Rhizoid‐mediated phosphate uptake and internal transport in the non‐vascular plant <i>Marchantia polymorpha</i>

Nutrient acquisition and internal transport are key challenges for land plants in terrestrial environments. Phosphate, an essential plant macronutrient, is often present at low concentrations in soils, requiring plants to evolve multiple adaptations for its uptake and mobilization (Nussaume et al., 2023). In vascular plants, nutrient acquisition and long-distance transport were greatly enhanced by the evolution of roots and vascular tissues – structures that facilitate the uptake and redistribution of water and a broad range of mineral nutrients, including phosphate – together with diverse nutrient transporter families that enable systemic distribution via xylem (Fig. 1a; Kenrick & Crane, 1997; Hetherington & Dolan, 2018; Roeder et al., 2022; Yang et al., 2024). By contrast, bryophytes lack these complex structures, and the mechanisms by which they acquire and internally redistribute phosphate remain poorly understood (Jones & Dolan, 2012). Here, using the liverwort Marchantia polymorpha as a model bryophyte, we investigate phosphate uptake and transport. We show that rhizoids exhibit strong expression of multiple phosphate transporter genes and actively absorb phosphate. The absorbed phosphate is efficiently translocated to meristems and asexual propagules. These findings suggest that rhizoids function analogously to root hairs in vascular plants and provide new insight into the evolution of land plant nutrient systems. In bryophytes, rhizoids are tip-growing filamentous structures that emerge from the gametophyte body and are both morphologically and genetically similar to the root hairs that develop on the sporophyte bodies of vascular plants (Breuninger et al., 2016; Proust et al., 2016). In vascular plants, root hairs play a crucial role in water and nutrient uptake, where they secrete compounds such as organic acids and peptides to facilitate the solubilization of inorganic nutrients (Libault et al., 2010). By contrast, bryophyte rhizoids have long been regarded mainly as anchoring structures (Jones & Dolan, 2012). Aqueous dyes have been shown to move rapidly between rhizoids and thallus cells via capillary action, indicating that rhizoids also participate in water transport (Jones & Dolan, 2012). Despite these observations, it remains unclear whether rhizoids also contribute to the active uptake of inorganic nutrients, including phosphate (Jones & Dolan, 2012). Marchantia polymorpha, a model bryophyte, possesses two types of rhizoids: smooth and tuberculate (pegged). Recent studies have shown that pegged rhizoids function as water-conducting cells (Lu et al., 2024). Whether rhizoids also mediate nutrient uptake, however, has yet to be demonstrated experimentally. Previous studies nevertheless provide suggestive evidence for rhizoid-mediated phosphate acquisition. Most M. polymorpha PHOSPHATE TRANSPORTER B (PTB) genes, which encode putative Na+/phosphate symporters, show predominant expression in rhizoids and are induced under phosphate starvation (Bonnot et al., 2017). In addition, one PHOSPHATE TRANSPORTER 1 (PHT1) gene was reported to be phosphate-responsive (Saint-Marcoux et al., 2015). More recent whole-thallus transcriptomic analyses identified coordinated upregulation of genes involved in phosphate acquisition – including PHT1 and PTB transporters and Purple Acid Phosphatases (PAPs) – as well as increased phosphatase activity around rhizoids under low-phosphate conditions (Rico-Reséndiz et al., 2020). Together, these findings suggest that rhizoids may play a more active role in phosphate acquisition than previously recognized. We performed transcriptome analysis of M. polymorpha rhizoids to assess their potential roles in nutrient absorption and biomolecule secretion. To obtain highly pure rhizoid samples, thalli were grown on cellophane, frozen, and rhizoids were carefully scraped off using tweezers (Supporting Information Fig. S1a). The collected samples contained both smooth and pegged rhizoids and were largely free of green tissues containing chloroplasts (Fig. S1b). To evaluate sample purity, we compared the expression of MpRSL1 (Mp3g17930), a transcription factor regulating rhizoid development (Proust et al., 2016), with MpCHLH (Mp8g10280), which encodes Chl synthase. MpRSL1 expression was significantly higher in rhizoids than in thalli, whereas MpCHLH expression was markedly