APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 1984, p. 639-646 Vol. 48. No.

3
0099-)'40/84/090639-08$02.00/0
Copyright C 1984, American Society for Microbiology

High-Gravity Brewing: Effects of Nutrition on Yeast Composition,

Fermentative Ability, and Alcohol Production
GREGORY P. CASEY. CAROL A. MAGNUS, AND W. M. INGLEDEW*
Departmeiet of Applied Microbiology a,zd Food Scien7ce, University of Saskatchewvan, Saskatooni, Saskatc-lIewan, Canada
S7N OWO
Received 16 April 1984/Accepted 26 June 1984

A number of economic and product quality advantages exist in brewing when high-gravity worts of 16 to
18% dissolved solids are fermented. Above this level, production problems such as slow or stuck fermentations
and poor yeast viability occur. Ethanol toxicity has been cited as the main cause, as brewers' yeasts are
reported to tolerate only 7 to 9% (vol/vol) ethanol. The inhibitory effect of high osmotic pressure has also been
implicated. In this report, it is demonstrated that the factor limiting the production of high levels of ethanol by
brewing yeasts is actually a nutritional deficiency. When a nitrogen source, ergosterol, and oleic acid are added
to worts up to 31% dissolved solids, it is possible to produce beers up to 16.2% (vol/vol) ethanol. Yeast viability
remains high, and the yeasts can be repitched at least five times. Supplementation does not increase the
fermentative tolerance of the yeasts to ethanol but increases the length and level of new yeast cell mass synthesis
over that seen in unsupplemented wort (and therefore the period of more rapid wort attenuation). Glycogen,
protein, and sterol levels in yeasts were examined, as was the importance of pitching rate, temperature, and
degree of anaerobiosis. The ethanol tolerance of brewers' yeast is suggested to be no different than that of sake
or distillers' yeast.

In traditional brewing, worts of 11 to 12% dissolved solids nitrogen (22), literature reports with normal-gravity worts
are fermented to produce beers of 4 to 5% (vol/vol) ethanol. have illustrated that nitrogen-induced problems, mimicking
Recently, high-gravity brewing at a limit of 16 to 18% those found in high-gravity brewing, are found when more
dissolved solids has become popular due to advantages such and more of the extract is substituted with adjunct (19. 22).
as increased plant efficiency; reduced energy, labor, and Oxygen is required by brewing yeasts for the synthesis of
capital costs; use of higher adjunct ratios; improved smooth- sterols and unsaturated fatty acids (1. 2). Such lipids are
ness, flavor, and haze stability of beer; and increased present in suboptimal concentrations even in normal-gravity
ethanol yields per unit of fermentable extract (15, 38). worts (7). In high-gravity worts, oxygen availability is dimin-
Attempts to ferment worts above 18% dissolved solids ished even further due to the decreasing solubility of oxygen
have proven to be difficult, largely because of problems with with increasing wort gravity (3). As reproductive growth
yeast viability and slow and incomplete fermentations (9, ceases once a limiting value of sterols is reached in yeasts
37). Both ethanol toxicity (9, 37) and high osmotic pressure (22), this lowered 02 solubility in high-gravity worts only
levels (29) have been implicated as the limiting factors. increases the probability of growth-related attenuation prob-
In recent studies with normal-gravity worts, it has been lems.
indicated that the factors described above could be related to Recently, our research has demonstrated that a combina-
nutritionally induced growth problems. In sharp contrast to tion of increased pitching rates (5) and nutritional supple-
the long-held belief in brewing that the bulk of wort attenua- mentation (6) can permit the rapid fermentation of worts up
tion is done by nongrowing cells (22), it is now clear that the to 28% dissolved solids. Specifically, supplementation of
specific rate of sugar utilization by growing yeast cells in these worts with yeast extract (as a source of assimilable
fermentation is substantially higher than that of nongrowing nitrogen)-ergosterol-oleic acid (to provide a sterol and unsat-
cells (22). Thus, when the period of new cell mass produc- urated fatty acid) at a pitching rate of 20 x 106 to 30 x 106
tion ceases during fermentation. the rate of attenuation also CFU per ml resulted in the production of up to 15% (vol/vol)
slows dramatically (by as much as 33-fold) (22). It therefore ethanol at 14°C within 5 days (compared with 2 weeks in
follows that in high-gravity brewing, both the length and unsupplemented worts). The improvements were a result of
level of new cell synthesis must be increased over the prolonged and increased production of yeast cell mass
amounts found in normal-gravity brewing to have rapid arising from supplementation (6). This demonstrates that
fermentation. It is our belief that the concentration of the brewers' yeasts, without any genetic manipulation or strain
nutrients most likely to limit growth (e.g., oxygen and improvement, are tolerant to at least 15% (vol/vol) ethanol.
assimilable nitrogen) must certainly be increased in such The results presented here describe continued research on
worts. the importance of nutritional supplementation in high-
Such nutritional concerns have not been considered in gravity fermentations.
brewing practice. The most widely used method to prepare MATERIALS AND METHODS
high-gravity worts has been the addition of corn syrups to
the kettle (15). Such syrups are virtually devoid of any Brewing yeast. Fresh slurries of a commercial lager yeast,
nitrogenous nutrients, and their use effectively decreases the Saccharoinvces livarimn (carlsbergensis). were collected just
proportion of all noncarbohydrate nutrients in the wort (20). before use.
Although worts made only of malt contain excess assimilable High-gravity worts. The 11.5°P (22.5% corn grit adjunct)
commercial wort used was autoclaved to remove protein and
* Corresponding author. to allow yeast cell mass assessment without interference
639
640 CASEY, MAGNUS, AND INGLEDEW APPL. ENVIRON. MICROBIOL.

