Microbiological Determination of Vitamins and Amino Acids Produced by Microorganisms, Using the Dialysis Cell1 VEIKKO NURMIKKO Laboratory of Valio, Biochemical Institute, Helsinki, Finland Received for publication November 21, 1956 Investigations on the associative interrelationships between microorganisms have engaged increasing attention in recent years. As is well known, these relationships play an important role in nature and also in many branches of industrial.microbiology. In view of the great significance of these associative phenomena in biology, and especially in microbiological ecology, this is undoubtedly a field in which great and important a b c 1954, 1955, 1956). This method affords opportunities for elucidating the biosynthetic pathways of microbial growth factors such as vitamins and amino acids. More recently, an approach to the study of the interrelationships among microorganisms has been made (Nurmikko, 1955b). A simple apparatus has been constructed in which the organisms under investigation are separated from each other by one or more dialysis FIG. 1. Three-compartment dialysis cell; a, b, and c. A, dialysis membrane; B, rubber gaskets; C, aluminum hoops; D, set screws; E, rubber packings. discoveries have yet to be made. However, new techniques will certainly be developed to solve the difficult problems accompanying the study of the biochemistry of mixed microbial populations. The classical pure culture methods are in most cases inapplicable to these studies, or yield only scant information on the chemical factors affecting associations of microorganisms. The author has previously shown that in certain cases the chemical factors affecting symbiosis among lactic bacteria are vitamins of the B complex and amino acids (Nurmikko, 1952). On the basis of the symbiotic interrelationships between different species of lactic acid bacteria, a method, designated the symbiotic technique, has been developed (Nurmikko 1 This paper deals with those growth factors which are secreted by microorganisms into the growth medium. Production of some growth factors with certain microorganisms, comparing especially the secreted part of growth factors to the amounts of those reserved by the cells, will be presented in a later paper. 160 E membranes in a cell divided into two or more compartments. In these experimental conditions microorganisms can also grow in associations composed of more than two species. In this paper a method is described for assaying such growth factors as vitamins and amino acids produced by microorganisms, using various dialysis cell systems. This method permits the assay directly in the growing cultures. MATERIALS AND METHODS Description of dialysis cell. The apparatus consists of two or more compartments made of Pyrex glass, each furnished with a glass tube open at the top, which can be plugged with cotton wool. Figure 1 shows a threecompartment dialysis cell. Each compartment has approximately a 12- to 15-ml capacity (not including glass tube). As shown in figure 1, dialysis membrane (A)2 is placed between the compartments, the surfaces 2 Cellulose No. 4465-A2, Arthur H. Thomas Co., Philadelphia, Pennsylvania.
1957] DETERMINATION OF VITAMINS AND AMINO ACIDS 161 FIG. 2. Six of the three-compartment dialysis cells assembled on a rack. f FIG. 3. Modification of dialysis cell with two compartments; f and c, growth chambers. :. :..ra...;. -ass FIG. 4. Modification of dialysis cell with two compartments; a and d, growth chambers. between the glass sections being ground smoothly, and held in place by rubber gaskets (B). The compartments are connected by aluminium hoops (C) equipped with set screws (D) and rubber packings (E). As the walls of the glass compartments and the rubber packings are slightly conical, the joints can be made water-tight by tightening the screws (D). Figure 2 shows six of the di FIG. 5. Modification of dialysis cell, a and c, growth chambers. Middle compartment consists of a 50-ml Erlenmeyer flask. NX....;E :... 0: d FIG. 6. Modification of dialysis cell with two compartments; d, growth chamber. three-compartment dialysis cells assembled on the rack for the experiment. o The three-compartment dialysis cell described above was employed mainly when lactic acid bacteria were used as test organisms in microbiological determinations. Figure 5 shows a modification of the same apparatus for the experiments in which a Neurospora mutant was used as test organism. This dialysis cell likewise consists of three compartments, but the middle compartment consisted of a 50-ml Erlenmeyer flask. Figures 3, 4, and 6 show other modifications of the dialysis cell with two compartments. The use of these cells will be described later in this paper. Cultures and inocula. The organisms used were Lactobacillus arabinosus strain 17-5, Streptococcus faecalis strain R, Escherichia coli, Oospora lactis, and Neurospora sitophila (ATCC 9276). The lactic acid bacteria stock cultures were transferred by stab inoculation into the glucose-citrate-tryptone-yeast extract agar medium of the following composition: 1 per cent glucose, 1 per cent sodium citrate, 0.5 per cent Bactotryptone (Difco), 1.5 per cent agar and 20 per cent (by volume) yeast extract. The yeast extract was prepared as follows: 1 kg of fresh yeast was suspended in 1 L of water, and kept for approximately 24 hr at 42 C. It was then centrifuged and the cell-free solution
