Biosynthetic origin of the nitrogen atom in cyanamide in Vicia villosa subsp. varia

Abstract

Natural cyanamide (NH₂CN) has recently been found in three Leguminosae plants: Vicia villosa subsp. varia, Vicia cracca and Robinia pseudo-acacia. As cyanamide has long been thought to be absent in nature, its physiological role and biosynthesis are totally unknown. In the present study, we demonstrated the incorporation of ¹⁵N from [¹⁵N]nitrate and [¹⁵N]ammonium into cyanamide using shoots of V. villosa subsp. varia, which ruled out the possibility that nodules are essential in cyanamide biosynthesis. We also applied [¹⁵N₂]cyanamide to shoots of V. villosa subsp. varia to monitor its turnover, and detected [¹⁵N₂]cyanamide in the leaves within 4 h; it was present without detectable degradation for more than 4 days. In contrast, maximum incorporation of ¹⁵N into cyanamide molecules was observed after 4 days of feeding the shoots with ¹⁵N-labeled inorganic ions and l-[amide-¹⁵N]-glutamine, indicating that these nitrogenous compounds are distant precursors of cyanamide. Although the guanidino group of l-arginine (-NH-C(NH₂)=NH) and urea (NH₂C(=O)NH₂) were candidate precursors of cyanamide on the basis of structural similarity, direct incorporation of the guanidino group of l-[¹³C₆,¹⁵N₄]-arginine and [¹³C,¹⁵N₂]urea into cyanamide was not observed. These results eliminated the possibility that cyanamide is biosynthesized by the addition of ammonia to an electrophilic carbon or by the conversion of the tested compounds that were structurally relevant to cyanamide.

Key words:

• biosynthesis
• cyanamide
• incorporation
• stable isotope
• Vicia villosa subsp. varia

INTRODUCTION

Cyanamide (NH₂CN), a synthetic substance with various agricultural and chemical applications, has recently been shown to be a natural product. Cyanamide has been isolated from a Leguminous plant, Vicia villosa Roth subsp. varia (Host) Corb., as a plant-growth-inhibitory compound (Kamo et al. 2003). This finding is surprising because cyanamide has long been believed to be absent in nature (Estermaier et al. 1992; Maier-Greiner et al. 1991). To quantify cyanamide contents in a variety of plant materials, a stable isotope dilution gas chromatography-mass spectrometry (SID-GC-MS) method, using [¹⁵N₂]cyanamide as an internal standard, has been developed (Hiradate et al. 2005). Using this accurate and convenient method, we have revealed that, out of the 553 species of higher plants tested, only three, V. villosa subsp. varia, Vicia cracca L. and Robinia pseudo-acacia L., contain cyanamide (Kamo et al. 2006b; Kamo et al. 2008). This shows the limited distribution of natural cyanamide in the plant kingdom, denying its general and fundamental roles in plant germination, growth, flowering and fructification.

Although some characteristics of natural cyanamide have been elucidated, others, such as its biosynthesis, remain to be examined. The first step toward a biosynthetic study was to clarify that cyanamide is biosynthesized by the plants and not by the nodules. Although we have already revealed that the shoots of V. villosa subsp. varia contain much larger amounts of cyanamide than the roots (Kamo et al. 2008), the possibility that cyanamide is biosynthesized in the roots and transported to the leaves has not been ruled out. Nor has the possibility that the nodules biosynthesize cyanamide; as this compound consists of two nitrogen atoms per carbon atom, nitrogen fixation could be a reasonable source of the nitrogen consumed during biosynthesis. In the second step, we followed the fate of exogenous [¹⁵N₂]cyanamide to confirm that there was no detectable ¹⁵N exchange that could arise from the degradation and regeneration of this compound in the plant during the period tested. The natural occurrence of cyanamide was shown in our previous study, in which the incorporation of ¹⁵N from [¹⁵N]nitrate into cyanamide was observed in 2-week-old seedlings of V. villosa subsp. varia (Kamo et al. 2006a). This finding, however, serves only to confirm that cyanamide is a natural product, and does not elucidate any biosynthetic details because nitrate is the starting substance for nitrogenous metabolism pathways in plants. Identification of the intermediate precursor of cyanamide would be of great help in a study of its biosynthesis. To evaluate the incorporation of ¹⁵N from possible precursors into cyanamide, the feeding experiment period needs to be much shorter than the half-life of cyanamide in the plants. Based on the high stability of cyanamide in plant tissues, we monitored the incorporation of possible precursors, such as ammonium, urea and amino acids.

