Expansion dynamics, metabolite composition and substance transfer of the primary root growth zone of Zea mays L. grown in different external nutrient availabilities


  • A. WALTER,

    Corresponding author
    1. Botanical Institute, University of Heidelberg, Im Neuenheimer Feld 360, 69120 Heidelberg, Germany
    Search for more papers by this author
  • R. FEIL,

    1. Botanical Institute, University of Heidelberg, Im Neuenheimer Feld 360, 69120 Heidelberg, Germany
    Search for more papers by this author
    • *

      Present address: Max-Planck-Institut für Molekulare Pflanzenphysiologie, Am Mühlenberg 1, 14476 Golm, Germany.


    1. Botanical Institute, University of Heidelberg, Im Neuenheimer Feld 360, 69120 Heidelberg, Germany
    Search for more papers by this author
    • Present address: Forschungszentrum Jülich, ICG III (Phytosphäre), 52425 Jülich, Germany.

Achim Walter, (present address) Stetternicher Forst, Institut fuer Phytosphaere (ICG-3), Forschungszentrum Juelich, 52425 Juelich, Germany. Fax: + 49 2461 61 2492; e-mail: awalter@fz-juelich.de


A combined analysis of growth and metabolite composition was performed in primary roots of Zea mays L. (Var. Alexander). The seedlings were hydroponically cultivated either in pure water or in complete nutrient solution. The overall root growth performance was similar in both treatments. Yet, digital image sequence processing methods resolved, that growth distribution and oscillatory movements within the growth zone depended strongly on external nutrient availability. Metabolite concentration profiles were similar in both treatments for most investigated metabolites, indicating a thorough mobilization of nutrient resources from the seed, but concentrations of glutamine, glutamic acid, NO3, NH4+, malate and citrate showed pronounced differences between treatments. No diurnal variations in metabolite concentrations were found. Deposition rate profiles were in general more similar to relative elemental growth rate profiles than concentration profiles and were not affected by the treatment. Major ions were deposited maximally in front of the centre of growth activity, while greatest hexose deposition was found behind that. Relative to their abundance in the root growth zone, net rates of transfer from mature tissue were highest for sucrose, glutamic acid and aspartic acid, whereas glucose, fructose and most amino acids inversely showed high net rates of transfer out of the root growth zone, indicating a high catabolic rate for those substances there. NO3, but not other nutrients, was transferred to a great extent from the root growth zone to the mature tissue in nutrient solution. Overall, the results show, that a careful analysis of growth dynamics allows quantifying and interpreting a number of important flux parameters in the growing organ and that the performance of the primary root does not depend strongly on external nutrient availability.


Root length and the relative elemental growth rate distribution within the root growth zone are affected directly by a number of external parameters such as temperature (Pahlavanian & Silk 1988) or water potential (Sharp, Hsiao & Silk 1990; Wu & Cosgrove 2000). Seedling roots have been used as a model system, since their handling is straightforward and since they provide a simple, reductionistic research system. In Zea mays, the root growth zone is typically 10 mm long; maximal relative elemental growth rates (REGR) of up to 50% h−1 are found some millimetres behind the tip. There, cell division activity ceases and the cells are passed through the zone of maximal cell elongation within some hours (Silk 1984).

Although nutrient availability plays a major role in guiding the distribution of biomass between shoots and roots (Scheible et al. 1997a) and strongly affects root architecture, its impact on REGR distribution or other parameters of expansion of the individual root growth zone have not yet been investigated in detail. Since roots can actively direct their growth towards regions of higher nutrient availability (Bilbrough & Caldwell 1995; Arredondo & Johnson 1999), it is conceivable that root growth dynamics might be affected in many ways by external nutrient availability. Circumnutations are known to occur not only in growing hypocotyls or in stems of vining plants, but also in growing root tips (Head 1965; Ney & Pilet 1981, Antonsen et al. 1995). Their frequencies are reported to be correlated with fluctuations in ion uptake (Shabala & Newman 1997a, b) and might thus also be affected by external nutrient availability.

Nutrient uptake patterns across the root growth zone surface were studied extensively during the past decade for many nutrients, especially for NO3 and NH4+ (Colmer & Bloom 1998). While significant findings were made in this field, especially with the use of vibrating microelectrodes (Felle 1998), little is known about the internal flux of substances within the root growth zone (Bret-Harte & Silk 1994). As root growth zones are important sinks within the plant, a quantification of their metabolite fluxes would be beneficial for many purposes. This quantification is only possible by means of a kinematic analysis in which both the concentration of the investigated substance and the growth of the organ are measured with an adequate spatial and temporal resolution (Silk 1984). For some carbohydrates and cations, concentration and deposition profiles along the growth zone have already been investigated, but never on the background of differing external nutrient availability (Silk et al. 1986; Zhong & Läuchli 1994 , Walter, Silk & Schurr 2000).

Studies of carbohydrates and some cations, performed in constant nutrient availability, show that concentrations of certain metabolites do not differ from day to day and that the dilution effect of growth is balanced on a daily time scale by the import or deposition of substances (Silk et al. 1986; Sharp et al. 1990). Although there are a large number of diurnally regulated physiological processes with importance for root growth, it is not known whether there are diurnal changes in composition or deposition of metabolites. For example, the uptake of nitrogen (Delhon et al. 1995), the delivery of carbohydrates from the shoot (Hansen 1977) or respiratory activity (Hansen 1980) change diurnally, but it is still unclear, whether the dynamics of those processes lead to concentration changes on a smaller time scale.

Another open question is, in which way spatial profiles of metabolite concentration and deposition interact with patterns of growth activity. It is known that the pH patterns on the interface between the root growth zone and the rhizosphere are correlated with REGR distributions (Peters & Felle 1999) as well as patterns of electrical potentials there (Toko et al. 1987; Souda et al. 1990). Overall growth activity does not change in a diurnal manner (Ijima et al. 1998, Walter et al. 2002), but small-scale spatial and temporal variations of the root growth zone are just about to be unravelled. Root growth analysis with classical techniques (Silk et al. 1986) or with single image processing techniques (Beemster & Baskin 1998; Walter et al. 2000) often does not provide the necessary spatio-temporal resolution or is too tedious to reveal the complex kinematics of the expanding plant organ.

