Feb., 19511 HARRISON 77 The Microbiological Assay of Growth Factors after Separation by Paper Chromatography BY J. S. HARRISON (Presented at the meeti.Pzg of the Biological Methods Group o n Thursday, M a y 26th, 1949) A general method is described for the separation, identification and estimation of growth factors by means of paper partition chromatography. The apparatus consists of two crystallisation dishes of different diameters, one placed within the other; the inner dish, containing a paper cylinder, is covered by a tall beaker or bell-jar. On completion of the chromatographic separation and after removal of the solvent by evaporation, the positions of the factors on the chromatogram are located by placing strips cut from the paper on a synthetic agar medium that is deficient in the growth factor and inoculated with a suitable organism.The identified zones are then cut out and extracted with water, made up to known volume and aliquots submitted t o microbiological assay. The technique described has been applied with success to some of the growth factors in molasses. MANY natural products contain mixtures of growth factors that have similar stimulatory effects on micro-organisms ; for instance, pyridoxine, pyridoxal and pyridoxamine all stimulate certain strains of Saccharomyces carlsbergensis to the same extent. These three compounds have been estimated in mixtures by Rabinowitz and Snel1,l who used a differential micro- biological assay with three organisms, and they have been separated qualitatively by Winsten and Eigen,2 who used paper chromatography.We are not aware that any accurate quantita- tive determinations of growth factors have been carried out after separation,* although various workers have estimated amino acid$** and other substances, including inorganic compoundsY6 by soaking the portion of the paper that contains the substance in water and carrying out chemical tests on the solution. This general method has been developed to separate and estimate growth factors by microbiological assays. In particular, the technique has been applied to molasses for the quantitative separation of pantothenic acid from p-alanine and of biotin from desthiobiotin. After chromatographic separation, the growth factors were identified qualitatively by an agar plate method that fixed the positions on the chromatogram compared with those of the chemically pure factors, which were run separately on the same sheet.Microbiological assays were then carried out on the extracts from the portions of the chromatogram containing the separated factors. EXPERIMENTAL SEPARATION- Mixed growth factors in artificial mixtures or in natural products were separated by single-dimensional paper partition chromatography. The apparatus consisted of two crystallisation dishes of different diameters, the larger holding water saturated with the solvent and the smaller, which was placed inside this, containing the solvent saturated with water. A tall inverted beaker or a bell jar was placed over the smaller dish and fitted inside the larger. A sheet of Whatnian No. 1 filter-paper, 15 inches square, treated with suitable quantities of the test sample, was rolled into a cylinder, which was held together with metal clips and stood in the smaller dish.On account of the small amounts of growth factors present in some natural materials, it was often necessary to run on the chromatogram much more raw material than could be applied as one spot. This difficulty was overcome by applying the test material in a line along the lower edge of the filter-paper by means of a capillary pipette in such a way that the * Since this paper was presented, various methods have been described in the literature, e.g., J. P. Bowden and W. H. Peterson, -1. Biol. Chevn., 1949, 178, 533; J. T. Holden and E. E. Snell, Ibid., 1949, 178, 799; and H. Yacowitz, L.C. Norris and G. F. Heuser, Proc. SOC. Ex$. B’iol. _Wed., 1949, 71, 372.78 HARRISOX : THE MICROBIOLOGICAL ASSAY OF GROWTH FACTORS [Vol. 76 total amount of solids deposited per inch was not greater than 2 mg, otherwise the concentra- tion of solids was high enough to interfere with the running of the chromatogram by holding back part of the growth factors and so causing tailing. Beyond the end of the line of test substance were placed separate marker spots of the growth factors under examination, and the chromatogram was run with the