Production of Fumaric Acid in 20-Liter Fermentors
R. A. RHODES, A. A. LAGODA, T. J. MISENHEIMER, M. L. SMITH, R. F. ANDERSON,
AND R. W. JACKSON
Fermentation Laboratory, Northern Utilization Research and Development Division,1 Peoria. Illinois
Received for publication June 28, 1961
ABSTRACT
RHODES, R. A. (U. S. Department of Agriculture,
Peoria, 111.), A. A. LAGODA, T. J. MISENHEIMER, M.
L. SMITH, R. F. ANDERSON, AND R. W. JACKSON.
Production of fumaric acid in 20-liter fermentors.
Appl. Microbiol. 10:9-15. 1962.-The conditions
necessary for the production of fumaric acid in 20-liter
fermentors by fermentation of glucose with Rhizopus
arrhizus strain NRRL 2582 were determined. Continuous neutralization of fumaric acid was necessary
for optimal yields. Yields of the calcium salt were
in excess of 65 g of fumaric acid from 100 g of sugar
consumed during fermentation of sugar concentrations
of 10 to 16 %. Conditions established for calcium
fumarate production include a simple mineral salts
medium, 0.5 v:v:min aeration rate, 300 rev/min agitation rate in a baffled tank, 33 C incubation temperature, CaCO3 to neutralize the acid formed, and a 4 to
5 % (v/v) vegetative inoculum. A suitable procedure
and medium for the preparation of a vigorous vegetative
inoculum were established. The tendency for calcium
fumarate fermentations to foam excessively was controlled with a proper antifoam agent added prior to
sterilization of the medium and again at daily intervals
during fermentation. The production of soluble sodium or potassium fumarates was inhibited when
the concentration of fumarates reached 3.5 to 4.0 %.
No means of overcoming this inhibition was found.
Starches and certain other grain-derived carbohydrates
were fermented to form calcium fumarate in flask
experiments with approximately the same efficiency
as was glucose.
MATERIALS AND METHODS
The 20-liter fermentors were constructed of stainless
steel and equipped with facilities for the continuous
control of pH by automatic addition of sterile alkali.
Detailed descriptions of these fermentors and their
pH control system have been published by Dworschack, Lagoda, and Jackson (1954, 1956). Fermentations were made at a 10-liter operating volume.
Agitation rates were established with a tachometer,
aeration rates by measurement of the effluent air flow.
Addition of sterile 5 N alkali was made from a reservoir
attached to each fermentor and was controlled by a
solenoid-type valve actuated through an automatic
titration device connected to pH electrodes in the
fermentor. This system maintained the pH of the fermentation medium within 0.1 of a pH unit. The pH
of the fermentation medium initially was 6.6 to 6.8.
Ordinarily, less than 2 hr of fermentation time were
required to achieve the optimal operating pH range
of 5.8 to 6.0; hence, initial adjustment with sterile acid
was not required.
Essentially the same medium previously found
optimum for flask fermentations was satisfactory for
the 20-liter fermentors. KH2PO4, 0.4 g; MgSO4 7H20,
0.4 g; ZnSO4.7H20, 0.044 g; iron tartrate, 0.01 g; and
corn steep liquor, 0.5 ml, were incorporated per 1,000
ml fermentation volume. A commercial corn sugar
closely approximating glucose monohydrate was employed as the fermentable carbohydrate. The fermentation medium containing the carbohydrate was
sterilized in the fermentor by steam injection at 121
C for 5 min. To produce calcium fumarate, CaCO3 was
added as a thick slurry to the fermentors and sterilized
with the medium. The (NH4)2SO4 used as a nitrogen
source was added aseptically as a sterile solution to
the prepared inoculum immediately before inoculation
of the fermentors.
R. arrhizus Fisher strain NRRL 2582 was used to
produce fumaric acid in all experiments reported.
