A. M. GAUDIN, A. L. MULAR, AN-D R. F. O'CONNOR

determination of nitrate and nitrite in soil. Analyst, 80,141-145.

COOPER, C. M., FERNSTROM, G. A., AND MILLER, S. A. 1944Performance of agitated gas-liquid contactors. Ind. Eng.Chem., 36, 504-509.

DONALD, C., PASSEY, B. I., AND SWABY, R. J. 1952 Bioassayof available trace metals from Australian soils. Austral-ian J. Agr. Research, 3, 305-325.

ENGEL, M. S. AND ALEXANDER, M. 1958 Culture of Nitr o-somnonas europaea in media free of insoluble constituents.Nature, 181, 136.

EYLAR, 0. R., JR. AND SCHMIDT, E. L. 1959 A survey ofheterotrophic micro-organisms from soil for ability to formnitrite and nitrate. J. Gen. Microbiol., 20, 473-481.

GUNDERSEN, K. 1955 Effects of B-vitamins and amino-acidson nitrification. Physiol. Plantarum, 8, 136-141.

MEIKLEJOHN, J. 1952 Minimtum phosphate and magnesiumrequirements of nitrifying bacteria. Nature, 170, 1131.

MEIKLEJOHN, J. 1953 Iron and the nitrifying bacteria. J.Gen. Microbiol., 8, 58-65.

MEYERHOF, 0. 1916a Untersuchungen uber den Atmungs-vorgang nitrifiizierender Bakterien. I. Die Atmtung desNitratbildners. Arch. ges. Physiol., 164, 353-427.

MEYERHOF, 0. 1916b Untersuchungen uber den Atmungs-vorgang nitrifizierender Bakterien. IL. Beeinflussungender Atmuing des Nitratbildners durch chemische Sub-stanzen. Arch. ges. Physiol., 165, 229-284.

STOJANOVIC, B. J. AND ALEXANDER, M. 1958 Effect of inor-ganic nitrogen on nitrification. Soil Sci., 86, 208-215.

ZAVARZIN, G. A. 1958 The efTect of heavy metals on nitri-fication (In Russian). Mikrobiologiya, 27, 542-546.

Separation of Microorganisms by Flotation

I. Development and Evaluation of Assay Procedures1

A. M. GAUDIN,2 A. L. MULAR,3 AND R. F. O'CONNOR3

Departmnent of Metallurgy, Massachusetts Institutte of 7'echnology, Canmbridge, Massachusetts

Received for publication August 10, 1959

The separation of bacterial cells from the growthmenstruum is a routine operation which continues todemand more than its fair share of effort from workersin such fields as physiology, immunology and cytology.Preparation of "clean" cell suspensions by centrifu-gation is arduous and yields are often low.The observation that masses of cells collected in the

foam above the liquid level in fermentors led Boylesand Lincoln (1958) to use vigorous aeration of theculture medium after growth as a means for removingand concentrating spores of Bacillus anthracis. Quanti-tative estimates of the degree of separation and/orconcentration were based on viable counts. Success inseparating spores of Bacillus cereus T (formerly calledBacillus cereus var. terminalis) from autolyzed cultureby frothing was reported by Black et al., (1958).The concept of the separation of particles in liquid

by foaming has long been known in the field of mineralengineering, and is the basis for "flotation" processes.Flotation studies on minerals have provided, in ad-dition to rapid, efficient separation techniques, muchbasic information concerning the structures and surfaceconfigurations of mineral particles. Flotation testing of

I This work was done at 'Massachusetts Inistitute of Tech-nology tiunder contract Nith the U. S. Army Chemical Corps.

2 Professor of 1lIineral Engineering, Malssachuisetts Instituteof Technology, Cambridge, MAassachusetts.

3 Staff member, Division of Sponsored Research, Massa-chusetts Instituite of Technology, Cambridge, 'Massachusetts.

Bacillus subtilis var. niger was undertaken because ofthe possibility that such investigation might lead tothe development of methods for separating and concen-trating bacterial cells, and might also provide a newtool for studying the composition and molecular orien-tation of cell surfaces.

This paper presents the results of some initial flo-tation testing with B. subtilis var. niger related to thedevelopment of two assay techniques for the evaluationof flotation tests. One assay is based upon a chemicalanalysis for a constituent of spores, pyridine-2,6-di-carboxylic acid (dipicolinic acid); and the other assayis based upon a procedure in which cells, spores, andother particles are visually counted.

