March, 19511 ASSAY OF ANEURINE IN FOODSTUFFS 133 The Assay of Vitamin B12 Part HI* Microbiological Estimation with Lactobacillus Zactis Dorner by the Plate Method BY W. F. J. CUTHBERTSON, H. F. PEGLER AND JOAN T. LLOYD (Presented at the meeting of the Biological Methods Group on Tuesday, May 23rd) 1950) The cup-plate method adapted by Bacharach and Cuthbertson to the rapid microbiological assay of aneurine and riboflavine has been found applicable to the estimation of vitamin U,, with a suitable strain of Lacto- bacillus lactis Dorner. The procedure involves the use of a medium similar to those found necessary for other strains of lactobacillus. Ascorbic acid is needed in the assay medium but not in that for preparation of the inoculum. The technique is similar to that described previously for the method.Factors affecting the “zone” response of this organism to vitamin B,, have been investigated; they include the effects of vitamin B12C and the deoxy- ribosides, times of incubation and standing before incubation, range of concentrations of the vitamin and density of the inoculum. As a result a standardised (2 + 2) assay design has been devised; this permits rapid estimates of potency over a range sufficiently wide and with fiducial limits sufficiently narrow for routine purposes, without the use of an inconveniently large number of assay plates. A GROWTH factor for Lactobacillus Zactis Dorner ATCC 8000 has been shown by Shorbf. to be present in highly refined liver extracts in amounts proportional to the anti-pernicious- anaemia activity. This factor has been isolated by Rickes, Brink, Koniuszy, Wood and Folkersz and by E.L. Smith,3 and shown to be responsible for the clinical activity of liver extracts against pernicious anaemia; it is now known as vitamin BIZ. We were unable to develop a satisfactory assay procedure based on the data published by Shorb. In our experience the growth responses of this organism were highly irregular and not capable of giving reproducible results in normal microbiological assay procedures. Lacto bacillus fermenturn P 36, which in our hands gave very irregular results in standard techniques, could be used to assay aneurine with fair precision by the cup-plate method of Bacharach and Cuthbertson.* We therefore attempted to apply this technique to the estimation of vitamin B,% with L.lactis. METHOD In the cup-plate method a suitable agar medium, deficient in the factor under investiga- tion and inoculated with the test organism, is poured into petri dishes and “cups” are cut out of the agar and filled with appropriate dilutions of standard or test solutions. On * For particulars of earlier papers in this series (not in The Analyst), see reference list, p. 140.134 CUTHBERTSON, PEGLER AND LLOYD: ASSAY OF VITAMIN B12 [Vol. 76 incubation under suitable conditions, growth of the organism may occur in sharply defined circular "zones of exhibition" around the cups. The diameters of these zones have been found to be related to the concentration of the ,added growth factor and can be measured and used in the determination of the amount of growth factor present in the test solutions.The method developed in the estimation of vitamin B, is very similar to the technique employed in the estimation of aneurine by Bacharach and C~thbertson,~ but modifications have been introduced in the light of observations by Cuthbertson and Lloyd5 on the mode of growth of L. Zactis and its response to vitamin B,,. MAINTENANCE OF ORGANISM- The culture of L. Zactis is maintained by weekly transfer in soya bean medium (10 per cent. suspension of whole soya beans macerated in a Waring Blenclor and strained through coarse muslin). The inoculum is prepared from this by subculture in 10 ml of complete basal medium. This medium is the same as that described in Table I except for the omission of agar and ascorbic acid and the addition of 0*OOc5 to 0-02 pg of vitamin B,, (as liver extract) per ml.The inoculum is allowed to grow for 15 -to 16 hours and is then centrifuged, washed once in saline and diluted with saline to corresponld to turbidity 4 on the Burroughs Wellcome turbidity scale. Cultures more than +7 to 18 hours old are frequently unsatisfactory for assay purposes. PREPARATION OF THE MEDIUM- The medium is of the composition shown in Table I. The ingredients, except salts, are dissolved in about 800ml of water; after adjustment of the pH to 6.8, the phosphates TABLE I COMPOSITION OF AGAR MEDIA DEFICIENT I N VITAMIN B,, 1 mg Acid casein hydrolysate . . 