Often in silico, rarely in vivo: characterizing endemic plant-associated microbes for system-appropriate biofertilizers

Abstract

The potential of phosphate-solubilizing microbes (PSMs) to enhance plant phosphorus uptake and reduce fertilizer dependency remains underutilized. This is partially attributable to frequent biofertilizer-farming system misalignments that reduce efficacy, and an incomplete understanding of underlying mechanisms. This study explored the seed microbiomes of nine Australian lucerne cultivars to identify and characterize high-efficiency PSMs. From a library of 223 isolates, 94 (42%) exhibited phosphate solubilization activity on Pikovskaya agar, with 15 showing high efficiency (PSI > 1.5). Genomic analysis revealed that the "high-efficiency" phosphate-solubilizing microbes belonged to four genera (Curtobacterium, Pseudomonas, Paenibacillus, Pantoea), including novel strains and species. However, key canonical genes, such as pqq operon and gcd, did not reliably predict phenotype, highlighting the limitations of in silico predictions. Mutagenesis of the high-efficiency isolate Pantoea rara Lu_Sq_004 generated mutants with enhanced and null solubilization phenotypes, revealing the potential role of "auxiliary" genes in downstream function of solubilization pathways. Inoculation studies with lucerne seedlings demonstrated a significant increase in shoot length (p < 0.05) following treatment with the enhanced-solubilization mutant, indicating a promising plant growth-promotion effect. These findings highlight the potential of more personalized "system-appropriate" biofertilizers and underscore the importance of integrating genomic, phenotypic, and in planta analyses to validate function. Further research is required to investigate links between genomic markers and functional outcomes to optimize the development of sustainable agricultural inputs.

Keywords: bacterial genome; biofertilizer; lucerne; mutation; phosphate solubilization; seed microbiome.

FIGURE 1

(a) Images of representative high-efficiency PSM (Lu_Au_030), a low-efficiency PSM (Lu_HR_012), and a non-PSM (Lu_Ed_008) after 10 days of growth, (b) phosphate solubilization capacity across bacterial isolates from nine commercial Australian lucerne cultivars. Isolates were categorized based on PVK screening results: negative (F – fungal) and negative (B – bacterial) indicate isolates with no zone of clearing, while 0 < PSI < 1.5 indicates isolates with low phosphate solubilization efficiency and PSI > 1.5 indicates high phosphate solubilization efficiency, respectively, based on clearing zone diameter relative to colony size, (c) description of colony morphologies and PSI of lucerne seed isolates chosen for sequencing.

FIGURE 2

Pigment production phenotype in wildtype and mutant strains. (a) Lu_Sq_004_WT at day 6 (left) and day 10 (right) on R2A, (b) Lu_Sq_004_WT at day 6 (left) and day 10 (right) on PVK, (c) first identification of Lu_Sq_004_1_2 in visual screen (left) and PSI screen (right), (d) Lu_Sq_004 mutant pigment production screen at day 6 (left) and day 10 (right).

FIGURE 3

Circos plot showing the distribution of point mutations across the chromosome (chr—blue) and plasmid (pls1—green and pls2—yellow) in high-efficiency mutant Lu_Sq_004_1_2 in the outer ring, low-efficiency mutants Lu_Sq_004_1_1 and Lu_Sq_004_4_2 in the middle ring and null mutants Lu_Sq_004_6_3, Lu_Sq_004_6_4 and Lu_Sq_004_6_5 in the inner ring.

FIGURE 4

Average shoot length of 40 lucerne seedlings per treatments. Treatments were undiluted and diluted bioinoculant solutions of the wild type Lu_Sq_004_WT, Lu_Sq_004_1_2 (gain of function mutant) and Lu_Sq_004_6_5 (loss of function mutant). Bars represent the mean shoot length (cm) of seedlings treated with neat or diluted inoculants. Error bars indicate the standard error of the mean (SEM). *Represents a significant difference (n = 280, 7 treatments, p < 0.05, Tukey’s HSD test). The control group represents PBS soaked seeds, used as a baseline for comparison.