In the present study, we used V. villosa subsp. varia because this species is commercially available in the form of seeds that germinate without any eliciting treatments and this species grows faster than V. cracca and R. pseudo-acacia (Kamo et al. 2008). The shoots were used for the experiments after the roots were cut from the seedlings. The origin of the nitrile group in the cyanamide molecule is not mentioned in the present study: it is indistinguishable from that of the amino group in a series of our analyses because the nitrile group and the amino group are converted into each other through tautomerization () as reported by Steinman et al. (1964).

MATERIALS AND METHODS

General

Preparative HPLC was conducted with an ODS column (Inertsil ODS-3, 250 mm × 4.6 mm internal diameter; GL-Sciences, Tokyo, Japan) eluted with water at a flow rate of 1.0 mL min⁻¹; detection was at 200 nm.

Plant materials

Three seeds of V. villosa Roth subsp. varia (Host) Corb. (commercially available as “hairy vetch” from Takii, Kyoto, Japan) were planted in each pot (6 cm diameter × 5.5 cm high) containing vermiculite and incubated in an illuminated growth chamber (LIB-301; Iwaki, Tokyo, Japan) without the application of fertilizer under an illumination cycle of light (16 h) and dark (8 h) at 20°C. For experiments conducted in short (< 7 day) and long (> 10 day) periods, 3-week-old and 2-week-old seedlings were used, respectively.

Chemical reagents

[¹⁵N]Ammonium nitrate (min 98 + atom%¹⁵N), ammonium [¹⁵N]nitrate (min 98 + atom%¹⁵N), [¹⁵N₂]urea (98 + atom%¹⁵N) and [¹³C,¹⁵N₂]urea (min 99 atom%¹³C, min 98 + atom%¹⁵N) were purchased from Aldrich (Milwaukee, WI, USA). Potassium [¹⁵N]nitrate (98 + atom%¹⁵N), [¹⁵N]ammonium [¹⁵N]nitrate (min 98 + atom%¹⁵NH₄, min 98 + atom%¹⁵NO₃), [¹³C]urea (99 atom%¹³C), l-[¹³C₆,¹⁵N₄]-arginine (min 98 + atom%¹³C, min 98 + atom%¹⁵N), l-[amide-¹⁵N]-glutamine (min 98 + atom%¹⁵N), l-[¹⁵N]-glutamic acid (min 98 + atom%¹⁵N), l-[¹⁵N]-aspartic acid (min 98 + atom%¹⁵N), l-[¹⁵N]-alanine (min 98 + atom%¹⁵N), [¹⁵N]glycine (min 98 + atom%¹⁵N) and [¹⁵N₂]cyanamide (min 99.1 atom%¹⁵N, in 51.5% aqueous solution) were purchased from Isotec (Miamisburg, OH, USA).

Preparation of the incubation solutions for the feeding experiments

For the [¹⁵N]ammonium and [¹⁵N]nitrate feeding experiments, the following solutions were prepared: (1) CaHPO₄·2H₂O (1,370 mg) and MgSO₄·7H₂O (986 mg) in 10 mL of water, (2) EDTA-Fe (226 mg), MnCl₂·4H₂O (18.0 mg) and H₃BO₃ (28.6 mg) in 10 mL of water; (3) ZnSO₄·7H₂O (11.0 mg) and CuSO₄·5H₂O (4.0 mg) in 50 mL of water, (4) (NH₄)₆Mo₇O₂₄·4H₂O (3.7 mg) in 100 mL of water. The solutions (1) (500 µL) and (2–4) (each 100 µL) were combined and filled up with water to 100 mL. We refer to this as the “fundamental solution”. For the other feeding experiments, the following solution was prepared: (5) KNO₃ (2,113 mg), Ca(NO₃)₂·4H₂O (1,889 mg), NH₄H₂PO₄ (230 mg) and MgSO₄·7H₂O (986 mg) in 10 mL of water. The solutions (5) (500 µL) and (2–4) (each 100 µL) were combined and filled up with water to 100 mL (Hoagland solution). The maximum concentration of nitrogenous compounds, including amino acids and urea, in the tested solutions was determined by the growth of shoots during incubation periods in the preliminary experiments.

