- Email: [email protected]
An improved semi-automated method for the colorimetric determination of triglycerides in serum
45
Clinica ~~i~~~~a
0 Elsevier
WA
Acta, 64 (1975) 45-50 Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlands
7265
AN IMPROVED SEMI-AUTOMATED METHOD FOR THE COLORIMETRIC DETERMINATION OF TRIGLYCERIDES IN SERUM P.N.M. DEMACKER, and A.P. JANSEN
J.B.H.A.
VAN OPPENRAAY,
H. BAADENHUIJSEN
Department of Internal Medicine, Division of Clinical Chemistry, Sin1 Radboudziekenhuis, University of Nijmegen, Geert Grooteplein Zuid 16 (The Netherlands) (Received
April 22, 1975)
Summary A semi-automated calorimetric determination for triglycerides in serum is described. The extraction is performed in sample cups and a part of the upper layer is sampled automatically. The use of modified reagents leads to greater stability and results in only one phase in the continuous flow system. The method is simple and inexpensive and therefore convenient for most laboratories with increasing requests for triglyceride determinations in serum. The method also shows good reproducibility and excellent correlation with the enzymatic method (Eggstein, M. and Kreuz, F.H. (1966) Klin. Wochenschr., 267-273).
Introduction Hyperlipemia is frequently accompanied by coronary heart disease. It is widely accepted that combined determination of cholesterol and triglycerides is the most reliable way to detect hyperlipemia in man [l].This is reflected in an increasing number of requests for serum triglyceride determinations in clinical laboratories. Hence, a simple and inexpensive automated method for this determination would be useful. Many semi- and fully automated applications have been described [ 2-71. However, most chemical methods, based on continuous flow, involve instable reagents, complex flow patterns and phase separations, while the enzymatic methods are expensive. Moreover, most semi-automated methods are laborious. A semi-automated chemical method is described here that meets most of the above-mentioned objections. The extraction of triglycerides is performed in large glass sample cups, based on the procedure as described by Royer and Ko [8]. The extraction is performed manually in the sample cups and then they are placed in the sampler. After separation of the two phases, an ahquot of the
upper layer is sampled automatically and the inter-samplta Lvasti Inkcs pluc*o from an external reservoir. In our system the extract is analysed in a similar manner to that used l)y ltoycr and Ko [ 81. The triglyceridta fatty ac*itls arc’ transesterificd and the glycerol liberated is oxidized to formaldrhytlc. whicsh is then condensed in the> IIantzsch reaction with acetylacetotlrl and ammonia to 3,5-diacetyl-1;t-dihydrolutidine. give a colored (and fluorescent) product. The use of a different periodate and acet.ylac*etone reagent [ 91 leads to a greater stability; also, the water content in the continuous-flow system is restricted, as compared to the systems previously described. There is only one phase in the continuous-flow system, and a stable baseline is obtained. The method has good reproducibility and shows excellent correlat.ion with the enzymatic method of Eggstein and Kreuz IlO] . Materials
and methods
Reagents 1. Extraction
solvent, n-nonane/isopropanol (2.0 : 3.5, by vol). It is possible to use the inexpensive nonane-fraction (Fluka A.G. 74260) and isopropanol (Baker 8119). 2. Dilute sulfuric acid, 40 mmol/l. 3. Saponification reagent, dissolve 3.40 g of sodium methylate (Baker 2812) in 50 ml of distilled water, dilute to 1 litre with isopropanol. Allow the solution to stand overnight then pour into a clean bottle. During the following days a slight precipitate becomes fixed on the inner wall of the bottle and the solution becomes clear. For best results the reagent should be prepared fresh every week. 4. Periodate reagent, dissolve 144 g of ammonium acetate in about 600 ml of distilled water. Add 100 ml of glacial acetic acid and 3.48 g of potassium periodate, dilute to 1 litre with distilled water and dissolve overnight by continuous stirring. The reagent is stable for at least three months [9]. 5. Acetylacetonc reagent, dilute 30 ml of acetylacetone (Merck 9600) to 1 litre with isopropanol. This remains stable for at least three months 191. 6. Standard. A stock standard contains 50 mmol of triolein (Sigma Chemical Co.) per litre of extraction solvent. A working standard is prepared by diluting with extraction solvent and remains stable for at least three months at room temperature. A Cenco* sampler was modified for automatic sampling of a part of the upper layer. The pick-up mechanism moves in mutual perpendicular directions from the sample cups to an external reservoir with wash solution. The sampler plate is high enough to carry the sample cups. A Technicon sampler II may also be used 16,111. As a wash solution, distilled water is used. The sample to wash ratio is 1 : 2, and thirty samples can be analysed per hour. The reaction temperature is 55°C. Colorimetry is performed with the flow scheme as shown in Fig. 1. A Cenco* pump and flow-through calorimeter [12] are used and a Vitatron recorder. The calorimeter is equipped with a 20-mm flowcell and a
* Crnco
Brrda.