Inocula of this organism were grown in the described
fermentation medium modified by increasing the
KH2PO4 concentration to 1.6 g per 1,000 ml and by
Fumaric acid is well established as an article of commerce in the United States. The estimated 1959 production of fumaric acid was about 10 million pounds,
most of which was utilized by the plastics industry
in polyester and alkyd resins, and the remainder went
to lesser volume uses such as rosin adducts, varnishes,
and foods. The fermentative production of fumaric
acid, if favored by a cheap source of sugar, could become economically important. A previous publication
(Rhodes et aL., 1959) reported that fumaric acid was
produced readily by Rhizopus arrhizus in shaken flask
1 Agricultural Research Service, U. S. Department of
Agriculture.
9
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
cultures with yields of 65 % of the sugar supplied.
Our initial investigation of the fermentation has been
extended to a semipilot-plant scale with 20-liter fermentors.
[VOL . 10(
RHODES, LAGODA, MISENHEIMER, SMITH, ANDERSON, AND JACKSON
the use of 4.0 g per 1,000 ml of (NH4)2S04 as a nitrogen
source. These amounts are approximately four and
two times, respectively, the concentrations optimal
for fumaric acid production. Two per cent glucose was
used as the carbon source and 3 g of CaCO3 was included per 1,000 ml of inoculum medium. This medium
allowed production of a vigorous vegetative inoculum.
The physical characteristics required of inoculum
for fermentors were not the same as those required
for flasks. Efficient fermentations required a finely
dispersed growth consisting of minute particles of
filamentous mold growth. Growth of the mold in
the fermentors had to be diffuse but not of a particle
size such as to adhere to fermentor baffles, air sparger,
impeller, pH electrodes, or fermentor surfaces. When
improper inocula, such as the delicate mycelial strands
of germinated spores, were used, virtually all of the
mold growth was removed physically from the medium
onto fermentor surfaces during the first hours of fermentation; indeed, the fermentation liquor became
almost sterile. Conversely, mold inoculum of heavy
vegetative growth was not readily dispersible in the
medium and resulted in mycelial pellets that induced
only slow fermentation. Many experiments were conducted to determine the procedure necessary for development of a suitable inoculum for fermentors.
The following procedure gave a vegetative inoculum
with capacity for rapid fermentation.
Stage 1. Spores of strain 2582 were washed from a
slant culture with 20 ml of sterile 0.06 % Tween 802
into 100 ml of inoculum medium contained in conically
indented 300-ml Erlenmeyer flasks and incubated 20
to 22 hr at 33 C on a rotary shaker.
Stage 2. Additional inoculum medium was inoculated with 20 % by volume of growth from stage 1 and
incubated 22 to 24 hr at 33 C. Indented flasks were
employed to supply the volume needed for our purpose but it was shown that use of properly agitated
fermentors resulted in similar mold growth.
During each stage in the development of mold
growth to be used as inocula, about 1.0 to 1.5 %o glucose
was consumed and the pH of the inoculum fell to
about 1.5; the amount of inoculum growth was controlled by the amount of CaCO3 supplied since the
mold does not grow appreciably at pH levels below
3.0. The vegetative growth of stage 2 was used to
inoculate the 20-liter fermentors. It was readily dispersed by agitation in the fermentors and remained
dispersed without pellet formation; it did not adhere to
fermentor surfaces and initiated immediate and vigorous production of acid. Furthermore, it could be
handled easily with common pipettes and laboratory
pumps. Although additional volume build-up of inoeAtlas Powder Company, Wilmington, Del. The mention of
products does not imply endorsement by the U. S. Department
2
of Agriculture
over other products
of
a similar nature.
ulum was possible by further transfer of the vegetative
growth using the same protocol, this kind of vegetative
inoculum was not as efficient for fumarate production
in flasks as was a germinated-spore inoculum or the
two-stage material.
Samples from the fermentations were taken aseptically and were held under refrigeration until analyzed.