DESCRIPTION AND DISCUSSION OF FLOTATION

Definitions. 1'lotation may be defined as a processfor separatinig finely divided solids from each other(Gaudin, 1957). The solids are suspended in waterthrough which gas bubbles are caused to flow. Sepa-ration takes place when particles of one type adhereto gas bubbles and are carried to the top of the liquidas a froth, whereas particles of other types adhere tothe liquid and remain in suspension.That product of a flotation operation which contains

the valuable or preferred constituent is called the"concentrate" (C) and usually is the froth or "float."The other product contains the worthless constituentanid is called the "tailing" (T) which usually is the non-

84 [VOL. 8

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

SEPARATION OF MICROORGANISMS BY FLOTATION. I

float. The initial suspension, which is divided by theflotation operation into the concentrate and tailing, iscalled the "feed" (F). The relationship which thesethree products bear to one another is expressed by theequation:

F =C+T (1)

This assumes no loss or gain in any component duringflotation-an assumption that is valid for most systems.This relationship may be applied to include the volumesand weights of the three products as in equation (1),or it may be applied to any one constituent, for exampleto the spore content in the form

Ff = Cc + Tt (2)

In this equation the lower-case letters denote the assayor percentage content of the component of interest ineach of the products denoted by capital letters. Theassays can be expressed on a weight per unit volume orweight per unit weight basis.

Equations (1) and (2), which are balance-of-materialsrelationships, are basic to any quantitative expressionof the separations achieved.

Indices for evaluation. In flotation, as in any physicalseparation process, evaluation of the effectiveness ofthe operation requires the quantitative measurement(or assay) of either the preferred or unwanted con-stituent in a product, or, better still, of both. Theseassays are translated into ratios or indices which makeclear the degree of success in separation. Two indicesof effectiveness commonly used in flotation are "grade"(or purity) and "recovery." The grade expresses thenumber of parts of a constituent per hundred parts ofdry product, for example, a concentrate containing 2 gof spores and 3 g of other solids would have a grade of40 per cent. The recovery is the percentage of the totalamount of a constituent from a feed which is found ina product or

100 Cc (3)Ff

A flotation test cannot be evaluated unless bothrecovery and grade are measured. A feed containing10 g of spores and 1 g of vegetative cells might yield aconcentrate containing 9 g of the spores, and the testwould be judged a success from the standpoint of re-covery. But if the concentrate also contained 0.9 g ofthe vegetative cells, the grade has not been altered.Since the object of flotation is the separation of solidsfrom each other, the analysis of a flotation test shouldbe made on the basis of both grade and recovery.

Equations (1) and (2) assume that during the sepa-ration there is neither increase nor decrease in thequantity of any component or in the sum of the com-ponents of the system. This is true of inanimate systemsbut it can be far from true in living systems where

new organisms are born and others die. In particular,they are likely to fail in microbiological systems wherethe life span is short, especially if the results are evalu-ated on the basis of viability tests only. This is becausethe separation operation may well include the use ofreagents that may be toxic or inhibitory to bacteria sothat their viability or germination will be affected. Insuch cases output will not equal input and recoverycalculations may be greatly in error. All this makes ithighly desirable to develop assay methods that do notdepend on viability, and to use these methods in ad-dition to viability assays.The two assays to be described here are independent

of viability, permit close balances of materials, andagree well with each other.

MATERIALS AND METHODSM1aterials

Media and preparation of cultures. The organismused throughout this study was Bacillus subtilis var.niger. Spore concentrates were obtained from the U. S.Army Chemical Corps, Fort Detrick, Frederick,Maryland, or were produced at Massachusetts Insti-tute of Technology (M.I.T.). Spores from Fort Detrickwere grown in a casein hydrolyzate medium (Boyles andLincoln, 1958) in metal fermentors and, after autolysis,were harvested by centrifugation at 50,000 X g. Thespore concentrate, or paste, was shipped both frozenand unfrozen, with and without an additional washingoperation in the centrifuge. Freezing was accomplishedby extruding the paste into an acetone-Dry Ice bath.Frozen preparations were stored at -20 C, whereasunfrozen preparations were kept at 4 C. Frozen prepa-rations contained about 50 X 1010 viable spores per gof dry wt.

