CA1276576C - Microbial co-culture production of propionic acid - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE
A simultaneous sequential anaerobic fermentation process for the in vitro production of propionic acid and acetic acid is disclosed. The process comprises employing an obligatory two-component co-culture which maintains a relatively constant ratio of species populations over multiple passages. A first co-culture component is a Lactobacillus or Streptococcus which homofermentatively converts the hexose to lactic acid. A second microorganism in the co-culture is a Veillonella which is metabolically incapable of assimilating the hexose and converts the lactic acid product to propionic acid and acetic acid. The co-culture is inoculated into a nutrient growth feedstock, such as whole whey or a clarified dairy whey lactose permeate, which contains a metabolizable source of a hexose, such as glucose.
lactose or sucrose.

Description

This invention relates to a process for the production of lactic acid or its salts and propionic acid and/or acetic acid or its salts by the catabolism of carbohydrate feedstocks utilizing a simultaneous two-stage bacterial fermentation process. In a first stage, carbohydrates are converted to lactic acid, e.g. by saccharolytic bacteria such as Lactobacillus casei subspecies rhamnosus. In a second . _ stage, the resultant lactic acid is fermented to propionic and acetic acids, carbon dioxide and hydrogen by a second bacterium which is adapted to grow in the presence of the first bacterium, e.g. by a lactic acid-catabolizing bacterium such as Veillonella criceti.

Propionic acid is used commercially as an esterifying agent, in the production of cellulose propionate (a thermoplastic), and as a naturally occurring bacterial fermentation metabolite in cheese and other dairy products. Salt forms of the free acid such as calcium or sodlum propionate are used as preservatives in food products to inhibit fungal growth, as well as and in manufacturing ester solvents, fruit flavors, and perfume bases.
., 7~ i7t~

Traditional means of producing propionic acid have been via bacterial fermentation using strains of the genus Pro~ioni-b_cterium~ e.g. see U.S. Patents 1,459,959; 1,865,146; ],875~401;
1.898,329; 1.913.346; 1,'332,755; and 3.067,107. In the past 30 to 40 years. chemical processes, e.g. cnndensation of carbon mono~ide and ethylene or ethanol. have proven economically feasible. However, recent increases in the cost of petrochemical feedstocks have resulted in a re-examination of biological and agricultural feedstocks for the manufacture of rnany chemicals including propionic acid.
Propionibacteria of the genus Propionibacterium have traditionally been used for propionic acid production by bacterial fermentation. However~ use of _ropionibacterium species in monoculture or in co-culture with strains of _actobacillus has generally resulted in either lower levels of propionic acid, long fermenter residence periods, or both. These limitations may result form a long lag period preceding the growth of Propio_ibacteria.-inhibition of growth of Propionibacteria by propionic acid or, in the case of co-cultivation with lactic acid-producing species of Lactohacillus, from the preferential cataholism of carbohydrate by Propionibacterium species over the fermentation of lactic acid to propionic acid.
Accordingly, it is an object of a general aspect of the present invention to provide a process for the microbial co-culture of propionic acid in high yields.

.~.:`' All object of another aspect of the present invention is o provide such a process whlch provides high vields in dramatical]y shorter fermenter residence periods.
All object of a further aspect of the preserlt invention is to provide such a process wherein the lactic acid metabolic product of one microorganism serves as a feedstock for the production nf propionic acid by a second microorganism.
An object of an additional aspect of the present invention is to provide such a process which eonverts a major portion of lactic acid from a variety of feedstoeks into propionic acid, An object of a more particular aspeet of the present inventioll is to provide a novel co-culture of microorganisms fo use in sueh a proeess.
Bv a broad aspeet of this invention, a simultaneous sequential anaerobic fermentation process is provided for the _ v o production of propionic acid and acetic aeid, which comprises: a) seleeting a stable, obligatory two-component co-culture which maintains a relatively-constant ratio of species populations over multiple passages, the co-culture consisting essentially of: i) a first microorganism component which homofermentatively converts a hexose to a first metabolic product consisting essentially of lactic aeid; and ii) a seeond microorganism eomponent from the genus Veillonella which is metabolieally incapable of assimilating the hexose and which converts the lactic acid metabolic product of the first ~ Z~7~'7~

microorganism to a second metabolic product consisting essentially of propionic acid and acetic acid: b) inoculating the co-culture into an assimilable nutrient growth feedstock containing a metabolizable source of the hexose; c) anaerobi-cally fermenting the feedstock with the co-culture, at a ferrnen-tation rate of at least five millimoles per liter per hour, for a period of time and under conditions sufficient to convert a rnajor portion of the lactic acid into a fermentation product consisting essentially of propionic acid~ acetic acid, salts and mixtures tllereof: d) maintaining the pH of the fermentation mixtllre such that the Veillonella continues to ferment the lactic acid being produced by the first microorganism for a period of tirne sufficient to accumulate the fermentation product; and e) recovering the accumulated fermentation product.
The hexose source may be selected rom the group consisting !~
of glucose, sucrose, lactose, and mixtures thereof, e,g, lactose, The feedstock may be whole whey or a clarified dairy whey lactose permeate.
The first microorganism may be a Lactobacillus or a Streptococcus, preferably Lact _acillus casei or Lactobacillus casei subc rhamnosus, The first Veillonella preferably is _eillonella criceti. The co-culture preferably is ATCC Deposit No. 3g~662 or a mutant having the identifying fermentation characteristics thereof.