Culture liquors were diluted to appropriate volume
and, when solid calcium fumarate was present, heated
to effect its solution; mycelium and residual calcium
carbonate were removed by filtration. Sugar content
was measured by the method of Shaffer and Hartmann (1921). Aliquots of the diluted liquors were
passed through a cation column (Dowex 50, H+ form3),
and a portion of the eluate was titrated to measure
the amount of total fermentative acidity. Fumaric
acid in the cation-treated material was determined by
adsorption of the fermentation acids on an anion
resin (Dowex 1, OH- form) and the selective elution of
fumaric acid. Other nonvolatile fermentation acids
present, chiefly malic and some succinic, were eluted
from the resin columns with 1.15 N acetic acid; fumaric
acid then was eluted with 0.1 N HCl. Eluted acids
were measured by titration with 0.01 N NaOH after
drying on a steam bath to remove the volatile eluting
acids. Analytical results were corrected for volume
changes in the fermentors determined on final samples.
Yields were calculated on an anhydrous weight basis
I
Dow Chemical Company, Midland, Mich.
TABLE 1. Factors affecting production of soluble fumarates
Fermentation conditions
Alkali
Inocu-
pH
NaOH
5
5.4
5
5
5
5
5
5
5
5
5
5
5
5.8
6.2
4
4
5
KOH
4
4
5
conGlucose
Total
O.l N
Templumglucose
used
C
%
Fermentation results
~
~~z
~
E
sumed
Fma
aid
Amt Yield
g/10O ml mg/100 g/lOo ml ml/100 ml glo0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
33
33
33
33
30
32
34
36
33
33
33
33
33
33
33
10
10
10
10
10
10
10
10
10
10
10
10
7.9
9.1
11.3
200
200
200
200
200
200
200
200
100
150
200
250
160
160
200
9.2
9.7
9.8
8.5
9.1
10.1
10.1
9.6
5.9
8.6
9.2
9.5
7.8
8.5
11.3
6.0
6.0
6.0
33
33
33
7.7
9.2
11.1
160
160
200
7.7
8.6
10.1
6.6
acid
712
950
925
710
880
930
950
750
790
900
860
860
850
730
3.1
4.0
3.7
1.9
3.7
4.0
3.8
3.0
3.6
3.6
3.6
3.5
3.6
3.0
1,000 4.1
34
42
38
22
42
40
38
31
61
41
39
37
46
35
36
720 3.1 41
720 3.3 39
830 3.4 33
Conditions: 0.5 v:v:min aeration 0 to 24 hr, 1.0 v:v:min
aeration 24 to 72 hr; 200 rev/min agitation rate.
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
10
FUMARIC ACID PRODUCTION
from the amount of fumaric acid produced and the
amount of carbohydrate consumed during the fermentation:
Per cent
yield
=
fumaric acid produced
carbohydrate consumed
'
EXPERIMENTAL RESULTS
Sodium fumarate production in 20-liter fermentors.
The sodium and potassium salts of fumaric acid are
much more soluble than is the calcium salt, which
precipitates at low concentration (<3 %) and causes
thickening of the medium. Also, the large amounts of
CaCO3 required in the fermentation are not easily
handled. For these and other reasons, fermentations in
20-liter fermentors were undertaken with continuous
neutralization of fumaric acid by the automatic addition of sodium or potassium hydroxide. It is apparent
from the data in Tables 1 and 2 that yields of soluble
fumarates comparable to those obtained in flasks with
CaCO3 neutralization were obtained only in the fermentation of limited amounts of sugar. Experiments
in flasks demonstrated that both growth and sporulation of R. arrhizus were restricted by sodium fumarate
concentrations above 3 to 4% (w/v) and were completely inhibited by concentrations above 8 to 10%
(w/v). Sporulation of the mold was more readily inhibited than was growth. Sodium chloride was tolerated in higher concentration than was sodium fumarate.
