61
Clinica Chimica Acta, 95 (1979)
61-67 Biomedical Press
0 Elsevier/North-Holland
CGA 1024
SEPARATION AND QUANTIFICATION OF URINARY DI- AND POLYAMINES BY GAS CHROMATOGRAPHY WITH ELECTRON CAPTURE DETECTION
J.M.
RATTENBURY
*, PAULINE
M. LAX,
K. BLAU
and M. SANDLER
**
Bernhard Baron Memorial Research Labomtories and Znstitu te of Obstetrics and Gynaecology, Queen Charlotte’s Maternity Hospital, Goldhawk Road, London W6 OXG (U.K.) (Received
December
9th,
1978)
Summary A sensitive and specific gas chromatographic assay procedure employing electron capture detection has been developed for the assay of free and total diand polyamines in human urine. Urine samples, hydrolysed with hydrochloric acid where necessary for the measurement of total amine output, were evaporated to dryness and, after the residues had been taken up in water, purified successively on Porapak Q and Dowex 50 X2 columns. Following evaporation of eluate, pentafluoropropionyl derivatives were made and analysed gas chromatographically using temperature programming. Di- and polyamines can be measured accurately at the picomole level and normal urinary output values calculated using this method agree well with those noted by other workers.
Introduction Diamines and polyamines are involved in growth processes in a wide range of living systems [ 11. Clinical interest in polyamines has centred on their increased rate of excretion in the urine of cancer patients [2] and their altered distribution in the blood of patients with cystic fibrosis [3]. Such interest has in turn stimulated the development of a wide range of approaches to their identification and measurement in biological materials, including radioimmunoassay, enzymatic methods, paper electrophoresis and a variety of chromatographic procedures (for review, see Ref. 4). Several gas chromatographic methods for di- and polyamine analysis have
* Present
address: Department of Paediatric Biochemistry, EHJ 1 LS. U.K. ** To whom correspondence should be addressed.
Royal Hospital for Sick Children. Edinburgh
62
been developed. Volatile free polyamines (excluding spermidine and spermine) can be resolved on columns of porous polymer [ 51 and N-trifluoroacetyl derivatives of polyamines from urine can be separated on a mixed silicone stationary phase [6]. Gas chromatography-mass spectrometry has also been used recently for the identification of trifluoroacetyl derivatised polyamines with the help of deuteriated internal standards [ 71. Electron capture is a very sensitive method for detecting compounds eluted from a gas chromatographic column [8,9], provided suitable derivatives can be prepared. Makita et al. [lo] prepared pentafluorobenzoyl derivatives of di- and polyamines but were unable to resolve them all with a single stationary phase. Furthermore, the spermine derivative could only be detected by flame ionisation, with consequent loss of sensitivity. We describe here a sensitive and specific method for measuring di- and polyamines as their pentafluoropropionyl (PFP) derivatives during a single gas chromatographic run, using electron capture detection. This procedure gave a detection limit of 10 pg and polyamines extracted from urine could be measured accurately and reproducibly at the picomole level. Materials and methods Di- and polyamines were bought from Sigma Chemical Co. Ltd., Poole, Dorset as the hydrochlorides and dieldrin from Shell Research Ltd., Sittingboume, Kent. Porapak Q, 150-200 mesh (Waters Associates, Framingham, MA, U.S.A.), was prepared for use as described by Niederwieser [ll]. Dowex 50X-2 resin, 50-100 mesh, was obtained from Sigma Chemical Co. and converted to the hydrogen form before use. Ethyl acetate was purchased from BP Chemicals International Ltd., London, and subjected to single fractional distillation. Heptane was redistilled from the heptane fraction “low in aromatics” purchased from Hopkin and Williams Ltd., Romford, Essex. 