chosen solvent. Tests with mixtures of amino acids, which gave easily identifiable colour reactions, showed that the line of substance being chromatographed travelled in a band parallel to the lower edge of the paper. IDENTIFICATION- After evaporation of the solvent, vertical strips above the marker spots were cut out and a narrow vertical strip of known width was cut from the centre of the portion of filter- paper holding the sample.These strips were placed on a large plate of synthetic agar medium deficient in the growth factor and inoculated with a suitable organism. The strips were removed after 5 minutes, by which time a proportion of the growth factors had been transferred to the medium. After incubation overnight, zones of growth marked the points to which the growth factors had travelled. This step was necessary because the RF values varied somewhat, and if the zones were near together it was important to locate them exactly. The zones given by the pure sub- stances were used to identify the naturally occurring factors. If the RF value for the natural factor and that of the pure substance agreed when several solvents were used, this was taken as evidence that the two were probably identical.ESTIMATION- Horizontal strips slightly wider than the zones of growth found with the sample under test were cut from the chromatogram. These were separately extracted with water, using a Waring blendor to homogenise the paper, which was then removed from the solution by filtering or centrifuging. The solution was made up to a known volume and aliquots were used to carry out a microbiological assay. BIOTIN AND DESTHIOBIOTIN The chromatographic method was first tested with biotin and desthiobiotin, because a reliable differential assay method was available to check the results.Most strains of yeast respond to both biotin and desthiobiotin, while a smaller number of yeasts, as well as certain other organisms, such as Lactobacillus arabinosus, respond specifically to biotin. By use of typical organisms in these two groups, the amount of biotin-like and desthiobiotin-like TABLE I RF VALUES OF BIOTIN AND DESTHIOBIOTIN RF value r A 7 Growth factor Phenol Butanol Collidine - lutidine Biotin, pure . . . . . . .. . . . . 0.95 0.23 0.70 and 0.85 29 in blackstrap molasses . . . . . . 0-95 0.33 - 33 in high test molasses . . . . . . __ 0-23 - Desthiobiotin, pure . . . . . . . . . . 0.86 0.48 91 in blackstrap molasses . . . . 0.88 0.52 9 9 in high test molasses . . .. - 0-38 2 7 in refiners’ cane molasses . . . . - 0.45 0.83 - substances in natural products such as molasses can be estimated.There has naturally been some doubt as to the precise identity of these substances, particularly those that react as desthiobiotin, in the absence of any test more specific than the microbiological response. Chromatograms run in butanol, phenol and a 1 + 1 collidine - lutidine mixture showed that the biotin-like and desthiobiotin-like substances in several types of beet molasses behaved in a similar way to biotin and desthiobiotin respectively. Table I gives the R F values obtained in these tests. It was found that both natural and synthetic biotin gave two distinct zones with RF values of 0.70 and 0.85 in collidine or a mixture of collidine and lutidine. Two- dimensional chromatography with the same solvents showed that the leading fraction (RF 0-85) changed progressively during the run to the slower fraction (RF 0.70).Feb., 19511 AFTER SEPARATION BY PAPER CHROMATOGRAPHY 79 Quantitative microbiological assays using S.carkbergensis or L. arabinosus as the test organism showed that biotin and desthiobiotin could be completely recovered from filter-paper after drying. After chromatography with butanol or phenol, the recovery was also 100 per cent. A sample of blackstrap molasses was tested after separation of the biotin and desthio- biotin, with the results shown in Table 11. The values quoted are the means of three or more assays. It will be seen that the biotin and desthiobiotin contents after chromatography agree well with those obtained by differential assay.TABLE I1 BIOTIN AND DESTHIOBIOTIN CONTENT OF BLACKSTRAP MOLASSES Growth factor content f 1 Differential Chromatographic* -7 Rutanol, 0 Pg/g Pg/g P d g Pg/g Growth factor Yeast, I.. avabznosus, Phenol, Biotin . . . . . . . . 1-4 1.1 1.4 1.3 Yesthiobiotin . . . . . . 