Communications to the Editor

Fermentation of o-xylose, 0-glucose,

r-arabinose Mixture by Pichia stipitis:
Effect of the Oxygen Transfer Rate on
Fermentation Performance

J. P. Delgenes and R. Moletta

lnstitut National de la Recherche Agronomique, Station d‘oenologie et de
Technologie des Produits Vegetaux, Boulevard du General de Gaulle 11 104
Narbonne Cedex France
J. M. Navarro*
Universite des Sciences et Techniques du Languedoc, Centre de Genie et
Technologie Alimentaires, Microbiologie In dus trielle, place E. Ba taillon
34060 Montpellier Cedex France
Accepted for Publication September 14, 1988

INTRODUCTION
Although the discovery of xylose-fermenting yeasts has
enhanced interest in the microbial conversion of renewable
lignocellulosic resources to ethanol, various problems oc-
curred in the development of an efficient fermentation; the
main problem is that these yeast strains exhibit low ethanol
productivities from D-xylose, compared to those obtained
from D-glucose with other microorganism^.'-^ It appears
that the limiting step of xylose metabolism is the oxido-
reductive pathway from xylose to x y l ~ l o s e . The
~ - ~ metabo-
lism of xylose leads to an overproduction of NADH, H’.
The presence of exogenous hydrogen acceptors, like
oxygen, is one of the keys of the xylose catabolism in
these yeasts; this regulatory mechanism is refered as the
Kluyver effect. lo
To improve the efficiency of xylose fermentation, it is
necessary to optimize the availability of oxygen to culture.
Although it has been demonstrated that oxygen stimulated
ethanol production from D-xylose, little is known about the
relation between the oxygen transfer rate and the fermenta-
tive parameters. This article attempts to study the effect of
oxygen transfer rate on the kinetic and stoichiometric
parameters during the ethanol production from a sugar
mixture by Pichia stipitis Y7124. The sugar mixture, with
D-xylose as the major component, was a model sugar
mixture of a hemicellulosic-derived hydrolysate from
cereal straw.
MATERIAL AND METHODS
Organism and Medium
Pichia stipitis NRRL Y7124 was obtained from the
Northern Regional Research Center, ARS culture collec-
* To whom all correspondence should be addressed.
tion, (Peoria, IL). The organism was cultivated in the me-
dia described by Slininger et al.’’ A mixture of D-xylose,
D-glucose, and L-arabinose in the respective proportions of
75, 20, and 5% was used as carbon source at a total
concentration of 20 g/L. The sugar solution was auto-
claved separately.
lnoculum
Inocula were grown aerobically in Erlenmeyer flasks at
26°C on a rotary shaker at 150 rpm for 24 h. The fermen-
tors were inoculated to an initial dry-cell concentration of
-0.7 g/L.
Fermentation Conditions
Fermentations were run at 30°C in a 2-L Biolafitte fer-
mentor with 1.5 L working volume. The pH was con-
trolled at 5 +O. 1. Dissolved-oxygen tension was followed
using a Biolafitte oxygen analyzer with an Ingold 0,
probe. For all studied fermentations the stirring speed was
800 rpm. For strict anaerobic conditions (noted 0) at ini-
tial time the culture medium was flushed with nitrogen.
For “anaerobic” conditions (noted 0’) at initial time the
culture medium was saturated with air. Under anaerobic
conditions (0 and 0’) each sampling was followed by
flushing aseptically nitrogen in medium. For aerated ex-
periments the studied aeration rates were 0.1, 0.05, 0.025,
0.01, and 0.001 vvm, giving respective oxygen transfer
rates of 20.3, 12.5, 8.4, 4.2, and 0.7 mmol/L h. This
correlation was determined through the oxygen uptake
rate (calculated from off-gas data’,). After inoculation,
dissolved-oxygen tension decreases rapidly from 100 to
near 0% saturation so P . stipitis culture reached a maxi-
mum and stable value of oxygen transfer rate (OTR) during
practically the whole cultivation.

