PSMs are widely distributed in bacteria, fungi, actinomycetes, and cyanobacteria (Rawat et al., 2021). They drive the “Mobilization” stage of the P cycle by converting fixed P into soluble P that can be directly utilized by microorganisms or plants. Inorganic P solubilization is chiefly driven by the pyrroloquinoline quinone-dependent glucose dehydrogenase (PQQ-GDH) encoded by the gcd gene, an enzyme that catalyzes the oxidation of sugars (e.g., glucose) to produce large quantities of organic acids. These organic acids not only acidify the rhizosphere microenvironment but also, through potent chelation, break down mineral lattices, thereby releasing immobilized inorganic P (Li G. et al., 2023; Li H. P. et al., 2023; Li Z. et al., 2023; Peng et al., 2025). For organic P mineralization, in response to the chemical heterogeneity of the organic P pool, PSMs have evolved a broad spectrum of extracellular enzyme systems (Sun et al., 2017; Stosiek et al., 2020; Zhang et al., 2025). Notably, these metabolic functions exhibit high transcriptional regulatory plasticity. PSMs can respond not only to P starvation induction via the classic Pho regulatory system but also utilize constitutive enzymes (e.g., PafA) to maintain basal mineralization activity in P-rich habitats. This diverse metabolic mechanism is a key adaptive strategy for coping with environmental P fluctuations (Lidbury et al., 2022) (Table 1).

PAOs’ retention and buffering capacity fundamentally rely on a reversible aerobic-anaerobic cycle (Table 2). Under aerobic conditions, PAOs oxidize intracellular carbon sources to generate energy, utilizing high-affinity Pst systems and low-affinity Pit systems to uptake phosphate against the concentration gradient (Mino et al., 1998). Concurrently, cells co-transport magnesium (Mg²⁺) and potassium (K⁺) to neutralize negative charges, thereby maintaining charge balance and osmotic stability (Acevedo et al., 2012). Ultimately, intracellular phosphate is converted into poly-P storage under the catalysis of polyphosphate kinase (PPK) (Crocetti et al., 2000). Conversely, under anaerobic conditions, PAOs hydrolyze glycogen and poly-P to energize volatile fatty acid (VFA) assimilation while releasing phosphate (Mino et al., 1998) (Table 2). Through this cycle of anaerobic release and aerobic uptake, PAOs effectively regulate environmental P levels. They occupy distinct ecological niches based on their carbon substrates and electron acceptors. For instance, Ca. accumulibacter utilize O₂ as an electron acceptor; while denitrifying PAOs (DPAOs) like Ca. dechloromonas utilize nitrate in an anoxic environment (Goel et al., 2005). Fermentative PAOs, such as Tetrasphaera, can ferment complex macromolecules like glucose and amino acids, gaining them a competitive advantage (Close et al., 2021). Recent studies also show that some Ca. accumulibacter strains can use light energy for P uptake under O₂, NO₃⁻, and NO₂⁻ deficient conditions (Carvalho et al., 2024). These findings highlight the metabolic diversity and adaptive flexibility of PAOs, underscoring the necessity for further research into their niche-specific mechanisms.

The MRB strategy has proven effective for sustainable P utilization at multiple scales. In agricultural, co-inoculating PSMs and PAOs reduces P leaching by 22.6% and increase the soil’s available P pool by 18.3% (Li et al., 2023c). PSMs enhance wheat P utilization by 91% through hydrolyzing poly-P-rich fertilizers (Khourchi et al., 2023). In composting systems, microbial biomass P increased by 83% with PSM and PAO inoculation, converting organic waste into high-quality bio-slow-release fertilizers (Zhan et al., 2023). PAO-rich biosolids from EBPR treatment of agricultural wastewater aid in recovering high-purity P through technologies like struvite crystallization (Kalpakchiev et al., 2025). Notably, using EBPR activated sludge as P fertilizer improves corn growth, with ~30% of PAOs surviving in the soil, forming a mutualistic network with plant growth-promoting rhizobacteria for sustainable P supply (He et al., 2025). Furthermore, a positive correlation has been found between PAO abundance and plant growth-promoting traits, stress resistance, and alleviation of salt stress (Srivastava et al., 2022). The MRB strategy holds significant potential for P pollution control, legacy soil P activation, and increased crop uptake and yield.

