326 Analyst, June, 1974, Vol. 99, @. 326-329 A Method for the Determination of Volatile Fatty Acids in the Blood Plasma of Ruminant Animals BY J. W. GARDNER AND G. E. THOMPSON (Department of Physiology, The Hannah Research Institute, Ayr, KA 6 5HL, Scotland) An improved method for the determination of volatile fatty acids in the blood plasma of ruminant animals is described. The acids are extracted as their sodium salts with isopropyl alcohol and then dissolved in 9 + 1 diethyl ether - formic acid for determination by gas - liquid chromatography. THE accurate determination of volatile fatty acids in human or animal blood is difficult to achieve and has been attempted by a variety of methods.1-9 Many of these methods have involved the use of steam distillation, but it has been shown that steam distillation of the small amounts involved here is an irreproducible process and that considerable losses may occur.10 In the present work it has been found possible to extract the volatile fatty acids from plasma, as their sodium salts, with isopropyl alcohol, thereby eliminating the need for steam distillation.Previous techniques for the determination of the acids after extraction have used either paper or gas - liquid chromatography. The paper-chromatographic methods have depended on the production of a coloured derivative of the acids, which can be determined by densito- metry. Tranger3 prepared the hydroxamates from volatile fatty acid methyl esters and determined them by measuring the intensity of the colour formed with iron(II1) chloride. Considerable losses were reported when using this method.The gas - liquid chromatographic analyses of volatile fatty acids have generally been carried out in dilute solutions of mineral acid, especially metaphosphoric acid, and it is very difficult to obtain good results with such methods.ll Ghosting1, and anomalous peak broadening13 have been observed, and poor peak shapes are frequently obtained.14 In order to overcome these difficulties, volatile fatty acid derivatives have been pro- duced and determined by gas - liquid chromatographic methods.l5J6 Preliminary efforts, made in this laboratory, to apply these techniques to the volatile fatty acids in blood have been unsuccessful, probably because the minute amounts of acids present make it difficult to convert all of the acid into the derivative.A simple way of overcoming the problem of the gas - liquid chromatographic analysis of solutions of volatile fatty acids is to inject them on to the column in an organic acid. Zerilli, Brambilla and Rimorinis have suggested 9 + 1 acetone - formic acid as a suitable solvent system. In the present work 9 + 1 diethyl ether - formic acid has been found to give sharper peaks and less solvent tailing. EXPERIMENTAL REAGENTS- Formic acid (Analar)-This acid was fractionally distilled from a 500-ml flask by using a 0-5 m x 20 mm lagged glass column, packed with Fenske helices and fitted with a partial take-off head. The first 40 ml of distillate collected were redistilled in semi-micro apparatus and the first 10 ml of distillate collected were again fractionally distilled in the same apparatus.The first 2 ml of this distillate were found to be pure on gas - liquid chromatographic analysis and contained no traces of acetic, propionic and butyric acid, which had been present in the original volume of formic acid. Isopropyl alcohol-Used after fractional distillation with a 1-m Vigreux column. Diethyl ether (May and Baker Ltd., Proaalys)-Used without further purification. Samples of acetic, propionic, isobutyric, butyric, isovaleric (3-methylbutanoic) and valeric acids were supplied by BDH Chemicals Ltd. and were of the highest purity available. n-Hexanoic acid (Sigma Chemical Co. Lid.)