Feeding experiments of stable isotopic labeled compounds to V. villosa subsp. varia shoots

Administration of [¹⁵N]nitrate and/or [¹⁵N]ammonium for 10 days

A 2-week-old V. villosa subsp. varia shoot was inserted into each tube containing 2 mL of the fundamental solution with KNO₃ (3.0 mmol L⁻¹) plus NH₄NO₃ (0.9, 2.3, 4.7 and 9.4 mmol L⁻¹), K¹⁵NO₃ (3.0 mmol L⁻¹) plus NH₄ ¹⁵NO₃ (0.9, 2.3, 4.7 and 9.4 mmol L⁻¹) or KNO₃ (3.0 mmol L⁻¹) plus ¹⁵NH₄NO₃ (0.9, 2.3, 4.7 and 9.4 mmol L⁻¹). The shoots were incubated in the growth chamber under an illumination cycle of light (16 h) and dark (8 h) at 20°C for 10 days and regenerated roots were removed every other day. The solution was replaced every 3 days.

Administration of [¹⁵N]nitrate plus [¹⁵N]ammonium for 20 days and 40 days

A 2-week-old shoot was inserted into each tube containing 2 mL of the fundamental solution with K¹⁵NO₃ (8.7 mmol L⁻¹) plus ¹⁵NH₄ ¹⁵NO₃ (0.5 mmol L⁻¹). The shoots were incubated in the conditions described above, regenerated roots were removed every other day and the elongated stem was cut every 10 days to maintain a length of 10 cm from the top. The solution was replaced every 3 days.

Administration of [¹⁵N]nitrate plus [¹⁵N]ammonium and l-[amide-¹⁵N]-glutamine

A 3-week-old shoot was inserted into a tube containing 2 mL of the fundamental solution with K¹⁵NO₃ (3.0 mmol L⁻¹) plus ¹⁵NH₄ ¹⁵NO₃ (4.7 mmol L⁻¹) and Hoagland solution with l-[amide-¹⁵N]-glutamine (12.4 mmol L⁻¹), respectively. After 24 h, the solution was changed to Hoagland solution and the shoot continued to be incubated under the same conditions for another 7 days. The shoot was sampled for GC-MS analysis every 24 h. For administration of l-[¹⁵N]-glutamic acid, l-[¹⁵N]-aspartic acid, l-[¹⁵N]-alanine and [¹⁵N]glycine, a 3-week-old shoot was treated in the same way.

Administration of [¹³C]urea, [¹⁵N₂]urea and [¹³C,¹⁵N₂]urea

A 2-week-old shoot was inserted into a tube containing 2 mL of the fundamental solution containing KNO₃ (3.0 mmol L⁻¹) plus NH₄NO₃ (0.9 mmol L⁻¹) with urea (5.0 mmol L⁻¹), [¹³C]urea (5.0 mmol L⁻¹), [¹⁵N₂]urea (5.0 mmol L⁻¹) and [¹³C,¹⁵N₂]urea (5.0 mmol L⁻¹), respectively. The shoot was incubated under the same conditions for 10 days.

Administration of l-[¹³C₆,¹⁵N₄]-arginine

A 2-week-old shoot was inserted into a tube containing 2 mL of the fundamental solution containing KNO₃ (3.0 mmol L⁻¹) plus NH₄NO₃ (0.2 mmol L⁻¹) with l-arginine (0.9 mmol L⁻¹) and l-[¹³C₆,¹⁵N₄]-arginine (0.9 mmol L⁻¹), respectively. The shoot was incubated under the same conditions for 10 days.

Administration of [¹⁵N₂]cyanamide

A 3-week-old V. villosa subsp. varia shoot was inserted into a tube containing 2 mL of Hoagland solution with [¹⁵N₂]cyanamide (12.4 mmol L⁻¹) for 24 h. The shoot continued to be incubated in the absence of [¹⁵N₂]cyanamide for another 3 days. The shoot was sampled for GC-MS analysis at 1, 2, 4, 8, 16, 24, 48, 72 and 96 h. After the stems were removed, only the leaves were used for the analysis.