Spores were grown at M.I.T. on a medium containingSheffield HY Case-SF (BBL),4 1 g; glucose, 2.5 g; yeastextract (Difco),5 5g; MnSO4, 0.1 g; FeSO4, 0.001 g;CaCl2, 0.05 g; and agar, 30 g in 1 L of tap water atpH 6.8. Pyrex baking dishes measuring 8% x 131/x 1½ in. fitted with 1½6 in. thick aluminum covers wereused as incubation vessels. Each vessel contained 500ml of medium. Incubation was at room temperature(26 to 30 C) for 5 days, at which time sporulation andautolysis were essentially completed. Spores werewashed off the agar surfaces with distilled water, andresuspended in cold distilled water. If further washingswere desired, they were carried out using a Servall6Model A centrifuge operated at 5000 X g.

Yields were of the order of 32 to 35 X 1010 cells pervessel, of which about 5 per cent were heat sensitive.

Viability studies. Plate counts were made by thespread plate technique, using a glass rod to spread 0.2

4 Baltimore Biological Laboratories, Baltimore, Maryland.5 D)ifco Laboratories, Ine., Detroit, Michigan.6 Ivan Sorvall, Inc., Norwalk, Connecticuit.

19601 85

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

A. MI. GAUDIN, A. L. MIULAR, AND R. F. O'CONNOR

ml of inoculum over the surface of a dry agar medium.Plates were poured and incubated at 34 C for 24 hr priorto inoculation. The medium was that of Church et al.,(1954) for Bacillus globigii. Dilutions were made in tryp-tose-saline containing tryptose, 1 g, and NaCl, 5 g per Lof distilled water. Dilutions were plated in quintuplicateand counts were made after 24-hr incubation at 34 C.The calculation of vegetative cells was made by sub-

tracting the number of cells viable after exposure to65 C for 30 min from the total count.

Preparation of flotation feeds. Flotation feeds wereprepared by dispersing either the spore paste or sporesuspension in distilled water. Dispersion was ac-complished by means of a malted-milk mixer. The pulpsgenerally contained from 1 to 4 g of spore solids per L.

Flotation cells. Flotation experimenits were carriedout, for the most part, in either a pnieumatic cell or asmall-volume subaerationi cell. Although the designi ofthe flotation cell may have an effect upon the testresults, it is considered pertinient here only to note thebasic features of a flotation cell of the types used, andto leave a more complete description of the apparatusto a later publication. Several hundred milliliters ofsuspension is agitated in a small vessel, while air isintroduced into the liquid at the rate of several L permin. The froth which forms is collected as it overflowsthe vessel.

Analysis for dipicolinic acid (DPA). The procedureof .Janssen et al. (1958) was followed, with the samplesbeing autoclaved in 10-ml amounits in 25-ml Pyrexgraduated cylinders. Followinig the addition of the INacetic acid, the meniscus was readjusted by the additionof distilled water to compensate for loss of voltumeduring autoclaving. A color standard containing 100 ,gof a,r-diaminopimelic acid (DPA)7 per ml was used.Optical density was measured in a Bausch and Lomb8Spectronic 20 spectrophotometer at 440 m,u. The weightof sample was determined after evaporating to drynessa measured volume (usually 20 ml) of suspension in anoven for 24 hr at 86 C.

Mllethods

Assay for spores based upon their DPA content. Anumber of workers have demonstrated that the devel-opment of heat resistance in bacterial spores is ac-companied by the appearanice of DPA (Collier andNakata, 1958) and that the compound is not presentin vegetative cells (Powell and Strange, 1953). Valuesreported for DPA range from about 4.3 per cent of thedry weight for Bacillus cereus (Janssen et al., 1958) toas high as 12 to 15 per cent for Bacillus megatherium.(Falcone, 1954). If the dry weight percentage of DI'A

7P'urified l)PA was obtainied from Aldrich Chemical Co.,Inc., Milwaukee, Wisconsini.