The first aim of this study was to analyse the effect of external nutrient availability on the expansion dynamics of a Zea mays L. (var. Alexander) seedling root growth zone. This was done both by comparison of steady-state treatments and by investigation of transitions. Time-lapse movies of growing roots were analysed with digital image sequence processing algorithms (Walter et al. 2002) in terms of the following parameters: (a) velocity of the root tip (vtip); (b) REGR-distribution along the root growth zone; and (c) oscillation frequency of the root tip perpendicular to the direction of growth. The second aim of this study was to examine the effects of external nutrient availability on the concentration patterns and the diurnal fluctuations of metabolites. Finally, results from growth and metabolite analysis were combined to calculate deposition patterns and to quantify the net rates of metabolite transfer between the root growth zone and the adjacent, mature tissue by a box model.


Plants, cultivation and fresh weight determination

Seeds of Zea mays L. (var. Alexander) were germinated in thoroughly watered sand pots. After 5 d, roots were approximately 2 cm long and 20 seedlings were transferred to a rhizotron. This set-up allowed optical monitoring of seedling root growth for several days and is described in more detail elsewhere (Walter et al. 2002). A plastic rhizotron base plate, inclined by 68° towards the horizontal plane, was used to hold the seedlings. Double-distilled water (‘pure’ water) or nutrient solution flushed constantly over the growing roots. The solution was collected in a plastic basin (volume 10 L) and was pumped up to the top of the base plate continuously. The stream was widened there to the width of the base plate, flowing downwards along the roots between base plate and a transparent plastic foil. The rhizotron was placed in a growth room with the lights switched on between 0900 and 2100 (150 µmol m−2 s−1 photosynthetically active radiation), 40% relative humidity and 26 °C air and water temperature during growth of the seedlings.

After 3 d incubation time, plants were harvested and analysed for their fresh weight. The apical 10 mm of 10 pooled roots were rinsed with pure water and sectioned with a multiple razor blade into millimetre-segments. The groups of homologous millimetre-segments were weighed sequentially, beginning with the most apical segments on a microbalance.

Addition of 0.1 mmol L−1 Ca2+ to pure water, as usually practised in hydroponic cultivations (e.g. Walter et al. 2000), was not performed, since (a) preliminary experiments did not show ameliorated performance of the plants, when Ca2+ was added, and (b) nutrient availability was aimed to differ as much as possible between treatments within physiological limits. In pure water cultivation, plants looked healthy and reached a final fresh weight of up to 2 g after 4 weeks incubation time. When nutrient solution was used, it contained the following amounts of nutrients in relation to nitrogen content, which was fixed at 10 mmol L−1: NO3 61.5%, NH4+ 38.5%, K+ 65%, PO43– 13%, Ca2+ 7%, Mg2+ 8.5%, SO42– 9%, Fe 0.7%, Mn 0.4%, Cu 0.03%, Zn 0.06%, B 0.2%, Mo 0.07%, Na+ 0.034%, Cl 0.033%. The nutrient solution was buffered at pH 6 (via K2HPO4 and KH2PO4) and had an osmolarity of 17 mosmol L−1– a value far below the osmolarity, which might lead to salt stress. Preliminary studies with diluted stock solution (nitrogen content 0.1 and 1 mmol L−1, respectively) led to intermediate results for root and shoot growth; hence the strength mentioned above was chosen.

Analysis of velocities and REGR by digital image sequence processing

Roots of mean length and straight habit were chosen for image sequence acquisition. The charge-coupled device (CCD)-camera (Sony XC75; DBS, Bremen, Germany), equipped with infrared diode-fields (940 nm) for constant illumination of the scene by day and night and with appropriate filter sets (Schott RG9; Schott, Mainz, Germany) was mounted in perpendicular orientation to the base plate onto an arrangement of two moving stages attached to a vibration free stand. Images (640 × 480 pixels) corresponding to an area of approximately 14 × 10 mm were acquired every minute from 2 to 3 d incubation time of the seedlings in the rhizotron on. Between 10 and 20% of the visible surface of the root growth zone showed enough structure for determination of velocities and relative elemental growth rates by the digital image sequence processing (DISP)-method. Tracking of the root tip with the camera and evaluation of the image sequences are described in more detail elsewhere (Schmundt et al. 1998; Walter et al. 2002).

Manual tip growth analysis

In addition to extraction of the tip growth velocity (vtip) from the DISP data, the vtip of each seedling was followed by marking the position of the tip at the outside of the covering foil in intervals between 4 and 24 h and measuring the distances between marks with a ruler.

Anions and cations

Concentrations of cations, organic and inorganic anions were analysed by capillary electrophoresis (Spectra Phoresis 1000; ThermoQuest Darmstadt, Germany) according to Bazzanella et al. (1997). The electrolyte for cation determination contained 6 mm imidazol (pH 4.5) and 2 mm 18-Krone-6. Separation was performed in a glass capillary (inner diameter 75 µm, length 43 cm) at 20 kV at a capillary temperature of 20 °C with 3–5 s injection time. The electrolyte for anion determination contained 7.5 mm salicylic acid, 15 mm TRIS, 0.5 mm DoTAOH, 0.600 mm Ca(OH)2. Separation was done on a glass capillary (length: 76 cm, inner diameter 75 µm) at −28 kV and a capillary temperature of 25 °C with 3–5 min injection time. Electropherograms obtained by indirect UV detection at 214 nm were analysed by a software package (TSP data system; ThermoQuest).

Carbohydrates and amino acids

Glucose (Glc), fructose (Fru) and sucrose (Suc) were analysed according to Jones, Outlaw & Lowry (1977) by enzymatic analysis in ethanolic extracts. Amino acids were determined by high-performance liquid chromatography after derivation of primary amino acids with o-phthalic acid dialdehyde (OPA). The derivation reagent contained 25 mg OPA in 500 µL methanol, 4.5 mL 800 mm borate buffer (pH 10.4), 50 µL mercapto-propionic acid. Derivation of 35 µL of diluted xylem sap was performed automatically by the autosampler (Kontron 465; Kontron, Eching, Germany). Separation was done after injection of 20 µL on a Hypersil ODSII column (Knauer, Berlin, Germany) at a flux rate of 0.8 mL min−1 with a step mixing gradient. Detection of separated derivated amino acids was done fluorometrically (SFM 25; Kontron) with 330 nm excitation at 450 nm (emission wavelength). Chromatograms were quantitatively evaluated by a data system (Kontron).