5 g Nicotinic acid .. .. Glucose . . .. .. .. l o g Riboflavine .. .. .. 2oopg Sodium acetate, A.R... .. 6€! Adenine . . .. .. 10 mg .. . . . . 10 mg L-Cystine .. .. .. 200mg Guanine . . .. .. 10 mg DL-Tryptophan . . .. .. 100mg Uracil . . Tomato juice . . .. .. 60 ml .. I g Pg Tween 80t .. Aneurine hydrochloride Biotin . . .. Folic acid .. 2 P.tg .. I g p-Aminobenzoic acid . . .. 100 p g Ascorbic acid .. Pyridoxamine .. . . . . 400 p g KH2P04 . . . . . . 0.6 g . . . . . . . . Salt solution C; . . .. 5 ml * . 20; pg Calcium-D-pantothenate . . 200 p g K,HP04 .. .. . . 0.6 g Water to make . . . . . . 1000 ml pH adjusted to . . .. .. 6-8 Agar (N.2.): . . .. .. to 15g * Salts C: a solution of 10 g of MgSO4.7H2O, 0-5 g of sodium chloride, 0.5 g of FeS0,.7H,O, 2.0 gof t This amount of Tween 80 may be replaced by 0.39 g of Estax 36 or 29 (Watford Chemical Co.). 3 From Davis Gelatine Ltd., 29, Mitre Street, E.C.3.MnS0,.4H20, and water to 250 ml. and salt solution are added. The precipitate formed during this procedure is removed by filtration through washed paper pulp on a Buchner funnel. The medium is made up to volume and adjusted to pH 6-8. The agar is added to this solution and dissolved by steaming for 1 hour. The medium is now filtered, while still hot, through washed paper pulp, distributed in 130-ml aliquots and kept in 250-ml flasks. The flasks are plugged and autoclaved a t 10 lb steam pressure for 10 minutes. At this stage the medium may be kept for several weeks in a refrigerator. PREPARATION OF ASSAY PLATES- The flasks of medium are steamed for 35 minutes to melt the agar and are then cooled to 4 5 O to 50" C in a thermostatically controlled e:nclosure at 45" C.One millilitre of 0.13 per cent. ascorbic acid solution is now added. Each flask is inoculated with 1 ml of a freshly prepared suspension of L. Zactis adjusted to opacity 4 on the Burroughs Wellcome scale. The inoculum is well mixed in the medium, which is then distributed in the petri dishes at the rate of 1 2 6 d per plate. The plates are cooled in a refrigerator for half an hour to harden the agar; they may be stored in a cold room overnight or even for a week if not neededMarch, 19511 WITH Lactobacillus Zactis DORNER BY THE PLATE METHOD 136 at once. The plates are cut with a sharp 8-mm No. 10 cork-borer; the resulting discs of agar are removed with a needle leaving the cups into which standard and sample solutions may be distributed.TYPE OF PETRI DISHES REQUIRED- For this technique it is essential to use perfectly flat petri dishes so as to obtain a layer of agar of uniform thickness. The diameters of the zones of exhibition vary widely with differences in thickness of the agar. TIME OF INCUBATION- After the plates have been prepared they are incubated a t 37" C overnight. Further incubation does not alter the size of the zones of exhibition, but on certain subnormal media further incubation may lead to some improvement in the density of growth within the zones. CONCEIGTRATION OF VITAMIN B12- The diameter of the zones of growth is proportional to the logarithms of the amounts of vitamin B,, placed in the cups; this is shown in Fig. 1. This relationship holds true over a wide range of concentrations.The clarity with which the rings are defined decreases 10 0005 0-02 005 0-1 0;2 0 Response of L. Eactis Dorner ATCC 8000 to Each point corres- Vitamin 812, pg per ml vitamin B,, in the cup plate assay. ponds to the average reading obtained from six cups Fig. 1. slightly as the concentration of vitamin B,, decreases; under our conditions well defined rings are obtained with concentrations of vitamin B,, in the range 0.01 to 0.5 pg of vitamin B,, per ml. TIME OF STANDING BEFORE INCUBATION- As in the assays of penicillin and streptomycin, the time the plates are allowed to stand before incubation but after distribution of sample and standard solutions in the cups affects the size of the growth zones obtained. This effect is variable from occasion to occasion, but typical results are shown in Table 11.EFFECT OF INOCULUM DENSITY- The inoculum used affects both the size and the appearance of the zones of exhibition. In general, the heavier the inoculum the more sharply are the rings defined and the smaller are the zones. The effect