lower in rhizoids, confirming the high purity of both the isolated rhizoids and the rhizoid-depleted thalli (Fig. S1c). We then conducted RNA sequencing on isolated rhizoid and thallus tissues. In parallel, organ-derived transcriptome data for thallus, midrib, gemma cup, antheridiophore, and archegoniophore were obtained from a MarpolBase Expression dataset (https://mbex.marchantia.info; Kawamura et al., 2022). Genes associated with external phosphate uptake and internal phosphate transport were identified and their expression profiles compared across these organs. PHT1, which belongs to the major facilitator superfamily (Nussaume et al., 2011), are the primary means by which angiosperms acquire phosphate from the environment. There are seven PHT1 homologs in M. polymorpha, hereafter referred to as MpPHT1;1 through MpPHT1;7 (Fig. S2a,b). Of these, MpPHT1;1 to MpPHT1;6 are clustered on chromosome 2, while MpPHT1;7 is located on chromosome 4 (Table S1). Among these, two PHT1 homologs, MpPHT1;1 and MpPHT1;2, are encoded in a single gene model, Mp2g20600 (Fig. S2c). Notably, MpPHT1;1–MpPHT1;6 were most highly expressed in rhizoids, with MpPHT1;6 showing the strongest expression. By contrast, MpPHT1;7 exhibited uniform expression across all organs, with slightly elevated levels in the midrib and gemma cup (Fig. 1b). PTB genes belong to the Pi TRANSPORTER (PiT) family, which is absent in vascular plants but broadly conserved across prokaryotes and eukaryotes including chlorophyte and charophyte algae and early-diverging land plants (Bonnot et al., 2017). Like PHT1s, PTBs are implicated in mediating phosphate uptake from the environment, but differ in ion coupling: PHT1 proteins function as H+/Pi co-transporters, whereas PTBs operate as Na+/Pi co-transporters (Bonnot et al., 2017). M. polymorpha possesses eight PTB genes, seven of which (MpPTB1, 2, 3, 4, 5, 6, and 8) were previously shown to exhibit higher expression in rhizoids than in thalli (Bonnot et al., 2017). Consistent with this trend, MpPTB1 to MpPTB7 in our dataset show rhizoid-enriched expression, whereas MpPTB8 displays a more systemic pattern (Fig. 1c), possibly reflecting differences in developmental stage or phosphate conditions between studies. In Arabidopsis, the PHOSPHATE 1 (PHO1) gene family includes multiple members, although only PHO1 and PHO1;H1 are clearly implicated in phosphate loading into xylem vessels (Hamburger et al., 2002; Secco et al., 2010). Other members contribute to phosphate deficiency responses and salt stress (Stefanovic et al., 2007; Li et al., 2019). Seven PHO1 homologs have been identified in M. polymorpha (Fig. S3), while four of these have been previously reported (Rico-Reséndiz et al., 2020), showing distinct expression patterns across tissues (Fig. 1d). Among them, MpPHO1;3 (MpPHO1a) has been implicated in the translocation of phosphate within the thallus (Pullagurla et al., 2025). Vacuolar phosphate transporters (VPTs) play a critical role in maintaining phosphate homeostasis in the cytoplasm (Liu et al., 2015; Yang et al., 2017). A search for Arabidopsis VPT homologs identified a single gene, MpVPT, which was most highly expressed in rhizoids (Fig. 1e). Finally, since PHT1 transporters depend on proton gradients generated by plasma membrane H+-ATPases, we also examined the expression of H+-ATPase genes. Okumura et al. (2012) reported eight H+-ATPases. Based on an updated search in MarpolBase v.7.1 (Tanizawa et al., 2025), we identified 11 further genes, bringing the total to 19 H+-ATPases. Consistent with their role in nutrient transport, several H+-ATPase genes were expressed more strongly in rhizoids than in other tissues (Fig. 1f). We next conducted 32P-based bioimaging to evaluate the overall phosphate absorption capacity of rhizoids. Due to the planar morphology of M. polymorpha thalli, it is technically challenging to analyze phosphate uptake via rhizoids and to track subsequent translocation. To overcome this limitation, we introduced a new isotope trace imaging system that integrates a fluorophore with a CMOS sensor. This system detects visible light (spectral peak c. 530 nm) emitted from a scintillator when excited by β-particles (high-energy electrons) from 32P, and the signal is captured directly by the CMOS sensor without the need for an optical lens. The direct coupling of scintillator and sensor markedly improves light-collection efficiency and reduces the integration time required for reliable