Concentrations are usually expressed on a fresh weight basis in this manuscript. For calculation of deposition rates, they have to be transformed to a segmental length basis, taking into account the differing fresh weight of the individual segments.

Deposition rates

Estimation of local deposition in growing tissues, which is influenced by fluxes and metabolic processes, needs to take into account concentration as well as expansion profiles. The deposition rate of a specific substance in growing tissue is given by the continuity equation (e.g. Silk et al. 1986):


where D is the local deposition rate in nmoles mm−1 root length h−1, C is the local density of the substance, expressed here as nmoles (of any metabolite) per mm root length, t is time, x is the distance from the root tip and v(x) is the local growth velocity in the root coordinate system.

Total amounts and total deposition rates

Total amounts and total deposition rates for the entire root growth zone were calculated for each metabolite by taking the sum of the concentrations or deposition rates for all 10 segments (integral over the growth zone). Those values were then summarized for all metabolites to give total amounts and total deposition rates for soluble sugars, amino acids, anion and cation equivalents. All ion equivalents were calculated under the assumption of constant pH within cellular compartments and constant compartmentation.

Net rate of transfer from growing to the adjacent, mature tissue

If the average concentration of a metabolite in the root growth zone does not change with time, the growth zone must gain the same amount of this metabolite as it loses. It will be shown (result section), that this condition is met here. Gain or loss of a metabolite into or out of the growth zone can occur via three different ways: (a) direct exchange with the nutrient solution; and/or (b) metabolic conversion of substances within the growth zone; and/or (c) transport from or to the adjacent mature tissue over the virtual interface at the end of the growth zone. This transport is performed by the vascular system, by apoplastic and symplasmic transport and by the passive transport of cells. The last mode of transfer (c) can be calculated from data acquired in this study. It represents the relative fraction of the investigated metabolite which is depleted from or acquired by the root growth zone across this interface per hour. From our data, it is not possible to distinguish between the different pathways that contribute to this net relative rate of transfer (vascular, apo/symplasmic, passive transport). For simplicity, we propose to call it ‘net rate of transfer’; its relative character is clearly expressed by its unit (% h−1). The net rate of transfer is given by the product of the relative difference (%) between the concentrations in the growth zone and the adjacent tissue times the mass transfer rate (mass flux (1/h), expressed as fraction of total growth zone mass) across this interface.


The concentration in the mature tissue adjacent to the root growth zone is approximated by the value of the final segment of the root growth zone (c10: amount per mg, m10: fresh weight of segment 10). This is admissible, since all concentrations reach an asymptotic value at the end of the growth zone (see Fig. 6). cRGZ is the average concentration (total amount divided by total fresh weight), mRGZ the fresh weight of the root growth zone. vtr is the transfer velocity across the interface, which equals vtip, as the length of the growth zone is constant with time.

Figure 6.

Concentration profiles of soluble sugars, amino acids, anions and cations. Exemplarily, soluble sugars are shown in µmol g−1 fresh weight (top row panels) and in µmol per segment (second row panels, calculated with root segmental fresh weight densities given in Fig. 2). Mean values and standard deviations from four replicates per treatment; each sample contained 10 1 mm segments pooled from different roots. ‘Total amino acids’ is the sum of Glu, Gln, Asp, Asn, Ser, Val, Gly, Ala, His, Phe, Tyr, Leu. Harvests were done at 1500 h.


General aspects of growth

Root elongation was significantly slower in nutrient solution (Fig. 1a): two to three days after incubation of the seedlings in the rhizotron, the average value for vtip in water was 2.6 mm h−1. In nutrient solution, an average value of 1.8 mm h−1 was reached (six independent populations of 20 seedlings for each treatment). Despite substantial variability between individuals, the differences were significant. Tip growth velocity increased during the first day after incubation (Fig. 1b). Diurnal fluctuations of vtip occurred neither in pure water nor in nutrient solution (data not shown here).

Figure 1.

Root tip growth velocity. (a) Mean values and standard deviation from day two to day three after incubation (120 seedlings per treatment). (b) Time series from two typical seedling populations (20 individuals per treatment).

Fresh weight of the growth zone was higher in nutrient solution than in pure water (Fig. 2). The average fresh weight per growth zone was 6.4 mg in nutrient solution and 5.3 mg in pure water. The fresh weight of millimetre-segments increased from the root tip to the base of the growth zone, due to the conical shape of the root growth zone. Segments from roots of higher external nutrient availability showed higher fresh weights than segments from cultivation in water throughout the whole growth zone.

Figure 2.

Fresh weight distribution within the root growth zone. Mean values and standard deviations are given for 1 mm segments from 100 roots per treatment.

Mass flux from the root growth zone to the differentiated zone of the primary root is given by the product of the fresh weight of the most basal segment of the growth zone and the growth velocity vtip. This product was 1.3 mg h−1 for roots from both nutrient solution (0.7 mg mm−1 × 1.8 mm h−1) and pure water cultivation (0.5 mg mm−1 × 2.6 mm h−1). Therefore, roots from both cultivation conditions grew with the same ‘fresh weight growth rate’ of 1.3 mg h−1. Yet, calculated as a relative share of the root growth zone, the relative growth rate of roots from water cultivation was higher than that of roots from nutrient solution, since the former were growing by 1.3 mg h−1/5.3 mg = 25% h−1; the latter only by 1.3 mg h−1/6.4 mg = 20% h−1.

As a consequence of the comparable fresh weight growth rates, roots from both treatments had a similar fresh weight at the end of incubation (Table 1). Shoot : root ratio was higher in plants from nutrient solution, due to a strongly increased fresh weight of the shoot. After 3 d, shoots from seedlings in nutrient solution had twice the fresh weight (300 mg) as shoots from water cultivation. Given an initial shoot fresh weight of approximately 60 mg at incubation, shoots grew with approximately 1 and 3 mg h−1 in pure water and nutrient solution, respectively, during the 3 d incubation.

