Introduction

The ammonium form (NH4+-N) of N fertilizers are easily transformed to the nitrate form (NO3−-N) due to the nitrification process in aerobic soils (Daims et al. 2015). But unlike NH4+, anionic NO3− cannot be adsorbed by negatively charged soil particles and thus is very mobile in soil. Therefore, NO3− in soil can be leached into water system and cause environmental pollution like water eutrophication, whereas denitrification produces various gaseous N forms (e.g., N2O, NO, and N2), among which, N2O is one of the greenhouse gases responsible for the global warming (Meinshausen et al. 2009). Both of nitrification and denitrification result in the reduction of NUE in agricultural systems (Gooding et al. 2012; Fowler et al. 2013; Subbarao et al. 2012). Thus, keeping soil N as NH4+ for an extended period may improve N recovery and maintain the ecosystem’s sustainability.

Application of synthetic nitrification inhibitors is one of the ongoing practices to control soil nitrification in agricultural soils. Nitrapyrin, 3,4-dimethylpyrazole phosphate (DMPP), and dicyadiamide (DCD) are well-known commercial synthetic nitrification inhibitors (Slangen and Kerkhoff 1984; Zerulla et al. 2001). However, synthetic nitrification inhibitors are not widely used by farmers due to their relatively high cost, associated environmental safety issue, and additional labor costs (Lam et al. 2017). Various nitrification inhibitory substances have been detected in the rhizosphere of plants and are termed biological nitrification inhibitors (BNIs) being responsible for biological nitrification inhibition (Subbarao et al. 2009, 2015, 2017). The presence and function of BNIs have been demonstrated in Sorghum bicolor (sorghum) (Zakir et al. 2008), Brachiaria humidicola (pasture grasses) (Subbarao et al. 2009), Oryza sativa (rice) (Sun et al. 2016), Triticum aestivum (wheat) (O’Sullivan et al. 2016), and Leymus racemosus (a wild relative of wheat) (Subbarao et al. 2007c). BNIs released from roots of these plant species can reduce the abundance of ammonium oxidation bacteria and archaea in soil (Nardi et al. 2013, 2020; Lu et al. 2018) and/or inhibit the ammonium oxidation activity of Nitrosomonas europaea in situ (Zakir et al. 2008; Sun et al. 2016; Subbarao et al. 2006a, b). Brachialactone secreted by roots of Brachiaria humidicola, a tropical pasture grass, can inhibit AMO (ammonia monooxygenase) and HAO (hydroxyl amino oxidoreductase) enzymes (Subbarao et al. 2009). Methyl 3-(4-hydroxyphenyl) propionate (MHPP) secreted by sorghum roots inhibits AMO pathway (Zakir et al. 2008; Subbarao et al. 2007a, 2013). Sorgoleone and sakuranetin are also secreted by sorghum roots, and inhibit the AMO and HAO, but sakuranetin did not show effective BNI function in soil assay (Alsaadawi et al. 1986; Subbarao et al. 2013). The recent research confirmed that 1, 9-decanediol from rice had a significant inhibitory effect on nitrification in alkaline tidal soil, neutral paddy soil, and acid red soil (Lu et al. 2018).

Interestingly, some BNIs also exhibited other functions besides inhibiting nitrification in soil. MHPP from sorghum roots could modify root system architecture by inhibiting primary root elongation and improving lateral root growth in Arabidopsis thaliana, and this may increase the total roots volume for the uptake of mineral nutrients in soil (Liu et al. 2016). Sakuranetin was found to act as flavanone phytoalexin against plant pathogen in rice leaves (Kodama et al. 1992). In addition, sorgoleone can function as a herbicide because it can inhibit photosynthesis of competitor plants by binding to the D1 protein and disturb mitochondrial electron transport, root H+-ATPase activity, and water uptake (Dayan et al. 2010).