Quantification of urea in 4-week-old V. villosa subsp. varia seedlings

Four-week-old V. villosa subsp. varia seedlings were used to quantify the urea content. The fresh weights (FW) of the shoots and roots of five seedlings were measured. Fresh plant material equivalent to 1.0 g FW was cut to 1 mm lengths and extracted with 4 mL of MeOH after the addition of 500 µL of [¹⁵N₂]urea aqueous solution (16.1 mmol L⁻¹) as an internal standard. The supernatant was concentrated using an evaporator at 30°C. The concentrated solution was diluted with 20 mL of water and extracted with n-hexane (20 mL × 3) and then EtOAc (20 mL × 3). The water layer was concentrated to 100 µL of the water solution and purified by HPLC to give a fraction containing urea (t _R 2.9–3.2 min). The fraction was again concentrated to 100 µL of the water solution and purified by HPLC to give urea (t _R 3.1 min). The water solution of the urea was concentrated to dryness and dissolved in 1 mL of acetone. A 2-µL portion of the sample solution was injected into GC-MS. The urea concentration was determined in triplicate. The urea content (µg seedling⁻¹) was calculated by multiplying the concentration (µg mg⁻¹ FW) by the fresh weight (mg FW seedling⁻¹).

Extraction and purification of cyanamide

For the experiment with 10-day incubation period, five shoots were combined and used as a source for the extraction of cyanamide (n = 1). For the experiment with 20-day and 40-day incubation periods, and with 24-h administration followed by 0–7-day incubations, one shoot was used for the extraction of cyanamide (n = 3–6). Fresh plant materials were cut to 1 mm lengths and grated using a ceramic mortar with the addition of 3 mL acetone and 10 µL dimethylsulfoxide. The supernatant was concentrated using an evaporator at a water-bath temperature of 30°C, and the concentrated solution was diluted with 200 µL of acetone and applied to a silica gel cartridge column (Bond Elut SI, 500 mg; Varian, Palo Alto, CA, USA) pre-conditioned with n-hexane. The column was eluted with mixtures of acetone and n-hexane (2:8, 3:7, 4:6; each 3 mL). The last two fractions were combined. A 2-µL portion of the sample was subjected to GC-MS analysis.

GC-MS analysis

The analytical conditions were as follows: GC-MS instrument, QP-5000 (Shimadzu, Kyoto, Japan); analytical column, CP-Sil 8 CB for amines (0.25 mm internal diameter, 30 m length, 0.25 µm thickness; GL Sciences, Tokyo, Japan); injector temperature, 250°C; interface temperature (ion source temperature), 250°C; ionizing voltage, 70 eV (EI-MS); signal sampling rate, 0.2 s; injection mode, splitless with 30 s sampling time; injection volume, 2.0 µL; column temperature, 50°C for the initial 5 min followed by increases of 15°C min⁻¹ up to 250°C, the temperature was then kept at 250°C for 3 min; carrier gas, He; total flow rate, 50 mL min⁻¹; column flow rate, 1 mL min⁻¹. Urea and cyanamide were detected at t _R 11.7 and 8.5 min, respectively. To quantify the urea content, the area on the mass chromatogram of the [M]⁺ ion derived from the natural urea was compared with that derived from the internal standard.

Table 1 Incorporation of 15N from [15N]nitrate and [15N]ammonium (10 days of administration) into cyanamide in Vicia villosa subsp. varia shoots

Nitrogen sources (mmol L^−1)^†				Relative area of ions (%)^‡			^15N/(^14N + ^15N) (%)
^14NO^− 3	^15NO^− 3	^14NH^+ 4	^15NH^+ 4	m/z 42	m/z 43	m/z 44
3.9	–	0.9	–	100.0	1.2	0.5	0.5
5.3	–	2.3	–	100.0	1.0	0.3	0.3
7.7	–	4.7	–	100.0	1.3	0.3	0.4
12.4	–	9.4	–	100.0	1.0	0.3	0.3
–	3.9	0.9	–	100.0	8.7	4.2	7.1
–	5.3	2.3	–	100.0	10.2	4.5	7.9
–	7.7	4.7	–	100.0	5.2	2.1	3.8
–	12.4	9.4	–	100.0	1.7	0.6	0.9
3.9	–	–	0.9	100.0	2.9	0.5	1.3
5.3	–	–	2.3	100.0	4.2	0.7	2.1
7.7	–	–	4.7	100.0	4.7	1.0	2.6
12.4	–	–	9.4	100.0	3.3	1.0	2.0
^†Nitrate and ammonium were administered as KNO3 plus NH4NO3, K^15NO3 plus NH4 ^15NO3 or KNO3 plus ^15NH4NO3 to 2-week-old shoots for 10 days. Five shoots were combined and used as the source for the extraction of cyanamide (n = 1).
^‡The areas of the peaks corresponding to isotopic cyanamides [t R; 8.5 min, MW: (^12C,^14N2)cyanamide; 42, (^13C,^14N2)- and (^12C,^14N1,^15N1)cyanamide; 43, (^13C,^14N1,^15N1)- and (^12C,^15N2)cyanamide; 44] were measured by gas chromatography-mass spectrometry analyses. The presence of natural ^2H was ignored owing to its small contribution to the results.