from precipitated protein (trub). Fermentations are unaffected Cellular analyses. Protein was estimated spectrophotomet-
by this step, as brewers' yeasts cannot use protein as a ni- rically at 750 nm (23), with bovine serum albumin (Sigma)
trogen source (19, 31). For the preparation of worts of higher used as the standard protein.
original gravity, Casco syrup no. 1636U was added (Canada Glycogen levels were determined spectrophotometrically
Starch Co. Ltd.). The manufacturer's carbohydrate specifica- at 660 nm by an idoine-potassium iodide staining method
tions for this syrup are 19.4% DP4+, 10.7% DP3, 39.0% DP2, (33). Rabbit liver glycogen (Sigma) was used as the standard.
and 30.9% DPi. The syrup is virtually nitrogen free, containing Total cell sterol levels were measured by the modified
only 0.03 to 0.06% (wt/vol) protein. In special instances, Liebermann-Buchard reaction (13). Ergosterol (Sigma) was
maltose or glucose was used in place of the syrup. used as the standard.
Chemicals. Ergosterol (Sigma Chemical Co., St. Louis, High-pressure liquid chromatographic analyses. A Beck-
Mo.) was dissolved in a mixture of ethanol-Tween 80 (1). man model 420 high-pressure liquid chromatographic system
Tween 80 (Sigma) serves as a source of unsaturated fatty was used. Detection was carried out with an Altex 156
acid, containing 0.6% free oleic acid (8). Yeast extract (Difco refractive index detector, and integration was carried out
Laboratories, Detroit, Mich.) was used as the nitrogen with a Hewlett Packard 3390A integrator. The heating block
supplement. (model CH-20; Scientific Systems Inc.) contained a ,u-
Fermentation conditions. All fermentations were in 1- and Spherogel column (Beckman Instruments, Inc.) designed for
2-liter jacketed Wheaton Celstir fermentors and stirred at 90 carbohydrate analysis. Column dimensions were inner diam-
rpm by a Wheaton model III Biostir 6. The temperature was eter, 7.5 mm; outer diameter, 10 mm; and length, 300 mm.
maintained at 14°C with a Haake model D-3G circulator. The packing was 7.5% cross-linked calcium-loaded sulfonat-
Unless otherwise stated, anaerobiosis, as experienced indus- ed polystyrene divinylbenzene resin. The operative tempera-
trially by CO, purging, was maintained by continually flush- ture was 80°C, with degassed, distilled, and filtered (pore
ing the headspace with nitrogen gas at a rate of 30 ml/min. In size, 0.45 [Lm) water used as the solvent. The flow rate was
special instances, however, semianaerobic fermentations 0.6 ml/min at 400 lb/in2. Sorbitol was used as the internal
were carried out where there was no nitrogen flushing, and standard at a constant concentration of 3.0 mg/ml. Sucrose
foam plugs were used in the sampling ports. Pitching rates of had a retention time identical to that of maltose, but as it
20 x 106 to 30 x 106 CFU per ml were normally used. only constitutes 1.9% of the DP2 in these worts (31), it was