of inoculum density is shown in Table 111.136 CUTHBERTSON, PEGLER AND LLO’YD: ASSAY OF VITAMIN B,, pol. 76 TABLE :[I RELATIONSHIP BETWEEN ZONE DIAMETER AND TIME INTERVAL BETWEEN Two experiments performed 0171 two different occasions FILLING PLATES AND PLACING I N INCUBATOR Concentration of vitamin B,, solution r 1 Time 0.05 pg per ml 0.2 pg per ml Time interval, Zone diameter, Zone diameter, interval, minutes mm mm hours 0 17.8 20.2 0 16 18.5 20.2 1.5 20 19.5 20.4 2-5 3.5 15-0 Concentration of vitamin B,, solution 0-05 pg per ml Zone diameter, Zone diameter, mm mm 16.12 19.0 16-5 19.12 17.62 21.0 17-75 20-9 26.0 30-0 A r \ 0.2 pg.per ml Each diameter is the average from three or four cups.TABLE ‘I11 EFFECT OF INOCU LUM DENSITY Volume of inoculum (turbidity 4) per plate (12.5 ml of agar medium), ml 0.0 1 0.02 0.06 0.1 0.2 Diameter o:E zone of exhibition for vitamin B, concentrations of- 7 A 3 0-05 pg per ml, 0-2 pg per ml, mm mm 21.2 24.4 20.8 23.8 19.9 23.5 19-6 23.6 17.9 20.4 Each diameter is the average from four cups. ASSAY DESIGN AND RESULTS For assay purposes, dilutions of standard vitamin B,, and test solutions are chosen to fall within the range 0.02 to 0.2 pg of vitamin B,,, per ml, over which the relationship between log dose and zone diameter has been found linear (Fig.1). Experiments have shown that the variance between zone diameters for the sa.me concentration of vitamin B, was much less when the cups used were on the same plate than when cups on different plates were compared; the “within plate” variance was about one-tenth of the “between plates” variance. Owing to the size of the petri dishes (9 cm dia.meter) and the size of the zones (13 to 25 mm) normally encountered, not more than 5 or 6 cups may conveniently be used on a plate. Consequently, for routine assays, two cups on each plate are reserved for standard solutions, the concentration of vitamin B,, employed being 0.02 and 0-2pg of vitamin B,, per ml. Two cups are normally used for the test solutions, which are diluted to fall within the range of 0.01 to 0-5 pg of vitamin B,, per ml.Two different concentrations of test solutions are used in the ratio of 1 to 10. The plates are fdled in the following order- (;) Lower test concentration. (ii) 0.02 pg of vitamin B12 per ml. (iii) 0-2 pg of vitamin B,, per ml. (iv) Higher test concentration. Three drops, 0-1 to 0.15m1, of solution, measured with a standard dropping pipette (see Bacharach and Cuthbertson*) are placed in each cup; the same dropping pipette is used for all dilutions of test and standard solutions. After the cups have been filled, the plates are left for a t least 10 minutes before being put in the incubator. This procedure has been adopted to minimise the effects of standing before incubation and of any differences between dropping pipettes.After incubation the zone diameters may be measured with a ruler (transparent celluloid rules are suitable) or callipers, or after projection on to a screen. The normal procedures (Irwine) applicable to the calculation of the results of assays involving a linear relationship between log-dose and response may be applied and fiducial limits may be determined.March, 19511 Plate 1 2 3 4 5 6 7 8 Average WITH LactobaciZZus lactis DORNER BY THE PLATE METHOD TABLE IV ASSAY OF VITAMIN B1, I N LIVER EXTRACT Variance analysis according to method of Irwin6 ANALYSIS OF VARIANCE- Standard vitamin B,, (3 drops per cup) 1 0.02 pg 0.2 I G 13-0 20.0 15.0 22.0 14.0 21.0 14.0 2 1.0 14-0 21.0 13.0 18.5 13.0 19.0 13.5 20.5 13.7 20-4 Sum of D.F.squares .. 1 374.1 Source of variation Common linear reaession .. Difference due to kbstances . . .. 1 9.6 Bias due to plates . . .. .. 7 16.6 Error . . .. .. .. .. 21 4.0 Total . . .. .. .. .. 31 405.1 Departure from parallel regression . . 