and productivity. Different studies noticed the positive influence of PGPB on plant photosynthetic pigments content [67][68][69]. In the present study, inoculation with some strains increased the amount of chlorophylls a and b in drought stressed plants, indicating a positive contribution to plant tolerance.
On the biochemical level, inoculation with PGPB induced different responses in maize plants. As mentioned by other studies, protein carbonylation is the most common protein modification caused by reactive oxygen species, being this biochemical parameter used as a marker of oxidative damage in proteins [70,71]. Overall, inoculation reduced protein carbonylation both under drought and control conditions. Half of the strains (A, D, F and G) were able to mitigate drought effects on membranes (LPO decrease). On the contrary Tiepo et al. [72], observed that all the strains used to inoculate neotropical trees (Cecropia pachystachya and Carinianna estrellensis) increased LPO compared to non-inoculated plants. The decrease in cell damage observed in the present study may be supported by the activity of antioxidant and biotransformation enzymes, which increased the capacity of cells to reduce ROS and to overcame the oxidative stress induced by drought in cells.
Accumulation of osmolytes, as proline and soluble sugars help cells to osmoregulate and reduce the effects of low osmotic potential on cell metabolism, proving to be crucial to both bacteria and plant [58]. In the present study it was observed that inoculation with PGPB markedly increased the proline content in plants, both in drought and control conditions. Similar results with plants exposed to drought stress were observed in other studies, with plant growth-promoting bacteria showing capacity to ameliorate the effects caused by drought, by increasing proline content [41,42,73]. However, in this study the proline content observed in the not-inoculated plants was substantially higher than in those studies, with proline content over 60 mg/g FW in both irrigated and drought conditions. For example, Moreno-Galván et al. [73] recorded proline contents in non-inoculated maize plants around 50 µg/g FW in irrigated plants and approximately 200 µg/g FW in plants exposed to drought. This evidences a constitutive characteristic (production of proline in high amounts) of the maize variety used in our study and that may justify, at least in part, the described drought tolerance.
The amount of soluble sugars on maize plants under drought stress was previously reported, and inoculation with PGPB increased it [39,44,[74][75][76][77]. Comparing non-inoculated plants, the maize variety used in this study showed higher ability to accumulate soluble sugars than other maize varieties [39,75], both in watered (around 2 fold) and drought conditions (by around 1.5 fold), evidencing a constitutive ability to accumulate osmolytes. The present study also shows different influences of strains on inoculated plants both in control and drought conditions. While two strains (R and T) were able to significantly increase the accumulation of soluble sugars under drought, three reduced it (A, F and G). Despite the observed reduction in soluble sugars in plants inoculated with strains A, F and G soluble sugars contents were still higher than in maize plants from other studies [39,44,75,76]. Evidencing once more that the ability of plants to accumulate osmolytes, in this case soluble sugars, is important for the tolerance of maize plants to drought.
The increase of protein content in plants under drought stress was already described [39,75,78,79]. It was reported that soluble protein prevents cellular components and molecules decomposition and denaturation, especially under stress [75]. In fact, some studies reported that when plants are inoculated, the content in protein is higher than in non-inoculated plants [39,75,78], corroborating the present results, where 50% of the strains significantly increased protein content. Most of the soluble protein in a cell are enzymes [80]. Thus, an increase in soluble protein means that the metabolism is increased, by induction of metabolic pathways involved in stress coping [81].
Plant mitochondria transform energy into forms, such as the phosphate nucleotide ATP, easily used by cells in metabolic processes crucial for cell adaptation to abiotic stresses. Under drought stress, mitochondria display the ability to defend themselves from excess of reactive oxygen species, through ROS detoxification, repairing ROS mediated damage and by keeping the electron transport chain oxidized [82]. ETS activity in mitochondria measurement provides information about respiratory potential and allows the estimation of plants general metabolic activity [83]. In the present study ETS activity was reduced under water stress by most strains, showing the influence of bacterial inoculation in plant cells energy use. This effect was also observed in other plant species, such as basil [84], Masson's pine [85] and potato [83].

Bacterial Strains
The bacterial strains used in this study were isolated from two plant species (Acacia albida and Amaranthus viridis) sampled from Sal Island, Cape Verde (16 • 35 33 N-22 • 55 29 W). After identification, 18 strains were obtained. The accession number associated to each strain is included in Table 1.