Similar patterns of inhibition were observed with
potassium fumarate and potassium chloride, although
considerably greater tolerance to both potassium salts
TABLE
Initial
glucose
supplied
hr
g/100 ml
g/100
24
10
12
13
13.5
Time
tion
36
Fumaric acid
Commercial glucose
Total
0.1 N
Initial
concn
hr
g/100 ml
11
Continuous*
10.7
7.5
5.4
Consumed
g/100
ml
0.9
0.3
1.1
0.2
acid
Amt
ml/100 ml g/100
37.5
25.0
32.5
20.0
Proportion
of total
acid
48
0.04
30
45
31
38
10
12
13
13.5
11
33
9
20
Continuous
10.7
7.5
5.4
45
72
3.5
2.3
3.3
1.2
280
165
245
160
1.1
0.6
0.5
0.6
65
64
35
59
31
26
15
60
50
Continuous
10.7
7.5
5.4
7.6
5.8
6.6
3.4
650
450
650
420
2.8
1.7
2.5
1.5
74
64
65
63
37
29
38
44
Continuous
10.7
7.5
5.4
10.1
8.5
7.5
5.2
796
730
790
650
3.9
2.8
3.0
2.6
71
67
65
68
39
33
40
50
*
Glucose added at 1-min intervals between 12 and 50 hr to
maintain sugar concentration in the fermentor at 1.5 to 2.0%;
initial concentration: 2.8%.
Inoculum and (NH4)2SO4 proportional to initial glucose:
10.7% glucose, continuous feed, 5% inoculum, 200 mg
(NH4)2SO4/100 ml; 10.7% glucose, initial batch, 5% inoculum,
200 mg (NH4)2SO4/100 ml; 7.5% glucose, 4% inoculum, 160
mg (NH4)2SO4/100 ml; 5.4% glucose, 3% inoculum, 120 mg
(NH4)2SO4/100 ml.
Conditions: 0.5 v:v:min aeration 0 to 24 hr, 1.0 v:v:min
aeration 24 to 72 hr; 200 rev/min agitation rate; pH 6.0; 33 C.
acid
Fumaric acid
Amt
ml/100 ml g/100 ml
Yield
%
6
7
8
10
12
10
5.8
6.1
6.3
6.2
6.8
6.2
4.6
5.0
4.3
3.4
3.5
3.7
700
681
435
532
435
322
2.7
2.8
1.6
1.8
1.4
1.1
59
56
37
53
40
30
6
7
8
5.3
5.7
6.0
10
8.2
9.3
7.4
6.8
7.0
6.3
1,036
1,188
1,020
1,162
1,000
850
3.9
4.6
3.5
3.2
4.1
2.6
48
50
47
47
59
41
5.6
7.8
7.5
7.4
6.8
6.7
64
74
76
67
69
73
7.0
9.0
71
80
10.0
91
8.3
8.1
65
65
76
10
12
13
13.5
6
7
8
10
12
10
5.2
4.5
5.4
5.7
5.8
5.8
8.8
10.5
9.9
11.0
9.9
9.2
1,305
1,750
1,729
1,715
1,628
1,635
6
7
8
10
12
10
4.0
4.1
4.7
5.7
5.8
5.7
9.9
11.3
11.0
12.7
12.5
11.9
1,575
2,020
2,318
1,900
1,925
1,980
6
7
8
10
12
10
4.0
4.0
4.1
5.7
5.8
5.3
10.1
12.0
12.2
13.0
13.0
13.3
1,452
1,868
2,130
2,100
2,040
2,172
10.3
8.7
9.1
62
70
74
79
67
68
8
4.0
5.8
6.2
4.4
12.7
13.2
13.3
13.9
2,004
2,010
1,932
2,190
8.8
9.0
8.4
10.1
69
68
63
73
4.0
4.3
12.8
15.1
1,920
2,280
8.9
10.5
70
70
10
12
13
13.5
10
12
13
13.5
13
13.5
16
123
g/lOO ml
ml
Total
0.1 N
10
16
96
Glucose
consumed
16
16
72
pH
5.8
5.8
6.2
16
21
CaCOs
12
Yield
ml
0.1
0.1
0.1
Fermentation results
Fermentation conditions
TABLE
Time of
fermenta-
3. Factors affecting calcium fumarate production
16
2. Effect of initial glucose concentration and
continuous glucose addition on sodium fumarate
production
11
10
12
10
13
8
16
10
9.1
6.3
8.4
9.0
Conditions: 0.5 v:v:min aeration; 300 rev/min agitation;
33 C; 180 mg (NH4)2SO4 supplied per 100 ml medium; 5%
(v/v) inoculum.