Table 1.  Shoot : root ratio of plants in the different treatments
Growth mediumShoot fresh
weight (mg)
Root fresh
weight (mg)
Shoot : root
  1. Mean value and standard deviation for fresh weight of 20 plants per treatment is given.

Pure water129 ± 14190 ± 90.68 ± 0.07
Nutrient solution305 ± 18217 ± 131.41 ± 0.05

Longitudinal distribution of REGR

The length of the growth zone was approximately 10 mm (Figs 3a & 5e). It did not differ significantly between treatments. Maximal REGR was found in a ‘plateau region’ 3–6 mm behind the root tip, often showing two separate maxima. The shape of the REGR profile differed with time and plant; leading to masked peak shapes when averages over time or over replicate roots were calculated. As an average value, maximal REGR was found 4 and 5 mm behind the tip for roots grown in pure water and nutrient solution, respectively. REGR was higher in roots from pure water throughout the whole growth zone with an average maximum of 47% h−1 (nutrient solution: 34% h−1). Differences in growth activity were least pronounced apical from the ‘plateau region’, where meristematic activity is found. In the basal part of the growth zone from 7 to 10 mm behind the root tip, roots from nutrient solution showed only 50–70% of the growth activity of roots from water cultivation.

Figure 3.

REGR-distribution along the root growth zone. (a) Mean values from six roots per treatment (image sequences between 1 and 24 h). (b) One-hour average values from typical roots.

Figure 5.

Distribution of transversal velocity along the root growth zone. (a) in water, and (b) in nutrient solution. Time interval between successive images: 1 min, image length: 14 mm. Transversal velocity of the root tip in different treatments: (c) in water, and (d) in nutrient solution (circles show average values of 12 min waves). Frequencies were calculated via Fourier analysis, performed with the software Origin (Origin Laboratory Corporation; Additive, Friedrichsdorf, Germany) . (e) Length of the root growth zone and colour-coded REGR distribution.

When two separate peaks were present, their relative height was affected by the treatment: In water cultivation, the apical one was usually higher, whereas in nutrient solution there was a higher probability for the basal peak to show the absolute maximum of expansion growth activity (Fig. 3b). A sign test, applied to 1 − h-average values of REGR distributions of roots from both treatments (Table 2), confirmed this observation (significance level P > 0.99).

Table 2.  Peak shape of REGR-distributions
Growth mediumSingle peakApical peak
Basal peak
  1. ‘Single peak’ indicates, that only one REGR-peak was recognizable (in 1 h average values). ‘Apical peak higher’ and ‘Basal peak higher’ means that two peaks with the respective weighting were present.

Pure water 741 5
Nutrient solution151332

The dynamics of the REGR distribution were very complex. As an example, the change of a typical REGR profile during the addition of nutrient solution is shown (Fig. 4). Two hours after addition of nutrients, growth activity was reduced mainly in the apical and basal flanks of the REGR distribution. After 5 h, two separate peaks of growth activity had developed. The basal peak was higher than the apical peak. The two peaks were still visible 7 and 15 h after addition of nutrients, but not as clearly as before. The experiment was repeated three times with comparable results. A complete investigation of short-term changes in the REGR profile is outside the scope of this manuscript and will be undertaken in future studies.

Figure 4.

REGR-distribution along a root growth zone during transition from water to nutrient solution. The root was exposed to water during the first 2 h of the experiment and to nutrient solution thereafter (1 h average values).

Transverse growth movements of the root tip

Roots were able to move freely in transversal direction during growth in the hydroponic setup. Often, this happened in a periodic manner (Fig. 5). The transverse movement of the root tip was not due to displacement of the root with the flowing medium, but to the changing curvature of the growth zone. The oscillating movement did not originate from a single point but covered the entire growth zone with diminishing activity towards the base of the growth zone. Maximal transverse velocities of up to 2 mm h−1 were measured near the root tip. The differentiated root had a straight shape and did not show any signs of an irregular morphology.

Nutrient availability also affected this feature of root growth dynamics. In pure water, transverse oscillation frequencies of approximately 1 h−1 were observed (period length: 40–70 min). In nutrient solution, frequencies were significantly and consistently higher with typical values of 5 h−1 (period length: 10–15 min). Under these conditions, the movement was often superimposed by a second, lower frequency component of approximately 1 h−1– the frequency corresponding to the one observed in pure water. In experiments where root growth was followed before and after the addition of nutrient solution, the change in transversal oscillation frequency occurred immediately after addition of nutrient solution.

Distributions of metabolite concentrations along the root growth zone

Pronounced concentration gradients were found for all three soluble carbohydrates (Fig. 6): Highest Suc concentrations were found in the root tip (up to 30 µmol g−1 FW); highest Glc and Fru concentrations were found at the base (120 and 40 µmol g−1 FW). Base–tip gradients were significantly stronger in roots from the pure water treatment. This effect of the treatment was not significant, when concentrations were expressed on the basis of root segmental length.

Total amino acids showed a homogeneous profile throughout the growth zone with roughly 20% higher concentrations in roots from nutrient solution (Fig. 6). Most amino acids showed comparable concentrations in roots from different treatments (Table 3). The difference between the two treatments mainly resided with glutamine (Gln), which was slightly compensated by glutamic acid (Glu) (Fig. 6). Gln showed an almost homogeneous distribution in roots from pure water but clearly decreasing concentrations from the tip (meristem) to the base of the growth zone in roots from nutrient solution. The meristematic concentration of Gln was more than twice as high in roots from nutrient solution in comparison with roots from pure water. Meristematic Glu concentration was lower in roots from nutrient solution.

Table 3.  Concentrations of major amino acids in roots from both treatments
MetaboliteConcentration (µmol g−1 FW)
Pure waterNutrient solution
  1. Harvests were done at 1500 h. Tip: mean value and SD of segment 1 and 2 (n = 6). Base: mean value and SD of segments 8 to 10 (n = 6).