8 Bausch and Lom ptical Co., Rochester, New York.

TABLE 1

Balance of weights and DPA*

Volume Wlt of Total TotalProduct of Pro- Solids Solids DPA/ DPA

duct per ml in Prod- mlt in Prod-uct uct

ml ??tg g pg Ag

Feed ................. 1750 2.02 3.535 43.7 7648

Concentrate ................ 260 2.78 0.723j75.7 1968Tailingg.................... 1490 1.85 2.7571 38.5 5737

Calculated Fee(d (Conceni-trate + Tailinig) ......... 1750 - 3.480 - 7705

* From flotation test No. 5.t D)PA = a,e-diaminopimelic aci(i.

in a spore preparation is known, an assay for DPA thenbecomes an assay for spores. Let

A = weight of DPA per unit weight of sporesB weight of DPA per unit weight of vegetative

cells and other nonspore constituents, collec-tively termed "debris"

C = weight of DPA per unit weight of sample ofunknowni composition

w = weight of unkniowni samplex = weight of spores in the unknown sampley= weight of vegetative cells and debris in the

unknowni sampleThen,

Ax + By = Cw.1 + ] = W

Solving for x, the following equation is obtained:

_ w(C - B)(A -B)

Where B = 0, the equationi becomes:

(,.X = w --

A

(4)

(5)

(6)

To use the measurement of DI'A as an assay forspore material, it is necessary to determine A for eachharvest or batch of spores, and w and C for each testsample. Spore suspensions approximately 98 per cent,pure were obtained by differential centrifugation after15 washings. The DPA content was found to be 5.19per cent. A preparatioin consisting entirely of vegetativecells and debris, similarly prepared, showed no meas-urable DPA. Cell suspensions which contained largenumbers of sporelike bodies, but which had no heatresistance, also showed no DPA.A convenient work sheet was developed to facilitate

the calculations for analysis, and to make possiblecalculations for balance of materials. A representativebalance of materials is showin in table l. This showsthat output of solids was 98.4 per cent of the input,

[VOL. 886

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

SEPARATION OF MICROORGANISMS BY FLOTATION. I

F'igure 1. Apparatus for spraying sample on microscopeslides.

whereas outpuit of DPA was 100.7 per cent of theinput.Assay for spores and vegetative cells based upon a

counting method. A direct counting method by whichassays of tomato bushy stunt virus were carried outhas been reported by Backus and Williams (1950).The method involved spraying suspensions containingmixtures of polystyrene beads (in known concentra-tion) and bushy stunt virus (in unknown concentration)on electron microscope grids, and determining theratio between the two types of particles. Since theconcentration of beads was known, the concentrationof virus could readily be calculated.An adaptation of this technique, in which a small

volume of sample was placed with a micro-dropperon the surface of an electron microscope grid andphotographed, was abandoned when it was learned thatthe particles concentrated at the periphery of the

droplet as it dried. Furthermore, because of the largersize of the particles being used, many more fields mustbe observed in the electron microscope if statisticallysignificant numbers of bacteria are to be counted.

The procedure which was finally adopted consisted

of mixing the sample with an aqueous dye solution

and spraying it on clean microscope slides, where the

droplets dried as circles almost instantaneously, with

no noticeable distortion in the arrangement of particles.1. Equipment. The spray apparatus consisted of a

tank of nitrogen equipped with a pressure regulator,a <4-in. spring-loaded needle valve, a glass nebulizer,pressure tubing, and a few support rods and fittings.Figure 1 shows the assembled apparatus. The glass

C-

0

GDEz

Circle Diameter (microns)

Figure 2. Frequency distribution of dried droplets

nebulizer or "gun"9 is the most critical piece of theapparatus and is drawn to scale.

Microscopic examination of the slides was made witha Bausch & Lomb binocular microscope, mounting a97X Bright M phase contrast objective (N.A. 1.25)and 15X compensated wide lens eyepieces. A greenfilter was used. The magnification was roughly 1450X.

2. Procedure. Three volumes of flotation productwere mixed with one volume of 4 per cent aqueousmethylene blue and one volume of polystyrene beadsuspension.10 The bead suspension as used contained5.33 X 109 beads per ml, and the avg bead diameterwas 1.171 ,u. The target slide was fixed at a distance of20 in. from the gun, and the mixture was sprayed at agas pressure of 30 psi. A spray period of 2 to 5 secsufficed to give a good circle density on the slide.Circles were viewed as they appeared in the microscopefield, no attempt being made to select them. In pre-liminary studies the diameter of the circle was measuredby means of an ocular micrometer, and the number ofparticles within the dried droplet was recorded.