Hydrophilic and hydrophobic BNIs

As root exudates, BNIs are synthesized and secreted by plant roots in the rhizosphere (Bais et al. 2001; Walker et al. 2003), and thus, they can modify the soil microbial community (Nardi et al. 2013, 2020). Usually, they are low molecular weight compounds, such as phenolics or other secondary metabolites. The BNIs are in general divided into two different categories. One is water soluble, hereafter referred as hydrophilic-BNIs, such as MHPP (Zakir et al. 2008). The other category is soluble in acidified-DCM (dichloromethane), hereafter referred as hydrophobic-BNIs, which could be obtained by washing roots with acidified-DCM (Subbarao et al. 2013). Due to their differential solubility in water, it is expected that hydrophobic BNIs may remain close to the root as they could be strongly adsorbed by soil particles within the rhizosphere (Subbarao et al. 2012). In contrast, the hydrophilic-BNIs may move farther away from the point of release due to their solubility in water, and this may amplify their capacity to control nitrification beyond the rhizosphere. Sorgoleone, a benzoquinone is a major component of the root-DCM wash and accounts for 80% of the hydrophobic-BNIs activity in sorghum roots (Subbarao et al. 2013). In contrast, sakuranetin, a flavanone, and MHPP, a phenylpropanoid, are hydrophilic, and both are released from sorghum roots (Zakir et al. 2008). In addition, brachialactone and its isomers and derivatives, belonging to the cyclic diterpenes, released by Brachiaria humidicola (Subbarao et al. 2009; Egenolf et al. 2020), and 1,9-decanediol, a fatty alcohol, released by paddy rice (Sun et al. 2016), are also hydrophilic-BNIs. The BNIs identified from plant root exudates are listed in Table 1.

Table 1 Biological nitrification inhibitors released from plant roots
Full size table

Processes of BNIs release from roots

For a long time, secretion of compounds from plant roots was primarily thought to be a passive process of diffusion across the plasma membrane, because small uncharged low molecular weight compounds were thought to pass through the lipid membranes freely (Guern et al. 1987). It is well-known that passive diffusion process depends on membrane permeability and concentration gradient of exudates between the cytoplasm and apoplast (Marschner 2012). In addition, many diffusion processes have been indicated to be mediated by aquaporins (aquaglyceroporins), which are integral membrane proteins that facilitate the transport of water and many other neutral molecules, such as glycerol, ammonia, and urea across cell membrane (Assmann and Haubrick 1996). However, most root exudates are electrically charged molecules or polarized ions (Bertin et al. 2003), and thus, simple diffusion across the plasma membrane bilayer cannot occur due to their low solubility in the lipid membrane. In some cases, the release of root exudates does not depend on its concentration in the root cells (Zhang et al. 2004; Zhu et al. 2005), suggesting that root exudation could be a tightly regulated process rather than simple diffusion (Pariasca-Tanaka et al. 2010). Further, some BNIs, such as sorgoleone, might be toxic alleochemicals, which cannot freely exist in the cytoplasm, but need to be sequestered in subcellular vesicles and released by exocytosis (Czarnota et al. 2003). Therefore, the release of BNIs from roots into the rhizosphere should be facilitated by membrane transport systems such as channels, pumps, carriers, and exocytosis. So far, the transporter of BNIs has not been identified by molecular and genetic evidence.

Involvement of plasma membrane H+-ATPase in the release of hydrophilic-BNIs under NH4 + nutrition

However, the recent two decades of research on BNIs provides us many clues to find possible transporter of BNIs. It has been suggested that NH4+ might act as a signal to trigger the release of BNIs (Subbarao et al. 2007b). Indeed the presence of NH4+ but not of NO3− in root medium could induce BNIs release from roots of various plants, such as Brachiaria humidicola (Subbarao et al. 2009), sorghum (Zakir et al. 2008; Subbarao et al. 2007b, c), wheat (Subbarao et al. 2007c), and rice (Zhang et al. 2019). NH4+ uptake in plant roots is coupled with H+ release and acidification of the rhizosphere, whereas NO3− cotransport with H+ across the plasma membrane results in the increase of rhizosphere pH (Marschner 2012). Moreover, NH4+ nutrition causes a stronger depolarization of plasma membrane as compared with NO3− nutrition (Schubert and Yan 1997). We found that the plasma membrane H+-ATPase activity is induced by NH4+ nutrition, which is associated with the acidification of rhizosphere (Zhu et al. 2009). Plasma membrane H+-ATPase is a universal electrogenic H+ pump, which generates H+ electrochemical gradient to provide driving force for the secondary influx or efflux of ions and metabolites across the plasma membrane (Palmgren 2001). In this way, it could facilitate the transport of hydrophilic-BNIs being putative anionic substances across the plasma membrane. In addition, the metabolism of NH4+ is coupled with H+ generation in the cytoplasm (Lewis et al. 1982; Marschner 2012), and thus, the enhanced activity of plasma membrane H+-ATPase can pump the excessive H+ out of the cell to maintain intracellular pH homeostasis (Schubert and Yan 1997; Zhang et al. 2021). At the same time, the released H+ could also be used as counterions for BNIs release.