Tot. amino acids 40 ± 7 51 ± 10 50 ± 8 51 ± 9
Gln8.7 ± 3.0 12 ± 3 24 ± 5 14 ± 2
Glu7.8 ± 2.41.7 ± 0.55.1 ± 1.21.8 ± 0.3
Asn3.2 ± 0.89.0 ± 1.86.0 ± 3.0 12 ± 4
Asp7.2 ± 3.10.9 ± 0.45.1 ± 1.10.9 ± 0.2
Ser2.2 ± 0.63.0 ± 0.72.1 ± 0.42.6 ± 0.8
Val1.1 ± 0.33.8 ± 0.81.3 ± 0.22.8 ± 0.6
Gly1.3 ± 0.44.2 ± 0.91.4 ± 0.34.2 ± 1.0
Ala3.3 ± 1.15.8 ± 1.62.6 ± 1.15.5 ± 1.0
His2.7 ± 1.42.5 ± 0.50.3 ± 0.21.1 ± 0.3
Phe0.4 ± 0.11.8 ± 0.40.3 ± 0.11.6 ± 0.3
Tyr0.2 ± 0.051.3 ± 0.30.2 ± 0.051.4 ± 0.3
Leu0.2 ± 0.051.5 ± 0.40.3 ± 0.071.4 ± 0.4
GABA0.8 ± 0.50.5 ± 0.20.3 ± 0.10.4 ± 0.1

Anions and cations showed stronger differences between roots of the two treatments (Fig. 6). The nitrogenous compounds NO3 and NH4+ had higher concentrations in roots from nutrient solution. NO3 was even below the detection limit in roots from pure water. In roots from nutrient solution, its concentration increased from 2 µmol g−1 FW in the meristem to more than 30 µmol g−1 FW at the basal end of the growth zone. NH4+ was found in roots from both treatments, but its concentration in roots from nutrient solution exceeded that in roots from pure water by a factor of two and by a factor of three in the meristem (40 and 15 µmol g−1 FW, respectively).

In contrast to this, the anions citrate, malate and Cl (which were not all supplied with the nutrient solution) showed higher concentrations in roots from pure water. The malate and Cl concentration profiles differed only in the zone of elongation. Despite strong differences in the availability from external solution, concentrations and profiles of SO42–, PO43–, K+, Ca2+, Mg2+ and Na+ did not differ significantly between treatments.

The high concentrations of Na+ and Cl were correlated with high amounts of those elements in the seeds (data not shown).

No significant diurnal effects were found for any of the investigated metabolites in the temporal variation of concentrations (e.g. Glc, Fig. 7). Strongest deviations from the temporal homogeneity were found for soluble sugars in roots from pure water with slightly increased nocturnal concentrations, but even in this case (Fig. 7), the deviations of measurements at two consecutive days were minimal and lay within the standard deviation of the replicates.

Figure 7.

Concentration profile of Glc at four different points in time. Mean values and standard deviations from two replicates per time; each sample contained 10 1 mm segments pooled from different roots.

Distribution of deposition rates

As a consequence from the absence of diurnal variations in growth and metabolite composition, the temporal term in the equation of continuity (see Materials and methods) was neglected for calculation of deposition rates. The necessary transformation of concentrations from fresh weight to segmental length basis changed concentration profiles slightly (compare sugars, Fig. 6), but did not alter distinct gradients, that can be extracted from Fig. 6 and Table 3.

From the continuity equation, it follows that metabolites with strongly increasing concentrations from tip to base have a maximum deposition rate behind the zone of maximal elongation, whereas substances with strongly decreasing concentrations have a maximum of deposition rate apical of the zone of maximal elongation. Hence, the slope of a concentration profile indicates, whether a metabolite is deposited maximally during the accelerating or the decelerating phase of cell expansion. A decreasing concentration profile from tip to base indicates that the deposition maximum occurs prior to the time of maximal cell elongation; an increasing concentration profile indicates that maximal deposition occurs after the cell has reached its maximal REGR.

Calculation of deposition profiles for extreme examples of actually measured concentration profiles showed, that the maximal deposition of practically all investigated metabolites occurred within the plateau region of the REGR distribution (Fig. 8). K+ showed a strongly decreasing; NO3 and Glc showed strongly increasing concentration profiles from tip to base. Hence, K+ was deposited predominantly apical, NO3 and Glc basal of the REGR-maximum. Gln was the metabolite that showed the strongest deviation in concentration profile between roots from nutrient solution and pure water. In roots from pure water, its maximum of deposition rate occurred later than in roots from nutrient solution.

Figure 8.

Profiles of deposition rate for K+, NO3, Glc and Gln. The profiles were calculated from the continuity equation with relative elemental growth rates and root segmental fresh weights as given in Figs 2 and 3.

Integrated values of concentrations and deposition rates

The total amount of soluble sugars was equal in both treatments (600 nmol) with minor shifts in the relative contribution of the three sugars (Fig. 9a, Table 4). Total deposition rate differed between treatments and was higher in pure water.

Figure 9.

Total amounts and total deposition rates of the different investigated metabolite classes (sum of all 10 segments that comprise the root growth zone). (a) soluble sugars; (b) amino acids; (c) anion charge equivalents; and (d) cation charge equivalents. W, pure water; NS, nutrient solution.

Table 4.  Total growth zone metabolite amounts, metabolite concentrations of the basal root growth zone segment, growth zone interfacial fluxes and net transfer rates
MetabolitePure waterNutrient solution
(nmol mg−1)
(nmol h−1)
Net rate
of transfer
(% h−1)
(nmol mg−1)
(nmol h−1)
Net rate
of transfer
(% h−1)
  1. Amount: sum of all 10 segments that comprise the root growth zone. c10: metabolite concentration of the basal root growth zone segment. Flux: product of average growth velocity of the root tip and metabolite concentration (nmol mm−1) in segment 10. Net rate of transfer: interfacial metabolite transfer from growth zone to adjacent, mature tissue.