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
1962]
RHODES, LAGODA, MISENHEIMER, SMITH, ANDERSON, AND JACKSON
was observed in the flask experiments. Since the use of
potassium hydroxide was not more successful in fermentations than was the use of sodium hydroxide
(Table 1), inhibition of acid production apparently
was more sensitive than was the inhibition of growth.
The mold could be adapted with difficulty to tolerate
greater concentrations of either sodium or potassium
fumarate by successive transfer on media containing
increasingly higher concentrations of the given salt.
However, these characteristics were not sufficiently
stable to provide a strain of 2582 which could be used
in fermentations. Inhibition of fumarate formation
occurred as the concentration of sodium or potassium
fumarate approached 3 %.
As shown in Table 1, the optimal pH range for
fumarate production was 5.8 to 6.2. Fermentations
employing alkali routinely were maintained at pH
6.0. A suitable temperature for this fermentation in
fermentors, as in flasks, was about 33 C (Table 1).
Yields of sodium fumarate obtained when different
concentrations of glucose were supplied in fermentors
are shown in Table 2. Continuous addition (each
minute) of glucose did not significantly increase yields
of fumarate compared to yields obtained from the
same amount of glucose supplied in a single initial
batch. However, a somewhat greater rate of sugar
utilization was obtained with continuous feed. Sterile
glucose was added dropwise from a 50 % solution
through a solenoid valve actuated by a timing device.
It was possible to predict the rate of sugar utilization
and thus to maintain glucose concentrations in the
fermentors within 0.5% over a 40- to 50-hr period
until the desired total amount was supplied.
Calcium fumarate production in 20-liter fermentors.
A portion of the results of experiments on the production of calcium fumarate in 20-liter fermentors is
summarized in Table 3. It is apparent that, as contrasted to sodium or potassium fumarate formation,
highly efficient fermentative production of calcium
fumarate was realized from a wide range of glucose
concentrations. Yields of 65 g of fumaric acid from
each 100 g of sugar fermented were obtained by the
fermentation of glucose concentrations between 10
and 16 %. The optimal amount of nitrogen supplied
as (NH4)2S04 was 34 to 42 mg per 100 ml of fermentation medium; the greater amount in the range often
was most suitable when higher sugar levels were fermented or when smaller amount of inocula were employed. For brevity, only the results obtained by the
use in the fermentation of 180 mg (NH4)2SO4 per 100
ml (38 mg N/100 ml) are included in Table 3.
The rapidity of the fermentation can be noted from
the data. The peak of acid production ordinarily occurred in the 24- to 48-hr fermentation period when
50 to 75 ml of 0.1 N acid per 100 ml of fermentation
liquid were produced each hour. Conversion of glucose
[VOL. 10
to fumaric acid sometimes approached 90 % during
this period. As insoluble calcium fumarate accumulated, the medium became thick, not unlike thin
mortar or plaster, and was thixotropic. Fermented
media containing appreciable calcium fumarate almost
immediately set to a gel on standing at room temperature but were quite fluid when agitated or heated.
Sufficient calcium fumarate was present, during much
of the fermentation of any of the sugar concentrations
used, to thicken the fermentation medium appreciably.
The mold growth was finely dispersed and intimately
intermixed with calcium fumarate and, in incomplete
stages, with CaCO3. This condition was indicative of
effective fermentation.
Diminution of efficiency in late stages of the fermentation suggests that the continuous presence of
solids in the medium and the thickened consistency
which developed with calcium fumarate accumulation
may result in less effective aeration. Higher rates of
air flow up to 1.0 v:v:min were beneficial in fermentations for the production of sodium fumarate, but
rates greater than 0.5 v:v:min were not useful with
the thicker medium encountered when calcium fumarate was formed. Agitation rates influence oxygen absorption more greatly than do aeration rates in fermentation media which contain particulate material
(Brierley and Steel, 1959). A 300-rev/min agitation
rate appeared optimal for calcium fumarate production,
whereas a 200-rev/min rate with higher aeration rates
was more effective for sodium fumarate production.