S o 6 - 4a -In one preferred embodiment of this invelltion. 5 moles of propionic acid are obtained from every 8 moles of fermented lactic acid.
The process preferably includes the additional step of drying the resultant product to form a free-flowing powder, preferably wherein residual microorganism are removed from the product prior to drying.
By another aspect of this :invention. a biologically-pure stable. -n vitro co-culture is provided of two microofganisms which are adapted anaerobically to grow together while maintaining a relatively-constant ratio of species populations over multiple passages such that neither microorganism overtakes the other, the co-culture consisting essentially of: i) a first microorganism component which homofermentatively converts a hexose to a first metabolic product consisting essentially of lactic acid; and ii) a second microorganism component from the genus Veillonella which is metabolically incapable of assimilating the hexose and which converts the lactic acid metabolic product of the first microorganism to a second metabolic product consisting essentially of propionic acid and acetic acid; the co-culture being adapted to the i_ vitro anaerobic production of propionic acid and acetic acid. The biologicallY-pure -n vitro co-culture preferably is of a Lactobacillus microorganism and a V illonella microorganisrn selected from the group consisting of ~TCC Deposit No. 39.622 and s~

i;7~
- 4h -mutallts having the identifying fermentation characteristlcs thereof.
Briefly~ in one emhodiment of the present invention, a process is provided for the i! vitro production of lact:ic acid bv fermentation of a nutrient growth feedstock containing a source of assimilable carbohydrates. e.g. a he.Yose or pentose~ with a first microorganism capahle of converting the carbohydla(ec to lactic acid under nutrient growth conditions. wherein the fermentation is conducted in the additional presence of a second microorganism which is adapted to grow in co-culture with 1he firct microorganism and which converts a major portion of the lactic acid ~ermentation product into a compound selected from the group consisting of propionic acid. acetic acid, and salts and miYtures thereof.
Suitahle feedctocks for the microbial production of lactic acid are well known in the art and include but are not limited to those described in the foregoing U.S. Patents and numerous other publications. e.g. see M. Brin, Biochem. Pre~ 61 11953);
S.C. Prescott et al., Indu-ctrial Microbiology (McGraw-Hill, New York. 3rd ed.. 195g) pp. 304-331; Andersen et al., Ind, Eng.
Chem. 34: 1522 (1942); and M. Brin et al., Ann N.Y, Acad, Scl.
119: ~51-1165 (1965). For commercial applications, feedstocks such as whey, cornstarch, potatoes, and molasses are generallv preferred, The presently preferred feedstocks comprise whole whey or a clarified dairy whey lactose permeate, especiallv that , . .

~7~57~i - /~c -described and claimed in PCT International Publication Nunlber W0 84/0110~ published March 2g, lg84.
The choice of a microorganism for producing lactic acid for use in accordance with process of broad aspects of the present invention will~ of course. depend on the particular feedstock components to be converted to lactic acid, for which many suitable microorganisms are well known in the art. Because clarified dairy whey lactose permeate is the presently preferred feedstock Eor use in the present invention, Lactob_cillus__case-i is the presently preferred microorganism for the first stage oE
the instant process~ especially _actobaclllus case_ suhsp.
rham1losus. Such strains are widely known and readily available to those skilled in the art~ e.g. from the American Type Culture Collection (ATC~), 12301 Parklawn Drive~ Rockville. Marylalld 20~52.
The choice of a second microorganism for converting the lactic acid metabolic product of the first microorganism into propionic acid requires the selection of a microorganism having the abilitv to ferment lactic acid into propionic acid alld oth~r end products. Several such bacteria are well known and widel~
available in the art for such purposes, e.g. as has beerl I described in British Patent 1,251,483; U.S. Patents 3~857,971 and 4~138.4g8; and by Huber et al. in Am. J. Vet Res. 7(5): 611-613 (1976). Such bacteria are well known and widely available to those skilled in the art. and include but are not limited to Meg~phaera elsdenii, Peptococcus asaccharolYticus. Selenomonas ,; ;

