Plant Physiol. (1987) 83, 232-234 0032-0889/87/83/0232/03/$0l1.00/0

ommulaton High Speed HPLC Analysis of in Plant Tissues' Received for publication August 19, 1986 HuGo J.-P. WALTER AND JAN M. C. GEUNS* Lab. Plantenfysiologie, Kard. Mercierlaan 92, B-3030 Heverlee, Belgium

ABSTRACT be much improved by using a Dual-Wavelength TLC Scanner with an activating wavelength of 360 nm and a fluorescent A high speed high performance liquid chromatography (HPLC) wavelength of 500 nm as used by Katoh et al. (3). This method method for the quantification of , , and spermine in also allows the use of an internal standard. biological samples is described. The dansylation is followed by a sample HPLC on conventional RP-columns was too slow for our cleanup. The isocratic HPLC-analysis with acetonitrile:H20 (72:28 vol- purpose (15-20 min analysis time) and required gradients caus- ume/volume) on 10 centimeter long, 3 micrometer octadecyl silica col- ing loss of time while re-equilibrating the column between runs. umns (3 millimeter inner diameter) takes only 4.5 minutes. By our method Gradients require expensive HPLC solvents and the solvent about 100 analyses can be done in 1 day. consumption per analysis is high. As we had no access to a fluorimetric HPLC-detector, we choose to work with dansyl derivatives that are measured at 337 nm, allowing the use of low cost solvents (benzoates as used by Redmond and Tseng [5] and Flores and Galston [1] would require very pure solvents to measure in the low UV). Gradients were avoided because they require more expensive and extensive Polyamines are known to be ubiquitous in plant and animal equipment, enhancing the risks of instrumental breakdown, and cells. Although there has been an increasing interest in polyam- they make re-equilibration necessary. ines during recent years as illustrated by the large number of survey articles on this subject, their exact function is still un- known (2, 8-10). MATERIALS AND METHODS Our interest in the content during the adventitious Extraction and Derivatization. Tissue (200-250 mg) frozen in root formation on the hypocotyls of intact mung bean seedlings liquid N2 is homogenized in 2 ml 4% HC104 containing 1,7- growing in cortisol solution (4) confronted us with a huge number diaminoheptane.2HCI (5 mg/l) as internal standard. After 1 h ofanalyses. Therefore, we needed a fast, inexpensive, and reliable at 4°C the homogenate is filtered by sucking over glass wool method for polyamine quantification. plugged into 5 ml pipet tips. To 0.2 ml homogenate 1 ml of Fluorometric quantification ofTLC eluates ofdansylated pol- carbonate buffer pH 9 and 1 ml of dansylchloride solution (10 yamines (6) did not give valuable results due to impurities as mg/ml acetone) is added. After vigorous shaking (Vortex) the revealed by fluorescence scanning of the eluates (similar peak mixture is heated for I h at 60°C in the dark, cooled and stored shapes were not always obtained). This throws doubt upon some in a refrigerator. The dansylated polyamines are extracted with results obtained by measuring the fluorescence at a fixed wave- 3 ml . length only, so often done by plant physiologists. Moreover, this Post Derivatization Cleanup. The cleanup step is based on the method is rather laborious. If two samples of one extract are method described by Seiler et al. (7), which was adapted for dansylated, the quantification of putrescine, spermidine, and allowing very quick routine analyses. The toluol extract is puri- spermine necessitates 2 x 3 scannings of the unspiked TLC fied over 0.5 g silica gel columns made in 5 ml pipet tips plugged eluates to check their purity, and at least 24 readings on the with glass wool at the constricted end (silica gel: 0.063-0.2 mm fluorimeter (one time for each unspiked TLC eluate and three diameter). Ten columns at a time are rinsed, under reduced times for each eluate spiked with three known amounts of pressure, with 5 ml toluol and 5 ml toluol-triethylamine (10:1 v/ standards). The use of an internal standard is excluded. v). These eluates are discarded. The rack with pipet tips is then Polyamine analysis was improved by using labeled dansylch- placed on a device to elute the dansylated polyamines under loride and measuring the radioactivity of the bands after chro- reduced pressure into 20 ml vials with 2 times 3 ml ethylacetate. matography on RP-TLC (C18)2 in acetone-water (70:30). How- The volume ethylacetate is reduced under a stream of N2 at ever, no internal standard can be used and the method is rather 50°C, transferred into I ml sample vials and completely dried. expensive. The sample cleanup takes approximately 1 d for 300 samples. The analysis and quantification of dansylated polyamines can HPLC Analysis of Dansylated Polyamines. Small columns of 10 cm length and 3 mm internal diameter filled with 3 gm 'Supported by Belgian National Fonds voor Wetenschappelijk On- reverse phase material of Chrompack-Nederland (C.P. Micros- derzoek and the Fonds voor Kolektief Fundamenteel Onderzoek. phere) were used. This material was selected because of its well 2Abbreviations: RP-TLC(C,g), reverse phase TLC on C,8 plates; ODS, suited retention characteristics to perform the analysis in an octadecyl silica; dansylchloride: 5-dimethylaminonaphthalene-1-sulfon- isocratic run within 4.5 min. The column was heated at 50 + ylchloride; Dns-putrescine: bis-dansyl-1,4-diaminobutane (putrescine); 0.1 C in a water jacket, so reducing the column back-pressure Dns-1,7-diaminoheptane, bis-dansyl-1,7-diaminoheptane; Dns-spermi- and the solvent viscosity and accelerating the separation while dine: tris-dansyl-spermidine; Dns-spermine: tetrakis-dansyl-spermine; the resolution remained sufficient for our analytical problem. AUFS, absorption units full scale. The solvent flow was 2 ml/min (acetonitrile:H20 72/28 v/v). 232 HIGH SPEED HPLC OF POLYAMINES 233 1