Despite the great potential of MRB strategies in agricultural P fixation, transformation, and wastewater P removal, limitations in reaction rate and stability remain (Dou et al., 2025). Traditional methods are hindered by a lagging understanding of core functional groups in complex habitats. These methods typically follow the “culture-first screen-second” model, which involves isolating and screening the PSMs or PAOs under laboratory, followed by screening strains with desirable phenotypes for validation (De Zutter et al., 2022; Fu et al., 2024). However, such strains often face genotype–phenotype mismatches, with poor colonization and unstable performance in environments (Coats et al., 2017; Raymond et al., 2021; Copeland et al., 2025). Therefore, there is an urgent need for new methods to screen strains with natural competitive advantages (Liu et al., 2024). Moreover, the presence of functional genes (e.g., pqqC) does not guarantee in situ metabolic activity, making it difficult to identify functional executors (Dai et al., 2020). In wastewater P removal, the unculturable nature of key PAO populations, such as Ca. accumulibacter, hinders progress and forces reliance on substitute strains that do not represent the real ecological niche, leading to misjudgments (Burow et al., 2008; Yan et al., 2024). More critically, existing mainstream methods are still limited to characterization at the population-average level. This low-resolution perspective not only masks metabolic heterogeneity and interspecies competition at the single-cell level, but also makes it difficult to precisely quantify the dynamic transformations of key intracellular polymers (e.g., poly-P) (Li G. et al., 2023). Overcoming this bottleneck lies in introducing in situ characterization techniques with single-cell resolution to unlock the metabolic “black box” of complex microbial communities.

SCRS is a non-invasive, label-free technique that analyzes the internal chemical composition of single cells by capturing molecular vibrations (Wang et al., 2017). In the 400–1800 cm⁻¹ wavenumber range, cells produce characteristic spectra known as the “fingerprint region” (Pan et al., 2025). These peaks reflect the metabolic and functional status of cells; allowing for the identification of cellular phenotypes (Li et al., 2012; Hatzenpichler et al., 2020; Cui et al., 2022).

PAOs can uptake P luxuriously, storing it as poly-P intracellularly (Ying et al., 2024). The accumulation of intracellular poly-P is a key feature of PAOs function (Bi et al., 2025). Majed et al. (2009) pioneered the use of Raman spectroscopy for in situ identification of intracellular poly-P at the single-cell level. Poly-P characteristic peaks are found at 690–700 cm⁻¹ (P–O–P bond stretching) and 1,168–1,177 cm⁻¹ (P–O–P bond stretching). The intensity of the latter is linearly correlated with poly-P concentration, serving as a reliable semi-quantitative indicator of PAOs’ P storage capacity (Majed et al., 2009). Furthermore, combining SCRS with fluorescence in situ hybridization (FISH) links phylogenetic identity to metabolic function, enhancing the identification of functional microorganisms in situ, which is critical for optimizing system stability in mixed-culture processes (Fernando et al., 2019; Petriglieri et al., 2022).

In contrast, PSMs convert insoluble P into bioavailable P extracellularly. Tracking P (³¹P) is challenging due to its single isotope, which prevents stable isotope probing (SIP) (Hatzakis et al., 2008). Recently, SCRS combined with deuterium (D) isotope probing (Raman-DIP) has enabled detection of cellular metabolic activity. The 2040–2,300 cm⁻¹ spectral region contains no intrinsic cellular signals. However, by providing deuterated substrates, active cells integrate environmental D into newly synthesized biomacromolecules, producing significant peaks from the conversion of C–H to C–D bonds. C–D bond intensity has become a universal indicator (Berry et al., 2015). This Raman-DIP strategy was applied in a P-limited culture system, where insoluble P was the sole source. Only phosphate-solubilizing cells can acquire P, maintaining metabolic activity. The study confirmed that the C–D ratio (CDR) was positively correlated with soluble P content and acid phosphatase activity (Li et al., 2019). This method converts phosphate solubilizing functions into recognizable metabolic signals, offering a powerful tool for quantifying the in-situ P solubilizing capacity of the soil microbiome. Also, beyond P-cycling organisms, SCRS has demonstrated extensive utility in characterizing diverse microbial phenotypes, such as those involved in carbon/nitrogen cycling and antibiotic resistance (Hatzenpichler et al., 2020; Pan et al., 2025).