Biomass  5.3 mg  0.5 mg mm−1  1.3 mg h−1 24.5  6.4 mg 0.7 mg mm−1  1.3 mg h−1  20.0
Glc425119156 12456 89114   5
Fru129 39 51 15111 20 25   3
Suc 42  4.0  5.3−12 33  2.4  2.9−11
NO3 0  0  0123 33 42  14
Citrate3– 18  3.5  4.6  1  0  0  0
Malate2– 42 10 13  7 17  1.5  1.9  −9
Cl 71 15 19  3 42  4.6  6.0  −6
SO42– 17  3.3  4.3  1 22  4.1  5.2   4
PO43– 36  5.7  7.4 −4 50  7.4  9.8  −1
K+416 63 83 −5435 65 84  −1
Ca2+ 17  4.1  5.3  7 21  2.5  3.1  −5
Mg2+ 19  2.6  3.3 −7 29  3.0  3.8  −7
Na+782136180 −2988132169  −3
NH4+ 72 13 17 −1182 24 31  −3
Li+ 16  2.4  3.2 −5 13  1.4  1.8  −6
Tot. amino acids225 51 66  5329 49 63  −1
Gln 60 12 15  1112 12.3 16  −6
Glu 16  1.5  2.0−12 17  1.6  2.1  −8
Asn 33  9.3 12 12 51  9.5 12   4
Asp 11  0.8  1.1−15 11  0.8  1.0−11
Ser 13  3.0  3.8  5 15  2.3  3.0   0
Val 13  3.8  4.8 13 15  2.8  3.5   4
Gly 15  4.3  5.6 13 23  4.3  5.5   4
Ala 22  6.5  8.4 14 29  6.1  7.7   7
His 10  2.3  2.9  5  6.0  1.3  1.6   7
Phe  6.3  1.9  2.5 15  7.4  1.4  1.8   5
GABA  4.4  1.5  1.9 19  7.5  1.4  1.8   4

For amino acids, the total amounts differed between treatments (Fig. 9b, Table 4) with the most prominent effect in Gln (110 and 60 nmol in nutrient solution and pure water, respectively). In contrast to the relations in soluble sugars, total deposition rate did not differ between treatments for all amino acids (60 nmol h−1). For Gln, the shift in orientation of the concentration profile (Fig. 6) led to a changed distribution profile (Fig. 8), but the total deposition rate within the entire root growth zone was nearly constant in both treatments (Fig. 9b), since deposition profiles differed only in the meristematic region.

The total amount of anion equivalents was lower in roots from pure water compared to roots from nutrient solution (Fig. 9c, Table 4). The total deposition rate of anion-equivalents did not differ between treatments and reached a value of 90 nmol h−1. The effect of increased NO3, SO42– and PO43– deposition in nutrient solution was compensated totally by the effect of decreased deposition of Cl, citrate and malate.

Cation equivalents also showed a lower total amount in pure water (Fig. 9d, Table 4). Total deposition rates were equal within both conditions (300 nmol h−1). The strongest differences for individual cations occurred for NH4+ and Na+. Total NH4+ deposition rate was higher in roots from nutrient solution; this effect was matched by a lower total deposition rate of Na+.

Net rate of transfer from growing to mature tissue

In roots from pure water, the net rate of transfer of Glc and Fru from the growth zone to the mature tissue was markedly positive with 12% h−1 (Glc) and 15% h−1 (Fru), respectively (Fig. 10, Table 4). The net rate of transfer is exemplarily calculated for Glc (values from Table 4): 100 × [119 nmol mg−1/(425 nmol/5.3 mg) − 1] × [(0.5 mg mm−1× 2.6 mm h−1)/5.3 mg] = 12% h−1. In roots from nutrient solution, the net rates of transfer were diminished to 5% h−1 (Glc) and 3% h−1 (Fru). For Suc, the net rate of transfer was negative with values of −12% h−1 and −11% h−1 in pure water and nutrient solution, respectively. For amino acids, net rates of transfer were mainly positive in roots from pure water and showed more neutral values in roots from nutrient solution. Glu, aspartic acid (Asp) and Gln differed in this respect from the majority of the amino acids, showing the lowest (most negative) values of all amino acids in both treatments; the values for Glu and Asp were comparable to the values reached by Suc. The investigated anions and cations showed in general more neutral net rates of transfer than Suc, Glu and Asp. Strongest effects were seen for NO3 and malate. NO3 was not present in roots from pure water but was transferred to a high extent from growing to mature tissue in roots from nutrient solution (14% h−1). For malate, the net rate of transfer was positive in roots from pure water (7% h−1) and negative in roots from nutrient solution (−9% h−1). This was the most pronounced difference between net rates of transfer in roots from both treatments of all investigated metabolites. The high positive value for Ca2+ in the pure water treatment might be an artefact, since that concentration profile did not show an asymptotic shape and hence calculation of the net rate of transfer as described in material and methods might not be appropriate.

Figure 10.

Net rate of transfer across the interface between root growth zone and adjacent, mature tissue. The schematic drawing shows the direction of the strongest net rates of transfer that were found for the major metabolites of this study.


Integrated values of concentrations and deposition rates

For the first time in this study, the total amounts and total deposition rates of the investigated metabolite classes were calculated in order to estimate their putative role as set points for root growth performance. The total amount of soluble sugars was not affected by external nutrient availability, whereas the total deposition rates, but not the total amounts, were the same in both treatments for amino acids, anion and cation equivalents. We propose, that for anions, cations and amino acids, the total deposition into a root growth zone is a property, which is kept constant as long as the endosperm reservoir allows. Soluble sugars do not depend in the same way on the endosperm reserves as the other substance classes. The fact, that their total deposition rate was higher in pure water than in nutrient solution may be caused by a decreased usage of synthesized carbohydrates by the shoot as a consequence of low shoot growth there. The detected total amount of sugars in growth zones of both treatments might represent the physiological limit for them. Any surplus of soluble sugars might contribute to the mucilage secreted via the calyptra. The difference of anion and cation equivalents was neither zero for the total amount nor for the total deposition rates. This charge ‘imbalance’ must have been compensated by, for example, the synthesis of negatively charged proteins that are abundant in the cell wall (Marschner 1998) or by exchange of protons with the rhizosphere (Weisenseel, Dorn & Jaffe 1979; Toko et al. 1987).

It would be desirable in future studies to investigate concentrations and deposition rates also in treatments that are not as harsh as the ones deployed in this study. Furthermore, it would be beneficial to analyse the content of all investigated metabolites not only in the root growth zone, but also in the external treatment solutions before and after cultivation.

