APPLIED AND ENVIRONMENTAL MICROBIOLOGY, OCt. 1987, p. 2496-2499 0099-2240/87/102496-04$02.

00/0 Copyright 1987, American Society for Microbiology

Vol. 53, No. 10

A Technique for Predicting the Solvent-Producing Ability of Clostridium acetobutylicum

HOWARD I. ADLER* AND WELDON CROW Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
Received 11 May 1987/Accepted 22 July 1987

Changes in colony morphology were associated with the degeneration of solvent-producing strains of Clostridium acetobutylicum. The most efficient solvent-producing strains gave rise exclusively to colonies with dense centers containing large numbers of spores. Many outgrowths of various morphologies developed from the perimeter of such colonies after several days of incubation. The most degenerate cultures did not produce solvents and gave rise to large diffuse colonies that did not contain spores. These diffuse colonies did not produce outgrowths. Intermediate colony types were also observed. These could be derived from liquid cultures that were relatively poor solvent producers or from the outgrowths of colonies of efficient solvent-producing strains. Some of these intermediate types produced spores but did so less frequently than the high-solventproducing strains. The spores of the intermediate types could not be distinguished from those of the most efficient solvent producers on the basis of heat sensitivity. The relationship observed between colony morphology and solvent production provides a method for predicting the solvent-producing potential of C. acetobutylicum cultures.
The rapid degeneration of Clostridium acetobutylicum and related species from solvent-producing organisms to ones that produce only organic acids has been observed by many workers. This phenomenon contributed to the problems encountered in the commercial application of the organisms for acetone and butanol production (4, 7, 10; E. McCoy and E. B. Fred, J. Bacteriol. 41:90-91, 1941). The loss of solvent-producing ability is usually accompanied by the loss of ability to produce spores. Therefore, it is frequently possible to predict the likelihood that a culture will produce solvents by observing its ability to form spores. However, it has been demonstrated that some strains that fail to produce mature spores but do reach the clostridial stage continue to make solvents (6, 9).In this report we present data demonstrating that colony morphology can be used as a reliable and sensitive indicator of solvent-producing ability. In addition, we established that degeneration can occur in a series of steps and that the appearance of degenerate types is a spontaneous event which does not require intentional exposure to chemical or physical agents.
MATERIALS AND METHODS

agar). Spores or cells were plated after appropriate dilution in NYG. Except when noted, plates were incubated for 3 to 5 days at 37C in Torball jars (model AJ-2; Torsion Balance Co., Clifton, N.J.) containing a mixture of 95% N2 and 5%
CO2.

Bacterial strains and growth conditions. C. acetobutylicum NRRL B643 was obtained from P. Rogers, University of Minnesota. C. acetobutylicum ATCC 4259 and ATCC 824 were obtained from the American Type Culture Collection, Rockville, Md. Several other strains used were soil isolates obtained in this laboratory (3). Except when noted, liquid cultures were grown in nutrient broth (Difco Laboratories, Detroit, Mich.) supplemented with yeast extract (0.1%) and glucose (0.4%) (NYG). Cultures were maintained in a potato medium consisting of approximately 5 g of diced Idaho potato, 0.15 g of CaCO3, and 12 ml of water in a screw-cap tube (16 by 110 mm). The oxygen-consuming membrane fraction previously described was used to produce anaerobic conditions in liquid cultures (1). NYG medium was solidified by the addition of 15 g of Bacto-Agar (Difco) per liter (NYG
*

Determination of colony morphology by microscopic observation. NYG agar plates containing 10 to 50 colonies were observed with a low-power microscope (magnification, x 10 to x 15). The light source was arranged so that light impinged obliquely on an opaque sintered-glass surface before being transmitted through the colonies. To achieve even greater contrast, a 1% aqueous solution of Coomassie blue R250 (Miles Laboratories, Inc., Elkhart, Ind.) was pipetted onto agar plates containing 3- to 5-day-old colonies. The dye was allowed to cover the colonies for 1 to 2 min, and then the excess was decanted. Determination of heat sensitivity of spores. Samples (0.5 ml) of sporulated cultures in potato medium were spread in sterile plastic petri dishes and allowed to air dry at 37C for 3 to 5 days before being suspended in NYG. A set of 0.2-ml samples of this suspension in thin-wall glass tubes was put into a 90C circulating water bath. Periodically, samples were withdrawn, diluted rapidly in NYG, and plated on the surface of NYG agar. Determination of solvent production. The cultures were assayed for ethanol, acetone, butanol, acetic acid, and butyric acid by injecting supernatants from centrifuged culture samples into a Shimadzu GC-9A gas chromatograph equipped with a flame ionization detector and a 2-m stainless steel column packed with Poropak Q (80/100 mesh). The samples were assayed under the following conditions: column temperature, 180C; injector temperature, 250C; detector temperature, 250C; carrier gas, argon; flow rate, 50
ml/cm2.

Corresponding author.
2496

Determination of cell size by flow cytometry. Cell size was determined with an Ortho 50H Cytoflyorograph (Ortho Diagnostics, Inc., Westwood, Mass.), interfaced to an Ortho 2150 computer, by previously published methods (12). Cell size was measured as a function of light scattered in the forward direction.