162 V. NURMIKKO [VOL. 5 used. The ph of the agar medium was 6.7. The medium was sterilized by autoclaving for 15 min at 112 to 115 C (Nurmikko, 1954). E. coli was maintained on Bacto Nutrient Agar (Difco) slants with transfers at biweekly intervals, incubated at 37 C for 24 hr. 0. lactis and N. sitophila were maintained on Bacto Neurospora Culture Agar (Difco) slants by biweekly transfer, with incubation at 30 C for 24 to 48 hr. The inocula of lactic acid bacteria and of E. coli were prepared by transferring the organisms from the stab culture to the glucose-citrate-tryptone-yeast extract medium. After incubation for 16 to 18 hr at 37 C, the cells were centrifuged out and washed with 0.9 per cent sterile saline. This process was repeated, the cells being washed 2 to 3 times. The cells were finally suspended in saline. One drop of barely visible suspension was used as inoculum for approximately 5 ml of the final medium in the dialysis cell and in the test tubes. 0. lactis and N. sitophila were transferred directly with a sterile platinum loop from the agar slants to saline, care being taken not to transfer agar medium together with the spores or mycelium. Assay procedure. The dialysis technique of microbiological assay for vitamins and amino acids produced by the microorganisms was as follows: Using lactic acid bacteria as test organism, the basal media of Hendersoni and Snell (1948) (in phenylalanine assay), and Anderson and Elliker (1953) (in nicotinic acid assay) in slightly modified form as described previously (Nurmikko 1954, 1955a) were used. In assays with Neurospora sitophila, Bacto Pyridoxine Assay Medium (Difco) was used. All basal media, free of the vitamin or amino acid to be determined, were prepared at twice their final concentration. The standard vitamin or amino acid solutions were added at increasing levels to the test tubes (Pyrex 14 x 155 mm) when lactic acid bacteria were used as test organisms, or to the Erlenmeyer flasks (50 ml) when the assay organism was N. sitophila. Duplicate tubes or Erlenmeyer flasks were set up to obtain the values for the standard curves. Using lactic acid bacteria, the standard vitamin or amino acid solution was added in the following amounts: 0, 0.5, 1.0, 1.5, 2.0, and 2.5 ml per tube. Quantities of 2.5 ml of the basal medium were then placed in the test tubes, and the total volume adjusted with distilled water to 5 ml. In pyridoxine assays with Neurospora, 15-ml quantities of the basal medium were placed in Erlenmeyer flasks. Up to 15 ml of standard pyridoxine solution were added and the volume adjusted with distilled water to 30 ml. In all cases, the basal medium used was diluted with an equal volume of distilled water and placed in the dialysis cells. After autoclaving at 112 C for 7 min, the dialysis cells and test tubes or Erlenmeyer flasks were cooled, then inoculated and incubated at 37 C (lactic acid bacteria as test organisms) or 30 C (Neurospora as test organism). The growth response of lactic acid bacteria was followed turbidimetrically. The extent of growth was recorded as galvanometer readings on a Klett-Summerson photoelectric colorimeter with a 660-mA filter. In vitamin B6 assay, dry mycelial weights of Neurospora were determined. The mycelium was harvested by filtering the culture medium through a filter paper placed in a Buchner funnel. After washing with distilled water, the mycelium was dried at 100 C for 3 to 4 hr and weighed. Also, the extent of the growth of E. coli and 0. lactis was measured by weighing the dried cells. The cells of E. coli were harvested by centrifugation, and the cells of Oospora by filtering the culture filtrates through a membrane filter. Both organisms were washed with distilled water and dried at 100 C overnight. A standard curve was prepared by plotting growth (dry weight of the cells) against micrograms of vitamin or amino acid per ml in the test tubes or Erlenmeyer flasks of the standard series. From this standard curve the amount of vitamin or amino acid per ml in the compartment of the assay organism was determined. Since the volume of the growth medium in the compartment of the test organism was exactly known, it was possible to calculate the amount of the vitamin or amino acid which was produced by the microorganism investigated. An