Net rate of transfer from growing to mature tissue

Net rates of transfer over the basal growth zone interface were also calculated for the first time in this study. The results showed, that within the root growth zone, catabolism of Glc and Fru was higher than anabolism. The smaller rates found in roots from nutrient solution might have been caused by an increase in respiratory activity (anabolism) there. The comparable negative net rates of transfer for Suc in both treatments point to a treatment-independent demand for metabolic processes or for excretion of carbohydrates via the calyptra. For most amino acids, root growth zones showed positive net rates of transfer, indicating a high production there. Gln was affected strongest by external nutrient availability, reflecting the differences in nitrogen assimilation between treatments, which are discussed below. For NO3, the net rate of transfer was positive in nutrient solution, indicating that NO3 was taken up to a significant degree by the growth zone itself. The difference between treatments for malate indicates, that malate can be synthesized or exported from the root growth zone as needed, depending on the external conditions (Marschner 1998). The small negative net rates of transfer for most investigated nutrients point to a continuous loss of those nutrients or on metabolic conversion, which can take place for some of the investigated nutrients (e.g. PO43–, NH4+).

Overall growth performance of root and shoot

The observed range of root tip elongation velocities was comparable to results obtained with other cultivation systems (Silk, Lord & Eckard 1989; Felle 1998). Nevertheless, the fresh weight of the root growth zone was a factor of two lower than in studies with soil-grown roots of Zea mays (Silk et al. 1986; Walter et al. 2000). This was possibly due to the decreased mechanical impedance in hydroponic cultivation, which leads to the formation of some aerenchymatic volume mainly in basal root zones, but also near the root apex (Konings & Verschuren 1980; Drew, He & Morgan 1989) and also to somewhat thinner roots. The proportion of aerenchymatic volume is reported to be higher in roots from pure water compared to nutrient solution, which might explain the lower fresh weight of the growth zone there.

Although root elongation was slower in nutrient solution, they gained the same amount of fresh weight per time as roots from pure water. The faster growth in low nutrient availability, which leads to longer, less rigid roots, might be an advantage in a natural system, since it increases the capacity to explore new soil regions. High external nutrient availability had a strong effect on shoot growth of the young seedlings. NO3, being the only macronutrient for which an uptake from nutrient solution could be indirectly shown via its net rate of transfer, is known to regulate shoot : root ratio (Scheible et al. 1997a) and branching processes of the root (Zhang & Forde 2000). It might also be the key element that induces either a stronger elongation of roots or a higher biomass deposition per root segment. The mechanisms which translate the NO3 signal into a phenotye in the case of shoot : root-ratio regulation (metabolism of organic acids and amino acids, Scheible et al. 1997b) are indirectly also shown to be affected by the treatments of this study. Net rates of transfer, and hence metabolism, of malate and most amino acids in the root growth zone were affected markedly by external nutrient availability.

Longitudinal distribution of growth

Since growth zone length was not affected by the treatment, the lower elongation velocity in nutrient solution consequently led to lower maximal values of REGR distribution there. This coincides with findings for different temperature regimes (Pahlavanian & Silk 1988; Walter et al. 2002), in which elongation velocity and maximal REGR value were also correlated.

The spatial distribution of REGR was in general very similar to the distributions already published in literature (Goodwin & Stepka 1945; Pahlavanian & Silk 1988; Morris & Silk 1992) with the exception of the transient occurrence of two separated growth peaks within the region of maximal growth activity (Walter et al. 2002). Since the two peak regions reacted differentially to a change in external conditions, we speculate, that their reaction indicates changes in physiological processes or anatomical structures. One possible candidate for such a feature is the differentiation of protophloem and protoxylem . In anatomical investigations of Phleum pratense it was shown that the formation of protophloem begins between the root tip and the zone of maximal growth activity and that the formation of protoxylem between the initiating region of protophloem and the zone of maximal growth activity starts a bit later (Goodwin & Stepka 1945). In the latter zone, both tissues are clearly distinguishable. An earlier differentiation of protophloem in pure water (compared with nutrient solution) could be connected with higher expansion rates in the zone of protophloem-formation (apical REGR peak). It is known, that phloem unloading within the root growth zone occurs symplastically directly into the cells of this zone (Farrar, Minchin & Thorpe 1995; Wright & Oparka 1997). A difference between treatments in the demand for protophloem and protoxylem is supported by the data of this study, as phloem-related total deposition of Suc is higher in pure water than in nutrient solution while, for example, the export of NO3, which occurs predominantly via the xylem is indirectly shown (via net rate of transfer) to be higher in nutrient solution than in pure water.

Transversal oscillations of the growth zone

A change of nutation frequency in response to external nutrient availability has not been reported previously. In this study, oscillations with low frequencies in pure water might have been coupled to internal, longitudinal ion fluxes, mediated by xylem and phloem. It is conceivable that oscillations with high frequencies were suppressed by the lack of an external nutrient supply.

It is known, that transverse oscillations of the root tip occur as a consequence of gravitropic reactions (Barlow et al. 1993) or circumnutations (Head 1965; Antonsen et al. 1995). The biophysical regulation of curvature movements in tropic reactions involves differential growth activities of different regions within the growth zone (Buff, Baake & Sievers 1987). Recently, the enhancement of lateral growth effects by enhanced expression of K+-transporter genes was shown to be an important regulatory mechanism in gravitropic reactions (Phillipar et al. 1999). This might also explain the change of nutation frequency with external ion availability.

Evidence for the coupling between nutation frequency and ion uptake kinetics in the root growth zone is also supported by a correlation between the periods of circumnutation frequencies of maize roots and the frequencies of their exchange of protons and Ca2+ ions with the surrounding nutrient solution (Shabala & Newman 1997a). A model describing root circumnutations associated with growth waves propagating around the root circumference, which are coupled to ion fluxes from the external solution was proposed (Shabala & Newman 1997b).

The existence of oscillating ion fluxes is also tightly connected with oscillations of the electrical potential difference around the root growth zone (Hecks, Hejnowicz & Sievers 1992), for which period lengths of 10 min were detected in different species (Lepidium and Phaseolus; Souda et al. 1990). It has long been known that the direction of growth is affected by the electrical field around the root growth zone (electrotropism, Brauner & Bünning 1930), which itself acts as an electrical dipole (Weisenseel et al. 1979; Toko et al. 1987). The connection between pulsing growth activity and ion fluxes or oscillations of the electrical field on a root flank might be mediated by an oscillatory tension of membranes, which induces oscillating activation of tension-inducible ion channels (Cosgrove & Hedrich 1991).