14C-Labelled putrescine, spermine and spermidine hydrochlorides were obtained from New England Nuclear, Dreieich, F.R.G.; acetonitrile was redistilled laboratory reagent grade (BDH Chemicals Ltd., Poole, Dorset). Pentafluoropropionic acid was bought from Koch-Light Laboratories Ltd., Colnbrook, Buckinghamshire, and converted to the anhydride by refluxing with phosphorus pentoxide for 16 h. It was redistilled before use. A Hewlett-Packard 5713A gas chromatograph was employed in this work, fitted with a pulse-modulated linear electron capture detector. The column was of “Pyrex” glass 30 X 0.25 cm, packed with 3% silicone OV-225 on 100-120 mesh Gas-Chrom Q (Applied Science Laboratories Inc., distributed by Field Instruments Co. Ltd., Twickenham, Middlesex). Extraction of amines from urine Urine samples (24 h) collected over 6 mol/l HCI (25 ml) were stored at -20°C. Di- and polyamines were extracted using a combination of previously described methods [12,13] with slight modifications. Two 5-ml portions of a urine sample were required for the measurement of free and two further 5-ml portions for total polyamines, 10 lug each of putrescine, spermidine and spermine being added to one of the pair. For the estimation of total polyamines, one sample with and one without internal standards were mixed with 5 ml
63
cont. HCl (spec. gravity 1.18) and hydrolysed at 110°C overnight in an autoclave. When cool, particulate material was removed by filtration through cotton wool which was then washed with distilled water (1 ml). Filtrate and wash were evaporated to dryness at 65°C under reduced pressure. For the estimation of free polyamines both remaining samples were evaporated to dryness without the hydrolysis step. The residue from each sample was dissolved in distilled Hz0 (4 ml) and transferred to a glass Porapak Q column (5 X 1 cm) previously equilibrated with 0.1 mol/l HCl. The column was washed with 0.1 mol/l HCl (6 ml) and the total eluate (10 ml) was loaded on to a Dowex 50 X2 (hydrogen form) glass column (3 X 0.2 cm). After application of the sample the column was washed twice with 0.2 mol/l HCl (10 ml) and the polyamines eluted with 5 mol/l HCl (10 ml). The eluate was evaporated to dryness under reduced pressure as before and the residue dissolved in 0.1 mol/l HCl (1 ml). This preparation was stored at -20”~ until required for derivatization. In experiments to measure recoveries after the individual extraction steps, trace amounts of 14C-labelled putrescine, spermidine or spermine were added to a urine sample at the beginning of the procedure and aliquots taken and counted after each step. Frepara tion and analysis of derivatives
A 20+1 portion of the final extract or the same volume of authentic standards dissolved in 0.1 mol/l HCl was freeze-dried in a 1 ml Reacti-Vial (Kontes) and pentafluoropropionic anhydride (100 ~1) and acetonitrile (10 ~1) added to the dried residue. The vials were closed with teflon-lined screw caps and heated at 65°C for 30 min. After cooling, unreacted reagent and solvent were removed by evaporation in a stream of nitrogen. The residue of prepared derivatives was redissolved in 20 ~1 ethyl acetate. Heptane (200 ~1) containing dieldrin (1 kg/ ml) as an injection standard was then added to the vial and mixed. After the heptane solution had been washed with 200 ~1 saturated Na2C03 solution, 0.3 ~1 was injected on to the column for gas chromatographic analysis. The column was temperature programmed at a rate of &C/mm from 130°C to 250°C and 2 min at 250°C. A carrier gas of argon-methane (95 : 5, v/v) was used at a flow rate of 60 ml/min. Results Fig. 1 shows the elution pattern of the PFP derivatives of di- and polyamines on a column of silicone OV 225. The elution temperatures for these amines are shown in Table I. Temperature programming enabled the amines with their widely differing volatilities to be analysed during a single chromatographic analysis. Other stationary phases tried, but found to be unsuitable, included 2% OV-17 with QF 1 [6] from which the spermine derivative could not be eluted, 1% OV-210 which gave a high rate of column bleed and OV-25 on which peaks were broad and asymmetrical Trichloroacetyl and pentafluorobenzoyl derivatives of these amines did not give satisfactory separations. Different conditions for the derivatization reaction with PFP were investi-
64
E?x 2x 2-s 19C 1x 1% IX Minutes Fig. 1. Gas chromatogram of PFP derivatives of some di- and polyamines, and temperature programme. Attenuation was X512 and other instrument settings and derivatization procedure were as described in the text. Key: A, 1.3-diaminopropane; 3, putrescine; C, cadaverine; 13, 1,6-diaminohexane; E, 1,7-diaminoheptane; F. l,&diarninooctane: G, l,lO-di~inodecane; H, 1,12-diaminododecane: I, spermidine; J, spermine.