0.9 - 1.2 1.2 * Assayed microbiologically with yeast as test organism. PANTOTHENIC ACID AND P-ALANINE Substances that react microbiologically as pantothenic acid and as ,5-alanine occur in molasses, but although pantothenic acid can be estimated specifically by L. arabinos~s and other organisms, a satisfactory method of assaying /?-alanine has not been found. The reason is that a-amino acids interfere with the /3-alanine activity,6 and many natural substances contain amino acids. The chromatographic method has been used to separate and estimate the two growth factors.Qualitative tests with three solvents showed that the growth factors in blackstrap cane molasses behaved chromatographically as /3-alanine and pantothenic acid. Butanol was found to give the best separation of the two factors, the RF value of /I-alanine being 0.02 and that of pantothenic acid 0.57. TABLE I11 Table I11 gives the values for various solvents. RF VALUES OF PANTOTHENIC ACID AND ,5-ALANINE RF value \ Growth factor Ghenol Butanol Collidine - lutidine Pantothenic acid, pure . . . . . . . . 0-73 0.57 0.67 3 1 - - in molasses . . . . . . 0.74 /3-Alanine, pure . . . . . . . . . . 0.63 0.02 0.2 1 75 in molasses . . . . . . . . 0.63 0.02 - Quantitative recoveries from paper before and after chromatography with butanol were close to 100 per cent., but this solvent was found to be unsatisfactory for the estimation of TABLE IV PANTOTHENIC ACID AND /3-ALANINE IN BLACKSTRAP MOLASSES Yeast assay Chromatographic : Chromatographic Growth factor direct, phenol, butanol, P=/g P6/g P 6 / 8 /3-Alanine .. . . .. . . c 70* 71 --t Pantothenic acid . . . . .. 32 30 2.5 * Obtained by correcting for inhibition due t o sc-amino acids. t a-Amino acids invalidate this estimation. 13-alanine in natural materials, because many a-amino acids had low RF values in butano and consequently the microbiological assay was invalidated. The most satisfactory results were obtained with phenol, which gave values for blackstrap molasses of 30.4 and 71.5 pg80 HARRISON [Vol.76 per g for pantothenic acid and /3-alanine respectively. The pantothenic acid content agreed with the value of 31-7 pg per g obtained by direct assay with S. cerevisiae, and the value for /3-alanine was of the same order as that obtained by differential assays corrected for the inhibitory effect of a-amino acids. Table IV shows the amounts of pantothenic acid and /3-alanine in molasses found by various methods. It should be noted that the growth factor measured is that available to the test organism. The total /3-alanine content of this sample of blackstrap molasses, determined after alkaline hydrolysis, was 170 pg per g. The chromato- graphic behaviour of the bound form has not been investigated. Preliminary tests show that pyridoxine, pyridoxal and pyridoxamine can be studied in a similar way, using butanol as the solvent.DISCUSSION OF RESULTS The technique described has already proved useful for the examination of some of the growth factors in molasses, and it is clear that modifications of the method could be used to clarify many points of a similar nature. For instance, it would be possible to examine cell contents by separating conjugates of growth factors, and to study the properties of these. If necessary, the information so obtained could be used to devise a suitable method of separating the factors on columns, when larger quantities would be available than in paper chromatography . The author wishes to thank the Directors of the Distillers Company Limited for permission to publish this paper. REFERENCES 1. 2. 3. 4. 5 . 6. RESEARCH AND DEVELOPMENT DEPARTMENT Rabinowitz, J. C., and Snell, E. E., J . Biol. Clievn., 1948, 176, 1157. Winsten, W. A., and Eigen, E., Proc. SOC. Exp. Biol. Med., 1948, 67, 513. Woiwod, A. J., Nature, 1948, 161, 169. Naftalin, L., Ibid., 1948, 161, 763. Arden, T. V., Burstall, F. H., Davies, G. R., Lewis, J. A., and Linstead, R. P., Ibid., 1948, 162, Hartelius, V., Cowipt. R z ~ . I-ab. Carlsberg., Sdv. Physiol., 1946, 24, 185. 691. THE DISTILLERS COMPANY LIMITED EPSOM, SURREY DISCUSSION DR. G. E. FOSTER asked whether Mr. Harrison had tried any chemical methods for detecting growth factors on the chromatograms. MR. HARRISON replied that he had used chemical methods only for /3-alanine. With biotin the total amount of growth factor present was only about 2 to 3 mpg, and he doubted if so small an amount could be detected chemically.