l’hr
Nrthcrlands
*solvatlex Fig.
1. Flow
diagram.
420~nm filter. Pump tubes and glass reaction coils are interconnected with polyethylene tubing (internal diameter 2 mm). After each run the system is washed with water. The saponification-reagent pump tube is changed once a month; the other pump tubes are changed every two months,
Extraction 0.2 ml of serum is pipetted
into cyiindric~ glass sample cups (40 mm X 14 mm, capacity 4.5 ml); 2.5 ml of extraction solvent and 0.5 ml of dilute sulfuric acid are added with a dispenser (Oxford). The cups are then closed with polyethylene caps. For the calibration, 0.2 ml of the working standards is pipetted into the sample cups. The solvent is evaporated under a stream of air and 0.2 ml of water is added to obtain similar conditions for standards and serum in the extraction step. Many standards and samples can be handled simultaneously. The cups are shaken manu~ly for one minute and the phases are allowed to separate. The cups are then placed in the sample plate and the analysis of the triglycerides in the upper phase is performed in the continuous-fI ow system. Results and discussion
Extraction Recently, fully automated applications for the calorimetric determination of triglycerides have been described [3,4]. However, the modification of the sampler into an extractor-sampler appears complicated, while the advantages of this modification are minimal. Tt does not save time, for the pipetting of the samples, which takes most of the time cannot be automated, since calibration with standards is then impossible. Manual extraction warrants a better control of the extraction step. The use of the ‘inexpensive nonane-fraction [4], instead of pure nonane, did not influence the specificity of the extraction. e-l-lecithin, bilirubin, glucose and glycerol showed no impo~ant interference (Table I).
TARLE:
I
INTERFERENCE
Extraction
IN
THE
TRIGLYCERIDF:
DE:Tl?RMINATIOh’
of:
Triglyccritle
value Exiwctr~d
F0UIld
0.1
ml
serum
+ 0.1
ml
distilled
water
1.39
1.43
0.7
ml
serum
+ 0.1
ml
glycerol
10
1.41
1.43
0.1
ml
serum
+ 0.1
ml
glucose
1.38
1:13
0.1
ml
serum
+ 0.1
ml
serum
0.2
ml
serum
+ 0.2
ml
u-L-lecithin
Experiments
12 with
mmolil mmolil high 10
bilirubin
110
~mol/l
nmolil
with triolein
showed recoveries
1.42
1.43
2.92
2.Xfi
of 93-1&I%
(IZ = 9).
Reaction conditions With the given optimal reagent concentrations a peak-plateau ratio of 0.74 was obtained. The peak-plateau ratio was not influenced appreciably by variation of the sample amount between 0.32 and 0.60 ml/min, but the peak height can so be regulated without the risk of disturbing the one-phase system. The peak height was also affected by the temperature; an elevation of the temperature to 60°C caused an increase of the peak height of 20%. Sample-wash conditions The peak height was proportional to the sample time between 20 and 45 s and the peak-plateau ratio increased from 0.27 to 0.82. For the inter-sample wash, water appeared superior to heptane or isopropanol. With a sample and
-
absorbance 10
08
1
-
STANDARDS
EXTRACTS
STANDARDS (mm01 /I)
I
06
04
1c 02 a5
0 !/
Fig.