3. Distribution of droplet sizes. Information con-cerning droplet-size distribution is to be found inNukiyama and Tanisawa (1939); Kolupaev (1941);and Bevans (1949). Figure 2 shows a distributioncurve for 1000 circles obtained in our laboratory. Forthis illustration, the circles have been grouped according

9 Obtained from Mr. L. H. Hinman at Cornell University,Ithaca, New York.

10 Kindly furnished by Dow Chemical Co., Midland, Michi-gan.

1960] 87

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

A. M. GAUDIN, A. L. MULAR, AND R. F. O'CONNOR

to categories of circles of which the diameters are inarithmetic progression. This curve bears close resem-blance to that obtained by Nukiyama and Tanisawa.

4. Distribution of particles relative to liquid. Samples

TABLE 2Spray-count assay

Ratio of beads to spores as a function of droplet volume

Circle Diameter Avg No. of Avg No. of Ratio of BeadsSpores per Circle Beads per Circle to Spores

27.8 2.96 8.5 2.8730.5 3.77 10.15 2.6733.3 4.30 11.6 2.8036.0 4.22 13.4 3.1738.9 5.27 15.3 2.9041.6 6.27 17.9 2.8544.4 5.73 17.2 3.00

Calculation of spores in

in which the suspended particles were simultaneouslyspores and beads were sprayed. The avg number ofspores and beads was determined for circles of thesame size, and the ratio of beads to spores was es-

sentially constant, as shown in table 2. Vegetativecells and debris particles also distribute themselveswithout regard to the volume of the droplets.

5. Ratio of number of beads to volume of sample.A suspension of beads was mixed with aqueous methy-lene blue such that the final concentration of the beadswas 4.28 X 109 beads per ml. After spraying, 500circles were measured, and the number of beads ineach was counted. The average number of beads per

100 circles was 886, thus the average volume repre-

sented by 100 circles was 886/(4.28 X 109) or 2.07 X10-7 ml. After counting the number of beads, spores,

cells and fragments in 100 circles (and assuming 886beads) it is only necessary to multiply each number by

TABLE 3Spray-count assay

flotation products, based upon counts of beads and spores in 200 circles

Spores per ~~~~~~~PerCent RecoveryExpe N Product Product Volume B eadSpores per ml Total Spores of Spores by:

No. Prdc rdc oue 200 Circles 200 Circles (Corrected of_Poc_ct_i_Prducfor Volume) X 108 X 1010- _____ Spray count DPA

ml

96D F 345 1691 704 738 29.83 102.87C 56 + 18 ml 1530 1733 2007 107.18 59.95 58.3 59.0

water*T 289 1678 357 377 15.23 44.09 - -

C + T 345 104.04

96E F 345 1627 703 766 30.96 106.71 - -C 66 + 8.5 ml 1513 1217 1425 65.03 42.92 40.2 39.6

water*T 279 1691 543 569 23.00 64.13

C + T 345 107.05

* Water added to wash out collection vessel.

TABLE 4Materials balances obtained with the DPA assay; effect of pH

upon balances

Initial pH of DPA in Con-Flotation DPA in Feed DPA in Con- DaiingCn-Feed centrate DP n Tailingcnrt n

mg mg mg mg

3.05 0.338 0.319 0.014 0.3333.90 0.387 0.267 0.067 0.3345.05 0.383 0.220 0.134 0.3546.00 0.367 0.192 0.173 0.3657.06 0.353 0.138 0.211 0.3497.88 0.394 0.148 0.252 0.4009.02 0.247 0.079 0.159 0.23810.00 0.355 0.188 0.176 0.36411.00 0.277 0.111 0.181 0.29211.92 0.377 0.171 0.207 0.378

TABLE 5Material balances obtained with the DPA assay; effect of reagent

addition upon balances

DPA inDPA in DPA in DPA in Concen-

Reagent Added Feed Concen- Taiin tratetrate Tlig Plus

Tailing

mg mg mg mg

di-Octyl amine ............... 0.364 0.350 0.045 0.395di-n-Hexyl amine ............. 0.355 0.188 0.176 0.364Acetic acid ................... 0.673 0.396 0.291 0.687Butyric acid ................. 0.290 0.138 0.149 0.287Octanoic acid ................ 0.298 0.123 0.179 0.302

88 [VOL. 8

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

SEPARATION OF MICROORGANISMS BY FLOTATION. I

4.83 X 106 (the reciprocal of 2.07 X 10-7), to get thenumber of spore cells and cell fragments per ml ofsuspension as sprayed. To convert these numbers toparticles per ml of product, a correction must be madefor dilution of the product with dye and bead sus-pension. In this case every number must be multipliedby 53.