Omission of baffles in the fermentors resulted in diminished fumaric acid production. The presence of baffles
and the proper rate of agitation were essential both for
TABLE 4. Production of calcium fumarate from commercial
sugars and starches in shaken flasks
Fumaric acid
Carbohydrate fermented
Type
Re-
maining
Consumed
g/100 ml g/1OO ml
Commercial glu-
0.1
8.7
Total
0.1 N
acid
Proportion
of total
Yield
ml/100 ml g/100 ml
1,324 5.8
76
67
1,150
4.9
73
61
1-1,162
4.9
72
61
-1,165
4.6
69
55
230
475
1,700
1,588
1.8
6.7
6.2
64
68
67
38
66
60
1,588
6.2
67
61
1,450
5.4
64
55
Amt
acid
cose
No. 70 corn 1.4 8.0
sugar
No. 80 corn 1.2 8.0
sugar
lst Greens
2.6 8.4
10.1 0.9
Hydrol
5.0 4.8
Xylose
Soluble starch 1.5 10.2
Batter process 1.7 10.4
starch
Turbo - milled 1.0 10.1
starch
Dextrin
2.6 9.9
Flask fermentation: 66-hr fermentation of sugars, 96-hr
fermentation of starches, 5% inoculum.
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
12
FUMARIC ACID PRODUCTION
effective aeration and for the dispersal and maintenance
of a suitable state of mold growth. The fermentor
baffles were submerged in the medium, ending just
under the liquid surface. A tendency was found for
inoculum growth to adhere and to proliferate on baffle
surfaces which extended above the medium. The
approximate oxygen absorption rate in bisulfite under
these conditions (0.5 v: v: min aeration, 300-rev/mmn
agitation, pipe sparger, baffled fermentor) was previously determined to be 1.6 mM oxygen absorbed
liter: min.
The production of calcium fumarate in aerated fermentors was accompanied by the formation of a tenacious foam. When foam formed, it was extremely
stable and appeared to be a rigid gel composed of
liquid, air, mycelial strands, and solid particles of
CaCO3 and calcium fumarate. This foam was difficult
to control. Antifoam agents based on liquid fat were
ineffective, possibly because of the formation of calcium soaps. Silicone antifoam agents were somewhat
more efficacious but could not be relied upon for routine
use. A polypropylene glycol, Polyglycol P2000,3 was
used routinely for foam control with marked success.
TABLE 5. Effect of ammonium sulfate concentration
on calcium fumarate production from starch
substrates in shaken flasks
0
_s
Carbohydrate
Total
Carbohydrate source
0.1
Re0
U
Amt
g/100 g/loo
%
%
3.9
4.4
4.4
4.4
1,010 3.4
870 2.7
63
70
67
69
57
54
49
55
55
55
42
820
820
830
900
840
870
2.0
2.1
1.2
2.7
1.9
2.4
42
44
25
51
39
48
26
27
16
35
24
31
25
50
u
100
0
150
200
0.7
0.7
0.7
0.7
0.7
0.7
8.0
8.0
8.0
8.0
8.0
8.0
1,070
1,070
1,140
1,100
25
50
100
150
200
0.4 7.8
0.5 7.7
0.5 7.7
0.5 7.7
0.4 7.8
0.5 7.7
0
50
100
150
200
50
100
150
200
250
g/JOO
Proportion of Yield
acid
mIlo
ml/100 mnl
250
Soluble starch
N
ac id
ml
Wheat mash
Batter process
starch
Cumed
ingsue
main-
ml
mg/100 ml
Corn mash
Fumaric acid
utilization
o
6.9
7.5
8.0
8.5
7.8
580
870
1,110
1,200
1,080
2.2
3.4
4.8
5.2
4.2
66
68
75
74
68
32
45
60
61
54
1.8
7.0
7.7
610 2.3
870 3.6
990 4.0
1,390 5.4
1,030 3.9
66
71
70
70
66
33
47
48
63
Flask fermentation: 90-hr
8.3
8.6
8.6
fermentation, 5% inoculum.
A small quantity (50 ppm, v/v) was added to the fermentors before sterilization of the medium; this amount
prevented foam formation during most of the fermentation. The manual addition at daily intervals of an
additional 10 to 20 ppm sometimes was necessary
during later stages of the fermentation.