example of the calculation: In a separate phenylalanine assay with a 3-compartment dialysis cell (figure 1; in table 5, dialysis cell no. 1), it was calculated from the standard curve that the production of L-phenylalanine by E. coli was 1.33,ug per ml. The volume of the growth medium in the middle compartment was 18.9 ml. Thus, the whole production of L-phenylalanine was 1.33 X 18.9 = 25.1,ug. The dry cell weight of E. coli (obtained from both the end compartments) was 16.7 mg. Consequently the production of L-phenylalanine by E. coli per mg of dry weight was 1.5,ug. RESULTS Control experiments. To test the accuracy of the dialysis method, a series of control experiments was carried out to determine the percentage recovery. In these experiments, nicotinic acid and phenylalanine assays were made with L. arabinosus 17-5 and folic acid assay with S. faecalis R as test organism, using a 3-compartment dialysis cell of the model shown in figure 1. In pyridoxine determinations, the test organism was N. sitophila (ATCC 9276) and the apparatus was the 3-compartment dialysis cell illustrated in figure 5. In all determinations, known amounts of amino acid or vitamin were added to the two end compartments after sterilization and cooling. The middle compartment was then immediately inoculated with the assay
1957] DETERMINATION OF VITAMINS AND AMINO ACIDS 163 TABLE 1. Nicotinic Acid of nicotinic acid pg,ag pg per cent 0.20 0.21 +0.01 105 0.20 0.19-0.01 95 0.20 0.19-0.01 95 0.20 0.25 +0.05 125 0.20 0.23 +0.03 115 0.40 0.39-0.01 98 0.40 0.40 0.00 100 0.40 0.41 +0.01 102 0.40 0.41 +0.01 102 0.40 0.42 +0.02 105 pg 0.40 0.40 TABLE 3. Pyridoxine pg 0.41 0.40 0.54 0.48 0.56 0.52 0.62 0.56 0.63 0.58 of pyridoxine pg +0.01 0.00 +0.04-0.02 +0.06 +0.02 +0.02-0.04 +0.03-0.02 per cent 102 100 108 96 112 104 103 93 105 97 Average... 104.2 Three-compartment dialysis cells. Lactobacillus arabinosus TABLE 2. Folic Acid of folic acid mpg mpg pg per cent 8.0 8.3 +0.3 104 8.0 7.7-0.3 96 8.0 8.3 +0.3 104 8.0 8.4 +0.4 105 8.0 9.2 +1.2 115 8.0 8.3 +0.3 104 8.0 7.8-0.2 98 8.0 7.6-0.4 95. Average.102.6 Three-compartment dialysis cells. Streptococcus faecalis strain R as test organism. organism. In the experiments in which L. arabinosus 17-5 was used, the time of incubation was 72 hr. In the Neurospora assays the time of incubation varied from 60 to 72 hr. An indication of the accuracy of the method may be obtained from the results of the recovery experiments presented in tables 1 to 4. The results show that the quantitative recovery of nicotinic acid, folic acid, pyridoxine and phenylalanine with mean average 102.9 per cent (when all determinations of these compounds were taken into consideration) was obtained when known amounts of these compounds were added directly to the dialysis cell used for assaying by the procedure described. The recoveries of the compounds tested were between 93 and 125 per cent. The greatest variation was found in nicotinic acid assays (ranging from 95 to 125 per cent) and the lowest variation in phenylalanine assays (ranging from 100 to 108 per cent). It was found that the variation was approximately ± 5 per cent of the average (102.9 per cent) recovery. Average...... 102.0 Three-compartment dialysis cells. Neurospora sitophila as test organism. TABLE 4. of L-phenylalanine L-Phenylalanine pg pg pg per cent 20.0 20.9 +0.9 104 20.0 21.7 +1.7 108 20.0 20.1 +0.1 100 40.0 40.3 +0.3 101 80.0 84.7 +4.7 106 80.0 83.2 +3.2 104 80.0 81.0 +1.0 101 80.0 81.0 +1.0 101 Average.103.0 Three-compartment dialysis cells. Lactobacillus arabinosus Experiments on the production of growth factors. In order to obtain information on the usefulness of the dialysis method, the production of nicotinic acid, folic acid, and phenylalanine by E. coli and the production of pyridoxine by 0. lactis were studied. The test organisms and the basal media in these determinations were the same as those used in the control experiments. Folic acid, phenylalanine and pyridoxine assays were carried out with a 3-compartment dialysis cell and nicotinic acid assay with a 2-compartment dialysis cell. Using the 3-compartment dialysis cell, the two end compartments were inoculated with the microorganisms to be investigated and the middle one with the assay organism. Using the 2-compartment dialysis cell in nicotinic acid assay, one compartment was inoculated with E. coli and the other with L. arabinosus 17-5. The results of these experiments are given in tables 5 to 8. As can be seen in table 5, the production of L- phenylalanine by E. coli in the experimental conditions used was 1.4 Ag per mg of cell dry weight. This is a mean value obtained from 5 individual determinations