It has been hypothesized that BNIs release is linked to the plasma membrane H+-ATPase activity by considering that pharmacological agents can stimulate (fusicoccin) or suppress (vanadate) plasma membrane H+-ATPase activities and this can affect the releases of BNIs from sorghum roots without NH4+ (under the treatment of fusicoccin) or with NH4+ (under the treatment of vanadate) (Zhu et al. 2012). This hypothesis was supported by the recent finding that H+ release and hydrophilic-BNIs release are stoichiometrically linked under various conditions, e.g., different rhizosphere pH, and a range of NH4+ concentrations in root medium (Di et al. 2018). Low rhizosphere pH caused by the uptake of cations, such as NH4+ or even K+, could trigger the activity of plasma membrane H+-ATPase. Recently, it was reported that 3-eip-brachialactone, one kind of BNIs from Brachiaria humidicola roots, is released via secondary transport depending on plasma membrane H+-ATPase generated proton motive force (Egenolf et al. 2021).

Further, the effect of methyl-ammonium, a non-metabolizable analogue of NH4+, was used to investigate whether NH4+ assimilation is also responsible for the stimulation of BNIs release under NH4+ nutrition. There was no significant effect of methyl-ammonium on plasma membrane H+-ATPase activity and BNIs release, suggesting the stimulatory effect of NH4+ on hydrophilic-BNIs release is functionally linked with NH4+ assimilation (Zeng et al. 2016). Apart from NH4+ uptake, NH4+ assimilation in root cells is also a critical factor to induce hydrophilic-BNIs release by triggering plasma membrane H+-ATPase activity to pump out excessive H+ generated by ammonium assimilation and avoid cytosolic acidosis (Zeng et al. 2016). This is also true that transcriptional and post-translational regulation of plasma membrane H+-ATPases was found to be involved in the stimulation of BNIs release under NH4+ nutrition (Zeng et al. 2016; Afzal et al. 2020). At least five plasma membrane H+-ATPase genes responded to NH4+ with a similar expression pattern (Zeng et al. 2016). Thus, the activation of plasma membrane H+-ATPase under NH4+ nutrition is at least partly due to transcriptional regulation of H+-ATPase genes. Recently, we found that NH4+ and low rhizosphere pH (pH 3.0) brought activation of plasma membrane H+-ATPase, based not only on transcriptional regulatory level that resulted in higher steady-state protein concentration of this enzyme, but also on the strongly up-regulated phosphorylation level of plasma membrane H+-ATPase, which further boosted up the H+ pumping activity that ultimately facilitated hydrophilic-BNIs release (Afzal et al. 2020) (Fig. 1).

Fig. 1
figure 1

Prospective pathways for BNIs transport in plant root cells. The uptake of NH4+ acidifies rhizosphere and assimilation of NH4+ in the cytoplasm of root cells produces H+, which in turn induces activation of plasma membrane H+-ATPase activity. The enhanced H+ gradient across the plasma membrane provides the driving force for hydrophilic-BNIs release through anion channels. If the anion channels were blocked, BNIs could also be possibly released by ABC transporters, while hydrophobic BNIs, such as sorgoleone, can be transported by vesicles and released out of cell via exocytosis

Full size image

Because hydrophilic-BNIs are putative anionic substances (Subbarao et al. 2006b; 2007a), and most of the membrane anchored anion channels are voltage gated and depend on membrane potential (Tyerman 1992), the release of the most of hydrophilic-BNIs may be mediated by the cooperation between corresponding anion channels and plasma membrane H+-ATPases that generate the electrochemical gradient (Fig. 1).