~7~ 6 ~, ( I
rum_ atium. and Veillon_lla_cri_eti. Especially suitable for ~Ise in broad aspects of the process of the present inventinn are those microorgallisms whicll preferentiallv use lactic acid as a source of assimilable carbohydrate: hecause they exhibit this property (with the exception of fructose~ V criceti appears incapable of fermenting carbohydrates directly, presumably due to the lack of hexokinase enzymes. and instead utilizes monocarboxYlic acids, e~g. lactic acid as a growth substratel arld do not exhibit a long lag period preceding the rapid growth phase in vitro. Veillonella _lceti is preferred.

i76 The sequential treatment of feedstock to first form lactic acid and then form propioni'c acid suffers from a number of inherent difficulties. In the first stage, the accumulation of lactic acid product eventually slows the feedstock conversion due to mass balance effects and lowering of the pH. While the latter can be adjusted, this introduces an additional risk of contamination whereas removal of lactic acid product involves removing both unconverted feedstock and the converting microorganism. In the second stage, an undesired lactate concentration will remain if the pH is not controlled, and introducing unconverted feedstock provides an opportunity for the microorganism to employ metabolic pathways leading to the formation of undesired products.

In accordance with the present invention, it has now been found that the above and other difficulties can be overcome by the catabolism of carbohydrates using a simultaneous two-stage bacterial fermentation process~ In the first stage, carbohydrates are converted to lactic acid by the saccharolytic bacterium, L. casei subspecies rhamnosus. In the second stage, the resultant lactic acid is fermented to propionic and acetic acids, carbon dioxide and hydrogen by V. criceti.
Propionate (and lactate) thus formed may be recovered by the use of an appropriate solvent extraction system, a distillation recovery process, cationic salt formation with precipitation, or by the concentration and drying of the fermentation broth medium (with or without the removal of the bacterial cells.) Formation of co-culture.

Parent strains of CLS917 (Lactobacillus casei subspecies rhamnosus) and 1218 (Veillonella criceti) were separately selected for spontaneous antibiotic resistant colonies growing on streptomycin and rifampicin (for L. casel) or rifampicin alone (for V. criceti). These mutant strains of L. casei and V. criceti, bearing genetic markers (i.e.
.
resistance to speclfic antibiotics), were then tested ior acid products of metabolism in a tryptone broth medium which has the following composition (values expressed as final concentrations on a ~a~'7~5'~f~

weight/weight basis): , tryptone, (1.0%); yeast extract, (1.0%); sodium lactate, (2.0%); cysteine hydrochloride, (0.5%); and sodium bicarbonate, (0.5%). Those mutant strains (CLSC~17 and 1218) demonstrating maximum acid production were then selected.

These broth cultures were incubated anaerobically for 24 hours at separate temperatures to determine the optimum for stable growth of both strains. The optimum temperature was determined to be 38C
estimated by the number of viable cells of each strain.

Viable and healthy bacterial cells of both mutant strains were added to a chopped meat broth culture medium prepared as described by Holdeman and Moore, Anaerobe Laboratory Manual, 4th edition, Department of Anaerobic Microblology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, 24061. Serial passage of the co-culture by transfer and subcultivation at 24 hour periods over a three day interval demonstrated a stable mixed two member bacterial culture (see Table 1.) The stability of the co-culture was determined by the enumeration of the two bacterial populations for each chopped meat broth culture. This was accomplished using the standard technique of a solid growth medium inoculated with serial dilutions of the broth cultures. The solid growth medium comprised the previously described tryptone broth medium with the addition of agar agar to a final concentration of 1.5%. The pH of the solid medium was adjusted before steam sterilization to 7.0 (and was observed to be 6.8 to 7.0 following sterilizaLion.) The diluent for preparation of the serial dilutions was that d~scribed in the Anaerobe Laboratory Manual, 4th edition.
Subcultures were incubated for 24 hours before the populations were enumerated. The co-culture is considered stable when the ratio of the two organisms remains constant within the limits of experimental error, demonstrating that neither organlsm is overtaking the other.

7~:i5~6 TABLE I
STABILITY OF CO-CULTURE

Population of Population of Ratio strain CLS917 str~ain 1218 of CLS917:1218 Subculture (values expressed as 10 cells per mllliliter) Transfer No.:
O (Initial inoculum) 78 142 0.55 1 850 1500 0.57 2 630 1000 0.63 3 730 1500 0.49 The size of the viable cell population of L. casei strain CLS917 in the inoculum at the start of the experiment was approximately one-half the size of the population of the V. criceti strain 1218. At the end of each of the three days, the ratio of these two populations was relatively constant. Further use of a single chopped meat broth co-culture as an inoculum for fifteen separate three-liter batch fermentations over a two month period resulted in mixed populations of similar viable cell proportions.

Maintenance of the co-culture.

