Functional mutants of Azospirillum brasilense elicit beneficial physiological and metabolic responses in Zea mays contributing to increased host iron assimilation

Abstract

Iron (Fe), an essential element for plant growth, is abundant in soil but with low bioavailability. Thus, plants developed specialized mechanisms to sequester the element. Beneficial microbes have recently become a favored method to promote plant growth through increased uptake of essential micronutrients, like Fe, yet little is known of their mechanisms of action. Functional mutants of the epiphytic bacterium Azospirillum brasilense, a prolific grass-root colonizer, were used to examine mechanisms for promoting iron uptake in Zea mays. Mutants included HM053, FP10, and ipdC, which have varying capacities for biological nitrogen fixation and production of the plant hormone auxin. Using radioactive iron-59 tracing and inductively coupled plasma mass spectrometry, we documented significant differences in host uptake of Fe^2+/3+ correlating with mutant biological function. Radioactive carbon-11, administered to plants as ¹¹CO₂, provided insights into shifts in host usage of 'new' carbon resources in the presence of these beneficial microbes. Of the mutants examined, HM053 exhibited the greatest influence on host Fe uptake with increased plant allocation of ¹¹C-resources to roots where they were transformed and exuded as ¹¹C-acidic substrates to aid in Fe-chelation, and increased C-11 partitioning into citric acid, nicotianamine and histidine to aid in the in situ translocation of Fe once assimilated.

Fig. 1. Radioactive Fe-59 reveals features for A. brasilense promotion of host assimilation and whole-plant transport of ferrous iron (Fe²⁺) and ferric iron (Fe³⁺).

A Bar graph of biological assimilation of radioactive ferrous ⁵⁹Fe²⁺ and ferric ⁵⁹Fe³⁺ after 3 h of incubation of maize roots using 0.74 MBq of tracer. Biological assimilation reflects a combination of root and microorganism assimilation of tracer. Data presented as weight normalized percentage of ⁵⁹Fe radioactivity administered. Asterisks indicate significant differences of treatment relative to control (*P < 0.05; **P < 0.01; ***P < 0.001). B Bar graph showing the extent of root-to-shoot transport of radioactive ⁵⁹Fe²⁺ and ⁵⁹Fe³⁺ after 3 h of incubation of maize roots using 0.74 MBq of tracer for the same five study conditions. Data presented as weight normalized percentage of the assimilated ⁵⁹Fe radioactivity. Asterisks indicate significant differences of treatment relative to control (*P < 0.05; **P < 0.01; ***P < 0.001). C Bar graph showing auxin biosynthetic rates in A. brasilense using [¹¹C]indole radiotracer reveals a trend of increasing rate of production with microbial BNF capacity. Correlations were made to BNF data measured using ¹³NN [4]. Data (±SE) reflects N = 4 biological replicates. D Correlation plot mapping ⁵⁹Fe²⁺ allocation against ⁵⁹Fe³⁺ allocation for the study conditions reveals systematic trends of clustering with treatments that correlate to auxin producing and N₂-fixing capacities of the beneficial microbes. E Principal component analysis correlates ⁵⁹Fe translocation to biological functions of the beneficial microbes and to chemical treatment using auxin.

Fig. 2. Radiographic imaging of plant tissues reveal different spatial patterning of radioactive ⁵⁹Fe as ferric (Fe³⁺) and ferrous (Fe²⁺) ionic forms.

Comparisons were made between non-inoculated control roots (A) and roots inoculated with HM053 (B), ipdC (C) and FP10 (D) strains of A. brasilense with ⁵⁹Fe³⁺ images are shown in the upper portion of each panel while ⁵⁹Fe²⁺ images are shown in the lower portion. ‘Hot’ spots of ⁵⁹Fe radioactivity (both for ⁵⁹Fe³⁺ and ⁵⁹Fe²⁺) are seen at lateral root junctions. Roots inoculated with the FP10 strain of A. brasilense only showed ⁵⁹Fe³⁺ ‘hot’ spots. Shoot images for HM053 inoculated plants are shown for ⁵⁹Fe³⁺ (E) and ⁵⁹Fe²⁺ (F). Translocation of ⁵⁹Fe³⁺ to shoots resulted in higher tracer accumulation in all of the leaf tips. Translocation of ⁵⁹Fe²⁺ to shoots resulted in a higher distribution of tracer in the younger developing leaf tissue.

Fig. 3. LA-ICP-MS and TEM reveal spatial patterning of Fe-56 in plant tissues that correlates with root cellular morphological changes due to beneficial microbes.