detection, enabling continuous second-scale imaging of 32P dynamics (Figs 2a, S4). Compared with the authors' previous lens-based design (Kanno et al., 2012), this configuration provides substantially higher temporal resolution and sensitivity, allowing real-time visualization of phosphate uptake at individual rhizoid tips and its rapid movement into the thallus. To localize phosphate uptake at rhizoid tips, we placed 32P-labeled agar medium directly in contact with the tips of several rhizoids and applied Vaseline to prevent lateral diffusion along the rhizoid surface (Fig. 2b). To test whether uptake occurs via PHT1-mediated active transport, we applied Protonstatin-1, an inhibitor of H+-ATPase activity, which is essential for generating the proton gradient required for PHT1 function (Yang et al., 2022). Quantification of 32P signal in the thallus body – reflecting phosphate translocated from the rhizoid tips – showed that Protonstatin-1 markedly reduced phosphate uptake relative to untreated controls, indicating that phosphate absorption by rhizoids is H+/ATPase dependent (Fig. 2c). However, Protonstatin-1 does not completely arrest the uptake, suggesting a possible contribution from PTBs. Notably, 32P signals in the thallus were detected in the first 20-s exposure image following application at the rhizoid tips (Fig. 2a). Because images were acquired as a series of 20-s integrated exposures, signal dynamics at shorter time scales could not be independently resolved. The rapid appearance of the signal nevertheless indicates that phosphate is promptly absorbed at the rhizoid tips and rapidly transported or diffused into the thallus. It should be noted that 32P imaging does not exclusively represent orthophosphate, as absorbed phosphate can be rapidly incorporated into a range of metabolites including ATP, sugar phosphates, phospholipids, and inositol polyphosphates. Inositol polyphosphates have been detected in green algae and bryophytes, including M. polymorpha, where they are thought to function in phosphate storage and signaling (Lorenzo-Orts et al., 2020; Ghosh et al., 2025). A previous study identified specific morpho-physiological responses to phosphate starvation in M. polymorpha, including thallus growth inhibition, accumulation of the red pigment auronidine, enhanced rhizoid elongation, and increased phosphatase activity near rhizoids (Rico-Reséndiz et al., 2020). To further confirm rhizoid-specific responses under phosphate-deficient conditions, we examined the developmental and transcriptomic changes that occur in the presence or absence of adequate phosphate. To assess developmental responses, WT thalli were grown upside down from gemmae on phosphate-sufficient (+P) or phosphate-deficient (–P) medium for 7 d, and rhizoid growth was quantified. Rhizoids extending vertically from thalli cultured under –P conditions were significantly longer and more numerous than those grown under +P conditions, indicating that phosphate deficiency promotes both rhizoid formation and elongation (Fig. S5). Next, transcriptome analysis of thalli grown under ±P conditions revealed that nearly all rhizoid-dominant MpPHT1 and MpPTB genes were strongly upregulated under phosphate-starved conditions, with the exception of MpPHT1;6 and MpPTB7, which showed little or no induction (Table S1) (Bonnot et al., 2017). In addition, several MpPAP genes (Fig. S6), especially MpPAP10, exhibited rhizoid-enriched expression and were strongly induced by Pi deficiency (Fig. 1; Table S1). These findings were consistent with those of previous studies (Saint-Marcoux et al., 2015; Rico-Reséndiz et al., 2020). In angiosperms, phosphate starvation is known to increase root hair density and induce PAP secretion to solubilize phosphate from organic matter (Del Pozo et al., 1999; González-Muñoz et al., 2015; Mehra et al., 2017). Our results suggest that similar adaptive responses occur in bryophyte rhizoids under phosphate-limited conditions, which may enhance phosphate acquisition from rhizoids. In vascular plants, root-absorbed phosphorus is distributed throughout the plant body via vascular tissue. By contrast, the dynamics underlying internal phosphorus redistribution in non-vascular bryophytes remain largely unknown. To address this gap, we examined temporal changes in phosphorus concentration across different regions of M. polymorpha thalli and assessed the potential for internal phosphorus transport and redistribution. In addition, we conducted