2.
method.
Recorder The
peaks
tracing were
of
peak
recorded
appearance from
right
of to
left.
standards
and
extracts
for
the
proposed
triglyceride
49 ne w method
(mmol/I
) --
12
10
6
. y: 104x r-0999 n: 71
. .*I 8.
4
6
8
-003
10 enzymatic
12 method
14 (mmol/l )
Fig. 3. Graphic comparison of the proposed method with the ennaxatic method.
wash time of 40 and 80 s, respectively, the recording continually returns to the baseline level, even after a sample with a very high triglyceride content. The carryover is negligible and the reproducibiiity of the lower concentratioIls is better then with a sample and wash time of 30 and 60 s (Fig. 2). Therefore, a sample and wash time of 40 and 80 s is preferred, although the capacity of the system is lower than in most of the systems previously described. Waste The method is inexpensive, but the pollution by the organic solvents high. In our opinion it is desirable to gather and burn the waste.
is
reproducibility and assay precision The withinday precision of the method for specimens with triglyceride values of 0.8 and 2.9 mmol/l was 1.5 and 2.3%, respectively, (n = 10) and the day-today precision was 4.4 and 3.3% (n = 20). The sample cups on the sample plate were not covered. The possible error due to evaporation of the upper layer appeared to be negligible. The weight loss of the nonane extract due to evaporation during the longest time on the sample plate, amounted to only 1.7% at 25*C.
Comparison of t,his method with thr enzymatic method Kreuz [10,12] gave a regression equation Y = 1.04 X --0.03, coefficient was 0.999 ( IZ = 71) (Fig. 3). In the enzymatic method values were corrected for fret glycerol (0.15 mmol/l).
of 1l:ggstrin antI the c,orrelation the triglycberidr
Acknowledgements
tance
We thank Mrs Mary and The Netherlands
Thijssen Paddenburg Heart Foundation
for the excellent technical for financial support.
assis-
References 1
Fredrickson,
2
Levv.
D.S.,
A.L.
Vol.
I. PP. H. and
3
Ko.
4
Bjiirksti.n.
5
Schmidt,
Levy,
and
R.I.
Keyloun,
497-502, Royer.
Mediad M.F.
F. (1972) V.F.H..
dnd C.
Inc..
(1972)
Clin.
P.S. in
White Acta
V.G.,
(1967)
Med. 40,
Dahl,
N.
Advances Plains.
Biochem.
Chim.
Schmidt,
Lees,
(1970)
Engl.
in
.l. Med.
Automated
276.
9,%~-103
Analysis.
‘l‘t~chnlron
6. 144-159
143--152 K.
and
Stork,
H.
(1972)
%. Klin.
Chem.
160-163 6
Holub,
W.R.
7
Bucolo,
G..
Yabut,
(1973)
Clin.
8
Rover,
M.E.
and
Ko,
9
Foster,
L.B.
and
Dunn,
M. and
Kreur,
Eggstein,
11
Schuttringer,
G. (1973)
12
.Jansen.
Peters,
13
Ttxt
memo
for
Yen
H. (1969)
10
A.P.,
Chem.
J. and
R.T.
19. Chap, Anal.
(1972)
1391
Clin. Klin.
Clin.
Chem.
19.
Triglycerides
and
Zelders, (I 972)
Clin.
Biochem.
(1966)
K.A.
1393
T. (1975)
E‘.H.
Svmposiutn.
N.Y.
29,
Chrm.
Chem.
19.
Abstract
T. (1970) Boehringer
420-424
338-340
Wochenschr., 673.
21,
405-416
Clin.
267-273 No. Chim.
Mannheim
179 Acta
27.
