The Medicago truncatula Vacuolar iron Transporter‐Like proteins VTL4 and VTL8 deliver iron to symbiotic bacteria at different stages of the infection process

Summary The symbiotic relationship between legumes and rhizobium bacteria in root nodules has a high demand for iron, and questions remain regarding which transporters are involved. Here, we characterize two nodule‐specific Vacuolar iron Transporter‐Like (VTL) proteins in Medicago truncatula. Localization of fluorescent fusion proteins and mutant studies were carried out to correlate with existing RNA‐seq data showing differential expression of VTL4 and VTL8 during early and late infection, respectively. The vtl4 insertion lines showed decreased nitrogen fixation capacity associated with more immature nodules and less elongated bacteroids. A mutant line lacking the tandemly‐arranged VTL4–VTL8 genes, named 13U, was unable to develop functional nodules and failed to fix nitrogen, which was almost fully restored by expression of VTL8 alone. Using a newly developed lux reporter to monitor iron status of the bacteroids, a moderate decrease in luminescence signal was observed in vtl4 mutant nodules and a strong decrease in 13U nodules. Iron transport capability of VTL4 and VTL8 was shown by yeast complementation. These data indicate that VTL8, the closest homologue of SEN1 in Lotus japonicus, is the main route for delivering iron to symbiotic rhizobia. We propose that a failure in iron protein maturation leads to early senescence of the bacteroids.

Growth of wild-type, 13U and vtl4 Medicago truncatula with and without nitrogen.     Table S1 Tnt1 insertions in Medicago truncatula lines NF17463 (vtl4-1) and NF21016 (vtl4-2).  Methods S1 Additional information on Materials and Methods.   Figure S4. Nodule fresh weight of complemented 13U lines. For each root transformation, nodules were pooled to determine the total fresh weight, divided by the number of nodules. The values are the mean of 3 independent transformations ± SE. Letters shared in common between the transformation indicate no significant difference (one-way ANOVA, Tukey's HSD tests). (a) Expression of the Sinorhizobium meliloti mbfA promoter is regulated by the Iron response regulator IrrA in Rhizobium leguminosarum. The PmbfA:lux reporter plasmid was transferred by conjugation into a wild-type strain and an irrAmutant strain of R. leguminosarum (Wrexler et al., 2003;Todd et al., 2006). Cells were grown in liquid medium with or without 40 µM FeSO 4 (as indicated) and luminescence was measured at an OD600~1. Values represent the mean ± SE of 3 bacterial cultures.
(b) R. leguminosarum strains as in (a), grown on TY agar plates (iron sufficiency). Luminescence was imaged with the NightOWL imaging system, and represented as an artificial colour scale from blue (low) to red (high).  13U mutant Figure S6. Iron staining of nodules using Perls' reagent. Medicago truncatula Jemalong nodules at 28 dpi from wild type (a) and 13U mutant plants (b) were stained for iron using potassium iron cyanide which gives a blue colour, also known as Perls' staining. To stain haem-iron in the infected cells of wild-type nodules, the ferricyanide salt was used instead of ferrocyanide, which results in a greenish-blue colour. The images are of 50-100 µm sections made with a vibratome, imaged with a Zeiss AXIO Imager Z2 (left) or Leica DM600 microscope (right). Scale bars are 0.1 mm.

Acetylene reduction assay
The activity of nitrogenase was assayed by testing the capacity of nodules to reduce acetylene to ethylene, another reaction catalyzed by the enzyme. Nodulated roots were harvested 28 dpi and those of two or three single plants were placed in 1.8 ml glass vials sealed with a rubber cap. To start with, dilutions of ~99.5% ethylene were injected into a Shimadzu 2010 GC gas chromatograph using a HP-PLOTQ (30 m x 320 m x 20 m) column to generate a calibration curve. Next, 180 l of acetylene was injected into each vial containing detached nodulated roots. After incubation at room temperature for at least 2 h, 100 l gas samples were taken and injected into the gas chromatograph to determine the amount of ethylene produced. After the assay, nodules were picked off the roots and weighed. The acetylene reduction activity (ARA) was calculated as picomole of ethylene per min per mg nodule weight.

Measuring bacteroid length
About 80 mg nodules were homogenized in 300 L ice-cold PBS, pH 7.4. Cell debris was removed by centrifugation (500 x g, 10 min, 4 o C) and supernatant was filtered through a 13U; and ≥ 400 from the dnf7-2 mutant (Horvath et al., 2015) was measured using ImageJ software. The relative distribution of different bacteroid size was plotted.

Yeast complementation
The yeast strain DY150, which is derived from W303, was used as wild type. The Δccc1 strain in this background carries a genomic deletion of CCC1, initially identified as Cross-Complements Ca 2+ 1, but later shown to mediate vacuolar iron transport (Li et al., 2001).
Plant genes were cloned into shuttle vector pYES2 under the control of the GAL1 promoter for galactose-inducible expression. The coding sequence of Arabidopsis VIT1