-This reagent was better than 99 per cent. pure, and gas - liquid chromatographic analysis indicated that it contained none of the C, to C, volatile fatty acids.As preliminary investigation had indicated that n-hexanoic acid was 0 SAC and the authors.GARDNER AND THOMPSON 327 not present in appreciable amounts in sheep or bovine plasma it was chosen as an internal standard for the method. Aqueous solutions of the acid were made up to contain lob2 mg ml-l and were used as described in the following section. METHOD SAMPLE EXTRACTION AND PREPARATION FOR GAS - LIQUID CHROMATOGRAPHY- Take a 10-ml sample of blood and centrifuge it at 3500 r.p.m. for 15 minutes. Transfer, by pipette, 6.0 ml of the supernatant plasma into another centrifuge tube and add 1.0 ml of the n-hexanoic acid standard solution. The volumes of sample and internal standard must be accurately measured.Adjust the mixture to pH 9 to 10 by adding 0.05 M sodium hydroxide solution dropwise, then add 40 ml of redistilled isopropyl alcohol and remove the protein precipitate by centrifugation at 3500 r.p.m. for 30 minutes. Decant the supernatant liquid and evaporate it to dryness on a rotary film evaporator, the final 1 ml of solution being taken to dryness in a small (10 ml) B14 test-tube. Seal the top of this tube with a layer of Para- film (Gallenkamp Co. Ltd.) and inject 200 p1 of freshly mixed 9 + 1 ether - formic acid (triple distilled) through the film with a syringe. Seal the resultant hole immediately with another layer of Parafilm. Agitate the contents of the tube vigorously on a Whirlimixer and allow them to stand for 15 minutes. Finally, withdraw 10 pl of the solution from the tube and inject it on to the gas - liquid chromatographic column.GAS CHROMATOGRAPHY- A Pye Unicam, Model 104, dual-column gas chromatograph, fitted with flame-ionisat ion detectors, was used, and separations were carried out by using glass columns that were 214 cm long by 6 mm o.d., packed with 17 per cent. neopentyl glycol adipate (Phase Separations Ltd.) in 3 per cent. orthophosphoric acid supported on Embacel (May and Baker Ltd.) of 60 to 100 mesh (R. S. Reid, personal communication). Condition the packed column at 150 "C over- night. For volatile fatty acid analysis use the following conditions : column temperature, 105 "C; detector temperature, 150 "C; and nitrogen flow-rate, 60 ml min-l. (Optimise the hydrogen and air flow-rates for maximum response.) Inject the mixture of acids directly on to the column packing.Under these conditions typical retention times in minutes of the acids are (retention times relative to n-hexanoic acid in parentheses) : acetic, 3.13 (0.095) ; propionic, 5.5 (0.167) ; isobutyric, 7.0 (0.212) ; butyric, 9.38 (0.284) ; isovaleric, 12.5 (0.379) ; valeric, 17.5 (0.53) ; and n-hexanoic, 33.0 (1.0). PREPARATION OF CALIBRATION GRAPH- Aqueous solutions of each of the C, to C, volatile fatty acids were prepared in a range of concentrations in the expected physiological ranges (Le., acetic acid from 1 to 4 mg per 100 ml and the other acids from 0.01 to 0-04 mg per 100 ml). These solutions were made up in cali- brated glassware and standard mixtures of the acids were prepared from them in order to calibrate the method.Known amounts of each mixture were then taken through the analytical procedure exactly as a plasma sample would have been. The peak areas given by the individual acids were measured relative to that of the internal standard. The use of peak height x width at half-height as an estimate of peak area was found to give the closest results to cutting and weighing the chart paper and the former method was used. RESULTS AND DISCUSSION- Fig. 1 shows the calibration graph obtained for acetic acid and Fig. 2 those for propionic, isobutyric, butyric, isovaleric and valeric acids. Each graph approximates to a straight line and it was possible to determine a calibration factor for each acid that was the slope of the line in the calibration graph.It was then possible to calculate the original concentration of each acid in every blood sample, knowing the area of each peak and the amount of internal standard that had been added, by applying the following equation: (1) Peak area of acid x mass of n-hexanoic acid Peak area of n-hexanoic acid x calibration factor Mass of acid = * * ' * A typical chromatogram of a sample of sheep plasma is shown in Fig. 3. Four experi- ments were carried out in order to test the completeness of recovery of known amounts of328 GARDNER AND THOMPSON : DETERMINATION OF VOLATILE FATTY ACIDS [Afialyst, Vol. 99 0 -0 0 m u .- .- 16 m 0 f 12 m Y m a --. 9 v 8 0 .w u .