Tolerance to Polyethylene Glycol (PEG)
PEG 6000 was purchased from Tokyo Chemical Industry (TCI) and used as received. According to the manufacturer, this polymer has an average molecular weight around 6000.
The tolerance to drought was determined following the method described by Sá et al. (2019). The strains were inoculated in Yeast mannitol broth (YMB) medium supplemented with peptone and different concentrations of PEG (0, 7.5, 10, 12.5, 15, 17.5, 20 and 25%). The tubes were incubated in an orbital shaker (150 rpm) at 26 • C for 15 h, and the optical density was measured at 620 nm. After determining the PEG concentration that inhibits growth by 50% (IC 50 ), each bacterial strain was grown at two osmotic conditions (0% PEG-control and % PEG inhibiting each strain 50%) for 15 h at 26 • C. Optical density was measured, cells were pelleted by centrifugation at 10,000 rpm for 10 min at 4 • C, and stored at −20 • C for further biochemical analysis.

Siderophore Production
For siderophore production, the methodology described by Arora and Verma [86] was followed with some modifications. The strains were inoculated (100 µL) in YMB medium supplemented with 1 g/L of peptone (control) and incubated for 15 h at 150 rpm in an orbital shaker. After growth, the optical density was measured, and 1.5 mL of growth medium was collected to a microtube. The microtubes were centrifuged for 5 min (12,000 rpm) at 4 • C and the supernatant used for siderophore quantification. Into a 96-well microplate, 130 µL of sample (supernatant), and 20 µL of chrome azurol S reagent (CAS) were pipetted. A blank was prepared with 130 µL of non-inoculated medium and 20 µL of CAS. The optical density was measured at 630 and 750 nm after 20 min incubation. The siderophore production (percent siderophore unit−PSU) was quantified by the formula described by Arora and Verma [86]: , where Ar stands for the absorbance of the blank and As represents the absorbance of the samples. The results were expressed in PSU/Optical density.

Alginate Production
Alginate produced by bacteria was determined following the method described by Johnson et al. [87], with some modification [36]. Bacteria were inoculated in modified YMB medium supplemented with 1 g/L of peptone for 15 h at 150 rpm in an orbital shaker. After the optical density was measured, 2 mL of each tube was collected to a microtube, centrifuged for 10 min (at 12,000 rpm) at 4 • C, and both pellet and supernatant collected in different microtubes for determination of alginate in two fractions: present in the medium and attached to the cell wall.
In order to determine the alginate present in the medium, after centrifugation 165 µL of the supernatant was transferred to a different microtube, to which was added 35 µL of 517 mM ethylenediamine tetraacetic acid (EDTA). The sample was incubated at 37 • C for 1 h.
The pellet was resuspended in 1 mL of dH 2 O and posteriorly centrifuged, and the supernatant discarded. To the pellet 200 µL of 100 mM EDTA were added and the microtubes were incubated at 37 • C for 1 h in an ultrasonic bath. After incubation of both samples, the microtubes were centrifuged and 100 µL of the supernatant were pipetted to the 96-well plate, followed by 100 µL of 120 µM dimethylmethylene blue (DMMB).
A standard curve was performed in a 96-well microplate to determine the amount of alginate in the samples, alginate diluted in distilled water in concentrations ranging from 0 to 500 µg/mL was used as standard. The absorbance was measured at 525 and 595 nm. The results were expressed in µg/mL/OD.

Indole Acetic acid Production
The methodology to evaluate the indole acetil acid (IAA) production was based on the protocol of Gordon and Weber [88] with some modifications [89]. Initially, 100 µL of inoculum was added to tubes with YMB medium supplemented with 100 µL of tryptophan (5000 µg/mL). The tubes were incubated for 15 h at 150 rpm in an orbital shaker. After growth, the optical density was measured, and 2 mL was collected to a microtube and centrifuged at 10,000 g for 10 min at 4 • C. In a 96-well microplate 100 µL of the supernatant and 200 µL of Salkowsky reagent were added, and after 25 min incubation at room temperature, the absorbance was measured at 530 nm. Two blank wells were performed, one with 100 µL of YMB medium with tryptophane and 200 µL of Salkowsky reagent and the other one with 100 µL of YMB medium without tryptophane and 200 µL of Salkowsky reagent.
A standard curve was performed (0 to 40 µg IAA/mL) and the amount of IAA in the samples was expressed as µg IAA/mL/OD.