RESULTS AND DISCUSSION

Biosynthesis of cyanamide in the shoots of V. villosa subsp. varia

The incorporation of ¹⁵N from [¹⁵N]nitrate and [¹⁵N]ammonium into cyanamide by the shoots is summarized in Table 1. The administration of non-labeled nitrate plus non-labeled ammonium gave non-labeled cyanamide (control): ¹⁵N in total N was 0.3–0.5% on the GC-MS analysis, consistent with the natural abundance of ¹⁵N in total N (0.38%). In contrast, the administration of [¹⁵N]nitrate plus non-labeled ammonium enhanced the ¹⁵N percentage; it reached 7.9% when 5.3 mmol L⁻¹ of [¹⁵N]nitrate was administered. This demonstrated the incorporation of ¹⁵N from [¹⁵N]nitrate into cyanamide by seedlings without roots. As nodules supply fixed ammonium to plants when symbiosis occurs, the starting nitrogenous substance of cyanamide biosynthesis could be derived from the nodules in individuals grown in fields. It was confirmed, however, that the foliar part of V. villosa subsp. varia has the ability to biosynthesize cyanamide.

Administration of [¹⁵N]ammonium at 4.7 mmol L⁻¹ gave the maximum incorporation of ¹⁵N into cyanamide. However, ¹⁵N reached no more than 2.6%, which was smaller than when 5.3 mmol L⁻¹ of [¹⁵N]nitrate was fed. Nitrate might be the more likely precursor for cyanamide than ammonium; alternatively, it could be more efficiently transported than ammonium to an unidentified subcellular organ where cyanamide biosynthesis takes place. The latter appears to be the case because nitrate is a preferable form to ammonium in the systemic transportation mechanism of nitrogenous sources from roots to other organs (Crawford et al. 2000).

Despite the administration of ¹⁵N for 10 days, the relative area of the ion peak at m/z 42 exceeded the areas of the peaks at m/z 43 and 44, as shown in Table 1. This observation implies that the amount of cyanamide that had been biosynthesized and accumulated in the tissues prior to the feeding experiment was large. We analyzed extracts from plants that had been incubated with ¹⁵N-labeled precursors continuously for 20 or 40 days. The 20-day [¹⁵N]nitrate plus [¹⁵N]ammonium feeding period resulted in large relative areas of ion peaks at m/z 43 and 44 on the mass chromatogram (Table 2). However, these relative areas revealed that (¹⁴N₂)cyanamide accounted for more than 30% of the total isotopic cyanamides; it constituted a portion of the isotopic cyanamides, even though the sole nitrogen source supplied was ¹⁵N during the 20 days of incubation. In the 40-day feeding period, (¹⁴N₂)cyanamide was still present, although at a smaller percentage than that observed in the 20 day experiment. These results suggested that cyanamide has high stability in plants. At the same time, it is also true that nitrate and ammonium are clearly incorporated into cyanamide when an ample supply is given in a sufficient time, indicating that they are certain, but distant precursors of cyanamide in its biosynthetic route.

Table 2 Incorporation of 15N from [15N]nitrate plus [15N]ammonium (20 days or 40 days of administration) into cyanamide in Vicia villosa subsp. varia shoots

Administration period (days)^†	Experiment No.^‡	Relative area of ions (%)^§			^15N/(^14N + ^15N) (%)	Percentage of (^14N2)cyanamide (%)^¶
		m/z 42	m/z 43	m/z 44
20	1	84.2	46.6	100.0	53.1	36.9
	2	100.0	68.7	98.0	49.3	37.9
	3	100.0	26.1	39.0	31.1	61.2
40	1	57.9	47.6	100.0	60.0	28.5
	2	26.6	27.5	100.0	73.6	17.5
	3	38.7	39.1	100.0	67.0	22.0
^†Nitrate and ammonium were administered as K^15NO3 (8.7 mmol L^−1) plus ^15NH4 ^15NO3 (0.5 mmol L^−1) to 2-week-old shoots for 20 or 40 days.
^‡Each treatment (20 or 40 days of administration) was carried out independently three times. A different shoot was used for each experiment.
^§The areas of the peaks corresponding to isotopic cyanamides [t R; 8.5 min, MW: (^12C,^14N2)cyanamide; 42, (^13C,^14N1)- and (^12C,^14N1,^15N1)cyanamide; 43, (^13C,^14N1,^15N1)- and (^12C,^15N2)cyanamide; 44] were measured by gas chromatography-mass spectrometry analyses. The presence of natural ^2H was ignored owing to its small contribution to the results.
^¶The percentage of (^12C,^14N2)cyanamide plus (^13C,^14N2)cyanamide to the total isotopic cyanamides.