- m 4 L m Y m p. 0 0 10 20 30 40 Mass of acetic acidbmass (1 0-2 mg) of n-hexanoic acid j I I I I Fig.1. Calibration graph for determina- tion of acetic acid. Slope = 0.44 added volatile fatty acids from plasma. Four separate plasma samples were each divided into two equal portions and one portion of each was used to determine the amounts of the individual volatile fatty acids present initially. A measured amount of a standard aqueous mixture of the acids was added to the remaining portion of each plasma sample and the resulting solutions were analysed. It was then possible to calculate the recovery of each acid, knowing the initial and final amounts. Results are shown in Table I as the individual acids present before and after addition of the standard mixtures. I I I I 1 I I I 0 5 10 15 20 I ' 35 Ti me/ m i nutes Fig. 3. A typical chromatogram of a sample of sheep hepatic portal vein plasma showing: 1, acetic; 2, propionic; 3, isobutyric; 4, butyric; 5, isovaleric; 6, valeric; and 7, n-hexanoic acid peaks. Attenuation factors: 1, x 2000; 2, x 500; 3, x 100; 4, x200; 5, x 100; 6, x 50; and 7, x200 A blank performed on 10 ml of distilled water contained no detectable amounts of any of the C, to C, acids and the results can therefore be taken as a measure of the accuracy of the method.The mean recoveries all lie in the range 92 to 106 per cent. There are evidently no substantial losses, except of propionic acid, but in this instance the mean is probablyJune, 19741 I N THE BLOOD PLASMA OF RUMINANT ANIMALS TABLE I RECOVERY OF KNOWN AMOUNTS OF VOLATILE FATTY ACIDS ADDED TO PLASMA 329 Original amount . . Amount added .. Total amount recovered Recovery, per cent. . . Original amount . . Amount added . . Total amount recovered Recovery, per cent. . . Original amount . . Amount added . . Total amount recovered Recovery, per cent. . . Original amount . . Amount added . . Total amount recovered Recovery, per cent. . . Mean recovery, per cent. Standard deviation . , Acetic acid/mg 0-182 0.210 0.380 96.9 0.146 0.210 0.362 101.7 0.168 0.420 0.575 97.8 0.185 0.420 0.580 95.9 98-1 2.5 Propionic acid/mg 0.0037 0.0020 0.0058 101.7 0.00 17 0.0020 0.0035 94.6 0.0045 0.0041 0.0082 95.3 0.0049 0.0041 0.0070 77.8 92.3 10.3 Isobutyric acid/mg 0.0003 0.0020 0.0025 108.7 0~0002 0.0020 0.0024 109.0 0.0006 0.0039 0.004 1 91.1 0.0015 0.0039 0.0057 105.5 103.6 8.5 Butyric acid/mg 0-0027 0.0019 0.005 1 110.8 0.0012 0.0019 0.0029 0.0025 0.0039 0.0065 93.5 103.1 0.0017 0-0039 0*0060 107.2 103.7 7.5 Isovaleric acid/mg 0.0006 0.0019 0.0027 108.0 0*0003 0.0019 0.0024 109.0 0.0008 0.0037 0.0049 108.9 0.0008 0.0037 0-0044 97.8 105.9 5.4 Valeric acid/mg 0*0008 0.00 13 0.0018 85.7 0.0003 0-0013 0.0015 93.8 0.0006 0.0026 0.0035 109.4 0.0004 0.0026 0.0034 112.3 100.6 13.0 artificially low, owing to the final recovery figure of 77.8 per cent.The over-all accuracy of the method could possibly be increased by the use of an integrator for the measurement of peak area. CONCLUSION The method described in this paper for the determination of volatile fatty acids in plasma has been shown to be quantitative and reproducible for the peripheral blood of ruminants. It has the distinct advantage, over published methods involving steam distillation, of relative speed and good recoveries.The gas-chromatographic technique is accurate and reliable and, after several months of continuous use, there is no evidence of column deterioration or the build-up of unstable products. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. REFERENCES Annison, E. F., Biochem. J., 1954, 58, 671. Erwin, E. S., Marcos, G. J., and Emery, E. M., J . Dairy Sci., 1961, 44, 1768. Tranger, M., Vet. Archiv., 1963, 33, 248. Storry, J. E., and Millard, D., J . Sci. Fd Agric., 1965, 16, 417. Tanaka, K., Budd, M. A., Effron, M. L., and Isselbacker, K. I., Proc. Natn. A cad. Sci. U.S.A ., 1966, Kurtz, D. J., and Levy, M. L., Clinica Chim. Acta., 1971, 34, 463. Ross, J. P., and Kitts, W. D., J . Dairy Sci., 1971, 54, 1824. Mahadevan, V., and Zieve, L., J . Lipid Res., 1969, 10, 338. Zerilli, L. F., Brambilla, E., and Rimorini, N., Atti. Soc. ITtaZ. Sci. Vet., 1971, 25, 173. Edwards, G. B., McManus, W. R., and Bigham, M. C., J. Chromat., 1971, 63, 397. Holman, R. T., Editor, “Progress in the Chemistry of Fats and Other Lipids,” Volume 12, Pergamon Geddes, D. A. M., and Gilmour, M. N., J. Chromat. Sci., 1970, 8, 394. Ackman, R. G., Ibid., 1972, 10, 506. OttensJein, D. M., and Bartley, D. A., Ibid., 1971, 9, 673. Downing, D. T., and Greene, R. S., Analyt. Chem., 1968, 40, 827. Umeh, E. O., J. Chronmt., 1970, 51, 147. 56, 236. Press, Oxford, 1972, p. 258. Received October Sth, 1973 Accepted December 31st, 1973