Viable counts. Viable counts were determined by the not considered separately from the maltose. The ratios of
membrane filtration technique (5). peak height responses of each individual sugar to the sorbitol
Total solids and dry weight determinations. Carbohydrate internal standard were plotted against the concentrations.
utilization was approximated by removing yeasts by centrif- The concentrations of the sugar in the wort and beer samples
ugation and assessing total solids in the supernatant (5). were obtained by calculating their ratios of peak heights
For dry weight determinations during manometric analy-
ses, cell pellets were washed three times in M/15 KH2PO4
buffer (pH 4.5) and then resuspended in the same solution. 30
These suspensions were used for manometry, and triplicate
3-ml samples of the suspensions and of the resuspended
KH2PO4 were transferred to preweighed aluminum foil pans.
Pans were dried to a constant weight at 105°C (corrected for 25
buffer weight), and the cell dry weight (in milligrams per
milliliter) was calculated. For dry weight determinations of
the fermentation samples, distilled water was used in the
place of M/15 KH2PO4 buffer. E 20
Ethanol assays. Ethanol was measured enzymatically with
alcohol dehydrogenase, as outlined in Sigma Technical Bul-
letin number 331 U.V.
FAN. Free alpha amino nitrogen (FAN) levels in wort and 0
beer were determined by the European Brewing Convention
ninhydrin method (11) at 570 nm. Glycine (Fisher Scientific
Co., Pittsburgh. Pa.) was used to prepare a standard curve,
and the results were calculated by linear regression.
Manometric analyses. The fermentative power of washed !210>_
ON N
yeast samples was determined in a Warburg respirometer by
the European Brewing Convention Analytica Microbiologica
manometric technique (12). The single alteration to the
previously published method was the continual maintenance 5
of anaerobic conditions (via nitrogen flushing) in the samples
from the time of sampling onward. Fermentative power
(QCO) was expressed as the number of microliters of CO,
given off under a nitrogen atmosphere per hour per milligram 0
(dry weight) of yeast sample; this was measured over a 90- 0 2 4 6 8 10
min period. To determine fermentative tolerance to ethanol, TIME (days)
(Qc/)j values were determined in the presence of 0, 5, 10, 15, FIG. 1. Dissolved solids versus time during the semianaerobic
and 2a0% (vol/vol) ethanol. From these, the concentration of fermentation of high-gravity worts that were unsupplemented (A),
ethanol resulting in a 50% reduction of the (QN .) value 1% yeast extract supplemented (-). 40 ppm ergosterol and 0.4%
found in the absence of ethanol (hereafter referred to as the Tween 80 supplemented (V), or 1% yeast extract with 40 ppm
IF,(0) was then calculated. ergosterol and 0.4% Tween 80 supplemented (0).
VOL. 48. 1984 HIGH-GRAVITY BREWING 641