Greenhouse Experiment
Zea mays seeds used in this work (certified trial seeds, Pioneer Optimum AQUAmax hybrids, variety P9911) were described as drought tolerant, and were kindly offered by Cooperativa Agrícola de Esposende and Pioneer.
Zea mays seeds were hydrated for 2 days in running water. The germinated seeds were sown in cups with washed and autoclaved river sand. The experiment included 2 conditions (watered on drought), each with 3 replicates (each cup was a replicate) with 3 seeds per cup. Each condition included a not-inoculated control (marked as Ctl) and inoculated with one of the 18 strains.
The maize was grown in a greenhouse, at 25 • C, with a photoperiod of 16:8. The 114 cups ((18 strains + 1 control) × 2 conditions × 3 replicates = 19 × 2 × 3 = 114) of the experiment were randomly distributed in a shelf. The luminous intensity hitting the surface sand was between 7500 and 9000 lux, measured with Testo 540.
The cups were initially irrigated with 40 mL of distilled water (dH 2 O). The drought condition was only watered once, with 5 mL of dH 2 O, and the control was watered 2 times, with 15 mL of dH 2 O.The assay had the duration of 6 days. The shoots and roots of control (non-inoculated) and plants inoculated were measured, weighted and used immediately (photosynthetic pigments) or frozen (−20 • C) for future use. Seedling emergence was determined when the non-inoculated cups, under irrigation, reached 90% of germination (4th day of growth). Biochemical analysis was performed in the bacterial strains that promoted seedling emergence (A, D, F, G, Q, R, S and T).

Photosynthetic Pigments
Chlorophylls and carotenoids were quantified in a dark cold room. Shoots (0.1 g) were homogenized with a mortar and pestle using 2 mL of cold acetone (80%). The homogenate was pipetted to a microtube and centrifuged at 10,000× g, for 20 min at 4 • C. In a 96-well microplate 150 µL of sample supernatant and 150 µL of acetone (80%) were pipetted to each well. A blank (300 µL of 80% acetone) was prepared. The absorbance was measured at 470, 645, 663 and 750 nm. The amount of chlorophylls a and b and carotenoids was determined using the equations proposed by Wellburn and Lichtenthaler [91], and the results were expressed in µg/g FW.

Extraction
The bacterial strains biomass grown in the presence and absence of PEG (4.2) was extracted in sodium phosphate buffer (50 mM sodium dihydrogen phosphate monohydrate; 50 mM disodium hydrogen phosphate dihydrate; 1 mM ethylenediaminetetraacetic acid disodium salt dihydrate (EDTA); 1% (v/v) Triton X-100; 1 mM dithiothreitol (DTT), pH 7.0). The volume (between 500 µL and 1500 µL) of the buffer used varied with the amount of cells (optical density of each sample). Microtubes were vortexed, and bacterial cells lysed in an ultrasonic probe for 30 s at 0.6 Hz, always keeping the samples in an ice bath to prevent overheating, and centrifuged at 10,000× g for 10 min at 4 • C. The supernatant was used to determine the activity of superoxide dismutase (SOD), catalase (CAT), glutathione S-Transferases (GST), protein content (PROT) and protein carbonylation (PC), and the pellet used for determination of lipid peroxidation. Both fractions were used immediately or frozen at −20 • C until used.
For proline extraction, the samples were suspended in 3% sulfosalicylic acid, lysed for 30 s in an ultrasonic probe and centrifuged at 12,000× g for 10 min at 4 • C. The supernatant was collected for further analysis.
Shoots of plants (0.2 g frozen material) grown in two water regimes and inoculated or not with a bacterial strain (2.4) were extracted with 0.5 mL of potassium phosphate buffer (50 mM sodium dihydrogen phosphate monohydrate; 50 mM disodium hydrogen phosphate dihydrate; 1 mM EDTA; 1% (v/v) Triton X-100; 1 mM DTT, pH 7.0) (1:2 w/v) and centrifuged at 10,000× g for 10 min at 4 • C. The supernatant and pellet were used immediately or frozen (−20 • C) until used. The supernatant was used to determine CAT, SOD, and electron transport system (ETS) activity, PC and PROT content. The pellet was used for determination of LPO.
For proline determination 0.2 g of frozen shoot was homogenized with 0.5 mL of 3% sulfosalicylic acid (1:2 w/v) and centrifuged at 10,000× g for 10 min at 4 • C. The supernatant was collected and frozen (−20 • C) until further use.