In summary, SCRS employs two complementary strategies, D₂O-labeled metabolic tracing and endogenous fingerprint imaging. These strategies overcome the challenge of phenotypic identification of key functional groups in the P cycle, providing a powerful in situ analytical tool for studying P metabolic flux at the micro-scale.

SCRS not only identifies PSMs and PAOs through functional peaks in Raman spectra but also provides in situ metabolic insights at single-cell resolution through non-destructive, real-time chemical imaging. For example, using the “Raman-D₂O” strategy in soil matrices, it was shown that under P limitation, PSMs enter metabolic dormancy to minimize wasteful energy consumption (Li et al., 2024). Under these conditions, their phosphate-solubilizing activity can only be activated and enhanced when exogenous organic carbon is supplied. This indicates that PSMs adopt a strategy characterized by “enhanced carbon metabolism in exchange for P accessibility”. In contrast, PAOs exhibit complex mechanisms for intracellular P chelation and allocation. SCRS was first applied to study EBPR-related metabolites (poly-P, PHA and glycogen) in PAOs, validating the anaerobic release and aerobic uptake metabolic model at the single-cell level (Majed et al., 2012). SCRS further revealed that under extreme starvation stress, PAOs preferentially consume glycogen for maintenance energy, only hydrolyzing poly-P once glycogen reserves are depleted (Bucci et al., 2012).

SCRS corrected previous models of P metabolism by revealing microbial phenotypic heterogeneity in situ. For example, the P-solubilizing activity of soil PSMs varied from 2% to 30% (Li et al., 2024). This suggests that in the same microenvironment, only a fraction of PSMs maintain high metabolic activity to dissolve P, possibly as a survival mechanism. This highlights the critical need to shift from abundance-based to activity-based analyses, focusing on functionally active microbial subsets. SCRS also reveals microbial metabolic diversity. Using FISH-Raman, it was confirmed that Tetrasphaera, a key PAO in wastewater treatment, lacks PHA (Singleton et al., 2022). SCRS studies indicate that microbial diversity in P metabolism is far more complex than anticipated and cannot be fully understood through population-level analyses.

Besides precise non-culture phenotypic identification, SCRS, combined with microfluidic chips and optical tweezers, enables the precise manipulation and sorting of targeted single cells. The sorted cells can be further studied in two ways: (i) coupling with low-bias nucleic acid amplification to generate high-coverage genomic data linked to metabolic phenotypes, which is particularly advantageous for elucidating the genetic functions of difficult-to-culture or rare taxa (Xu et al., 2020; Wang et al., 2021; Zhang et al., 2023), and (ii) direct single-cell cultivation targeting in situ metabolic activity, which helps avoid competitive contamination and improves the success rate of isolating in situ functional microorganisms (Jing et al., 2022; Jing et al., 2025).

This strategy has successfully isolated highly active PSMs from various habitats. For instance, using Laser-Induced Forward Transfer (LIFT) for single-cell sorting, cells with high phosphate-solubilizing activity were isolated. Subsequent 16S rRNA sequencing and metagenomic analysis identified previously overlooked low-abundance soil taxa, including Bacillus marmarensis, Moraxella osloensis, and Stenotrophomonas maltophilia, further revealing the genomic-level metabolic mechanisms of soil PSMs in the P cycle (Li et al., 2024). A “screen-first culture-second” strategy via single-cell Raman-activated sorting and cultivation (scRACS-culture) successfully isolated efficient PSMs such as Comamonas spp., Acinetobacter spp., and Citrobacter spp. from wastewater (Jing et al., 2022). Notably, PSMs activity under in situ conditions was 1–2 times higher than in pure culture, highlighting the significant value of this strategy in reflecting microbial in situ functions (Jing et al., 2022). Similarly, the scRACS-Culture workflow was used to isolate highly active PAOs, including Acinetobacter spp., Micrococcus luteus, and Bacillus spp. Notably, Micrococcus luteus, a novel PAO, showed ~14.8% of its poly-P accumulation in pure culture compared to in situ (Jing et al., 2025). This significant “phenotypic degradation” explains why it is often missed in traditional screenings. When applied to an anaerobic-anoxic-aerobic reactor treating municipal wastewater, this strain significantly improved P removal efficiency from 45% to 89%. These findings indicate that in situ metabolic activity-based screening can uncover key functional strains overlooked by traditional methods, thereby providing valuable resources for wastewater ecosystem engineering (Figure 1).