gated. At room temperature spermidine and spermine failed completely to derivatize within an hour and at temperatures above 65°C there was an increase in the amount of interfering material on the chromatogram. The addition of 10 ~1 of acetonitrile or ethyl acetate to the PFP reaction mixture gave a greater yield of derivatives at 65°C for 30 min than when PFP alone was used at 65°C
TABLE I GAS CHROMATOGRAPHIC POLYAMINES ON OV-226
ELHTION
TEMPERATURES
FOB PFP DERIVATIVES
Compound
Elution temperature (“cl
1,3-Di~noprop~e Futrescine (1,4diamlnobutane> Cadaverme (l.bdiaminaPentane) 1,bDiaminohexane 1.7.Diammoheptane 1 ,I-Diaminooctane 1 ,l ODlaminodecane 1,12-Diaminododecane Spermidine (~-(339minopro~~l)_l,4-diam~ob~tane~ Spermme (N,N’-bi9(3-amlnoPropyl~l.4-diaminobntane)
147 163 168 172 177 183 194 205 214 250
OF DI- AND
65
TABLE
II
PERCENTAGE
RECOVERIES
OF
DI-
AND
POLYAMINES
AFTER
INDIVIDUAL
EXTRACTION
STEPS Recoveries were calculated by addition of trace amounts of “C-labelled
amines to a lo-ml urine sample.
Percentage reCo”ery
Stage during extraction at which recovery was measured
Eluant from Porapak Q column Eluant from Dowex column Final extract in 0.1 mol/l HCl
Spermine
Spermidine
Putrescine
81 17 77
100 81 75
100 89 86
for 6 h and was therefore adopted in the definitive procedure. Pentafluoropropionic acid, produced by hydrolysis of the anhydride, elutes in a broad peak at about 2OO”C, and was a serious source of interference in the initial stages of this work. We found that it was absorbed by the teflon cap-liner of the vial during derivatization, and then released back into the solvent used to dissolve the derivatives to make the final solution for injection. This problem was solved by replacing the liner with a fresh one after derivatization was complete. Recoveries after the individual extraction steps are shown in Table II. Adequate recoveries were achieved by elution of the Dowex columns with 5 mol/l HCl without the necessity of recourse to the lengthy procedure of McGregor et al. [13]. The overall recoveries shown in Table III were calculated for 12 normal urine samples by running each in parallel with a duplicate containing known amounts of authentic compounds. These recoveries were constant for a particular urine sample but not when different samples were compared. Therefore, for accurate quantification, it is necessary to run the samples in duplicate, with known amounts of standard added to one. The pure compounds can be detected at a concentration as low as 10 pg and the method may be used to measure di- and polyamines accurately and reproducibly at the 1 picomole level. A chromatogram of a normal urine extract is illustrated in Fig. 2. When the PFP polyamine derivatives were redissolved in ethyl acetate a number of spurious peaks were present on the chromatogram. Heptane gave a much clearer chromatogram but failed to dissolve the spermine derivative. It was found that washing with saturated Na2C03 eliminated impurities without loss of polyTABLE III OVERALL GAS CHROMATOGRAPHIC FROM 12 NORMAL URINE SAMPLES Compound
RECOVERIES
Recovery (%) Mean
Range
Putrescine Spermidine
69 II
spermine
73
58-96 60-96 51-96
OF DI- AND
POLYAMINES
CALCULATED
T *C
/ 2
I 4
I 6
I 6
I 10
I 12
I 14
j 1
Minutes
Fig. 2. Gas chromatogram of PFP derivatives of some di- and polyamines extracted from a normal urine, and temperature programme. Extraction, derivatization and instrument settings are as described in the text. Attenuation was X256. Key: B, putrescine: I, spermidine: J. spermine; K. dieldrin (injection standard).