Stability of [¹⁵N₂]cyanamide administered to the shoots of V. villosa subsp. varia

We incubated 3-week-old shoots in the presence of 12.4 mmol L⁻¹ of [¹⁵N₂]cyanamide for 24 h and in its absence for another 72 h. We confirmed that the shoots continued growing; the fresh weights of the shoots were 102.7 ± 7.1, 106.4 ± 5.2, 111.8 ± 12.3 and 117.5 ± 10.8% (mean ± standard deviation, n = 6) at 24, 48, 72 and 96 h, respectively. Leaves of 4-week-old seedlings of this species grown in an illuminated chamber contain 15.0 ± 5.0 mmol L⁻¹ of natural cyanamide (Kamo et al. 2008). These observations indicate that this condition caused no serious damage to the shoots.

As shown in Fig. 1, the relative area of the ion at m/z 44 corresponding to (¹⁵N₂)cyanamide rapidly increased within 4 h, and remained at 43.0–61.0% until 96 h. In contrast, the area at m/z 43 corresponding to (¹³C)cyanamide and (¹⁵N)cyanamide was almost constant at approximately 2% over the entire 96 h. This result shows that cyanamide is hardly metabolized in the period tested, and that the ¹⁵N groups in the ¹⁵N-labeled cyanamide molecule are stable enough not to be exchanged with other ¹⁴N groups in plant tissues through degradation and regeneration. Cyanamide's high stability in the tissues suggests that it is possible to incubate the shoots after the administration of labeled compounds for a sufficient period to monitor distinct ¹⁵N incorporation.

Figure 1  Absorption of [15N2]cyanamide by shoots of Vicia villosa subsp. varia. (a) [15N2]Cyanamide (12.4 mmol L−1) was administered to a 3-week-old shoot for 24 h. (b) The shoot was incubated in the absence of [15N2]cyanamide for 72 h after the administration. The shoot was sampled for gas chromatography-mass spectrometry analysis at 1, 2, 4, 8, 16, 24, 48, 72 and 96 h. The vertical axis represents the relative area of the ion at m/z 43 (•) and 44 (○) to that at m/z 42. The ion percentage at m/z 43 was calculated by subtracting the relative area assignable to (15N1)cyanamide contained in the commercial [15N2]cyanamide from the area observed. A different shoot was used for each analysis. Values are mean ± standard deviation (n = 6).

Comparison of ¹⁵N-labeled inorganic ions and amino acids as precursors of cyanamide

In biosynthetic studies, it is helpful to compare the periods required until incorporation is observed after labeled precursors are administered because an immediate precursor is supposed to be most rapidly incorporated into its target molecule. To investigate this, we periodically sampled the plant materials after the 24-h administration of [¹⁵N]nitrate plus [¹⁵N]ammonium and l-[amide-¹⁵N]-glutamine, respectively, and monitored the ¹⁵N incorporation into cyanamide using GC-MS analyses. In both treatments, the relative area of the ion at m/z 43 increased slowly and peaked at between 2 and 3% after 4 days (Fig. 2). The relative area at m/z 44 remained less than 0.4% throughout the period when l-[amide-¹⁵N]-glutamine was administered, reaching 1.0% at 5 days in the [¹⁵N]nitrate plus [¹⁵N]ammonium administration. These results imply that l-glutamine is also a poor precursor, as are nitrate and ammonium. The percentages in Fig. 2are apparently lower than those in Tables 1 and 2, but the differences only result from the shorter administration period (24 h) of the labeled precursors in the figure than in Table 1(10 days) and Table 2 (20 and 40 days).