Since the mean droplet diameter is affected by thesurface tension, viscosity, and density of the materialbeing sprayed, the sizes of dried droplets will differfor feed, concentrate, and tailing. To compensate forthis volume difference, equal numbers of beads aretaken as representing equal volumes of sample. Inthe examples given in table 3 the adjustment of volumesis made such that comparison is made between thenumber of spores in volumes of 1772/(4.28 X 109)or 4.14 X 10-7 ml from each product. Correspondencebetween recovery of spores calculated by this methodand by the DPA method is excellent. Material balancesare usually within 2 per cent.

TABLE 6Comparison of spore recoveries in the concentrate as calculated

by the DPA and spray-count assays

Per Cent Recovery of Spores in Concentrate by:Test No.

DPA assay Spray-count assay

103-A-5 59.5 59.6103-B-4 91.9 92.2103-B-5 52.4 53.4103-B-6 46.9 43.5103-B-7 30.7 32.5103-C-5 54.3 51.5103-D-5 45.0 42.0103-E-4 32.9 26.7103-E-5 38.8 34.0103-E-6 45.2 40.8103-E-7 30.7 30.3103-F-5 32.6 27.6

RESULTSDipicolinic acid assays. The DPA assay has been

used for calculating results in over 100 flotation testswith satisfactory balances of materials. As shown intables 4 and 5, the assay is very little affected by drasticpH variation in the flotation feed, or by the additionof various flotation reagents.

In a limited number of tests, where calculations were

made by both the viable count and DPA assays thetwo methods showed fair to good agreement with