Biotechnology and Bioengineering, Vol. 34, Pp. 398-402 (1989)

0 1989 John Wiley & Sons, Inc. CCC 0006-35921891030398-05$04.00
Analytical Procedures
Dry cell mass was determined gravimetrically: 10-ml
culture samples were filtered, washed, and dried to con-
stant mass at 100°C. Ethanol, sugars, and xylitol were deter-
~ gases (N2, CO,, and
mined as described e 1 s e ~ h e r e . lOff
0,) were analyzed by gas chromatography using a column
filled successively with Silicagel (80-100 mesh) and with
molecular sieve (80-100 mesh) and using a catharometric
detector; the gas carrier was helium. l4
The determination of the molecular formula of P . stipitis
Y7124 was made on an elemental analyzer mod 1106 ac-
cording to the instructions of the manufacturer; the value
for the percentage of each element was the average value
of two determinations.
Ethanol Evaporation
Ethanol evaporation was evaluated in conditions similar
to the most aerated culture except that sterile medium with
5 g/L ethanol was used. Ethanol concentration was mea-
sured periodically, and results showed that ethanol evapo-
ration was negligible during the course of studied cultures.
RESULTS AND DISCUSSION
Influence of Culture Environment on
Fermentative Behavior of Pichia stipitis Y7124
The fermentative behavior of P. stipitis Y7124 grown on
a sugar mixture in aerated batch cultures (0.001 vvm) is il-
lustrated in Figure la,b.
Inoculated on the sugar mixture, P. stipitis consumed
first glucose before assimilation of xylose (Fig. la). The
strain started to utilize L-arabinose so when all the other
sugars were exhausted. At maximum ethanol concentration,
the percentage of consumed arabinose reaches 5% at OTR
0.7 mmol/L h and was comprised in a range of 20-30% at
higher oxygen transfer rates. The sequential utilization of
sugars occurred at all studied oxygen transfer rates except
in anaerobic conditions where L-arabinose was not assimi-
lated. The catabolite repression by glucose was also ob-
served with other xylose-fermenting yeasts grown on sugar
mixtures. ''-I7 The preferential utilization of glucose
compared to xylose could include, apart from catabolite
repression, another regulatory mechanism, i.e., selective
transport as proposed by Slininger. I7

FERMENTATION TIME (H)

(a)

30 40 50
FERMENTATION TIME (H)

(b)

Figure 1. Sugar, ethanol, oxygen and biomass concentrations (a), oxygen transfer rate
(OTR), specific growth rate (w,) and specific ethanol production rate (qp)(b) vs. time in
course of sugar mixture fermentation by Pichia sripitis. (aeration rate 0.001 vvm)