The pattern of acid formation in this fermentation
is readily ascertained from the data in Table 3. The
decreased yield of acid sometimes noted in samples
taken at the end of the fermentations represents principally the correction applied for volume change in the
fermentor. No actual measurement of the volume of
liquid in the fermentor was possible during the fermentation. Concentration changes thus may have
altered the apparent yields based on samples taken
during the course of the fermentation. Final volumes
of fermentation liquid were determined and the results
for final samples corrected accordingly. At times, the
entire fermentor contents were collected at completion
of fermentation, measured, made to a corrected volume, and a sample then taken for analysis.
Flask fermentations. The fermentation of molasses
besides commercial dextrose was described in a previous publication (Rhodes et al., 1959). Brief trials
were conducted to determine the feasibility of producing fumaric acid as the calcium salt by the fermentation
of still other carbohydrates. The results are illustrated
by the data shown in Table 4. With the exception of
hydrol, these grain carbohydrate substrates were converted to fumaric acid with reasonable efficiency;
it is possible that further adjustment of the fermentation medium would make them even more amenable
to fermentation. Hydrol was not readily utilized, probably because of the high salt content resulting from
neutralization of the acid used for starch hydrolysis.
R. arrhizus strain NRRL 2582, in addition to utilizing glucose and sucrose readily, possesses an amylase
34
1.7
1.1
0.6
0.1
0.8
1.1
0.5
0.2
0.2
13
4.5
12 _
2400
E
Total Acid
>10=
80
Glucose
1600
.%
6
¢ \ EA"
.Fumaric Acid
O
X
E
0
4
2000-E
100
%~~~~~~~~80
go
%
Time, Hours
FIG. 1. Calcium fumarate fermentation in 20-liter fermentors
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
1962]
RHODES, LAGODA, MISENHEIMER, SMITH, ANDERSON, AND JACKSON
by which starches also were quite readily hydrolyzed
and then fermented to form fumaric acid (Table 4).
For efficient fermentative conversion of any carbohydrate substrate to fumaric acid, the supply of nitrogen
mtust be restricted. R. arrhizus apparently has only
limited proteolytic ability and thus can ferment efficiently commercial starch products which contain up
to 3 % protein. Fermentation of whole corn or whole
wheat mash was less efficient than was that of the
starch fractions, probably because of the presence of
excessive, available nitrogen.
Starch substrates were thinned to make them fluid
at fermentation temperatures. Mild acid hydrolysis
was not suitable for this purpose when applied to
starches which contained protein because the acid
hydrolysis procedure made excessive nitrogen available to the mold and resulted in a diminished yield of
product. Liquefaction of starch with enzymes was
most satisfactory. A heat-stable commercial liquefying
enzyme (Rhozyme DX or H-394) was used at recommended concentrations to produce a nearly water-thin
liquid medium containing between 1 and 3 %s concentration of free glucose. The effect of varying nitrogen
levels on the fermentation of starches (Table 5) indicates that less nitrogen was required when impure
starch substrates were fermented than when relatively
pure starches were used.
DIscussIoN
Fumaric acid can be produced efficiently by the
fermentation of glucose with R. arrhizus in 20-liter
fermentors. The fermentation followed a regular
pattern. Initially, there was slow utilization of carbohydrate accompanied by the development of mycelial growth and initiation of acid production. A
progressive increase in the rate of acid production and
the virtual completion of mycelial development occurred between 12 and 24 hr. During the ensuing 24to 48-hr period, the duration depending on available
sugar supply, acid production, and sugar utilization
were nearly linear (Fig. 1). Subsequently, as the concentration of sugar diminished to less than 1 %o and
as the CaCO3 was exhausted, the fermentation became
more erratic and slowed considerably. Residual reducing action equal to 0.5 % glucose always was noted;
it is not known whether this represented unfermented
glucose or whether it resulted from the presence of
other reducing material.
Although a theoretical yield of fumaric acid from
hexoses may be considered to be 64.4 % (mole of fumaric acid per mole of glucose fermented), the data
indicate that during the most rapid phases of fermentation, nearly complete utilization of glucose carbon
to form fumaric acid was obtained. Such rates of conI
Rohm and Haas Company, Philadelphia, Pa.