125-l
32
Klin.
Biochem.
10,
Clinica ~~i~~~~a
0 Elsevier
WA
Acta, 64 (1975) 45-50 Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlands
7265
AN IMPROVED SEMI-AUTOMATED METHOD FOR THE COLORIMETRIC DETERMINATION OF TRIGLYCERIDES IN SERUM P.N.M. DEMACKER, and A.P. JANSEN
J.B.H.A.
VAN OPPENRAAY,
H. BAADENHUIJSEN
Department of Internal Medicine, Division of Clinical Chemistry, Sin1 Radboudziekenhuis, University of Nijmegen, Geert Grooteplein Zuid 16 (The Netherlands) (Received
April 22, 1975)
Summary A semi-automated calorimetric determination for triglycerides in serum is described. The extraction is performed in sample cups and a part of the upper layer is sampled automatically. The use of modified reagents leads to greater stability and results in only one phase in the continuous flow system. The method is simple and inexpensive and therefore convenient for most laboratories with increasing requests for triglyceride determinations in serum. The method also shows good reproducibility and excellent correlation with the enzymatic method (Eggstein, M. and Kreuz, F.H. (1966) Klin. Wochenschr., 267-273).
Introduction Hyperlipemia is frequently accompanied by coronary heart disease. It is widely accepted that combined determination of cholesterol and triglycerides is the most reliable way to detect hyperlipemia in man [l].This is reflected in an increasing number of requests for serum triglyceride determinations in clinical laboratories. Hence, a simple and inexpensive automated method for this determination would be useful. Many semi- and fully automated applications have been described [ 2-71. However, most chemical methods, based on continuous flow, involve instable reagents, complex flow patterns and phase separations, while the enzymatic methods are expensive. Moreover, most semi-automated methods are laborious. A semi-automated chemical method is described here that meets most of the above-mentioned objections. The extraction of triglycerides is performed in large glass sample cups, based on the procedure as described by Royer and Ko [8]. The extraction is performed manually in the sample cups and then they are placed in the sampler. After separation of the two phases, an ahquot of the
upper layer is sampled automatically and the inter-samplta Lvasti Inkcs pluc*o from an external reservoir. In our system the extract is analysed in a similar manner to that used l)y ltoycr and Ko [ 81. The triglyceridta fatty ac*itls arc’ transesterificd and the glycerol liberated is oxidized to formaldrhytlc. whicsh is then condensed in the> IIantzsch reaction with acetylacetotlrl and ammonia to 3,5-diacetyl-1;t-dihydrolutidine. give a colored (and fluorescent) product. The use of a different periodate and acet.ylac*etone reagent [ 91 leads to a greater stability; also, the water content in the continuous-flow system is restricted, as compared to the systems previously described. There is only one phase in the continuous-flow system, and a stable baseline is obtained. The method has good reproducibility and shows excellent correlat.ion with the enzymatic method of Eggstein and Kreuz IlO] . Materials
and methods
Reagents 1. Extraction
solvent, n-nonane/isopropanol (2.0 : 3.5, by vol). It is possible to use the inexpensive nonane-fraction (Fluka A.G. 74260) and isopropanol (Baker 8119). 2. Dilute sulfuric acid, 40 mmol/l. 3. Saponification reagent, dissolve 3.40 g of sodium methylate (Baker 2812) in 50 ml of distilled water, dilute to 1 litre with isopropanol. Allow the solution to stand overnight then pour into a clean bottle. During the following days a slight precipitate becomes fixed on the inner wall of the bottle and the solution becomes clear. For best results the reagent should be prepared fresh every week. 4. Periodate reagent, dissolve 144 g of ammonium acetate in about 600 ml of distilled water. Add 100 ml of glacial acetic acid and 3.48 g of potassium periodate, dilute to 1 litre with distilled water and dissolve overnight by continuous stirring. The reagent is stable for at least three months [9]. 5. Acetylacetonc reagent, dilute 30 ml of acetylacetone (Merck 9600) to 1 litre with isopropanol. This remains stable for at least three months 191. 6. Standard. A stock standard contains 50 mmol of triolein (Sigma Chemical Co.) per litre of extraction solvent. A working standard is prepared by diluting with extraction solvent and remains stable for at least three months at room temperature. A Cenco* sampler was modified for automatic sampling of a part of the upper layer. The pick-up mechanism moves in mutual perpendicular directions from the sample cups to an external reservoir with wash solution. The sampler plate is high enough to carry the sample cups. A Technicon sampler II may also be used 16,111. As a wash solution, distilled water is used. The sample to wash ratio is 1 : 2, and thirty samples can be analysed per hour. The reaction temperature is 55°C. Colorimetry is performed with the flow scheme as shown in Fig. 1. A Cenco* pump and flow-through calorimeter [12] are used and a Vitatron recorder. The calorimeter is equipped with a 20-mm flowcell and a
* Crnco
Brrda.