Spatial distribution of metabolites

The concentration profiles for Glc, Fru and Suc are very similar to results previously published (Sharp et al. 1990). The finding that the ratio of Suc to hexoses declines from the root tip to the base of the growth zone is in agreement with measurements of invertase-activity (Toko et al. 1987), which is minimal within the meristem and increases towards the end of the growth zone. Hence, the highest Suc : hexose-ratio is expected within the meristem. Profiles and absolute values of all soluble sugars, of most amino acids and even of most investigated macronutrients (PO43–, SO42–, K+, Ca2+, Mg2+ and Na+) are not affected significantly by the external nutrient availability. Thus, it can be concluded, that mobilization of all metabolites that are required as root growth substrates can be performed from the endosperm reserves alone during the analysed phase of seedling development.

Amino acid profiles show that Gln and Glu are affected by the treatment. Although Gln showed an elevation in its meristematic concentration by a factor of two in pure water cultivation compared to cultivation in nutrient solution, meristematic Glu concentration was lower in roots from pure water compared to nutrient solution. Those two meristematic effects can be interpreted as direct physiological responses to the differing external nutrient availability: In high nutrient availability, assimilation of nitrogen takes place within the meristem itself, whereas in pure water, there is only an import of already synthesized nitrogen compounds (Huppe & Turpin 1994). At low assimilation of nitrogen in a tissue, the amination of Glu by glutamin-synthetase is slow and hence the concentration of Glu is high, while the concentration of Gln is low. When nitrogen is available externally in the form of NH4+, the activity of glutamine-synthetase and glutamine-2-oxo-glutarate-amino-transferase is rising. As a consequence, the concentration of the substrate Glu is lowered while the concentration of the product Gln is increased. The effect of external nutrient availability on concentrations of NO3 and NH4+ is not surprising, as it has been shown recently, that NO3 and NH4+ are taken up via the whole root growth zone surface (Colmer & Bloom 1998). The almost linear increase of NO3 concentration along the growth zones in nutrient solution might also reflect increasing vacuolization of the cells as vacuoles are the predominant compartment storing NO3.

The concentration profiles of K+, Mg2+ and Ca2+ decreased from the root tip towards the base of the growth zone. This observation is consistent with the results of many other studies (Scott, Gulline & Pallaghi 1967; Silk et al. 1986; Walter et al. 2000). The high Ca2+ concentrations in roots from pure water show, that Ca2+ must have been translocated from the seed, most probably via apoplastic transport. Ca2+ is not very tightly bound to the cell wall, which makes it ‘exchangable but not removable’ (Clarkson 1984; Cramer et al. 1987). Other studies showed as well, that the Ca2+ found in root growth zones of young seedlings must not be supplied from the external solution alone (Walter et al. 2000).

The concentrations of citrate and malate were lower in roots from nutrient solution compared to pure water. Differences in malate concentration can be explained by the role of malate for the regulation of charge balance and ion uptake in more natural growth conditions, where malate is exuded into the rhizosphere to increase the availability of PO43– (Martinoia & Rentsch 1994) and NO3 (Touraine, Muller & Grignon 1992). With high probability, such an exudation and the corresponding synthesis of malate also occurred in the nutrient deficiency conditions of this treatment. Moreover, it is known, that malate is synthesized on demand in the root in order to compensate charge surplus, that results from the imbalanced uptake of anions and cations (Osmond & Popp 1983; Martinoia & Rentsch 1994).

Diurnal changes

Contrary to other observations that many physiological processes within the root show diurnal rhythmicity, e.g. respiratory activity (Hansen 1980), NO3 uptake and assimilate influx (Hansen 1977), our measurements did not indicate diurnal changes in metabolite concentrations within the root growth zone. It has to be concluded that either metabolism buffered the metabolite concentrations or that the export rate of certain metabolites was very closely coupled with the import rate of nutrients and assimilates. The regulation of this striking homogeneity is largely unexplored and presents an interesting aim for future studies, elucidating the dynamics of metabolism and transport, in which high-resolution growth measurements and transport concepts such as the net rate of transfer will hopefully play a vital role.

Profiles of deposition rates

The fact that practically all deposition rate profiles, but no concentration profile, showed a similar shape to the growth rate profile indicates very strongly that deposition rates and not absolute metabolite concentrations are the properties that are regulated during the growth process. From a biophysical point of view, growth requires a slight imbalance between turgor and cell wall extensibility (Lockhart 1965). The data of this study indicates, that this imbalance is achieved biochemically through differences between the position of the maximal deposition rates of ions (increasing the turgor) and hexoses (acting as substrates for cell wall and thus decreasing cell wall extensibility). While the major ions were deposited in front of the centre of growth activity, the maximum of hexose deposition lay behind the position of maximal growth activity.


External nutrient availability affects many dynamic parameters of root growth, such as elongation velocity, REGR distribution and transversal oscillation frequency, but does not change biomass gain of the seedling root. Ameliorated access to external nutrient sources favours shoot growth. This reaction is probably mediated by uptake of NO3 and by changes in the fluxes of some sugars, amino acids and organic acids. Fluxes and concentrations of individual metabolites that are affected by the treatment are often compensated in a way that leads to comparable total amounts or total deposition rates for metabolite classes within the root growth zone. The kinematic approach taken in this study to reveal fluxes and deposition rates of metabolites by combining sensible growth analyses with analyses of metabolite concentrations might well lead the way to future studies revealing the molecular control of growth processes by combinations of growth and DNA analyses, performed with an appropriate resolution.


We thank Wendy Silk for helpful discussions during the experiments and during preparation of this manuscript. We appreciate critical reading of this manuscript by Barry Osmond and we are grateful for the excellent critical comments of the unknown referees and the subject editor, which improved the manuscript a great deal. The experiments were supported by the Deutsche Forschungsgemeinschaft (SFB 199). A.W. was supported by a Feodor-Lynen-Fellowship from the Alexander-von-Humboldt Stiftung during preparation of this manuscript. The work was partly supported by the USDA-grant USDA/NRICGP 00-35100-9531.