VOL. 53, 1987

SOLVENT PRODUCTION AND COLONY MORPHOLOGY

2497

FIG. 1. Photographs of Coomassie blue-stained 4-day-old colonies of C. acetobutylicum. (A) Type I; (B) type II; (C) type III; (D) type IV.

RESULTS
A typical 4-day-old colony derived from a vigorous solvent-producing, sporeforming culture of strain NRRL B643 is shown in Fig. 1A. All of the colonies produced from cells or spores taken from such a liquid culture had the dark centers and lighter outgrowths seen in Fig. 1A. If these colonies were observed after only 1 to 2 days of incubation, they appeared as dark, roughly circular structures 1 to 2 mm in diameter. No outgrowths were visible. When liquid cultures were repeatedly transferred in a rich medium (NYG) or were allowed to incubate for many days, cells producing colony types such as those seen in Fig. 1B, C, and D appeared. These same secondary colony types could be

obtained from the various outgrowths of a colony such as that shown in Fig. 1A. A similar range of colony types was observed for cultures of strains ATCC 4259 and ATCC 824 and several isolates from soil samples. The colony type associated with vigorous, solvent-producing, sporeforming cultures was designated type I, and the derived types were designated types II, III, and IV. Some properties of the four types are given in Table 1. It is interesting that the degeneration phenomenon always occurred from type I toward type IV. No example of spontaneous reversion was observed, although many thousands of type II, III, and IV colonies were examined for outgrowth that would indicate this. Except for the frequency of sporangia, cells from type I, II, and III colonies or their corresponding liquid cultures

2498

ADLER AND CROW

APPL. ENVIRON. MICROBIOL.

TABLE 2. Spore and solvent production by several isolates of each of the various colony types" Amt (g/liter) of solvent Colony type from
which liquid culture was initiated

TABLE 1. Characteristicsa of C. acetobutylicum colony types

Type Morphology of colony Type(s) of outgrowths produced

produced

% Spores

II III IV

Diameter, 2-3 mm; very dark center; many outgrowths produced (approx 6-8/ colony) Diameter, 3-5 mm; gray-brown centers; fewer outgrowths than type I (approx 24/colony) Diameter, 4-6 mm; brown centers; relatively smooth perimeter; outgrowths rare (less than 0.01/colony) Diameter, 4-8 mm; light brown centers, edges indented; striae present, particularly near perimeter; no outgrowths

II, III, and IV III and IV IV

Acetone 1.425 5.328 3.901 4.051 4.618

Butanol 6.011 6.469 8.988 7.372 10.80

61 50 46 52 51
0 0

II
None

1.005 1.031 0.000 1.101 1.630 1.822 0.377 0.000 1.548 0.000

2.629 3.984 0.176 3.085 6.995

5.194 2.164 0.614 6.462 0.018

0
8 16
16 0

Descriptions apply to 4-day-old colonies observed by the technique described in Materials and Methods before Coomassie blue staining. More than 200 colonies of each type were examined.

III

<0.1
26 0

were indistinguishable when observed by phase microscopy or flow cytometry. However, cells from type IV cultures were readily distinguished from the other three types, particularly when observed during the late log or early stationary phase. Type IV cells were relatively long and slightly thinner than cells of the other types. No clostridial or sporangial forms were observed. This observation, first made microscopically, was confirmed by quantitative flow cytometric data (Fig. 2). The average late log, type IV cell was twice the size of the average type I, II, or III cell. In addition, the range of cell sizes was greater for type IV cells. When observed by phase microscopy, the spores produced by either type II or III cultures were indistinguishable from those produced by type I cultures. They were also indistinguishable on the basis of heat sensitivity. Approximately 90% of the spores of each type were inactivated by heating at 90C for 3 min. However, when type II or III spores were plated, they gave rise only to type II or III colonies, respectively. They did not produce type I colonies. Quantitative data on some properties of liquid cultures derived from the four types of colonies are presented in Table 2. A correlation between colony type, solvent produc1998

0 0.000 0.000 0 0.000 0.000 0 0.000 0.000 0 0.000 0.000 in potato medium at 37C. The Results are for 5-day-old cultures grown presence of spores was determined by microscopic observation of the IV
frequency of phase-bright, oval structures. At least 200 cells and spores were scored for each culture.

tion, and the ability to form spores was observed. It is this correlation that forms the basis of the method for predicting solvent production based on colony morphology. A similar correlation was observed for all strains of C. acetobutylicum tested. DISCUSSION Changes in bacterial colony morphology associated with altered biological and biochemical properties have been observed frequently. At one time, this phenomenon, usually referred to as dissociation, occupied the attention of many
1

A
C

B
c E L L

E
L L
N U

N B E R

N U N B E R

* 49 298 299 -868e lo9 698 669 se8 408 e e l9e9 RELATIVE CELL SIZE RELRTIVE CELL SIZE FIG. 2. Relative cell size distribution for type I (A) and type IV (B) cells as determined by flow cytometric analysis. A total of 77,885 cells were analyzed for type I, and 74,491 cells were analyzed for type IV.
1