l’hr
Nrthcrlands
*solvatlex Fig.
1. Flow
diagram.
420~nm filter. Pump tubes and glass reaction coils are interconnected with polyethylene tubing (internal diameter 2 mm). After each run the system is washed with water. The saponification-reagent pump tube is changed once a month; the other pump tubes are changed every two months,
Extraction 0.2 ml of serum is pipetted
into cyiindric~ glass sample cups (40 mm X 14 mm, capacity 4.5 ml); 2.5 ml of extraction solvent and 0.5 ml of dilute sulfuric acid are added with a dispenser (Oxford). The cups are then closed with polyethylene caps. For the calibration, 0.2 ml of the working standards is pipetted into the sample cups. The solvent is evaporated under a stream of air and 0.2 ml of water is added to obtain similar conditions for standards and serum in the extraction step. Many standards and samples can be handled simultaneously. The cups are shaken manu~ly for one minute and the phases are allowed to separate. The cups are then placed in the sample plate and the analysis of the triglycerides in the upper phase is performed in the continuous-fI ow system. Results and discussion
Extraction Recently, fully automated applications for the calorimetric determination of triglycerides have been described [3,4]. However, the modification of the sampler into an extractor-sampler appears complicated, while the advantages of this modification are minimal. Tt does not save time, for the pipetting of the samples, which takes most of the time cannot be automated, since calibration with standards is then impossible. Manual extraction warrants a better control of the extraction step. The use of the ‘inexpensive nonane-fraction [4], instead of pure nonane, did not influence the specificity of the extraction. e-l-lecithin, bilirubin, glucose and glycerol showed no impo~ant interference (Table I).
TARLE:
I
INTERFERENCE
Extraction
IN
THE
TRIGLYCERIDF:
DE:Tl?RMINATIOh’
of:
Triglyccritle
value Exiwctr~d
F0UIld
0.1
ml
serum
+ 0.1
ml
distilled
water
1.39
1.43
0.7
ml
serum
+ 0.1
ml
glycerol
10
1.41
1.43
0.1
ml
serum
+ 0.1
ml
glucose
1.38
1:13
0.1
ml
serum
+ 0.1
ml
serum
0.2
ml
serum
+ 0.2
ml
u-L-lecithin
Experiments
12 with
mmolil mmolil high 10
bilirubin
110
~mol/l
nmolil
with triolein
showed recoveries
1.42
1.43
2.92
2.Xfi
of 93-1&I%
(IZ = 9).
Reaction conditions With the given optimal reagent concentrations a peak-plateau ratio of 0.74 was obtained. The peak-plateau ratio was not influenced appreciably by variation of the sample amount between 0.32 and 0.60 ml/min, but the peak height can so be regulated without the risk of disturbing the one-phase system. The peak height was also affected by the temperature; an elevation of the temperature to 60°C caused an increase of the peak height of 20%. Sample-wash conditions The peak height was proportional to the sample time between 20 and 45 s and the peak-plateau ratio increased from 0.27 to 0.82. For the inter-sample wash, water appeared superior to heptane or isopropanol. With a sample and
-
absorbance 10
08
1
-
STANDARDS
EXTRACTS
STANDARDS (mm01 /I)
I
06
04
1c 02 a5
0 !/
Fig.