isotope tracer experiments using 32P to visualize phosphate dynamics within the thallus. First, we quantified phosphate concentrations in defined thallus regions at multiple developmental stages. We defined the basal and apical regions of the thallus as representative older and younger tissues, respectively (Fig. S7a; Solly et al., 2017). We designated the tissue located immediately adjacent to the gemma cup on the proximal side as the basal region, whereas the distal tip of the thallus beyond the gemma cup was defined as the apical region. We next measured the phosphate concentration of thalli grown from gemmae for 2, 3, and 5 wk. At all developmental stages examined, phosphate levels were significantly higher in the apical region than in the basal region. Notably, the phosphate concentration in the basal region declined over time, whereas that in the apical region remained relatively constant (c. 15 μmol g−1 fresh weight) (Fig. S7b). In M. polymorpha, thallus growth occurs through cell division at the apical meristem, with new cells added toward the basal end (Solly et al., 2017). Consequently, the basal region comprises older tissue that remains relatively unchanged during growth. Taken together, these findings suggest that phosphate is actively translocated from older basal to younger apical tissue, thereby maintaining stable phosphate levels in the actively growing region of the thallus. Since gemmae may maintain phosphate homeostasis independently of surrounding thallus tissue, we next examined the phosphate concentrations of clonal propagule gemmae within cups during thallus development. While a phosphate concentration gradient was established between the basal and apical thallus regions (Fig. S7c,d), gemmae within cups consistently retained high phosphate concentrations regardless of their position within the thallus. Next, we directly assessed phosphate translocation by performing 32P pulse-chase tracer experiments. Thalli grown from gemmae for 1 wk were incubated in 32P-containing medium for 24 h and then transferred to 32P-free conditions. The distribution of 32P was subsequently imaged over the following 7 d. Immediately after labeling, 32P signals were detected throughout the thallus (Fig. 2d, day 0), with nearly no change observed after 20 h. By day 3, however, signals became more concentrated in the apical region than in the basal region (Fig. 2d, day 3), suggesting directional movement of phosphate. By day 7, 32P signal had prominently accumulated in the gemma cup area (Fig. 2d, day 7), suggesting active or preferential redistribution of phosphate into developing gemmae. In vascular plants, phosphate uptake from the soil primarily depends on PHT1 family transporters. Interestingly, the non-vascular liverwort Marchantia polymorpha also possesses multiple-copy genes of both PHT1 and PTB-type transporters, although their precise physiological functions remain to be fully characterized. Our results, including the rhizoid-dominant expression of multiple PHT1 genes compared with other thallus tissues, together with the strong reduction of phosphate uptake upon inhibition of H+-ATPase activity by Protonstatin-1, suggest that phosphate is at least partially absorbed by rhizoids via a the other the that transport under reduced H+-ATPase activity that the PTB transporters may also contribute to phosphate acquisition via rhizoids. In addition, rhizoids to secrete which enhance the solubilization of phosphate from organic root hairs in vascular plants, bryophyte rhizoids may water and active nutrient uptake, as for phosphate acquisition. We that the integrated functions of tip-growing filamentous structures – water and nutrient absorption – may have in the of bryophytes and vascular at the specific functions and mechanisms of these transporters provide insight into the and of phosphate acquisition in land bryophytes homologs of the (Pullagurla et al., and VPT gene both of which mediate internal phosphate transport. these transporters contribute to directional phosphate through mechanisms to nutrient in a vascular may a model for the evolution of nutrient transport across early-diverging plant The of Marchantia polymorpha was for all experiments. were cultured from gemmae on medium with plant tissue agar to with et al., To a phosphate-deficient medium we agar in these experiments contained only trace of phosphate. were grown under continuous light at an of μmol by a at thalli to a new we to prevent