Lipid Peroxidation
The quantification of thiobarbituric acid reactive substances (TBARS) was performed according to the procedure described by Buege and Aust [92], using the bacteria and plant pellets. To the pellet 200 µL of 0.5% 2-thiobarbituric acid (TBA) and 150 µL of 20% trichloroacetic acid (TCA) were added. The blank was also prepared (200 µL of TBA and 150 µL of TCA). The microtubes were pierced and incubated at 96 • C for 25 min, and the reaction was stopped by placing the samples in ice. To a 96-well microplate 250 µL of each microtube were transferred to each well and the absorbance read at 532 nm. The LPO was expressed in pmol of MDA equivalents per million cells in bacteria (pmol/M cells), and nmol of MDA equivalents per g of fresh weight (nmol/g FW) in plants.

Superoxide Dismutase
The methodology used was based on the procedure described by Beauchamp and Fridovich [93]. In a 96-well microplate, to 25 µL of supernatant, 250 µL of reaction buffer (50 mM Tris-HCl (pH 8.0), 0.1 mM DTPA, 0.1 mM Hypoxanthine) with tetrazolium salt (NBT) and 25 µL of xanthine oxidase (XO), were added and incubated for 20 min at room temperature with orbital rotation. Blanks were also prepared (by 300 µL reaction buffer without NBT). The absorbance was measured at 560 nm and the results were expressed in µU per million cells (µU/million cells) in bacteria and U per g of fresh weight (U/g FW) in plants. One unit of enzymatic activity (U) refers to a 50% inhibition of NBT reduction.

Catalase
This methodology was based on the one described by Johansson and Borg [94]. In a 96-well microplate was added 25 µL of sample, followed by 125 µL of reaction buffer (1 M K 2 HPO 4 and 1 M KH 2 PO 4 , pH 7.0), 37.5 µL of methanol and 25 µL of 35.28 mM hydrogen peroxide (H 2 O 2 ). After 20 min of incubation at room temperature, was added 37.5 µL of 10 M potassium hydroxide (KOH) and 37.5 µL of 34.2 mM Purpald, followed by 10 min of incubation at room temperature, and 12.5 µL of Potassium periodate (KIO 4 ). After an incubation of 5 min, the absorbance was read at 540 nm. The catalase production was determined using a standard curve with increasing concentrations of formaldehyde (0-300 µL) and the results were expressed in µU/million cells in bacteria and mU/g FW in plants.

Glutathione S-Transferases
The activity of glutathione S-transferases in bacteria was determined following an adaptation of the method described by Habig et al. [95]. Sequentially, on a 96-well microplate was added 100 µL of supernatant, 200 µL of reaction solution (0.1 M potassium phosphate buffer, pH 6.5, 10 mM GSH and 60 mM CNDB) and the absorbance was read at 340 nm during 15 min in intervals of 15 s. The results were expressed in U/million cells, where U represents the quantity of enzyme that catalyses 1 µmol of thioether formed per min.

Protein
Protein was determined following the method described by Robinson and Hogden [96]. To a 96-well microplate was added 25 µL of sample and 275 µL of Biuret reagent. After 10 min of incubation in the dark, the absorbance was read at 540 nm and the final concentration of protein was calculated using a standard curve with increasing concentrations of BSA solution (0, 5, 10, 20, and 40 mg/mL). The results were expressed in mg/million cells in bacteria and mg/g FW in plants.

Protein Carbonylation
Carbonyl groups in proteins were determined using the methods described by Mesquita et al. [97] and Udenigwe et al. [98]. Into a 96-well microplate was added 120 µL of sample or blank, followed by 120 µL of 2,4-dinitrophenylhydrazine (DNPH). After 10 min of incubation at room temperature, was added 60 µL of sodium hydroxide (NaOH) and the reaction was incubated during 10 min. The absorbance was read at 450 nm, and the results were expressed in µmol/million cells in bacteria and µmol/g of FW in plants.

ETS
The ETS activity was measured based on the method described by Kind and Packard [99]. In a 96-well microplate was added 35.7 µL of supernatant, 107 µL of BSS buffer (0.13 M Tris-HCl and 0.3% Triton X-100, pH 8.5), 35.7 µL of NAD(P)H and 71.4 µL of 8 mM p-IodoNitroTetrazolium (INT). The absorbance was read at 490 nm during 15 min in 25 s intervals. The results were expressed in nmol/min per g of plant FW.