The co-culture can be maintained in sealed vials in aliquots of 0.2 milliliters containing equal volumes of sterile ~lycerol and tryptone broth culture medium (previously described) at a temperature of -oO~C.
Each aliquot should contain equal numbers of both strains (approximately one hundred million bacterial cells each) previously harvested from healthy growing cultures. This co-culture of Lactobacillus casei subsp. rhamnosus and Veillonella criceti has been deposited with the American Type Culture Collection on April 9, 1984 and has been designated ATCC Deposit No. 39,662. The strains may be revived following storage by ~inoculation of the contents into either tryptone or chopped `meat broth medium and incubated under anaerobic conditions at 38C for 24 to 48 hours.

.

~L 2~ ~ 7~i Production and use of metabolic products.
I

The metabolism of a ~ermentable carbohydrate in a suitable nutrient medium by strains of Lactobacillus results in the production of low .
levels of acetate and high levels of lactate. The lactate is then rapidly metabolized by strains of Veillonella present in the co-culture into propionate, acetate, carbon dioxide and hydrogen. Table 2 illustrates part of the array of carbohydrate substrates that can be fermented into propionate, acetate, carbon dioxide, hydrogen and lactate by the co-cultivation of strains of Lactobacillus and lG Veillonella.

ACID PRODUCTION (MG/ML) AFTER 72 HOURS OF CULTIVATION*

V. criceti (1218) L. casei (CLS917) CO-CULTURE
ACET PROP LACT ACET PROP LACT AC~T PROP LACT
CARBOHYDRATE:
CONTROL (WATER) 0.34 0.79 ND ND ND 0.30 0.46 0.95 ND
CELLOBIOSE 0.36 0.75 ND 0.18 ND 7.40 3.11 3 . 92 1 . 03 FRUCTOSE 2.24 2 . 85 ND ND ND 0. 26 2 o 94 4 . 02 0 . 47 GALACTOSE 0.37 0.70 ND ND ND 8.00 2.73 4.70 0.01 20 GLUCOSE 0.34 O. 68 ND ND ND 0.34 2.77 4.73 0.01 GLUCONATE 0.51 0.88 ND 0. 26 ND 0.37 3.01 2.45 ND
LACTOSE 0.34 0.76 ND ND ND 9.40 2.58 4.32 0.77 MANNITOL 0.36 0.74 ND ND ND 1.40 1.65 4.26 0.03 RHAMNOSE 0.34 0.75 ND 0.72 ND 2.90 1.80 4. 20 ND
SORBITOL 0.35 0.73 ND ND ND 0.80 1. 66 3 . 84 ND
TREHALOSE 0.35 O. 80 ND ND ND 8 . 40 2. 66 4.51 0.05 * The metab-olic acid products are designated as follows: ACET =
acetic acid; PROP = propionic acid; and LACT = lactic acid. ND = Not Detected (Less than 0.01 mg/ml) Volatile and non-volatile fatty acids were determined by gas-liquid chromatographic procedures as described by Holdeman and Moore (Anerobe Laboratory Manual, 4th edition, Department of Anaerobic Microbiology9 Vlrginia Polytechnlc Institute and State University, Blacks~urg, Virginia, 24061 ) .

:~ ~'7~ 6 The presently preferred best mode of this lnvention is ~he cultivation of the two bacterial strains in a nutrient growth medium sufficient to provide for a stable co-culture using the two broth media prevlously described or in the media described in the following Examples.

Speclfically, a mixed culture oE two bacterial strains (L. casei CLS917 and V. criceti 1218 is cultivated ln a growth medium containing a carbohydrate substrate that stra:Ln CLS917 can ferment. Such substrates include but are not limited to mono- and dl-saccharldes and complex polysaccharldes. Addltlonally, lt ls preferable that the growth medium also contain a source of vitamins and/or amino acids as are present in a 0.1 to 2.0~ solution of the extract of yeast cells. A
low concentration of a non-inhibiting, nontoxic salt of carbonic acid is preferably added to a 60 millimolar final concentration.

The conditions of the fermentation include: a temperature range generally between 20-40C, but preferably between 35-40C; a means of agitating the fermentation mixture at speed of up to 400 revolutions per minute, but preferably in the range of 150-250 revolutions per minute; and a pH generally in the range of 4.0 to 9.0, but preferably in the range of 5.5 to 6.0 for optimum propionic acid production.
Additionally, the growth medium should preclude dissolved oxygen by preventing air flow or exchange of gases which could result in an increase in the concentration of dissolved oxygen and thereby interfere w'~h ~erobic metabolism. Generally, the rate of the fermentation of substrate is approximately 1 to 10 millimoles per hour, preferably at least 5 millimolés per hour.

Without further elaboration, it is believed that one skilled in the art can, using the preceding description, utilize the present invention to its fullest extent. The following preferred specific embodiments are, therefore, to be construed as merelY illustrative-In the following Examples, the temperatures are set forth uncorrected ln degrees Celslus; unless otherwise indicated, all parts andpercentages are by weight.