COMMUNICATIONS TO THE EDITOR 399

 The dissolved-oxygen tension decreased rapidly within
the initial 15 h, and the culture was oxygen limited there-
after. The nitrogen and oxygen levels in the gas effluent
95 1
reached the culture maximum and stable values of, re-
spectively, 41 and 1.9%, giving an oxygen uptake rate of
0.022 g/L h (OTR 0.7 mmol/L h). After 5 h culture P.
stipitis exhibited simultaneously maximum specific growth
and ethanol rate (Fig. lb). At higher oxygen transfer rates,
a lag between the growth and ethanol production process OXYGEN TRANSFER RATE h r n o l / l h )
was observed: the specific growth rate peaks at starting Figure 2. Pichia stipifis growth (Yx,J and product (Y,,,) yields, volumet-
culture and the specific ethanol productivity. At OTR ric (Q,) and specific (9,) ethanol productivities, and maximum specific
20.3 mmol/L h., the maximum values of specific growth growth rate (p,,,)vs. oxygen transfer rate. Open symbols refer to andero-
and ethanol production rates occurred, respectively, at 3 bic culture. noted 0'.
and 9 h after inoculation.
The observed ethanol yield was comparable to those
As mentioned by other xylose fermenting P.
reported with other P. stipitis strains grown anaerobically
stipitis Y7124, in the presence of oxygen, utilized pro-
on D-XylOSe. '6,21,22 Under the anaerobic conditions, small
duced ethanol for growth after assimilation of xylose. The
amounts of xylitol accumulated; the xylitol yield (Y,,,, , of
variation of the oxygen transfer rate affected significantly
0.07 g/g) was lower than those obtained with other xylose
the specific rate of ethanol consumption and in a lesser ex-
fermenting yeasts. 17,*' A low transfer rate of oxygen (OTR
tent the cell yield from ethanol (Table I).
0.7 mmol/L h) permitted circumventing the imbalance of
An increase in the oxygen supply involved an increase
NAD+/NADH that occurred in anaerobic condition^.^^
in the specific rate of ethanol consumption by the strain
However, it may be probable that the accumulation of xyli-
(Table I). This dependence could contribute, in addition to
to1 resulted, not uniquely from the presence or from the
regulation mechanisms by oxygen to the utilization of car-
transfer rate of oxygen, but also from the ratio between the
bon from sugars, to a decrease of produced ethanol with an
rate of xylitol production (strictly related with the rate of
increase in oxygen transfer rate (Table I). Furthermore, the
xylose consumption) and the rate of oxygen utilization.
assimilation of produced ethanol by the strain probably
This one determines the rate of regeneration of NAD' re-
may also occur during the phase of ethanol production
quired for the conversion of xylitol to xylulose.
from sugars, suggesting a concurrent production and con-
Increasing the oxygen transfer rate tended to favor the
sumption of ethanol.
production of cells and was detrimental to ethanol produc-
tion (Fig. 2). At the highest oxygen transfer rates, the car-
Effect of Oxygen Transfer Rate on Fermentation bon flows preferentially through the tricarboxylic acid
Parameters
cycle. It showed that the catabolism of studied sugars may
The oxygen transfer rate affected the relative formation be regulated, at least, under a mechanism similar to the
of each fermentative product and the rate of these produc- Pasteur effect, with Saccharomyces cerevisiae growing on
tions (Fig. 2). As produced ethanol was metabolized by glucose. At the highest oxygen transfer rate, the carbon
the strain in the presence of oxygen, fermentative parame- flows Preferentially through the tricarboxylic acid cycle.
ters of aerated cultures were calculated at maximum accu- The percentage of sugar carbon distributed between the
mulated ethanol concentration ( E ) . fermentative pathway and the tricarboxylic acid cycle was
The highest ethanol yield near 0.42 g/g was obtained in stable during the fermentation since the ratio between all
both anaerobic cultures. Furthermore, the kinetic parame- ethanol concentrations ( X / P ) was constant in the course of
ters between these two cultures do not differ significantly. fermentations. When OTR = 20.3 mmol/L h, substrate

Table I. Influence of oxygen transfer rate on growth yield from ethanol (Y+), specific rate of
ethanol utilization ( r , ) , and production of ethanol by Pichia sripitis Y7124.a

Oxygen transfer 20.3 12.5 8.4 4.2 0.7 O+ 0

rate (mmol/L h)

Yd,,(g/g) 0.76 0.80 0.82 0.79 0.69 0 0

r, (g/g h) 0.040 0.025 0.017 0.011 0.003 0 0
E (g/L) 3.3 4.1 4.6 6.0 1.4 5.9 6.0
tf ( h ) 13.0 15.5 18.5 23.0 49 397 397

a Symbols: YXc, yield of produced cells from ethanol after ,f and corrected with regard to ara-

binose consumption; r,, quantity of ethanol consumed after t, divided by period of ethanol utiliza-
tion ( 1 , ) and average cell concentration during t , .