164 V. NURMIKKO [VOL. 5 TABLE 5. Production of L-phenylalanine by Escherichia coli Dialysis Cell Growth of E. coli L-Phenylalanine L-Phenylalanine No. (Dry Wt of Cells) Produced Produced per Mg (Dry Wt of Cells) mg pg pg 1 16.7 25.1 1.50 2 13.6 20.0 1.47 3 10.4 16.6 1.60 4 11.9 15.4 1.29 5 13.4 17.6 1.31 Average.1.43 Three-compartment dialysis cells. Lactobacillus arabinosus TABLE 6. Production of nicotinic acid by Escherichia coli Dialysis Cell Growth of E. coli tinic Acid Nicotinic Acid Pro- No. (Dry Wt of NPcodced duced Per Mg (Dry Cells)Prdued Wt of Cells) Mg pg pg 1 8.0 0.324 0.040 2 9.2 0.376 0.041 3 10.1 0.431 0.431 4 7.1 0.270 0.038 5 9.6 0.445 0.046 6 10.0 0.340 0.034 Average.0.040 Two-compartment dialysis cells. Lactobacilluts arabinosuts TABLE 7. Production of folic acid by Escherichia coli Dialysis Cell Growth of E. coli Folic Acid Produced No. (Dry Wt of Folic Acid Produced per Mg (Dry Wt of No. ~~~Cells) Cells) mg mpg mpg. 1 17.7 4.62 0.260 2 17.1 4.65 0.274 3 18.4 4.73 0.257 4 18.1 4.99 0.276 5 20.3 4.73 0.233 Average... 0.260 Three-compartment dialysis cells. Streptococcuts faecalis strain R as test organism. TABLE 8. Production of vitamin B6 (pyridoxine) by Oospora Dialysis Cell Growth of Oospora Vitamin Bo Vitamin B6 Pro- No. (Dry Cells) rue Wt of Cells) WtCof Produced duced per Mg (Dry Mg pg mpag 1 67.7 0.63 9.3 2 39.3 0.39 9.9 3 52.8 0.56 10.6 4 53.5 0.54 10.1 5 60.0 10.0 Average.10.0 Three-compartment dialysis cells. Neurospora sitophila (ATCC 9276) as test organism. (using 5 replications). The variation of the individual determinations is approximately + 8 per cent of the average value. Tables 6 and 7 show that the production of nicotinic acid and folic acid by the same organism was 40.0 m,ug and 0.26 m,ug, respectively, per mg of the dry weight of the cells. In nicotinic acid assay (6 replications), the variation is approximately 4-10 per cent and in folic acid assay (5 replications) i 5 per cent of the average value. In all these experiments with E. coli, the time of incubation was 72 hr. The separate determinations showed that this organism can grow very rapidly in the basal media used in this study. The organism reached its. maximum growth after approximately 10 hr. When a nicotinic acid assay was made using an incubation time of 92 hr, practically the same nicotinic acid production was observed as in the experiments with an incubation time of 72 hr. In table 8 it can be seen that the production of vitamin B6 (revealed as pyridoxine) by 0. lactis was 10.0 m,ig per mg of dry weight of the cells. This mean value was obtained using 5 replicates. The variation of the individual determinations was approximately -+f 4 per cent of the average value. An incubation time of 120 hr was used in this experiment. DISCUSSION The principal purpose of the work reported here was to develop a simple microbiological method for the quantitative estimation of vitamins and amino acids directly in growing microbial cultures. The dialysis cell technique described in this paper, which permits the determination of these compounds in the manner desired, is based on their known action as growth factors in associations of microorganisms. The use of dialysis cell systems affords the following advantages. Firstly, the performance of the determinations with the dialysis cell apparatus is easy and time-saving, because this method is designed to obviate the usual necessity of making separate microbiological or chemical determinations from the culture filtrates. Secondly, because the vitamin or amino acid investigated in the dialysis cell system passes through the dialysing membrane to the adjoining compartment containing the test organism, which utilizes this compound immediately for growth, the possibility of the growth factor produced being to any extent inactivated remains very meager. In addition, the fact that the compound tested cannot accumulate in the culture fluids in the dialysis cell system (because this compound will be utilized by the test organism as rapidly as the other microorganism in the adjoining compartment is capable of producing it) may have a considerable influence on the amount of the growth factor produced. Figures 1 to 6 illustrate 5 modifications of the dialysis cell used for the determination of the vitamins and