1 0.2 Test preparation (3 drops per cup diluted as shown) - 1/750 1/75 14.0 21.0 15.0 22.0 15.0 22-0 15.0 23.0 15.0 22.0 14.0 20.0 14.0 21.0 15.0 22.0 14.6 21.6 Mean Variance square ratio Significance 374-7 19-72 significant 9.6 50.52 significant 0.2 1.05 not significant 2.37 12.47 significant 0.190 137 Activity found: 21.64 pg per ml. Fiducial range (P = 0-95) : 19.49 pg per ml t o 24.20 pg per ml; 90 per cent. t o 112 per cent. A typical assay result is shown in Table IV, together with the calculated fiducial limits.On average, fiducial limits of 415 per cent. (P = 0.05) may be expected in assays of this type employing six plates, although narrower limits can be obtained with a greater number of plates. Assays by the plate method and by a tube technique employing L . Zeichmannii 313 (Lees and Emery') have given comparable results (see Cuthbertson, Lloyd, Emery and Lees3). INTERFERENCE FROM DEOXYRIBOSIDES Under our conditions L. Zactis responds to the deoxyribosides as well as to vitamin B, (see Smith and Cuthbert~on~). The response to thymidine is shown in Table V. Zones TABLE V RESPONSE OF LactobaciZZus Zactis TO THYMIDINE (PLATE TEST) Diameter of zone of Thymidine concentration, exhibition P*.g Per ml 1 6 10 20 50 no zone 24 mm approx., indistinct zone 25.9 mm 28.1 mm 34 mm Each diameter is the average from four cups of exhibition of less than about 25 mm are not formed and only solutions containing 6 pg or more of deoxyriboside per ml give zones of exhibition. Up to 5 pg of deoxyribosides per ml cause only very slight interference in the plate test.Table VI shows the diameters of the zones of exhibition produced by mixtures of thymidine and vitamin B12. Deoxyriboside zones are readily distinguished from vitamin B,, zones, for the growth is much fainter and more diffuse. If vitamin B,, is present an inner zone of dense growth may be seen, but when the concentration of thymidine is high this zone is not well defined and cannot be measured accurately.138 CUTHBERTSON, PEGLER AND LLOYD : ASSAY OF VITAMlN B12 [vol. 76 Samples containing deoxyribosides are readily detected by the appearance of the growth zones and also by the slope of the log-dose - response curve, which is much greater for deoxy- ribosides than for vitamin B,, as should be clear from Table VI and Fig.2. TABLE VI RESPONSE OF L. Zactis TO MIXTURES OF THYMIDINE AND VITAMIN B,, (PLATE TEST) Composition of solution r \ Vitamin BIZ, Thymidine, 0.02 0 0-02 1 0.02 6 0.02 10 0.02 20 0.02 60 0.2 0 0.2 1 0.2 5 0-2 10 0.2 20 0.2 50 A Per PLg Per ml Diameters of zones of exhibition 7 A \ Inner zone due to Outer zone due to vitamin B12, thymidine, rnm mm 13.1 no zone 13.45 no zone 13.25 24.1 not measurable 26.6 99 28.0 99 30.1 17-05 no zone 17.1 no zone 17.2 23.2 not measurable 25.9 99 28.0 99 29.9 USE OF CHROMATOGRAPHY Ih’ THE PLATE ASSAY Interference due to the deoxyribosides may be eliminated by a technique previously indicated by Smith and Cuthberts~n.~ Micro-drops of about 1 to 4 4 , containing 0.004 to 0.04 pg of vitamin BI2, of test and standard solutions are placed about 1 inch from one edge of a square sheet of Whatman No.4 filter-paper and about 1 inch apart. Micro-drops of saturated riboflavine solution are placed just beneath but not touching the positions occupied 1 10 20 50 Thymidine,yg per ml 2o 5 Fig. 2. Response of L. Zactis Dorner ATCC 8000 to thymidine in the cup plate assay. Each point corresponds to the average reading obtained from four cups 4 Vitamin BI1. pg Fig. 3. Response of L. Zactis Dorner ATCC 8000 to vitamin B,, after separation of deoxyribosides by partition chromatography.Each point represents the average zone diameter obtained with four aliquots applied to the filter paper by the standard and test spots. The paper is dried in the incubator and then rolled into a cylinder and held in this way by metal paper-clips. The filter-paper cylinder is now placed in a beaker containing a layer of water-saturated n-butanol about 1 cm deep. Upward development is allowed to take place for about 1Q to 2 hours, i.e., until the riboflavine has passed through and about 0.5 cm beyond the sites at which the vitamin B, and test solutionsMarch, 1951 J 139 had been placed. The paper is then removed and dried in the incubator, and when dried it is cut along the front reached by the riboflavine. During this brief development the vitamin B,, does not move from its site of application, while the deoxyribosides (except cytosine deoxyriboside) travel with and beyond the riboflavin e.When a strip of paper to which the test and vitamin B,, solutions have been applied is placed on an inoculated agar test medium and incubated overnight, circular zones of exhibi- tion are produced by the vitamin B,,. The type of response obtained is shown in Fig. 3. With only four 2 - 4 drops of each standard and test solution, fiducial limits (P = 0-05) of 60 to 150 per cent. were obtained in the assay of samples containing 3 to 5 pg of vitamin B,, per ml. WITH Lactobacillus