Figure 2  Incorporation of 15N into cyanamide during incubation following 24-h administration of the 15N-labeled precursors. The vertical axis represents the relative areas of the ions at m/z 43 (○; [15N]nitrate plus [15N]ammonium, •; l-[amide-15N]-glutamine) and m/z 44 (□; [15N]nitrate plus [15N]ammonium, ▪; l-[amide-15N]-glutamine) to those at m/z 42. (a) [15N]Nitrate plus [15N]ammonium (3.0 mmol L−1 of K15NO3 plus 4.7 mmol L−1 of 15NH415NO3) and l-[amide-15N]-glutamine (12.4 mmol L−1) were administered to a 3-week-old Vicia villosa subsp. varia shoot for 24 h. (b) The shoot was incubated in the absence of the labeled precursors for another 7 days. A different shoot was used for each analysis. Values are mean ± standard deviation (n = 4).

Four other amino acids, l-[¹⁵N]-glutamic acid, l-[¹⁵N]-aspartic acid, l-[¹⁵N]-alanine and [¹⁵N]glycine, were also administered and sampled after 4 days, giving 2.63 ± 0.24, 2.59 ± 0.27, 2.32 ± 0.39 and 2.63 ± 0.47 (mean ± standard deviation; control 1.55 ± 0.19) for the relative areas of the ions at m/z 43 compared with those at m/z 42, respectively. These incorporations were comparable to that of l-[amide-¹⁵N]-glutamine, suggesting that the four amino acids are also distant precursors of cyanamide. As shown in Fig. 1, shoots of V. villosa subsp. varia are able to absorb [¹⁵N₂]cyanamide and deliver it to the leaves within a couple of hours. Therefore, inorganic nitrogenous ions and amino acids would be transported to the foliar parts because they are supposedly the fundamental materials for nitrogenous metabolism in plant tissues. Thus, it would be more appropriate to attribute the low incorporation of nitrate, ammonium and the amino acids into cyanamide to the long biosynthetic distance between the nitrogenous sources and the cyanamide, rather than to the efficiency of the absorption and/or transportation processes.

Table 3 Contents of urea in 4-week-old seedlings of Vicia villosa subsp. varia grown in an illuminated chamber

	Shoot	Root
Fresh weight (mg seedling^−1)^†	201 ± 72	427 ± 51
Concentration of urea (mg kg^−1 fresh weight)^‡	357 ± 120	6.0 ± 1.7
Content of urea (µg seedling^−1)	71 ± 24	2.8 ± 0.8
^†Values are mean ± standard deviation (n = 5).
^‡Values are mean ± standard deviation (n = 3).

Occurrence of urea in the seedlings of V. villosa subsp. varia

We observed a structural similarity between the guanidino group of l-arginine [-NH-C(NH₂)=NH] and that of cyanamide (NH₂CN). Urea [NH₂C(=O)NH₂] is also discussed in the present context as well as l-arginine because, as is well established, the biosynthesis of urea in plants is generated by the elimination of the N-C-N skeleton of the guanidino group of l-arginine in the urea cycle (Jones and Boulter 1968; Kasting and Delwiche 1957, 1958). If we were to examine the incorporation of the guanidino group of l-arginine and urea into cyanamide, we would find that these feeding experiments are meaningful only when these possible precursors are present in the seedlings of V. villosa subsp. varia. While l-arginine is a ubiquitous amino acid in the construction of normal peptides, it is not known whether or not urea is present in this plant. Thus, we quantified the urea contents in the seedlings by GC-MS analysis with an internal standard ([¹⁵N₂]urea). As shown in Table 3, natural urea was present in the seedlings. Its concentration was much higher in the shoots (357 ± 120 mg kg⁻¹ FW) than in the roots (6.0 ± 1.7 mg kg⁻¹ FW); approximately 96% of the total urea is contained in the leaves plus the stems. The presence of urea in this plant is consistent with a report in which a marine diatom, Thalassiosira pseudonana, was found to possess the urea cycle as indicated by mitochondrial genome analysis (Armburst et al. 2004). The metabolic pathway found in this unicellular alga suggests a wide distribution of urea in the plant kingdom.