VOL. 53, 1987

SOLVENT PRODUCTION AND COLONY MORPHOLOGY

2499

bacteriologists (2). A theoretical treatment of the development of colony outgrowths based on differential growth rates of individual cells and their progeny has been presented (11). Our current observations are an extension of earlier work and provide a practical method for predicting the solventproducing potential of C. acetobutylicum cultures. The observations also provide a way of monitoring the degeneration of cultures and may eventually be useful in efforts to explain and control this phenomenon. Liquid cultures of C. acetobutylicum that produced only type I colonies consistently produced high levels of acetone and butanol, whereas cultures that yielded only type IV colonies produced little or no solvents. Of the cultures that produced type II or III colonies, some produced significant levels of solvents, whereas others did not. Liquid cultures that contained mixtures of the colony types produced intermediate levels of acetone and butanol. In at least one previous report, an outgrowth from a C. acetobutylicum colony, probably corresponding to our type IV, was noted (5). However, it was not recognized that several types of outgrowths occur and that they represent various stages in the degeneration of a culture. The use of colony type to predict solvent-forming efficiency was not suggested. Earlier workers have also reported the isolation of sporulation and solvent "mutants" after treatment with mutagenic agents but failed to note that these same mutants can be isolated from the edges of all mature type I colonies (8, 9). It is possible that the chemical agents used as mutagens were acting only as selective agents in these earlier experiments. In addition to the use of colony morphology to predict solvent-forming efficiency, our results help in predicting the likelihood that spore selection will "regenerate" a good solvent producer. If a liquid culture contains spores that give rise to a mixture of type I, II, and III colonies, it will probably not be possible to reestablish an efficient solventproducing strain by repeated heat treatment or other selection for spores. This is because the heat sensitivity of all three types of spores is equivalent and they will each continue to produce their respective cell types. It would, however, be possible to select an efficient solvent producer by plating the spore mixture and establishing a liquid culture from the center of an immature type I colony. The detailed mechanism accounting for degeneration is yet to be described. It probably involves spontaneous events affecting the efficiency of sporulation. In a nutrient-rich anaerobic environment, the cells with the least tendency to sporulate will eventually predominate. We hope that the

ability to correlate solvent production, colony morphology, and spore production will eventually lead to further understanding of degeneration.
ACKNOWLEDGMENTS

We gratefully recognize the contributions made by Russell Hand, who obtained the flow cytometry data presented, and Richard Machanoff, who contributed experimentally to the early stages of this work and has continued to be a valuable discussant. This research was sponsored by the Office of Health and Environmental Research, U.S. Department of Energy, under contract DE-AC05-840R21400 with Martin Marietta Energy Systems, Inc. LITERATURE CITED 1. Adler, H. I., W. D. Crow, C. T. Hadden, J. Hall, and R. Machanoff. 1983. New techniques for growing anaerobic bacteria. Biotechnol. Bioeng. Symp. 13:153-161. 2. Braun, W. 1947. Bacterial dissociation. A critical review of a phenomenon of bacterial variation. Bacteriol. Rev. 11:75-114. 3. Crow, W. D., R. Machanoff, and H. I. Adler. 1985. Isolation of anaerobes using an oxygen reducing membrane fraction. J. Microbiol. Methods 4:133-139. 4. Finn, R. K., and J. E. Nowrey. 1959. A note on the stability of
5.

6.

7. 8.

clostridia when held in continuous culture. Appl. Microbiol. 7:29-32. Jones, D. R., J. R. Webster, and D. R. Woods. 1980. The formation of simple fruiting body-like structures associated with sporulation under aerobic conditions in Clostridium acetobutylicum. J. Gen. Microbiol. 116:195-200. Jones, D. T., A. van der Westhuizen, S. Long, E. R. Allcock, S. J. Reid, and D. R. Woods. 1982. Solvent production and morphological changes in Clostridium acetobutylicum. Appl. Environ. Microbiol. 43:1434-1439. Kutzenok, A., and M. Aschner. 1952. Degenerative processes in a strain of Clostridium butylicum. J. Bacteriol. 64:829-836. Largier, S. T., S. Long, J. D. Santangelo, D. T. Jones, and D. R. Woods. 1985. Immobilized Clostridium acetobutylicum P262 mutants for solvent production. Appl. Environ. Microbiol. 50:

477-481. 9. Long, S., D. T. Jones, and D. R. Woods. 1984. The relationship between sporulation and solvent production in Clostridium acetobutylicum P262. Biotechnol. Lett. 6:529-534. 10. Prescott, S. C., and C. G. Dunn. 1959. Industrial microbiology, p. 262-263. McGraw-Hill Book Co., New York. 11. Shinn, L. E. 1939. Factors governing the development of variational structures within bacterial colonies. J. Bacteriol. 38:5-12. 12. Tyndall, R. L., R. E. Hand, Jr., R. C. Mann, C. Evans, and R. Jernigan. 1985. Application of flow cytometry to detection and characterization of Legionella spp. Appl. Environ. Microbiol. 49:852-857.