A Representative Fe-56 ion signals (in red) from the laser ablation of 100 μm root sections (taken 1 cm from the root tip) from non-inoculated control plants, and plants inoculated using HM053, ipdC and FP10 bacteria. B Data from ablation tracks was averaged across N = 6 biological replicates, and presented graphically as % distribution of the integrated Fe-56 ion signal for the different regions of the root cross section (epidermis, cortex, endodermis and inner pith region). These regions are identified on the inset anatomical key. HM053 treatment resulted in significantly higher iron levels in the endodermis than controls. C Absolute concentrations of Fe-56 (ppm) in leaves and roots were determined from direct injection ICP-MS of digested tissues and compared against NIST standards. D Transmission electron microscopy images on endodermal root cells reveal morphological differences between cell walls of non-inoculated control roots and those from HM053. HM053 inoculated roots appeared to have a more defined Casparian band in endodermal radial root walls than in control plants (highlighted by black arrows). E Representative Fe-56 signals from 40 μm diameter x 2 mm laser tracks rastered across freeze dried leaf tissues harvested from non-inoculated, HM053, ipdC and FP10 inoculated plants. The periodic patterning of the iron ion signals (red lines) presumably are due to the higher disposition of iron in leaf cell walls after the freeze drying process. The white lines reflect an average ion signal across the laser track. HM053 elicits a unique patterning with lower iron signal in and around vascular tissues that was not seen with the other beneficial microbes. F The ion signal for Fe-56 was integrated across the laser ablation track and plotted as a function of biologic type. HM053 showed a higher integrated iron signal than non-inoculated control, as well as ipdC and FP10 inoculated plants. G Relative percent of the Fe-56 ion signal in vascular tissue was plotted as a function of biologic type. HM053 inoculation decreased the relative percentage of iron in leaf vascular tissue suggesting greater mobility. Data (±SE) reflected N = 4–6 biological replicates. Asterisks indicate significant differences of treatment relative to control (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig. 4. Carbon-11 aids in mapping maize physiological and metabolic responses to bacteria inoculation.

A Leaf tissue was normalized to ¹¹CO₂ fixation and presented as % ¹¹C-activity in the pulse applied to the leaf cuvette. B Leaf export and root allocation of ¹¹C-photosynthates (measured at 3 h post ¹¹CO₂ pulse) presented as % fixed ¹¹C-activity by the plant. C Root exudation of ¹¹C-substrates (A = acidic substrates; N = non-acidic substrates measured at 3 h post ¹¹CO₂ pulse) presented as % fixed ¹¹C-activity by the plant. D Metabolic landscape reflecting the partitioning of ‘new’ carbon (as ¹¹C) into different metabolic pools of the load leaf tissue. E [^11/12C]-histidine, [^11/12C]-nicotianamine and [^11/12C]-citric acid specific activities (SA) levels measured in leaf tissue 20 min. after tracer administration and represented as % total ¹¹C-activity/μmol substrate gfw⁻¹ of tissue. Data (±SE) reflects N = 7–9 biological replicates. Asterisks indicate significant differences relative to control (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig. 5. Shared metabolic branch point in the biosynthesis of Auxin and DIMBOA.

The biosynthesis of auxin precursors such as indole-3-glycerol phosphate, indole and L-Trp takes place in plastids and are generated via the shikimate pathway. Four putative L-Trp pathways for auxin biosynthesis are shown. Enzymes known to operate these pathways are shown in italics. Solid pathway arrows reflect active processes. Dashed arrows are suggested to exist, but have not been proven. Benzoxazinoid biosynthesis yielding DIMBOA and its conjugate analogs share a common metabolic branch point with auxin as indole. A Gravitropic responses of root tips are associated with auxin biosynthesis and/or sensing. Inoculation of roots using HM053, ipdC and FP10 bacteria reduced root gravitropic responses relative to controls. B Auxin and ethylene biosynthesis are synergistic – inoculation of roots using HM053, ipdC and FP10 bacteria reduced root ethylene emission rates relative to non-inoculated controls. C Root indole emission rates were significantly elevated relative to controls in roots inoculated with HM053, ipdC and FP10 bacteria. D Root DIMBOA concentrations were significantly increased relative to control in roots inoculated with HM053 and FP10 bacteria and unchanged for ipdC bacteria. Data (±SE) reflects N = 6–8 biological replicates. Asterisks indicate significant differences relative to control (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig. 6. In vitro chemotaxis assays to examine effects of DIMBOA on bacteria growth.

A DIMBOA stability in the bacteria growth medium was examined using 0.5 mM DIMBOA. Gas chromatography peak area units (PAU) reflecting DIMBOA levels at different time points spanning 27 h shows a ~15% loss of DIMBOA integrity. B–D DIMBOA dose response effects on bacteria growth at 24 h as measured spectrophotometrically where OD₆₀₀ = 1 corresponds to 10⁸ CFU mL⁻¹. E–G Influence of an acute 0.5 mM DIMBOA dose on bacteria growth after 48 h as measured spectrophotometrically where OD₆₀₀ = 1 corresponds to 10⁸ CFU mL⁻¹. Data (±SE) reflected N = 3 biological replicates at each DIMBOA dose. Asterisks indicate significant differences of bacteria growth with different DIMBOA doses relative to 0 mM DIMBOA control (*P < 0.05; **P < 0.01; ***P < 0.001).