Introduction

Although organic compounds containing NH2 functional groups (e.g., amino acids, amino sugars) account for only a minor proportion of dissolved organic N in soils (Roberts et al. 2007; Warren 2017), they have attracted considerable interest as contributors of labile N in natural ecosystems due to their rapid turnover and intact uptake by plants and microorganisms (Jones et al. 2005a; Nannipieri and Paul 2009). Living and dead roots, microbial exudation, litter inputs (Jones et al. 2005a; Moe 2013; Holz et al. 2017), and microbial debris (Roberts et al. 2007) are key sources of amino acids and amino sugars in soil. Rhizodeposition accelerates microbial activity in the rhizosphere, resulting in stiff competition for amino acids as the most available N source in the root zone (Moe 2013; Holz et al. 2017; Blagodatskaya et al. 2021). Plants, in turn, are able to recapture large amount of exogenous amino acids from soil (Jones et al. 2005b). Therefore, plant roots might be both a major source and sink of amino acids.

Numerous methods have been developed to assay total and specific amino acid/sugar concentrations in soil. For example, a chromogenic dye assay using ninhydrin (Moore and Stein 1954) and a fluorometric assay using OPAME (O-phthalaldehyde and β-mercaptoethanol) have been adapted for the measurement of free amino acids (Jones et al. 2002; Darrouzet-Nardi et al. 2013) and amino sugars (Roberts et al. 2007) in soil. The balance between amino acid production and consumption determines the total amino acid concentration measured in a homogenized soil sample (Hill et al. 2019). Due to high soil heterogeneity, greater microbial activity and faster nutrient turnover rates occur in soil hotspots enriched with easily degradable organic matter (Kuzyakov and Blagodatskaya 2015; Ghaderi et al. 2022; Hao et al. 2022), where dynamic inputs of amino acids (either plant or microbial origin) are expected (Inselsbacher et al. 2011; Hill et al. 2019; Blagodatskaya et al. 2021). Indeed, compared to non-rhizosphere soil, larger concentrations of amino acids have been detected in the rhizosphere (Inselsbacher et al. 2011), biopores, and detritusphere hotspots by a microdialysis approach (Hill et al. 2019). Despite the successful determination of amino acids in soil microhabitats, the latter method only determines amino acid fluxes and cannot quantify the spatial distribution of NH2-containing compounds in the soil microenvironments.

The aim of the present study was to visualize and quantify the spatial distribution of amino–N compounds at the root-soil interface of Zea Mays L. We employed a new technique, termed time-lapse amino mapping (TLAM), that combined the principle of membrane soil zymography (Spohn and Kuzyakov 2013; Sanaullah et al. 2016) with a time-lapse approach (Guber et al. 2021). The technique involved applying an OPAME reagent (O-phthalaldehyde and β-mercaptoethanol) saturated membrane on the soil and at the soil-root interface. The OPAME reagent reacted with NH2 functional groups in compounds at the membrane-soil interface to generate a fluorescent product visible under UV light. Development of the fluorescence signal in the membrane over time was recorded using a DSLR camera (time-lapse approach). We hypothesized that amino-N compounds are mainly produced by roots and root-associated microorganisms, i.e., the rhizo-microbiome. Therefore, their concentrations were expected to be greater in the rhizosphere than in non-rhizospheric soil.

Materials and methods

The soil was collected from a depth of 0–50 cm of a loamy Haplic Phaeozem in an agricultural area planted with oilseed rape (Vetterlein et al. 2021) near Schladebach, Saxony-Anhalt, Germany (51.3087° N, 12.1045°E). The soil had a loam texture (sand 33%, silt 48%, and clay 19%) with 8.6 g kg−1 total organic carbon, 0.84 g kg−1 total nitrogen, and pH of 6.4 (CaCl2).

Soil and seeds were prepared as described by Vetterlein et al. (2021) as part of the DFG Priority Program 2089 “Rhizosphere spatiotemporal organization—a key to rhizosphere function.” The soil was air-dried, passed through a 2 mm sieve, and kept at room temperature. The soil was treated with 50 mg N (NH4NO3), 50 mg K (K2SO4), 25 mg Mg (MgCl2 6H2O), and 40 mg P (CaHPO4) per kg dry mass, and a 1 mm sieve was used to uniformly distribute the fertilizer. To visualize the spatial distribution of amino-N at the root–soil interface, a rhizobox setup was established to grow maize plants. Sieved and fertilized soil was packed in transparent rhizoboxes (3 × 8.8 × 17.8 cm, H × W × L; Clickbox® Germany) to a final bulk density of 1.26 g cm−3. Sterilized maize (Zea mays L.) seeds were sown 1 cm below the soil surface in three replicate rhizoboxes. Soil water content was maintained at ~ 22% (v/v) in the rhizoboxes during the growth period (21 days), and distilled water was added regularly to the soil to compensate for evapotranspiration losses. During the experiment, the rhizoboxes were inclined at 50° to the horizontal to direct root growth along the front rhizobox walls, as suggested by Spohn and Kuzyakov (2013) and Sanaullah et al. (2016).

For amino mapping, the OPAME reagent was prepared according to Jones et al. (2002). Briefly, 50 mg of O-phthalaldehyde was dissolved in 5 ml of methanol, then 100 µl of mercaptoethanol was added, and the mixture was vortexed in a fume hood (OPAME concentrate). To prepare the working solution, 200 ml of borate buffer (0.02 M, pH 9.5) was added to the OPAME concentrate. Various membrane types (cellulose, regenerated cellulose, polyamide, polycarbonate, and cellulose nitrate) were tested to determine the best membrane for amino mapping (see the supplement for details). Consequently, 140 µm thick cellulose nitrate membranes with 0.2 µm pores size were chosen as the most appropriate for amino mapping due to their hydrophilicity, non-auto fluorescent properties, and high signal-to-noise ratio.