Soluble Sugars
The determination of soluble sugars in plants was performed based on the method described by Dubois et al. [100]. In a microtube was added 10 µL of supernatant, 100 µL of 5% phenol and 600 µL of 98% sulfuric acid (H 2 SO 4 ). After 30 min incubation at room temperature, 300 µL from each microtube were transferred to a 96-well microplate and the absorbance was read at 492 nm. The final concentration was calculated using a standard curve with increasing concentrations of glucose (0, 0.1, 0.4, 1, 2, 3, 4 and 5 mg/mL). The results were expressed in mg/g plant FW.

Proline
The determination of proline content was performed following the method described by Bates et al. [101]. To 50 µL of sample were added 50 µL of acid ninhydrin and 50 µL of glacial acetic acid. The microtubes were pierced and incubated for 1 h at 100 • C, and the reaction was stopped by placing the samples in ice, and 125 µL of the samples were added to a 96-well microplate. The blanks used were constituted by 100 µL of sodium phosphate buffer, 100 µL of acid ninhydrin and 100 µL of glacial acetic acid. The absorbance was measured at 520 nm and proline standards (0-1.5 mg/mL) were used. Results were expressed in milligrams of proline per million cells (mg/million cells).

Statistical Analysis
All parameters assessed were submitted to hypothesis testing. Parameters were analysed following a one-way hierarchical design, with PEG conditions (absence and presence) as fixed factor. The matrix resemblance was obtained using Euclidean distance. The effect of PEG on samples was determined using pair-wise Permutational Multivariate Analysis of Variance (PERMANOVA), using PRIMER version 6.1.16 for Windows.
For bacteria, the null hypothesis tested was for each bacterial strain no significant differences exist between control (no PEG) and IC 50 . For plants the null hypothesis assessed was no significant differences exist between non-inoculated (control) and inoculated plants under watered and drought conditions. Significant differences were considered for p ≤ 0.05 and were identified in figures with single asterisks (for p ≤ 0.05) and double asterisks (for p ≤ 0.01).
A matrix combining the variables (LPO, SOD, CAT, PC, PROT, proline, GST, ETS and soluble carbohydrates) and the difference between osmotic stress (PEG for bacteria, drought for plants) and control non-osmotically stressed was performed, and the Euclidean distance similarity matrix was calculated after data normalisation. The similarity matrix was submitted to ordination analysis, and a Principal Coordinates Ordination (PCO) was performed. Based on the Pearson correlation vectors of biomarkers obtained (r > 0.50 for bacteria and r > 0.80 for plants) it was possible to identify the biochemical parameters that imposed more differences when organisms (bacteria or plants) were exposed to stress.

Conclusions
The study of soil bacterial diversity in Africa is limited, and the research on topics like application of bioinoculants to promote plant growth is very scarce. This study gives information on the culturable bacteria associated to plants growing in arid zones conditions, and that help plants survive in environments with limited water availability. These bacteria could be used as effective inoculants helping crops cope with water shortage and reducing the impacts of drought on yields. Some strains were able to promote the root growth of drought stressed plants, through the production of phytohormones, siderophores and moderate phosphate solubilization. The induction of osmolytes, also emerged as a factor contributing to plant tolerance to drought. Finally, the influence of bacterial inoculation on the antioxidant response also bestowed plant tolerance to drought. Evidencing that this bacterial tripartite influence can further increase the tolerance of drought tolerant varieties, and can be of major value to protect plants from water scarcity, increase water management efficiency, minimize yield losses and contribute to food security.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/plants11212912/s1, Table S1: Results obtained for biochemical parameters of bacterial strains grown in osmotic stress (% PEG inhibiting growth 50%) and control (no PEG added); Table S2: Maize plants grown for 7 days in watered and drought conditions. Morphometric parameters (fresh weight and length of plants) and variation of photosynthetic pigments in inoculated (A, D, F, G, Q, R, S, T) and non-inoculated (control-Ctl) plants; Table S3: Maize plants grown for 7 days in watered and drought conditions. Biochemical parameters evaluated in inoculated (A, D, F, G, Q, R, S, T strains) and non-inoculated (control-Ctl) plants.