400 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 34, JULY 1989

carbon was distributed as 48% for cells, 21% for ethanol,
and 31% for carbon dioxide. When the oxygen transfer
rate was low, the strain utilized more intensively the fer-
mentative pathway for the oxidation of NADH cofactors;
at OTR = 0.7 mmol/L h, 50.5% of the carbon source was
incorporated into ethanol, 15.5% into cell mass, and 23%
into carbon dioxide.
An increase of the oxygen transfer rate, in addition to
an increase of the cell yield (Fig. 2), led to a gradual in-
crease of the volumetric and specific rates of cells pro-
duction (Fig. 2). At oxygen transfer rates higher than
4.2 mmol/L h, the volumetric rate of ethanol production
was maximum in relation to a more rapid and large pro-
duction of cells. The specific rate of oxygen uptake in-
creases gradually when the oxygen transfer rate increases;
the values of maximum specific oxygen uptake rate are
0.018 and 0.155 g/g h at studied OTR limits.
The specific ethanol productivity was dependent on the
oxygen transfer rate. It was 4.5 times higher in the culture
run at OTR = 0.7 mmol/L h than in the anaerobic fer-
mentation (Fig. 3). In these conditions P. stipitis exhibited
the highest fermentative performances with an average
specific ethanol productivity of 0.09 g/g h and a qp,max of
0.24 g/g h. Watson et a].” using Pachysolen tannophilus
Y2460 grown on D-xylose obtained a qP,,,, of 0.07 g/g h
at oxygen transfer rates of 0.09 and 1.18 mmol/L h.
When the oxygen transfer rate exceeded 4.2 mmol/L h ,
the specific ethanol productivity declined (Fig. 2), although
the regeneration of NAD required during the conversion of
xylulose would be easier; a significant supply of oxygen
induces, in the strain, a deviation of the pyruvate flow
from the fermentative pathway to the TCA cycle, as re-
ported by Slininger et al.I7 with P. tannophilus.
Stoichiometry of Ethanol Production by Pichia
stipitis Y7124
The general equation of ethanol production could be de-
scribed as shown:
aCH20 + PNH3 + yo,+ 8C,Hb0,N, + &C2H6O
+ xH,O + TCO,
where a , P , y , 8, 8, x, T are stoichiometric coefficients
and C,H,O,N, refers to the formula for cells of P. stipitis
Y7124. The nitrogen source was presumed to be ammonia.
Experimental data lead to the determinations of a , y , 6, E ,
and T . The coefficient /3 and x can be calculated through
the nitrogen and hydrogen balances. The empirical formula
of P. stipitis’ cells was determined during the cultures run
with OTR = 20.3 mmol/L h; the C , H, 0, and N percent-
ages were, respectively, 42.96, 6.42, 34.5 1, and 8.41,
giving a molecular formula CHI 7900 60NO 17.
With aeration, in hypothezing that, first, the only source
of carbon was the sugar mixture and, second, the different
sugars of the medium were assimilated according to a simi-
lar way, the experimental equation for the production of
ethanol from the sugar mixture by P. stipitis Y7124, would
be, in moles (OTR = 20.3 mmol/L h)
2.06CH20 + 0.460, + 0.17NH3 --+
+ 0.23C,H60 + 0.73H20 + 0.66C0,
CHl,,900,60No~17
Carbon and oxygen balances between the first and sec-
ond terms of the equation do not differ at most from 4%.
CONCLUSION
Using a glucose, xylose, and arabinose mixture, an il-
lustration and quantification of the dependence on oxygen
of P. stipitis Y7124 was attempted. The ethanol yield was
maximized in anaerobic conditions, where the flow of
pyruvate through preferentially the fermentative pathway
did not generate a significant loss of substrate carbon into
the cell mass. However, P. stipitis showed a low ethaqol
productivity in relation to a metabolic deregulation of the
redox balance in the catabolism of the major sugar since
xylitol accumulated. An oxygen supply stimulates ethanol
productivity and diminished the formation of xylitol. In
these conditions the respiratory chain was operating and,
by generating a loss of carbon source into cells, minimized
the ethanol yield. The results were as follows: Aeration con-
ditions that support the ethanol yield were different from