Three cellulose nitrate membranes (circle, plain, 0.2 µm, 50 mm ∅, Whatman®, Cytiva 10401314) were soaked in OPAME reagent. The rhizoboxes were opened from the root side and placed in a dark chamber with 15 W blue-black ultraviolet lamps (Erolite® Germany) as sources of UV light. The OPAME-saturated membranes were placed directly on soil-root surfaces for amino mapping. A transparent glass sheet was placed over the membranes to keep them in contact with the soil and prevent evaporation of the substrate from the membranes. A Nikon D3500 DSLR camera equipped with an AF-P DX NIKKOR 18–55 mm f/3.5–5.6G VR lens (Nikon Inc., Melville, NY) was used to capture images. The object distance, lens aperture, and shutter speed were 210 mm, f/6.3, and 1/125 s, respectively. Membrane images were captured during a 10-min time series (every 10 s for the first minute and then every 30 s during the next 9 min).

Similarly to fluorometric microplate assay, L-leucine was used as a standard for amino mapping. L-leucine has been recommended as an affordable, representative standard, and the units of L-leucine equivalents have been shown to provide an approximation of the true total free primary amine concentration (Darrouzet-Nardi et al. 2013). For quantitative analysis of amino-mapping images (aminograms), 5 µl of an L-leucine standard solution of concentration 0, 0.1, 0.5, or 1 mM (corresponding to 0, 7, 35, and 70 ng of L-leucine-N, respectively) and 10 µl of OPAME were applied to the membranes, then they were covered with a transparent glass sheet to mimic the amino-mapping settings. Membranes saturated with the standard solutions were photographed under UV light with the same camera settings as used for TLAM.

The standard solution images were used for calibration and processed using ImageJ software (Schindelin et al. 2012) to determine the relationship between image brightness and equivalent N content in image pixels. The calibration images were split into red, blue, and green channels and converted to 8-bit format. The brightness in the image pixels across the calibration membranes was determined by subtracting average grayscale values (GSVs) of images with zero L-leucine content from the GSVs of the calibration membranes separately for each channel. Linear regression analysis between the applied N content and GSV sums in the pixels of the calibration images acquired 5 min after application of the L-leucine standards was used to calculate the calibration coefficient (a) from

$$\begin{array}{cc}{M}_{i}=a\sum_{1}^{n}{G}_{i}^{j}{F}_{i}^{j}& 1<j<255\end{array}$$

where Mi is the N amount applied to the membrane (ng), \({G}_{i}^{j}\) is the GSV in pixels, \({F}_{i}^{j}\) is the number of pixels in each grayscale group j, and a is the calibration coefficient (ng of N G-value−1 pixel−1). The best calibration was obtained for the green channel with a = 1.015 × 10−5 ng G−1 pixel−1 (Fig. S5). As this type of calibration depends on the individual imaging process and camera settings, it should be re-evaluated for each experimental setup. To acquire aminograms, the time series amino mapping was processed using a modified time-lapse procedure proposed by Guber et al. (2021). Specifically, the N-amino content in the membranes was calculated based on the maximum G values detected in individual pixels of the green channel during 10 min of amino mapping. The rhizosphere extent of the N-amino compounds was analyzed using the ImageJ Plot Profile tool from cross-sections drawn across selected roots perpendicular to their axes by determining the distance between a region with high amino-N concentration in the soil (mean soil concentration + SD) and the root border. The average GSV of areas corresponding to non-rhizosphere soil was defined as the mean soil concentration of amino-N.

Results and discussion

Application of OPAME-saturated membranes enabled visualization of the spatial distribution of amino-N, which was mainly associated with the maize roots and rhizosphere (Fig. 1). A higher concentration of amino-N compounds was detected in the rhizosphere than in non-rhizospheric soil (Fig. 1 a–c), most likely due to root exudation and high microbial activity in the rhizosphere. Amino acids are known to predominate in root exudates, which are released in vast quantities from roots to soil (Jones et al. 2005a; Moe 2013).

Fig. 1
figure 1

Aminograms (a–c) of maize roots (Zea Mays L.). Daylight photographs of maize roots (d–f) taken before TLAM application. The red arrows show the root tips

Full size image

The concentration of amino-N compounds decreased with increasing distance from the root surface (Fig. 2) and extended up to a 2 mm zone around the center of roots (Fig. 2). The amino-N concentration was 10 times higher in the vicinity of roots than in the non-rhizosphere soil, supporting our initial hypothesis. Despite the approximately two times smaller radius of the root tips than the average size of seminal roots, the area of elevated amino-N concentration was almost doubled around the former. Thus, the relative rhizosphere extent (as a ratio of the root radius) was about 5 times smaller at seminal roots (1.5) than at their tips (7.6). The higher concentration of amino-N at the root tips (Fig. 2) provided visual evidence supporting the common assumption that amino acids are mainly released from the root tips of Zea Mays L. (Jones and Darrah 1994). Previous observations of high amino acid concentrations at a fine scale in biopores and the detritusphere (Hill et al. 2019) suggest that high amino-N content may indicate soil hotspots.

Fig. 2
figure 2

Distribution of amino-N compounds measured from the center of a maize (Zea Mays L.) root tip or seminal root to non-rhizospheric soil: zero distance denotes the center of the maize root. RE indicates the extent of the rhizosphere

Full size image

Conclusions

An amino-mapping approach was successfully developed and tested to visualize the two-dimensional distribution of amino-N compounds at an undisturbed root-soil interface. However, further work is required to standardize the method and test it for different soil types, plant species, vegetation stages, and physiological situations, such as drought stress or nutrient deficiency.