;57~;

Example 1 pH Control of Fermentatlon i The pH of the fermentation must be maintained within the range of generally pH 5.0 to 9.0 or preferably 5.3 to 7.3. Any of several compounds acting as a Lewis Base can be used for this purpose.
Hydroxides of ammonium (this includes ammonia gas, which hydrates in aqueous solutions to form ammonium hydroxide), sodium, calcium, or potassium salts can be used without deleterious effects on the ; fermentation. However, the divalent inorganic metal oxides, e.g.
calcium hydroxide (and the corresponding oxides which hydrate in aqueous solution to form the corresponding hydroxide) appear to function better as a pH control agent than the monovalent cation hydroxides.

In this experiment, the two strains (CLS917 and 1218) were cultivated In three liters of a medium containing (as final concentrations based on the medium): lactose, (2%), supplied as either whole whey or ultrafiltered whey permeate; yeast extract, (1.0%); and 60 millimolar of carbonate buffer of the same cation as the hydroxide (except in the case of ammonium hydroxide and ammonium gas, where calcium carbonate was used). The conditions of the fermentation included a temperature of 38C, maintenance of pH between 5.5 and 6.0 by the automatic or manual addition of pH control agent, and continuous stirred agitation in a New Brunswick Fermenter at a speed of 200 revolutions per minute. Aliquots were removed at the beginning of the fermentation and at 4, 8~ 12, 24, and 48 hours during the fermentation. These samples were analyzed for acetic (ACET), propionic (PROP) and lactic (LACT) acids using the methodology previously described. Concentrations (in mg/ml) of these metabolic acid products from the 24 llour samples are presented in Table 3.

LEWIS BASE USED ACET PROP LACT
TO MAINTAIN pH **************units in mg/ml****************
(units in mM) AMMONIUM HYDROXIDE 4.60 7.89 0.22 AMMONIA GAS 5.75 8.28 0.23 CALCIIM HYDROXIDE 7.32 11.48 0.02 POTASSIUM HYDROXIDE 4.54 7.18 0.24 SODIUM HYDROXIDE 4.62 8.36 0.20 Ex_mple 2 Maintenance of Anaerobiosis Both strains CLS917 (L. casei) and 1218 (V. criceti) are anaerobic bacteria in that they do not grow (1218) or grow well (CLS917) in the presence of oxygen. However, under general conditions of fermentation in which a freshly steam sterilized medium is allowed to equilibrate to the temperature of the fermentation prior to inoculation with microorganisms, oxygen is effectively precluded. Thus a reducing agent, a chemical compound that can complex with dissolved oxygen, was not required for the effective fermentation of lactose substrate to propionic acid. Additionally, the use of an inert gas (nitrogen, helium, carbon dioxide, etc.) was not required to fill the headspace (that space above the surface of the fermenta~ion medium to the top of the vessel) to prevent the dissolution of atmospheric oxygen into the fermentation medium. Inert gas is not required during the fermentation of these two strains because strain 1218 regularly produces 1.5 moles of carbon dioxide for each mole of dissaccharide (and 0.75 moles for each mole of monosaccharide) fermented. Because strain 121B (V.
criceti) is an obligate anaerobe, care must be taken not to oxidize the fermentation medium by spraying or bubbling air or oxygen.
In this experiment the two stralns ~CLS917 and 1218) were cultlvated in three liters of a medlu~ containing lactose, (2%), supplied as either whole whey or ultrailtered whey permeate; yeast extract, (1.0~); and 60 millimolar calcium carbonaLe. The conditions ~ ~7~S76 of the fermentation included a temperature of 38C, malntenance of p~l between 5.5 and 6.0 by the automatic or manual addition of pH control agent, and continuous stirred agitation in a Cell Stir Jar (Bellco Glass Co.) at a speed of 200 revolueions per minute. Aliquots were removed at the beginning of the fermentation and at 4, 8, 12, 24, and 48 hours during the fermentation. These samples were analyzed for acetic (ACET), propionic (PROP) and lactic (LACT) acids using the methodology previously described. Concentrations of these metabolic acid products from the 24 hour samples in which either a reducing agent, inert gas or neither were used.are presented in Table 4. It is evident from these daea that standard fermentation preparations are sufficient under the conditions employed to prevent toxic concentrations of dissolved oxygen in the fermentation medium.