00

90

80

70~~~~~~~~~~

x

60 /

50

~~~~~~~~~x

a.40-> ~~~~~~x

~30 xE

Dormant Spores X-X

20 Germinated Spores 0-o

Vegetative Cells

10

0 2 4 6 8 10 12 14 16Time of Flotation (minutes)

Figure S. Cumulative per cent recovery of dormant spores,

germinated spores, and vegetative cells in function of flotationtime.

TABLE 7Cumulative per cent flotation recovery of dormant spores, germinated spores, and vegetative cells as a function of time. Butyric acid

added to feed as flotation reagent with resulting pH of 5.05

Dormant Total Germinated Total Vegetative Total Cumulative per cent Recovery of:Spores per Dormant Spores per Germinated Cells per VegetativeTime Product Product ml of Spores in ml of Spores in ml of Cells inInterval Volume Product Product Product Product Product Product Dormant Germinated VegetativeX 108 X 1010 X 108 X 1010 X 108 X 101" spores spores cells

min ml

Feed 322 Not assayed 100 100 1000-3 Concentrate 1 82 26.35 21.61 10.49 8.60 18.51 15.18 29.1 42.0 56.83-6 Concentrate 2 18 28.10 5.06 11.64 2.10 35.88 6.46 35.89 52.26 81.06-9 Concentrate 3 12 51.56 6.19 22.91 2.75 32.98 3.96 47.21 65.68 95.849-12 Concentrate 4 5 138.70 6.94 46.13 2.31 10.67 0.53 56.53 76.96 97.8312-15 Concentrate 5 22 37.39 8.23 7.90 1.74 1.99 0.44 67.63 85.46 99.48-_ | Tailing 183 14.41 26.37 1.63 2.98 0.06 0.11 32.37 14.54 0.52

Sum of all concen- 322 74.40 20.48 26.68trates plus tailing

19601 89

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

A. M. GAUDIN, A. L. MULAR, AND R. F. O'CONNOR

the values for recovery being always slightly higherwith the DPA method. This occurred probably becausethe DPA method was including nonviable spores aswell as viable. This conclusion was borne out by thefact that when both viable and nonviable spores werecounted by the spray-count method, values were ob-tained which were in agreement with DPA values.The disadvantages inherent in the DPA assay are:

(1) the requirement for at least 1 X 108 spores perml in order to have sufficient DPA for reliable colordevelopment, and (2) the average DPA content ofpure spores must be determined for each new batchof spore material so that the grades of flotation prod-ucts can be calculated.

Spray-count assays. About 35 flotation tests havebeen analyzed with the spray-count assay. Balancesof materials have, in general, been good; but somedifficulty has been encountered in this area because ofthe use of a polystyrene bead suspension which hadbeen frozen and thawed, resulting in slight clumping ofthe beads.

Table 6 shows the agreement between spore re-coveries in the concentrate, as calculated by DPA andspray-count assays. The average deviation in recovery,as calculated by the two methods is about 3 per centrecovery units.The spray-count method has inherent advantages

over both the DPA and viable count methods in thatit (1) permits counting viable and nonviable organisms,(2) assays for vegetative cells and germinated sporesas well as for dormant spores, and (3) should be appli-cable to all types of bacteria, not sporeformers alone.The use of the spray-count assay to determine re-

coveries of dormant spores, germinated spores (whichtake up methylene blue) and vegetative cells is shownin table 7. The data, when plotted in figure 3, showclearly the different rates of flotation for the three celltypes. It can also be seen that a product, essentiallyfree of vegetative cells can be obtained after only 15min of flotation.

ACKNOWLEDGMENTS

The authors wish to express appreciation to Drs.Ralph E. Lincoln and R. S. Hutton of the U. S. ArmyChemical Corps for their support and encouragementin the conduct of this work; to Mrs. Dorothy Dais andMr. H. C. Greenlaw, Jr. for their invaluable help insetting up methods of analysis, and to Profs. Harlyn 0.Halvorson of the University of Wisconsin and James

H. Brown of Massachusetts Institute of Technologyfor their helpful suggestions and criticism.

SUMMARYFlotation testing of Bacillus subtilis var. niger has

been undertaken on a systematic basis to gain infor-mation on surface properties of the organism, and todevelop methods by which separation of cells from thegrowth menstruum can be carried out quickly.Two assay methods have been devised to aid in

evaluating flotation tests. A chemical method basedupon the dipicolinic acid (DPA) content of spores isused for the quantitative determination of spores,whereas a direct count method assays for dormantspores, germinated spores, and vegetative cells. Dataare presented to show that the three cell types may bemade to have different rates of flotation, and thatvetetative cells can be effectively removed from themixture.

REFERENCESBACKUS, R. C. AND WILLIAMS, R. C. 1950 The use of spray-

methods and of volatile suspending media in the prepara-tion of specimens for electron microscopy. J. Appl. Phy.,21, 11-15.

BEVANS, R. S. 1949 Mathematical expressions for drop sizedistribution in sprays. Prepared for the conference on fuelsprays, University of Michigan, March 30-31. Published atMassachusetts Institute of Technology, Cambridge,Massachusetts.

BLACK, H., MAcDONALD, R. E., AND GERHARDT, P. 1958Permeability of dormant spores of Bacillus cereus var.terminalis. Bacteriol. Proc., 1958, 41-42.

BOYLES, W. A. AND LINCOLN, R. E. 1958 Separation andconcentration of bacterial spores and vegetative cells byfoam flotation. Appl. Microbiol., 6, 327-334.

CHURCH, B. D., HALVORSON, H., AND HALVORSON, H. 0. 1954Studies on spore germination: its independence fromalanine racemase activity. .1. Bacteriol., 68, 393-400.

COLLIER, R. E. AND NAKATA, H. M. 1958 Heat resistanceas correlated with dipicolinic acid synthesis in sporeforming bacteria. Bacteriol. Proc., 42-43.

FALCONE, G. 1954 Spectrophotometric tests on substancesliberated from bacterial spores during germination. Boll.ist. sieroterap. milan, 33, 460-463.

GAUDIN, A. M. 1957 Flotation, 2nd ed. McGraw-HillBook Co., Inc., New York, N. Y.

JANSSEN, F., LUND, A., AND ANDERSON, L. 1958 Colori-metric assay for dipicolonic acid in bacterial spores.Science, 127, 26-27.

KOLUPAEV, P. G. 1941 Atomization of heavy fuel oil. Sc.D.Thesis, Massachusetts Institute of Technology.

NUKIYAMA, S. AND TANASAWA, Y. 1939 An experiment onthe atomization of liquid (3rd report). Trans. Soc. Mech.Engrs., 5, 63.

POWELL, J. F. AND STRANGE, R. E. 1953 Biochemical changesoccurring during the germination of bacterial spores.Biochem. J., 54, 205-209.

90 [VOL. 8

on Decem

ber 28, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from