TABLE
IV
24-h URINARY SUBJECTS Compound
EXCRETION
OF SOME FREE
Mean f S.D. (pmo1/24
AND TOTAL
h)
DI- AND POLYAMINES
Range (pmo1/24
IN 12 NORMAL
h) -___
Free Putrescine Spermidine Spermine
1.57 0.51 0.71
t 1.05 i 0.16 * 0.57
Total
Free
Total
21.9 t 7.6 8.4 * 2.1 2.5 + 1.2
0.41-3.98 0.33-0.90 0.26-1.91
12.6-39.2 5.4-13.6 0.65.0
.-__
amines. In order to prevent the injection of large amounts of water, which is soluble in ethyl acetate, on to the gas chromatographic column, polyamine derivatives were dissolved in a small volume of ethyl acetate (20 ~1) and a larger volume of heptane (200 ~1) was added. This heptane solution could then be washed with saturated Na&03 which eliminated impurities with minimal water uptake. This procedure enabled chromatograms to be interpreted with greater ease and accuracy. The daily urinary excretion of polyamines, free and total, was measured in 12 normal subjects (6 male and 6 female) and the means and ranges for the individual amines are shown in Table IV. Discussion The gas chromatographic method described here separates urinary di- and polyamines in a single chromatographic analysis. Gas chromatography is a highly specific method and electron capture detection enables all the urinary
67
di- and polyamines to be measured at the picomole level. This level of sensitivity is provided in very few other published methods. Radioimmunoassay, although very sensitive, has problems of cross-reactivity and is dependent on the availability of specific antibodies for each of the amines. The only other method providing this sensitivity is gas chromatographymass spectrometry [ 71. The values for normal 24-h urinary excretion of polyamines measured using the described method are in general agreement with those reported by other workers [6,14-171. This method, because of its great specificity and sensitivity, would therefore be ideal for measuring polyamines in biological materials where only small amounts of sample are available. References 1
Bachrach.
2
Russell,
U. (1973)
3
Cohen.
4
Seiler.
5
Casselman,
6
Gehrke.
D.H. L.F.,
7
Smith, New &&rd,
9
Wang,
10
and
C.W., R.G..
K.C.,
23,
1519-1526
G.,
Morris.
R.A.B. D.H.,
Jr.
(1976)
and
pp.
Blood
D.P. D.,
48,
T.P.
Raven
(1978)
Daves,
in
G.D.
Academic Growth,
New
Press,
York
new
York
in
Normal
3340 (1973)
Press.
in
New
Advances
and
Press,
Raven
469475
88.
Waalkes,
343353,
Grettie, Bartos,
Polyamines.
and Neoplastic
J. Chromatogr.
R.W. ed.),
and
D.R.,
P.M.
(1974)
Zumwslt,
(Russell.
Daves, R.A..
Farrell,
Chem.
Occurring
in Normal
and
Bannard,
Kuo,
Growth
of Naturally
Polyamines D.W.
Clin.
A.A.
(Campbell. 8
(1973)
Lundgren,
N. (1977)
Neoplastic
Functions
(ed.)
in
Bartos,
Polyamines
and
York Polyamine
F.,
eds.),
pp.
Research, 23-35,
Vol.
Raven
12
Press,
York E. and K.P.,
Makita,
M.,
11
Niederwieser,
12
Pegg,
13
McGregor,
14
Abdel-Monem,
Hankey,
Ruthven. Yamamoto, A.
A.E.,
(1971)
search.
Vol.
Raven
Press.
New
Townsend,
R.M..
16
Durie,
B.G.M.,
17
Heby,
0.
and
Acta
Kono,
Stand.
M. (1973)
M. (1975) 54,
Atkinson, K.,
R.A..
3110-3119 Chim.
Chim.
Acta
Acta
61,
47.
215-222
403405
215-223 H.G.
(1970)
M. and Johnson,
Newton, Morris,
23,
Clin.
Clin.
and Williams-Ashman.
M.S.,
Ohno,
Chem.
Sandier.
J. Chromatogr.
2 (Campbell,
15
and
D.H.
Sharon, M.M.,
(1969)
S. and
Lockwood, R.F.,
A.
C.R.J.
N.E. D.R.,
and
Weeks,
Bartos.
D.,
Biochem.
D.E.
(1976)
C.E.
Daves,
(1978) G.D.
J. 117,
Prep.
in Advances and
Bartos,
York Banda.
Salmon, Andersson,
P.W. S.E.
and and
Marton,
Russell.
G. (1978)
L.J. D.H.
(1976) (1977)
J. Chromatogr.
Cancer Cancer
145,
73-80
38, Res.
2088-2092 37,
17-31
Biochem.
214-221
F.,
6, 403419 in Polyamine eds.),
pp.
3749,
Re-