3

2

4

vil L i Ea ax3 i i 1 2 3 4 5 min

FIG. 1. of reference HPLC-chromatogram dansylated compounds. 3 Column: 10 cm x 3 mm i.d. ODS (3 ,m). Solvent: acetonitrile:H20 (72:28) at 2 ml/min (200 bar). Column, solvent and tubing hold at 50'C. 1I Dns-putrescine; 2, Dns-diaminoheptane; 3, Dns-spermidine; 4, Dns- spermine.

The samples were dissolved in 50 Al methanol. The injection volume was 10 ,I. RESULTS AND DISCUSSION 1 In Figure I a typical chromatogram of dansylated reference compounds is shown (Dns-putrescine, Dns-diaminoheptane, 4

Dns-spermidine, Dns-spermine). The most retained compound 2 of our mixture (Dns-spermine) is completely eluted within 4.5 min, and this by an isocratic solvent run at 50C. This is much faster than any published method. At room temperature (20°C) the analysis time was more than 10 min. The reduction in retention time by placing the column in a water jacket at 50°C i and by insulating the tubing did not affect the peak resolution a . k very much, but had a positive effect on the lower detection limit 0 1 2 3 4 5 min as the peak heights increased. The higher temperature also low- ered the column back-pressure from 400 kg/cm2 to 200 kg/cm2 FIG. 2. HPLC-chromatogram of a dansylated leaf extract of 3 d old due to the lower viscosity of the solvent, and thus allowing a etiolated mung bean seedlings. Same conditions and numbering as in higher solvent flow (up to 2 ml/min). Figure 1. The detector response was linear for polyamine concentrations between 0 and 100 AM, thus covering the endogenous concentra- tion range in plants. Figure 2 shows a chromatogram of a dansylated extract of leafs from 3 d old etiolated mung bean seedlings (Phaseolus aureus Roxb.), grown at 25°C. The first large peak mainly consisting of dansylammonia and dansyldi- Table I. Retention Times and Amounts ofPutrescine, Spermidine, and methylamine (7) is sufficiently separated from putrescine to allow Spermine in 3 Day Old Etiolated Mung Bean Seedlings a correct integration. In our mung bean seedlings putrescine, Mean + SE of the mean of 10 different extracts. spermidine and spermine were detected, but no . As Polyamine Retention Time Amount we used a spectrophotometer equipped with 8 Ml flow-through cells as UV-detector, the practical sensitivity limit was 0.1 AUFS. Leafs Cotyledons This restricted the lower detection limit to about 1 nmol/g fresh min wt weight for putrescine and spermidine, and to about 10 nmol/g nmol/gfresh fresh weight for spermine. By using a modem UV-detector or Dns-putrescine 0.65 + 0.01 519 ± 31 80 ± 6 fluorescence detector the lower detection limit will be improved Dns-heptanediamine 0.94 ± 0.01 at least 50 times. With Dns-heptanediamine as internal standard Dns-spermidine 1.54 ± 0.01 1229 ± 37 389 ± 22 ± ± straight calibration curves were obtained up to 100 AM (putres- Dns-spermine 4.19 0.02 552 26 81 7 234 WALTER AND GEUNS Plant Physiol. Vol. 83, 1987 