OXYGEN PROTECTION EFFECTS ON
METABOLIC ACID PRODUCTION

ACET PROP LACT
OXYGEN PROTECTION *********units in mg/ml*********
Cystein hydrochloride (0.05%, final concentration) and the use of carbon dioxide to fill the headspace 5.84 8.28 0.13 Cystein hydrochloride only (0.05% final conc.) 6.14 8.69 0.89 No cysteine hydrochloride or carbon dioxide 5.15 7.31 0.05 Example 3 Fermentation Parameters Affecting Metabolic Acid Products Strains of V. crlceti demonstrate a sensitivity to low pl~, dlssolved oxygen, or temperatures exceedlng 40C in the fermentation medium. Thus, for example, an adJustment of the pll to values below 7~

about 5.3 - 5.5, either by the addition oE a Lewis acid or by the failure to maintain pH at a level greater than about 5.3 - 5.5, the metabolism of the Veillonella strains precludes further oxidation of lactic acid to propionic acid and acetic acid with the elaboratlon of carbon dioxide and hydrogen gases. Under conditions in which the co-culture is exposed to a pH of less than about 5.3 - 5.5, temperatures exceeding 40C, or dissolved oxygen, the Lactobacillus continue to ferment the carbohydrates substrate to lactic acid but the Veillonella are unable to oxidize the lactic acid to additional products.
,:
Any desired ratio of propionic acid to lactic acid can be produced by manipulatlon of the conditions of the fermentation, with or without the subsequent inclusion of additional substrate. This Example demonstrates the effects on the production of metabolic acid products by the variation of pH. In thls experiment, the ~wo strains (CLS917 and 1218) were cultivated in three liters of a medium containing lactose (2%) supplied as whole whey; yeast extract (1.0%); and 60 millimolar calcium carbonate. The conditions of the fermentation included a temperature of 38C, maintenance of pH between 5.5 and 6.0 during the first 24 hour period by the automatic or manual addition of ammonium hydroxide, and continuous stirred agitation in a New Brunswick Fermenter at a speed of 200 revolutions per minuce. Aliquots were removed at the beginning of the fermentation and at 8, 24, 32, and 48 hours during the fermeneation. The results are summarized in Table 5:

~;~7~576 TABI.E 5 EFFECT~ OF pll ON ME.TABOLIC ACID PRODUCTS

ACET P~OP LACT
(mg/ml) Time of Sample:
(Hours~
O 0.09 0.01 0.24 ô 0.83 1.04 0.00 24 2.05 3.78 6.90 (additional lactose substrate added;
the p~l was allowed to drop to 5.1) 10 32 1.88 3.51 6.90 (subsequent propionic and acetic acids are produced at a lower concentration than was initially observed) 48 2.00 4.01 28.04 (the final ratio of proplonic acid to lactic acid is 0.14 compared with 48.8 found under optimal conditions as demonstrated in Example 4) Example 4 Use of Whole Sweet Whey as Substrate Whole sweet cheese whey was used in a fermentation medium to which the two strains CLS917 and 1218 were added. In 250 liters of a medium containing lactose, (4%), supplied as whole whey (5%, with 2.5% added in the beginning of the fermentation and 2.5% added after 24 hours);
yeast extract, (1.0%); and 60 millimolar calcium carbonate. The conditions of the fermentation included a temperature of 38C, maintenance of pH between 5.5 and 6.0 by the automatic or manual addition of calcium hydroxide, and continuous stirred agitation in a New Brunswick Fermenter at a speed of 200 revolutions per minute.
Aliquots were removed at the beginning of the fermentation and at 16, 24, 32, and 48 hours during the fermentation. These samples were analyzed for acetic (ACET), propionic (PROP) and lactic (LACT) acids using the methodology previously described. Concentrations of these metabolic acid products from the 24 hour samples are presented in Table 6.

FERMENTATION OF LACTOSE IN WHOLE SWEET WHEY

ACET PROP LACT
(mg/ml) Tlme of Sample:
(Hours) O (initial concentration of lactose 2%) 0.11 0.00 1.35 16 7.05 9.58 0.35 24 (additional lactose 2%)8.02 10.80 0.12 32 9.76 12.96 18.7 ~18 14.72 20.00 0.41 Example 5 Industrial Production of Prl~ionates from Whole Sweet Whey A pilot plant production of calcium propionate by the fermentation of whole sweet whey was conducted in the following medium: lactose, (4%), supplied as whole sweet whey present initially at 2.5% with a !'` subsequent addition of 2.5% after 16 hours; yeast extract (Amber 510), (1.0%); and calcium carbonate (Huber-Carb S-3tm), (0.6~). The conditions of the fermentation included: maintenance of the pH between 5.5 and 5.7 by the addition of calcium hydroxide manually as necessary; temperature of 38C (+ 1.0 degree); and agitation of approximately 200 revolutions. per minute. Fermentation was conducted in 35 liter working volume stainless steel J sterilized in place fermenters.