1957] DETERMINATION OF VITAMINS AND AMINO ACIDS 165 amino acids produced by microorganisms. The dialysis cell to be used must be chosen according to the organism employed. For example, with an organism which grows best on the surface of the medium, it would be preferable to use compartments of type c or d (figures 3 to 6). In these cases especially, shaking of the dialysis cells, and also the solutions of the standard series, may be necessary in order to attain rapid dialysis of the growth factors. The modifications shown in figures 1 to 4 were designed for use with lactic acid bacteria as test organisms. Because these organisms are microaerophilic, they must be inoculated into compartments a, b, or f. When Neurospora mutants are to be used as assay organisms, it is preferable to choose one of the dialysis cells shown in figures 3 to 6. In this situation, the mold inoculum would be inoculated into compartments c or d. In determinations of the growth factors described in this paper, the variation of the average value using 5 or 6 replications per growth factor tested was of the order found in microbiological assays, that is, from 5 to 10 per cent. It is of special interest that the average recoveries of the 3 added vitamins and 1 amino acid from the different dialysis cell systems was 102.9 per cent. This suggests that the test organism can really take up the compound tested quantitatively from the adjoining compartment, through the dialysis membrane. In the course of developing the dialysis method, it was found to be necessary to take into special consideration the following precautions. In separate experiments it should be verified that no inhibitory growth effect exists in the dialysis cell system between the test organism and the organism whose growth factor production is to be investigated. If the growth of this organism and its growth factor production was especially vigorous, the dialysis cell system was constructed with a view to allowing sufficient nutrient medium for the maximum growth of the test organism (at least at the level of the maximum growth value of the standard series). For example, in pyridoxine determination the compartment of the type a (figure 5) was used for Oospora. The small air area on the surface of the nutrient medium is the factor limiting the growth of this organism, and thus also indirectly the production of pyridoxine. In these growth conditions, the nutrient medium was sufficient for the assay organism Neurospora. Further, it is preferable that the volume of the nutrient solution in the tubes of the standard series and in the compartment of the test organism should be the same. If it seems probable that the activity of the compound tested will decrease during autoclaving, the compound can be added aseptically to the tubes of the standard series after autoclaving of the basal medium. In this case the sterilization of this compound can be accomplished by filtration. Finally, it should be pointed out that the determination described here is to be regarded as mainly illustrative. Using as examples the determinations of phenylalanine, folic acid, nicotinic acid and vitamin B6, an attempt has been made to give a picture of the general procedure of the dialysis technique. Thus, the estimation of other vitamins and amino acids and the use of other microorganisms require other basal media and modifications of the growth conditions described in this paper. ACKNOWLEDGMENT The author wishes to thank Professor Artturi I. Virtanen, Director of the Biochemical Institute, Helsinki, Finland, for his encouraging interest in this work and for his valuable discussions. SUMMARY A rapid microbiological procedure is described for the determination of vitamins and amino acids produced by microorganisms. Primarily, this method was developed for the investigation of biochemical factors of importance in microbial associations. Using a dialysis cell system especially constructed for this purpose, it was possible to make analyses directly in the growing cultures. Various types of dialysis cell apparatus are described. experiments were carried out with phenylalanine, folic acid, nicotinic acid, and vitamin B6. The production of these growth factors with Escherichia coli and Oospora lactis by means of this method was also studied. REFERENCES ANDERSON, A. W. AND ELLIKER, R. P. 1953 The nutritional requirements of lactic acid streptococci isolated from starter cultures. I. Growth in a synthetic medium. J. Dairy Sci., 36, 161-167. HENDERSON, L. M. AND SNELL, E. E. 1948 A uniform medium for determination of amino acids with various microorganisms. J. Biol. Chem., 172, 15-29. NURMIKKO, V. 1952 Chemical factors affecting associations of lactic acid bacteria. Acta Chem. Scand., 6, 1258-1264. NURMIKKO, V. 1954 Symbiosis experiments concerning the production and biosynthesis of certain amino acids and vitamins in associations of lactic acid bacteria. Ann. Acad. Sci. Fennicae Ser. A II, 54, 1-58. NURMIKKO, V. 1955 Application of the symbiosis phenomenon among lactic acid bacteria to the study of the biosynthetic pathways of growth factors. Ann. Acad. Sci. Fennicae Ser. A II, 60, 216-225. NURMIKKO, V. 1955a Phenylalanine as a precursor in the biosynthesis of folinic acid (citrovorum factor) in lactic acid bacteria. Suomen Kemistilehti B, 28, 62-66. NURMIKKO, V. 1955b The dialysis technique in the study of the vitamins and amino acids affecting associations of microorganisms. Acta Chem. Scand., 9, 1317-1322. NURMIEKO, V. 1956 Biochemical factors affecting symbiosis among bacteria. Experientia 12, 24-249.