Zactis DORNER BY THE PLATE METHOD ASSAY OF VITAMIN B12 CONCENTRATES FROM FERMENTATION LIQUORS In the assay of extracts from streptomyces the results by the plate and tube methods (Lees and Emery’) often differed widely, the plate technique almost invariably giving higher results; in some instances they were three times those of the tube assay.It was noticed that the zones of exhibition obtained with these materials were not as sharp as those normally encountered and it was at first suspected that the interference was due to some substance other than vitamin B12; however, it was not possible to separate the vitamin B,, activity from the postulated interfering agent. It has recently been shown that another red cobalt-containing clinically active compound, in addition to the previously described members of the vitamin B,, group, is present in streptomyces fermentation products (Smithlo). This substance, designated BlZc in accordance with the suggestion of Buchanan, Johnson, Miles and Todcl,ll itself yields poorly defined zones of exhibition in the plate test under conditions used for vitamin B,, assay; weight for weight it appears to be 3 to 4 times as active as vitamin B,,, although in tube assays and clinically (Ungley, Mollin and Dacie12) its activity is approximately the same as that of vitamin B12.We have not been able to overcome the difficulty that thus appears to be due to vitamin 1312c. Until a suitable technique becomes available for doing this, results of plate assays on fermentation liquors may be very unreliable. FAILURE TO ASSAY SOLID MATERIALS CONTAINING ADSORBED VITAMIN B,, Vitamin B,, is firmly adsorbed by fuller’s earth. The assay of such adsorbates presents some difficulties.It was not found possible to elute completely all the vitamin B,, taken up by this material. When fuller’s earth adsorbate was placed on plates prepared for vitamin B,, assay, growth was noticed around the sites of application. Attempts were therefore made to develop an assay from this observation. Plates were prepared and cut in the normal way and aliquots of aqueous dispersions of the adsorbate were placed in the cups. After incubation the zones of growth around the cups containing the fuller’s earth were found to be closely comparable in qualitative appearance to the zones around the cups in which standard vitamin B,, had been placed; further, the assay data showed that the nature of the response curves obtained with the fuller’s earth adsorbate were not different from those obtained with aqueous solutions of vitamin B,, and that assays were statistically valid when tested by the usual mathematical analysis.The results were, however, much lower than those obtained by chemical assay (Fantes and Ireland13) or by elution techniques, e.g., one adsorbate gave 25, 48 and 50 pg per g respectively when tested by the plate assay, an elution technique and the chemical method. These observations are reported because with adsorbates the plate method gives results that are very consistent and appear to be valid when checked by methods in frequent use; for instance, another sample of adsorbate was tested on two separate occasions and gave results all within the range of 11 to 13 pg per g, although on other grounds it was believed to contain at least 30 pg of vitamin B,, per g.A fuller account of the assay of adsorbates is to be reported later. DISCUSSION OF THE METHOD The main advantages of the cup-plate technique are its speed and simplicity. Zones of exhibition are formed overnight, but they appear and may be measured after incubation for as little as 6 hours in favourable circumstances, whereas the tube methods require a t least overnight incubation for L. Zeichmanniil* and 4 days for EugZena gracilis.ls With the cup-plate method samples need not be sterilised; in the tube methods strict asepsis is140 CUTHBERTSON, PEGLER AND LLOYD: ASSAY OF VITAMIN I312 essential and all samples must be sterilised by filtration through collodion or sintered glass.Autoclaving or seitz filtration may lead to loss of vitamin B,,. Samples containing less than 50 rnpg per ml cannot be assayed satisfactorily, whereas the tubc methods employing L. leichmannii and Euglena graczlis may be used for concentrations as low as 2-0 mpg or 2 ppg of vitamin B,, per ml, respectively, in favourable circumstances. Thymidine and the other deoxyribosides may replace vitamin