Transporters potentially involved in hydrophilic-BNIs release

Because two commonly used anion channel blockers, anthracene-9-carboxylate and niflumic acid, could not inhibit (on the contrary, enhanced) the release of hydrophilic-BNIs from roots of three sorghum genotypes (Di et al. 2018), and these two anion channel blockers cannot inhibit the plasma membrane H+-ATPase activity (Zhu et al. 2005), we hypothesized that hydrophilic-BNIs were also transported through other transporters besides anion channels, such as ATP-binding cassette (ABC) transporters, and the multi-drug and toxic compound extrusion (MATE) transporters. It is well-known that ABC transporters mediate diverse cellular transport processes, such as the excretion of potentially toxic compounds, lipid translocation, and flavonoids secretion (Balzi and Goffeau 1994; Higgins 1995; Maathuis et al. 2003; Buer et al. 2007). ABC transporters are widely distributed in plant species (Martinoia et al. 2002). The model plant Arabidopsis contains 129 ABC transporter genes, seven of which spanning four subfamilies (MRP, PDR, ATH, and PGP) have been identified to participate in the secretion of phytochemicals from plant roots (Badri et al. 2008). More than one ABC transporters are involved in the secretion of a given phytochemical, and one ABC transporter can secrete more than one kind of secondary metabolites (Badri et al. 2008). Because BNIs release is totally suppressed in sorghum roots by vanadate, an inhibitor also of ABC transporters in plants (Rea 2007; Coskun et al. 2017), it is possible that BNIs could be released through one or more ABC transporters located in the plasma membrane of root cells (Fig. 1).

Transporters belonging to the multi-drug and toxic compound extrusion (MATE) family have been indicated to be responsible for citrate exudation under aluminum toxicity (Furukawa et al. 2007). The transportation of anions mediated by MATE depends on the gradient of counter ion, which is usually H+ in plant cells (Shen et al. 2005; Doshi et al. 2017). In addition, citrate released from proteoid roots of white lupin is also related to the activity of plasma membrane H+-ATPase (Yan et al. 2002; Zhu et al. 2005). The citrate efflux proteins in white lupin are characterized by electronic patch clamp (Zhang et al. 2004). Although the exact transporters involved in citrate release from proteoid roots of white lupin have not been identified, they were considered to be similar to MATE proteins (Zhang et al. 2004). MATE proteins can also transport benzoxazinoids, artemisinin, juglone, phenolics, alkaloids, and flavonoids (Zhao and Dixon 2009). MHPP and sakuranetin, two BNIs identified from sorghum, are phenolic and flavonoid, respectively (Subbarao et al. 2013). The BNIs release from sorghum roots depends on the plasma membrane H+-ATPase (Zhu et al. 2012; Zeng et al. 2016). Taken together, MATE transporters may be also involved in BNIs release, although further studies are required to verify this hypothesis (Sivaguru et al. 2013; Doshi et al. 2017).

Release of hydrophobic-BNIs

Unlike hydrophilic-BNIs, hydrophobic-BNIs release does not depend on rhizosphere pH in several sorghum varieties (Di et al. 2018). In addition, the relationship between the plasma membrane H+-ATPase activity and hydrophobic-BNI activity is not correlated (Di et al. 2018). Since most of the hydrophobic-BNI activity is attributed to sorgoleone (Subbarao et al. 2013), the release of this compound seems independent of plasma membrane H+-ATPase generated proton motive force or membrane potential. In addition, due to the toxic property of sorgoleone, its excretion by roots might involve vesicular transport (Battey and Blackbourn 1993). Like many phenylpropane and flavonoids, sorgoleone may be synthesized on the surface of the endoplasmic reticulum and secreted to the extracellular space through vesicles (Winkel-Shirly, 2001) (Fig. 1).

Environmental factors in soils affecting the BNIs release from plant roots

The physical, chemical, and biological properties of soils are tremendously heterogeneous and dynamic, and they strongly affect the activities of plants and microorganisms. Therefore, the release of BNIs in soils should be influenced by various soil environmental factors. Soil fertility is one of the important factors that decide morphological and physiological adaptations of plant roots. For the BNI plants, NH4+ availability is a major factor to stimulate hydrophilic-BNIs release, although it is difficult to measure the released BNIs in situ. Our previous experiments using split roots indicated that sorghum roots incubated in the solution with NH4+ caused stronger release of BNIs as compared with that incubated with nitrate (Zhu et al. 2012). In addition, the release rate of BNIs from sorghum roots depended on NH4+ concentration (Zakir et al. 2008; Zeng et al. 2016). Further, the hydrophobic BNIs, such as sorgoleone, were also found positively correlated to exchangeable NH4+ content in soil (Sarr et al. 2020). Therefore, it may be important that the release of BNIs was triggered at the root zones, where NH4+ is available in soil. It is thus important for plants root to exudate BNIs to protect NH4+ from nitrifiers directly in situ in order to utilize the N source in soil with high efficiency.