from the underlying medium and to stress during Thalli were first cultured on placed on the surface of the thalli were transferred along with these to subsequent of the was performed as of membrane were in and at under for 1 were then with water by at under for 20 was from MarpolBase and These were then using & with by using et al., 2020). The multiple of the are with et al., The and were by et al., for the PHT1, and PAP were using the in et al., with and were using The of PAP is on the following et al., 2002; et al., 2010). were first cultured for wk under conditions on placed on plants were in and rhizoids were carefully with to obtain rhizoid The plant from which rhizoids had been were as rhizoid-depleted thallus total RNA was first from samples using a Next, was using a with were performed as the was then performed using a and expression levels were to those of et al., 2015). The following were MpCHLH and MpRSL1 and and was performed using and total RNA was also from samples using an with all following the RNA was then with and using an were using the RNA for and on or to generated in this study were in the under was performed with was assessed using were using were to the Marchantia polymorpha through using et al., with the Other were at their including to possible The were to and using et al., were then to using the from the et al., 2010). were then obtained using et al., with the and the M. polymorpha primary were to were not and expression analyses were performed using et al., 2010). phosphate uptake we applied of medium to phosphate and of with 32P to the rhizoid tips of to M. polymorpha et al., immediately after the plant and medium were in the integrated images were acquired over a The imaging system of an CMOS with an integrated and CMOS sensor were acquired using then quantified using the et al., and functions were performed in To Protonstatin-1 was first added at a concentration of in this was then applied to rhizoids 15 imaging experiments. phosphate tracer gemmae were first grown on agar with for d. were then transferred to agar medium containing μmol Pi and 32P 24 h of plants were transferred to fresh 32P-free We then imaged the distribution of 32P in at 3, and 7 after to the 32P-free were acquired as integrated using an imaging image were using a imaging system were were placed on phosphate-sufficient (+P) or phosphate-deficient (–P) and were incubated upside of the thalli were then imaged with an with a in thalli were from images using et al., 2012). the rhizoid in thallus was measured using were for were cultured for under conditions on placed on phosphate-sufficient (+P) were then transferred to phosphate-deficient (–P) or fresh +P as previously and growth cultured for an 24 h. RNA was then from all plants, and analysis was as previously was performed using two To evaluate gene expression changes under –P relative to +P conditions, change on were and using et al., 2010). were grown on agar medium with for 2, 3, 4, and 5 wk. and basal regions of thalli, or gemmae from apical and basal gemma were and in using a sample was in of pure water and The was then at for 7 and at 20 for at The was collected and at for further inorganic phosphate was quantified using a et al., We first the following 5 and These four were then in a of 5 1 immediately The was added to samples at a 1 the Pi concentration of a sample the sample was or with pure and the was for at was measured at using a phosphate was as the phosphate was using to The and for We also our to and for insight and for the this We also the for and of for our experiments. This study was in by to and and and and to to the for no and from the by for and and and experiments. and performed experiments. and the and the with from all have and the of this data that the findings of this study are within the and in and Table data generated in this study are on under Fig. for transcriptomic analyses in Marchantia Fig. PHT1 homologs encoded in Marchantia polymorpha Fig. analysis of PHO1 homologs in the Marchantia polymorpha Fig. 32P imaging and signal in Marchantia Fig. development in to phosphate in Marchantia Fig. analysis of PAP homologs in the Marchantia polymorpha Fig. across different thallus regions of Marchantia Table H+-ATPase and PAP genes in Marchantia polymorpha and their under phosphorus is not for the or of Information by the than should be to the The is not for the or of by the than should be to the for the The remains with to in and in