Table 4 Incorporation of 13C and/or 15N labels of urea and l-arginine into cyanamide by Vicia villosa subsp. varia shoots

Administered compound^†	Relative area on mass chromatogram (%)^‡
	m/z 42	m/z 43	m/z 44	m/z 45
Urea	100.0	1.0	0.3	0.0
[^13C]Urea	100.0	1.8	0.2	0.0
[^15N2]Urea	100.0	9.5	5.1	0.0
[^13C,^15N2]Urea	100.0	12.1	3.8	0.0
l-Arginine	100.0	1.2	0.2	0.0
l-[^13C6,^15N4]-Arginine	100.0	3.0	0.6	0.0
^†The tested compounds were administered at 5.0 mmol L^−1 for urea and 0.9 mmol L^−1 for l-arginine to 2-week-old shoots for 10 days.
^‡The areas of the peaks corresponding to isotopic cyanamides [t R; 8.5 mm, MW: (^12C,^14N2)cyanamide; 42, (^13C,^14N2)- and (^12C,^14N1,^15N1)cyanamide; 43, (^13C,^14N1,^15N1)- and (^12C,^15N2)cyanamide; 44, (^13C,^15N2)cyanamide; 45] were measured by gas chromatography-mass spectrometry analyses.

Administration of ¹³C-labeled and/or ¹⁵N-labeled l-arginine and urea to shoots of V. villosa subsp. varia

To test whether l-arginine or urea is the direct precursor of cyanamide, we administered ¹³C-labeled and/or ¹⁵N-labeled l-arginine and urea to the shoots. The guanidino group of l-arginine, and particularly urea, appeared at first to have been incorporated into cyanamide (Table 4). However, inspection of these results led us to a correct understanding: l-arginine and urea were first metabolized to other, smaller molecules, such as ammonium and carbon dioxide, which were then incorporated into cyanamide. The GC-MS analysis of the cyanamide isolated from the shoots treated with l-[¹³C₆,¹⁵N₄]-arginine gave ions at m/z 42, 43 and 44, but not at 45. Administration of [¹³C,¹⁵N₂]urea also did not result in an ion at m/z 45. If l-arginine or urea had been directly converted to cyanamide in the plant, the administration of l-[¹³C₆,¹⁵N₄]-arginine or [¹³C,¹⁵N₂]urea must have given [¹³C,¹⁵N₂]cyanamide; the validity of this argument is supported by the high stability of cyanamide in the tissues (Fig. 1). However, in both treatments an ion peak at m/z 45, corresponding to the ion of [¹³C,¹⁵N₂]cyanamide, was not detected on the mass chromatogram. This implies that the ion peak at m/z 44 observed when [¹⁵N₂]urea was administered was also derived from other ¹⁵N-labeled molecules, such as [¹⁵N]ammonium, which could occur as a metabolite of [¹⁵N₂]urea. Therefore, neither l-arginine nor urea is the immediate precursor of cyanamide.

Biosynthetic route for cyanamide

We monitored the incorporation of ¹⁵N from ¹⁵N-labeled compounds into cyanamide using shoots of V. villosa subsp. varia after removing the roots, and clarified that both nodules and roots are essentially irrelevant in cyanamide biosynthesis. The shoots of this species absorbed [¹⁵N₂]cyanamide within 4 h and maintained it for 96 h without any isotopic exchange, highlighting the structurally high stability of cyanamide in plant tissues. Despite a series of trials, the immediate precursor of cyanamide could not be identified. We revealed that nitrate, ammonium and glutamine required 2 days to be converted to cyanamide, denying the possibility that any of these nitrogenous compounds are the direct precursor of cyanamide. l-Arginine and urea, whose molecules are structurally similar to that of cyanamide, were not directly converted to cyanamide, but were probably metabolized to small molecules and then used in cyanamide biosynthesis.

Theoretically, cyanamide could be biosynthesized by any of four routes. First, ammonia as a nucleophile could attack an electrophilic carbon atom. This route is unlikely given that ammonium is a distant precursor of cyanamide. Second, structurally similar compounds, such as urea, could be converted to cyanamide. This route also appears unlikely. Third, cyanide could attack a primary amine or an amide. This route remains possible, but we found no evidence of amine or amide possessing a functional group that leaves easily in a nucleophilic substitution. Finally, an unidentified compound larger than cyanamide could be metabolized to give cyanamide or its immediate precursor. This route is also possible, and a concrete candidate is necessary for further investigation. Thus, the next step toward addressing the biosynthesis of cyanamide is to identify its possible precursors.

ACKNOWLEDGMENT

This work was supported in part by a Grant-in-Aid from the Japan Ministry of Education, Culture, Sports, Science and Technology (19780086 to T. Kamo).