cine: y = 0.027x; spermidine: y = 0.0 19x; spermine: y = 0.0 lx). Acknowledgment-The authors wish to thank Ingrid Franken for skillful tech- nical assistance. As we encountered some problems with the reproducibility of the dansylation of some plant extracts, we found the yield of these dansylations to be strongly influenced by the pH. Therefore, LITERATURE CITED we used a carbonate buffer at pH 9 as also used by Rosier and 1. FLORES HE, AW GALSTON 1982 Analysis of polyamines in higher plants by Van Peteghem (personal communication). In our further work high performance liquid chromatography. Plant Physiol 69: 701-706 this carbonate buffer at pH 9 is always used thus avoiding all our 2. GALSTON AW 1983 Polyamines as modulators of plant development. Bio- previous problems with reproducibility. We have also experi- Science 33: 382-388 enced difficulties with the derivatization itself. Often the dansy- 3. KATOH Y, HASEGAWA, T SUZUKI, T FuJ) 1985 Changes in the amounts of putrescine, spermidine and spermine in hiproly barley callus after auxin lation is not carried out under strictly defined conditions (e.g. withdrawal. Agric Biol Chem 49: 1027-1032 overnight at room temperature). We have found varying results 4. LOEYS MJ, JMC GEUNS 1978 Cortisol and the adventitious root formation in due to the duration and temperature of derivatization, 1 h at mung bean seedlings. Z Pflanzenphysiol 87: 211-224 60°C being the best condition for quantitative and reproducible 5. REDMOND JW, A TSENG 1979 High-pressure liquid chromatographic deter- mination ofputrescine, cadaverine, spermidine and spermine. J Chromatogr results. By working under controlled conditions we obtained very 170: 479-481 reproducible results. The retention times calculated for 10 differ- 6. SEILER N, M WIECHMANN 1967 Mikrobestimmung von Spermin und Sper- ent plant extracts (3 d old etiolated mung bean cotyledons and midin als dimethylamino-naphthalin-5-sulfonsaure-Derivate. Z Physiol Chem 348: 1285-1290 leaves) are given in Table I. In this table the reproducible 7. SEILER N, B KNODGEN, F EISENBEISS 1978 Determination ofdi- and polyamines quantitative results are also demonstrated, the SE varying between by high-performance liquid chromatographic separation of their 5-dimethy- 3 and 8.6%. The use of this high speed isocratic HPLC analysis laminonaphthalene- I-sulfonyl derivatives. J Chromatogr 145: 29-39 with a peak detection at 337 nm enabled us to work with low 8. SMITH TA 1980 Plant . In: EA Bell, BV Charlwood, eds, Secondary Plant Products. Encyclopedia of Plant Physiology (New Series), Vol 8. cost solvents that could even be redistilled and reused. About Springer-Verlag, Berlin, pp 433-460 100 samples can be analyzed in 1 d. Then, however, an automatic 9. SMITH TA 1985 Polyamines. Annu Rev Plant Physiol 36: 117-143 injector becomes very desirable. 10. TABOR CW, H TABOR 1984 Polyamines. Annu Rev Biochem 53: 749-790