2.
method.
Recorder The
peaks
tracing were
of
peak
recorded
appearance from
right
of to
left.
standards
and
extracts
for
the
proposed
triglyceride
49 ne w method
(mmol/I
) --
12
10
6
. y: 104x r-0999 n: 71
. .*I 8.
4
6
8
-003
10 enzymatic
12 method
14 (mmol/l )
Fig. 3. Graphic comparison of the proposed method with the ennaxatic method.
wash time of 40 and 80 s, respectively, the recording continually returns to the baseline level, even after a sample with a very high triglyceride content. The carryover is negligible and the reproducibiiity of the lower concentratioIls is better then with a sample and wash time of 30 and 60 s (Fig. 2). Therefore, a sample and wash time of 40 and 80 s is preferred, although the capacity of the system is lower than in most of the systems previously described. Waste The method is inexpensive, but the pollution by the organic solvents high. In our opinion it is desirable to gather and burn the waste.
is
reproducibility and assay precision The withinday precision of the method for specimens with triglyceride values of 0.8 and 2.9 mmol/l was 1.5 and 2.3%, respectively, (n = 10) and the day-today precision was 4.4 and 3.3% (n = 20). The sample cups on the sample plate were not covered. The possible error due to evaporation of the upper layer appeared to be negligible. The weight loss of the nonane extract due to evaporation during the longest time on the sample plate, amounted to only 1.7% at 25*C.
Comparison of t,his method with thr enzymatic method Kreuz [10,12] gave a regression equation Y = 1.04 X --0.03, coefficient was 0.999 ( IZ = 71) (Fig. 3). In the enzymatic method values were corrected for fret glycerol (0.15 mmol/l).
of 1l:ggstrin antI the c,orrelation the triglycberidr
Acknowledgements
tance
We thank Mrs Mary and The Netherlands
Thijssen Paddenburg Heart Foundation
for the excellent technical for financial support.
assis-
References 1
Fredrickson,
2
Levv.
D.S.,
A.L.
Vol.
I. PP. H. and
3
Ko.
4
Bjiirksti.n.
5
Schmidt,
Levy,
and
R.I.
Keyloun,
497-502, Royer.
Mediad M.F.
F. (1972) V.F.H..
dnd C.
Inc..
(1972)
Clin.
P.S. in
White Acta
V.G.,
(1967)
Med. 40,
Dahl,
N.
Advances Plains.
Biochem.
Chim.
Schmidt,
Lees,
(1970)
Engl.
in
.l. Med.
Automated
276.
9,%~-103
Analysis.
‘l‘t~chnlron
6. 144-159
143--152 K.
and
Stork,
H.
(1972)
%. Klin.
Chem.
160-163 6
Holub,
W.R.
7
Bucolo,
G..
Yabut,
(1973)
Clin.
8
Rover,
M.E.
and
Ko,
9
Foster,
L.B.
and
Dunn,
M. and
Kreur,
Eggstein,
11
Schuttringer,
G. (1973)
12
.Jansen.
Peters,
13
Ttxt
memo
for
Yen
H. (1969)
10
A.P.,
Chem.
J. and
R.T.
19. Chap, Anal.
(1972)
1391
Clin. Klin.
Clin.
Chem.
19.
Triglycerides
and
Zelders, (I 972)
Clin.
Biochem.
(1966)
K.A.
1393
T. (1975)
E‘.H.
Svmposiutn.
N.Y.
29,
Chrm.
Chem.
19.
Abstract
T. (1970) Boehringer
420-424
338-340
Wochenschr., 673.
21,
405-416
Clin.
267-273 No. Chim.
Mannheim
179 Acta
27.
125-l
32
Klin.
Biochem.
10,