those that optimize the ethanol productivity. In the develop-
ment of a process for an efficient production of ethanol by
P. stipitis, the aeration strategy results in a compromise.
Results indicated that P. stipitis Y7 124 exhibited a maxi-
mum specific ethanol productivity of 0.09 g/g h at the
lowest studied oxygen transfer rates. Considering ethanol
yield and ethanol productivity, the fermentation run with an
oxygen transfer rate of 0.7 mmol/L h appeared to be more
efficient than the anaerobic one since for near ethanol
yields, the specific ethanol productivity was, compared to
the anaerobic culture, 4.5 times higher in the culture con-
ducted with OTR = 0.7 mmol/L h. At these conditions,
the strain exhibits an average specific rate of oxygen uptake
of 0.010 g/g h. with a maximum of 0,018 g/g h; these
data may be of significance for the improvement of the low
observed ethanol productivity.
NOMENCLATURE
maximum accumulated ethanol concentration (g/L)
oxygen transfer rate (mmol/L h)
average volumetric ethanol productivity (g ethanol/L h)
maximum volumetric ethanol productivity (g ethanol/L h)
average specific ethanol productivity (g ethanol/g cells h)
average rate of cell production (g cells/L h)
average specific rate of ethanol consumption (g ethanol/
g cells h)
period corresponding to oxydation of produced ethanol (h)
time at which produced ethanol was maximum (h)
ethanol yield (g produced ethanol/g utilized substrate)
cell yield from ethanol (g produced cells/g utilized ethanol)
cell yield (g produced cells/g utilized substrate)
xylitol yield (g produced xylitol/g utilized xylose)
specific growth rate (h-I)
maximum specific growth rate (h-l)
COMMUNICATIONS TO THE EDITOR 401
References
1. K. J. Lee, D. E. Tribe, and P. L. Rogers, Biotechnol. Lett., 1, 421
(1979).
2. K. J. Lee, M. L. Skotnicki, D. E. Tribe, and P. L. Rogers, Biotech-
nol. Lett., 2, 339 (1980).
3 . P. L. Rogers, K. L. Lee, and D. E. Tribe, Biotechnol. Lett., 1, 165
(1979).
4. B. L. Maiorella, H. W. Blanch, and C. R. Wilke, Biotechnol. Bio-
eng., 26, 1003 (1984).
5 . M. Chakravorty, L. A . Veiga, M. Bacila, and B. C. Horecker, J .
Biol. Chem., 237, 1014 (1962).
6. P. M. Bruinenberg, P. H. M. De Bot, J. P. Van Dijken, and W. A.
Scheffers, Eur. J . Appl. Microbiol. Biotechnol., 18, 287 (1983).
7. C. Verduyn, J. Frank, J. P. Van Dijken, and W. A. Scheffers, FEMS
Microbial. Lett., 30, 313 (1985).
8. C. Verduyn, R. Van Kleef, J. Frank, H. Schreuder, J. P. Van Dijken,
and W. A. Scheffers, Biochem. J., 226, 669 (1985).
9. P. M. Bruinenberq, P. H. M. De Bot, J . P. Van Dijken, and W. A .
Scheffers, Appl. Microbiol. Biotechnol., 19, 256 (1984).
10. A . P. Sims and J. A . Bamett, J . Gen. Microbiol., 106, 277 (1978).
1 1 . P. J. Slininger, R. J. Bothast, J. E. Van Cauwenberge, and C. P.
Kurtzman, Biotechnol. Bioeng., 24, 371 (1982).
12. C. H. Cooney, H. Y. Wang, and D. I. Wang, Biotechnol. Bioeng.,
19, 55 (1977).
402 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 34, JULY 1989
13. J . P. Delgenes, R. Moletta, and J . M. Navarro, in Proceedings ofthe
Second Catalan Congress on the Renewable Solar Energies, 5-7 Oc-
tober 1987 Gerona.
14. J . L. Uribelarrea, Doc. Ing. Thesis, INSA Toulouse, 1980.
15. K.F. Deverell, Biotechnol. Lett., 5, 475 (1983).
16. J . C. Du Preez, M. Bosch, and B. A. Prior, Appl. Microbiol. Bio-
technol., 23, 228 (1986).
17. P. J. Slininger, P. L. Bolen, and C. P. Kurtzman, Enz. Microbiol.
Technol., 9, 5 (1987).
18. T. W. Jeffries, Biotechnol. Lett., 3, 213 (1981).
19. R. Maleszka and H. Schneider, Appl. Environ. Microbiol., 44, 909
(1982).
20. A . Margaritis and P. Bajpai, Appl. Environ. Microbiol., 5 , 1039
(1982).
21. J . C. Du Preez, M. Bosch, and B. A . Prior, Enz. Microbiol. Technol.
8, 360 (1986).
22. H. Dellweg, M. Rizzi, H. Methner, and D. Debus, Biotechnol. Lett.,
6, 395 (1984).
23. J. C. Du Preez and B. A . Prior, Biotechnol. Lett., 7, 241 (1985).
24. J. P. Van Dijken and W. A . Scheffers, FEMS Microbiol. Rev., 32,
199 (1986).
25. N. E. Watson, B. A. Prior, J. C. Du Preez, and P. M. Lategan, Enz.
Microb. Technol., 6, 447 (1984).