The soil pH also plays an important role in the BNIs release by roots. In general, NH4+ uptake by roots causes strong acidification of rhizosphere (Marschner 2012; Zhu et al. 2009). Low pH increased the release of BNIs from roots of Brachiaria humidicola (Subbarao et al. 2007b), sorghum (Zakir et al. 2008; Zhu et al. 2012), and rice (Zhang et al. 2019), while pH above 7 depressed the release rate of BNIs by Brachiaria humidicola or sorghum (Subbarao et al. 2007b; Zhu et al. 2012). It was found that the nitrification inhibitory rate of Brachiaria humidicola root exudates was higher in an Andosol soil with pH 5.9 than in a Combisol soil with pH 6.9 (Gopalakrishnan et al. 2009). Further, it was found that light soils such as Alfisols of the semi-arid tropics India or sandy loams of West Africa are better suited to develop acidic rhizosphere (pH < 6.0) for the BNI function of sorghum, when compared to Vertisols soil (pH > 7.5) (Subbarao et al. 2013). In acid red soils (pH 4.26), the nitrification inhibition potential of the BNIs from rice root exudates was significantly stronger than that in neutral paddy soils (pH 6.25) or alkaline fluvo-aquic siols (pH 7.92) (Lu et al. 2018). These results indicated that the low soil pH enhanced the BNIs release and/or their activity. Since the suppression of nitrification in soil by BNIs can further improve NH4+ availability for plant roots, which in turn leads to the higher acidification of rhizosphere due to the uptake of NH4+, thus acting as a feedback loop for the efficient utilization of soil N.

The bulk density (BD) is an important soil property, which is an indicator of the amount of pore space in the soil (Dam et al. 2005). Bulk density can thus decide the difficulty of root penetration into the soil, and also the content of soil water, which can affect root growth and microbial activity. It is possible that bulk density is critical for the plant roots to release BNIs.

The soil moisture influences not only the plant growth and microbial activity but also the movement of various soluble substances in soil. Xeric moisture regimes are dry and may limit the movement of released BNIs from roots, which may cause the accumulation of BNIs in the rhizosphere and inhibit the root growth and subsequently the release capacity of BNIs. It has been documented that the sorgoleone content in soil is negatively correlated to the soil moisture (Sarr et al. 2020). The growth of terrestrial plants is inhibited under hydric regimes. Therefore, the mesic moisture regimes should be optimal for root growth of terrestrial plants and the release of BNIs in soil.

Besides the above factors, the stability and mobility of BNIs in soil should be considered. Since BNIs are mainly low molecular weight organic compounds, they could be absorbed by surface reactive soil particles, and this can reduce BNIs efficiency. BNIs could also be used as C sources by various soil microorganisms. It is still not clear whether the released BNIs could be degraded by some specific microorganisms. BNIs from root exudation of Brachiaria humidicola showed nitrification inhibitory ability after 60 days in Cambisol and Andosol soils (Gopalakrishnan et al. 2009). The nitrification inhibition in soils could last for 2.3 months by continuous release of sorgoleone from sorghum roots (Sarr et al. 2020). Thus, it seems that the most of BNIs in the soil are relatively stable and efficient. The duration of the nitrification inhibition of various BNIs in soils is worth further evaluation in the future.

Conclusions and perspectives

Controlling nitrification is critical to improve N retention in the soil. BNIs are active where plant root systems deliver powerful BNIs at nitrifier sites, and BNI is one of the best strategies to develop N-efficient production systems. Plants with high BNI-capacity root systems can be produced by using modern breeding tools and approaches. In addition, BNI-producing plants such as Brachiaria pasture grasses can be incorporated into soils (e.g., as green manures) (Subbarao et al. 2012, 2015), and BNI-producing plants can be inter-cropped or rotated with other crops to manage soil nitrifier activity to improve NUE of production systems (Karwat et al. 2017).