[VOL. 10
version occurred after mold growth was completed
and may indicate the existence of a 2-carbon utilization pattern of the type established in bacteria by
Kornberg and Krebs (1957). During the 24- to 48-hr
interval after completion of growth, nearly 100% of
the weight of glucose fermented appeared as fumaric
acid. For example, in the fermentor trial shown graphically in Fig. 1, slightly more than 4 g of fumaric acid
per 100 ml were formed from the fermentation of 4.2
g of glucose during the 36- to 48-hr period. Similarly,
in another fermentor of the same run, 4.18 g of fumaric
acid were formed from 3.87 g of glucose fermented.
As may be noted from Table 3, these results were not
unique; the weight yield of fumaric acid often exceeded
64.4 % during the most rapid phases of acid production
although yields of fumaric acid for the fermentation
as a whole were about 65 %.
The formation of soluble fumarates was not practical. At very low glucose concentrations, reasonable
conversion efficiencies sometimes were obtained in
relatively short fermentation intervals; however,
continuous addition of glucose did not alleviate the
inhibition which became apparent as fumarate concentrations reached ca. 3 %. The development of a continuous fermentation to overcome fumarate inhibition
probably would be difficult. Maintenance of sufficient
vegetative mycelial growth in the fermentors would be
especially troublesome with a nonseptate mold such as
R. arrhizus because the mycelium, upon transfer to
fresh medium, tends to increase in particle size and to
form pellets without fragmentation.
No extensive investigation was made of the recovery
of fumaric acid from fermentation liquors. A system
based on the recovery of fumaric acid as the relatively
insoluble free acid (0.7 g per 100 ml) was tried. First,
the pH of the medium was adjusted to 3.5 with HCl
to destroy any residual CaCO3, and the calcium fumarate was put into solution with heat; calcium fumarate
was not appreciably hydrolyzed at this pH. The mycelium which agglutinated during heating was separated readily by coarse filtration. Addition of HCl to
the filtrate to ca. pH 2 resulted in a heavy precipitation of finely divided fumaric acid particles, which
continued as cooling occurred. The relatively great
differential in the solubility of fumaric acid at 100 C
(9.8%) and 25 C (0.7 %) made possible the solution
and recovery of fumaric acid from a minimal amount
of water. A substantial amount of fumaric acid remained adherent to, or contained in, the heated
mycelium after filtration. Most of this acid could be
recovered by further washing and heating of the mycelium in the residual liquor from the initial fumaric
acid precipitation.
LITERATURE CITED
BRIERLEY, M. R., AND R. STEEL. 1959. Agitation-aeration in
submerged fermentation. II. Effect of solid disperse phase
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
14
FUMARIC ACID PRODUCTION
absorption in a fermentor. Appl. Microbiol. 7:57-61.
DWORSCHACK, R. G., A. A. LAGODA, AND R. W. JACKSON. 1954.
on oxygen
Fermentor for small-scale submerged fermentations. Appl.
Microbiol. 2:190-197.
DWORSCHACK, R. G., A. A. LAGODA, AND R. W. JACKSON. 1956.
The control and recording of pH in small-scale fermenta-
tions. Proc. Instr. Soc. Am. 11:paper no. 56-17-1.
KORNBERG, H. L., AND H. A. KREBS. 1957. Synthesis of cell
15
constituents from C2-units by a modified tricarboxylic
acid cycle. Nature 179:988-991.
RHODES, R. A., A. J. MOYER, M. L. SMITH, AND S. E. KELLEY.
1959. Production of fumaric acid by Rhizopus arrhizus.
Appl. Microbiol. 7:74-80.
SHAFFER, P. A., AND A. F. HARTMANN. 1921. The iodometric
determination of copper and its use in sugar analysis. II.
Methods for the determination of reducing sugars in blood,
urine, milk and other substances. J. Biol. Chem.
45:365-390.
Downloaded from https://journals.asm.org/journal/am on 16 May 2022 by 37.132.183.104.
1962]