Following fermentation, the bacterial cells were removed by microfiltration through a Romican hollow fiber membrane having a 50,000 dalton cutoff pore size and the permeate was decolorized in an activated carbon slurry, concentrated by flash evaporation, and spray dried in a tower type drier (air inlet temperature 150 - 160C; air outlet temperature ~0 - 100C) to a free flowing, off-white powder.
The chemical and physical characteristics of this fermentation product are shown in Table 7.

5~

TYPICAL ANALYSIS OF FERMENTATION PRODUCT

Bulk Density (grams/cubic centimeter) 0.03 Moisture (%) 8.1 pH of 1% Solution 6.34 Solubility (grams/100 ml water) @ 25C 33.00 Solubility (grams/200 ml water) @ 70C 32.40 Crude fiber content (%) ~0.10 Acid detergent fiber content (%) ~0.10 Ash (%) 55 3 Crude fat (%) cl.O
Crude protein (%) 11.10 Subtotal 66.40 Carbohydrate (by difference) (~) 33.60 Soluble carbohydrate (by gas-liquid chromatography) (%) Fructose ND
Glucose ~1.00 Galactose ~1.00 Lactose e1.00 Sucrose ND
Short chain fatty acids (volatile) Calcium acetate (measured as acetic acid) (%) 37.0 Calcium propionate (as propionic acid) (~) 42.20 Short chain fatty acids (nonvolatile) Calcium lactate (as lactic acid) (%) C2.00 Calcium succinate (as succinic acid~ (%) ND
Vitamins (milligrams/100 grams) Thiamine <0.10 Riboflavin ~0.10 Pyridoxine CO.10 Cobalamin CO.Ol Niacin ~0.10 Minerals (%) Calcium 16.69 Phosphorous 0.19 Sodium 1.15 Magnesium 0.29 Subtotal 13.32 57~

Minerals (parts per million) Aluminum 65.48 Barlum 3-45 Boron 6.58 Chromium 3.57 Copper 2.91 Iron 31.34 Manganese 3.98 Strontium 78.76 Zinc 8.07 Subtotal 204.15 Example 6 Fermentation of Ultrafiltered Sweet Whey _ The lactose present in ultrafiltered sweet whey was also used as substrate in the fermentation of propionic acid by a co-culture of strains CLS917 and 1218. A fermentation was conducted in the following medium: lactose (2%) supplied as dried sweet whey permeate (ultrafilter size exclusion of 30,000 daltons) and yeast extract tl.O%) and CaC03 (0.6%). Fermentation conditions included a temperature of 38C, maintenance of pH between 5.5 and 6.0 by the automatic or manual addition of ammonium hydroxide, and continuous stirred agitat-ion in a New Brunswick Fermenter a~ a speed of 200 revolu~ions per minute.
Ali~uots were removed at the beginning of the fermentation and at 4, 8, and 24 hours during the fermentation. These samples were analyzed for acetic (ACET), propionic (PROP) and lactic (LACT) acids using the methodology described in the preceding example. Concentrations of these metabolic acid products from the 24 hour samples are presented in Table 7.

PRODUCTION OF PROPIONIC ACID FROM
ULTRAFILTERED SWEET WHEY
ACET PROP LACT
(mg/ml) Time of Sample 0 0.17 0.05 1.17 4 0.92 0.93 1.68 8 2.59 3.67 4.12 24 7.50 9.56 0.25 Example 7 Use of Cellobiose as Feedstock . .
While lactose is the presently preferred substrate for use in the present invention, many other carbohydrates can serve as substrates for the production of metabolic acid products. One such carbohydrate that is frequently found in nature from non-dairy sources is cellobiose, a disaccharide resulting from the partial digestion of cellulose.
Following the general procedures described with reference to Table 2 demonstrates the co-cultivation of strains CLS917 and 1218 on cellobiose with concomitant production of propionic, acetic and lactic acids. The results are summarized in Table 9. Propionic acid is not produced to a significant concentration by either strain alone.

FERMENTATION OF CELLOBIOSE BY CO-CULTIVATION

V. criceti (1218) L. casei (CLS917) CO-CULTURE

ACET PROP LACT ACET PROP LACT ACET PROP LACT
CARBOHYDRATE:
CONTROL (WATER) 0.34 0.79 ND ND ND 0.30 0~46 0.95 ND
CELLOBIOSE 0.36 0.75 ND 0.18 ND 7.40 3.11 3.92 1.03 ii7~i Example 8 Use of Propionate _ rmentation Product in Bakery _udies Calcium propionate has traditionally been used to increase the shelf-life of breads and other bakery products by inhibition of the growth of molds. A fermentation of whole sweet whey in whlch the cells were removed and the resulcing filtrate was concentrated and spray dried analagously to the process of Example 4 was used in bakery studies to determine the effects of the calcium propionate content to inhibit the growth of mold contamination. In this study, the spray dried and free-flowing powder was used in two bakery recipes. In one recipe the dried product was added to the bread dough to a final concentration of 0.5% (on a flour weight basis) which resulted in a final calcium propionate concentration of approximately 0.25%.
Additional active ingredients included 0.1~ monocalcium phosphate and 0.9~ Teklactm (a food grade lactose product.) The second recipe contained only the fermentation product a~ a final concentration of 0.5% with a calcium propionate concentration of approximately 0.25% as the only active ingredient. A third dough was prepared which was used as the control and did not contain any active ingredients.