B,, in the nutrition of L. ZactisfG or L. Zeichmannii.17 In the tube techniques growth due to these substances is not readily differentiated from that due to vitamin BI2, but in the plate method the appearance of the growth zones due to the different types of substances is a t once discernible, because the deoxyribosides give wide diffuse zones of exhibition , while vitamin B,, under our conditions gives dense sharply defined zones of growth; hence confusion does not arise and interference is noticed a t once.Similarly, though antibiotics and toxic substances in general may interfere in any microbiological test, such interference is readily detected in the plate test by the appearance of zones of partial growth or rings of growth inhibition, Pol. 76 The cup-plate method is relatively insensit-be. . SUMMAKY Lactobacillus lactis Dorner ATCC 8000 requires vitamin B,, for its normal nutrition and has been foundsuitable for assay of the vitamin. by the cup-plate method. Vitamin B, and the deoxyribosides interfere with the procedure. The effect of the latter, but not of the former, can be eliminated by combining the method with payer chromato- graphy.The presence of deoxyribosides is generally apparent from the type of zone produced. The combined chromatographic and microbiological procedures make it possible, with a (2 + 2) assay design, dose ratios of 10 to 1 and incubation overnight, to attain a satis- factory degree of precision on a few microlitres of vitamin B,, solution. The method gives results in good accord with those given by the tube method based upon growth of Lactobacillus leichmannii 313. Our thanks are due to Dr. T. G. Brady, Trinity College, Dublin, for a very generous gift of thymidine and to Mr. B. Basil, Glaxo :Laboratories Ltd., Greenford, for statistical calculations. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. REFERENCES Shorb, M. S., J . Biol.Chem.. 1947, 167, 455. Rickes, E. L., Brink, N. G., Koniuszy, F. R., Wood, T. R., and Folkers, K., Science, 1948, 107, Smith, E. L., Nature, 1948, 161, 638. Bacharach, A. L., and Cuthbertson, W. F. J., Analyst, 1948, 73, 334. Cuthbertson, W. F. J., and Lloyd, J. T., J . (Yen. Microbiol., 1951, 5 , 416. Irwin, J. O., J . Roy. Statist. SOC., 1937, Suppl. 4, 1. Lees, K. A., and Emery, W. B., Biochem. J., 1949, 45, ii. Cuthbertson, W. F. J., Lloyd, J. F., Emery, W. B., and Lees, K. A., J . Pharnt. Pharmacol., 1949, Smith, E. L., and Cuthbertson, W. F. J., Bicchem. J., 1949, 45, xii. Smith, E. L., PYOC. Roy. SOC. Med., 1950, 43, 535. Buchanan, J. G., Johnson, A. W., Miles, J. A,, and Todd, A. R., Chem. and Ind., 1950, 426. Ungley, C. C., Mollin, D. L., and Dacje, J. V., Lancet, 1950, 1, 353; Proc. Roy. SOC. Med., inthe Fantcs, K. I<., Ireland, D. M., and Green, N., Biochcnt. J . , 1950, 46, xxxiv. Hoffmann, C. E., Stokstad, E. L. R., Franklin, A. L., and Jukes, T. H., J . Biol. Chem., 1948, Hutner, S. H., Provasoli, L., Stokstad, E. L. R., Hoffmann, C. E., Belt, M., Franklin, A. L., and Kocher, V., and Schindler, 0.. Intern. Rev. Vzt. Res., 1949, 20, 441. Snell, E. E., Kitay, E., and McNutt, W. S., .J. Biol. Chem., 1948, 175, 473. 396. 1, 705. press. 176, 1465. Jukes, T. H., Proc. SOC. Ex@. Biol. Med., 1949, 70, 118. NOTE-Reference 5 is to Part I of this series. GLAXO LABORATORIES LTD. GREENFORD, MIDDLESEXMarch, 19511 WITH Lactobacihs lacfis DORNER BY THE PLATE METHOD DISCUSSION 141 MR. I<. A. LEES confirmed that when vitamin BIZ, was assayed by the L. lactis plate technique the values obtained were 3 to 4 times as high as those obtained by the tube methods or physical chemical procedures. In Mr. Lees’ laboratories the zones of growth with vitamin BIzc in plate tests employing L. leichmannii 413 were poorly defined. The type of growth obtained could, however, be readily differentiated from the responses given by thymidine, although the edges of the zones were not sharp, because the growth within the zones was as dense as that observed with vitamin BI2. To overcome these disadvantages a plate method employing an E. coli mutant1 has been developed and all pure factors so far tested give clearly defined zones. The E. coli technique has the further advantagc that a very simple medium is employed. REFERENCE TO DISCUSSION 1. Bessell, Christine J., Harrison, Eleanor, and Lees, I<. R., “Assay of Vitamin B, with a Mutant of Escherichia Cola,” Chem. and Ind., 1950, 561.