The results of the s~udy demonstrated that the dried fermentation product, when used at a concentration based upon the calcium propionate content traditionally employed in bread dough recipes, was effective at inhibiting the growth of mold contamination. The first and second recipes provided loaves of br~ad that did not demonstrate any mold growth after storage under standard conditions up to and including thirty days, at which time the study was terminated. The third recipe which served as the control for the experiment provided loaves of bread which demonstrated mold growth after 7 to 10 days of storage.

It is clear from this example that the use of this invention in the fermentation of the lactose present in whole sweet whey by the co-cultivation of strains CLS917 (L. casei s~bspecies rhamnosus) and 1218 (V. criceti) resulted in the production of calcium propionate.
This fermentat1on product when used (in elther the drled or liqu~d ~ ~7~7~

form) at a calcium propionate concentration of 0.25% (final concentration on a ilour weight basis) can effectively inhibit the growth of mold contamination in bread under standard storage conditionsO

As can be seen from the present specification and examples, the present inventlon is industrially useful in providing a method for producing calcium propionate which has a varlety of known industrial applications.

Claims (14)

1. A simultaneously sequential anaerobic fermentation process for the in vitro production of propionic acid and acetic acid, which comprises:
a) selecting a stable, obligatory two-component co-culture which maintains a relatively-constant ratio of species populations over multiple passages, the co-culture consisting essentially of:
i) a first microorganism component which homofermen-tatively converts a hexose to a first metabolic product consisting essentially of lactic acid; and ii) a second microorganism component from the genus Veillonella which is metabolically incapable of assimilating said hexose and which converts the lactic acid metabolic product of the first microorganism to a second metabolic product consisting essentially of propionic acid and acetic acid;
b) inoculating said co-culture into an assimilable nutrient growth feedstock containing a metabolizable source of said hexose;
c) anaerobically fermenting said feedstock with said co-culture, at a fermentation rate of at least five millimoles per liter per hour, for a period of time and under conditions sufficient to convert a major portion of the lactic acid into a fermentation product consisting essentially of propionic acid, acetic acid, salts and mixtures thereof;

d) maintaining the pH of the fermentation mixture such that the Veillonella continues to ferment the lactic acid being produced by the first microorganism for a period of time sufficient to accumulate said fermentation product; and e) recovering the accumulated fermentation product.

2. A process according to claim 1, wherein said hexose source is selected from the group consisting of glucose, sucrose, lactose, and mixtures thereof.

3. A process according to claim 2, wherein said hexose source is lactose.

4. A process according to claim 3, wherein said feedstock is whole whey or a clarified dairy whey lactose permeate.

5. A process according to claim 1, wherein said first microorganism is a Lactobacillus or a Streptococcus.

6. A process according to claim 5, wherein said first microorganism is a Lactobacillus casei.

7. A process according to claim 6, wherein said first microorganism is Lactobacillus casei subs. rhamnosus.

8. A process according to claim 1, wherein said Veillonella is Veillonella criceti.

9. A process according to claim 1. wherein said co-culture is ATCC Deposit No. 39,662 or a mutant having the identifying fermentation characteristics thereof.

10. A process according to claim 1, wherein 5 moles of propionic acid are obtained from every 8 moles of fermented lactic acid.

11. A process according to claim 1, further comprising the step of drying the resultant product to form a free-flowing powder.

12. A process according to claim 11, wherein residual microorganism are removed from said product prior to drying.

13. A biologically-pure, stable in vitro co-culture of two microorganism adapted anaerobically to grow together while maintaining a relatively-constant ratio of species populations over multiple passages such that neither microorganism overtakes the other, said co-culture consisting essentially of i) a first microorganism component which homofermentatively converts a hexose to a first metabolic product consisting essentially of lactic acid; and ii) a second microorganism component from the genus Veillonella which is metabolically incapable of assimilating said hexose and which converts the lactic acid metabolic product of the first microorganism to a second metabolic product consisting essentially of propionic acid and acetic acid;
said co-culture being adapted to the in vitro anaerobic production of propionic acid and acetic acid according to the process of claim 1.

14. A biologically-pure, in vitro co-culture of a Lactobacillus microorganism and a Veillonella microorganism selected from the group consisting of ATCC Deposit No. 39,622 and mutants having the identifying fermentation characteristics thereof.