relative to that of the sorbitol internal standard and then by mentable carbohydrates (i.e., glucose, fructose, maltose,
referring to the calibration plot and multiplying by the and maltotriose) is complete in fully supplemented worts.
dilution factor. Low levels of maltose and maltotriose (0.85 and 0.99 g per
100 ml) remained in this beer. These levels are normal after a
RESULTS primary fermentation and would be reduced further during
Evidence for the stimulatory role of oxygen under semian- subsequent aging (31). In the unsupplemented fermentation,
aerobic conditions and for the importance of nutritional however, considerable amounts of fermentable carbohy-
supplementation is presented in Fig. 1. Under semianaerobic drate, especially maltose (2.65 g per 100 ml), still remained
conditions, the added yeast extract was stimulatory (it was even after 8 days.
not stimulatory anaerobically [6]), as 0 was available for The cellular levels of glycogen and sterols in yeasts
lipid synthesis under semianaerobic conditions. Anaerobi- obtained from the end of 27% dissolved solids wort fermen-
cally, lipid addition (40 ppm ergosterol and 0.4% Tween 80) tations were found to be significantly influenced by the use
enhanced fermentation (6), whereas under the conditions of anaerobic or semianaerobic conditions (Table 1). Al-
used in this study, the same added lipids were not stimula- though protein levels were similar in both sets of fermenta-
tory (as oxygen was being introduced into the unsupplement- tions, the levels of cellular glycogen were significantly lower
ed wort), and the wort attenuated at the same protracted rate and the levels of cellular sterols were higher in each of the
as under unsupplemented conditions. With complete supple- semianaerobic fermentations.
mentation, however, the fermentation time of this 27% In the anaerobic fermentations. the yeast crop from the
dissolved solid wort was reduced to 4 days, and a synergistic fully supplemented wort had the greatest level of cellular
effect of the two supplements was seen as described previ- sterols (0.56%) but the lowest level of glycogen (10.5%).
ously (6). When followed throughout the course of fermentation, sup-
Analysis of the levels of FAN before and after fermenta- plementation resulted in higher maximum levels of sterols
tion illustrates that in the presence of added lipids, useable (0.85 versus 0.63%), and as growth continued, the sterol
nitrogen is, indeed, growth limiting and that the simulta- concentration was diluted. The final levels reached were
neous presence of preformed lipids or oxygen is required for higher in the supplemented yeasts (0.61%) than in the
full utilization of such nitrogenous constituents. Anaerobi- unsupplemented yeasts (0.48%). In the case of glycogen,
cally, only 106 and 107 mg of FAN per liter were utilized in however, peak levels were higher in yeasts from the unsup-
the unsupplemented and 1% yeast extract-supplemented plemented fermentation (21.0 versus 17.4%) and remained at
worts, respectively (despite the ca. 400-mg/liter increase in higher levels by the end of the fermentation (19.8 versus
FAN levels arising from yeast extract addition). In the lipid- 12.5%).
supplemented wort, 165 mg of the original 213 mg of the The possibility remained that nutritional supplementation
available FAN per liter was utilized. However, when both of was also enhancing the rate and extent of fermentation by
the supplements were present, 289 mg of FAN per liter was increasing the tolerance of yeasts to ethanol. Unsaturated
utilized (36% more FAN than the starting total in unsupple- fatty acids, in particular, have been demonstrated to play a
mented wort). Increased cell growth (data not shown) and a role in the alcohol tolerance of yeasts (35). Table 2 records
significantly reduced fermentation time resulted. the (Qcc&) values of yeasts sampled throughout the fermen-
Semianaerobically, however, when there is contact with tation of unsupplemented and fully supplemented worts, and
air, more of the available FAN was utilized in the unsupple- Table 3 provides yeast cell mass and other fermentation
mented wort (128 versus 106 mg/liter), as well as in the yeast data. In virtually every instance, regardless of the concentra-
extract-supplemented wort (229 versus 107 mg/liter) than tion of ethanol present during the assay, the fermentative
anaerobically, which explains the stimulation seen in the rate power of unsupplemented yeasts was equal to or greater
of fermentation. Maximum utilization of FAN (286 mg/liter) than that of the yeasts in the supplemented fermentations.
was only seen when both the yeast extract and the lipids Average IFs( values for the unsupplemented yeasts were
were present (73% more FAN than the starting total in 11.8% (vol/vol) ethanol (±1.20X%), as compared with 11.4%
unsupplemented wort). (vol/vol) for the supplemented yeasts (40.5%).
High-pressure liquid chromatographic analysis of the Previously, 14.2% (vol/vol) ethanol was produced in fully
worts and beers from the fermentations described above supplemented wort when maltose, which can be entirely
(data not shown) demonstrated that attenuation of the fer- fermented, was used as the brewing adjunct instead of corn

TABLE 1. Percent glvcogen, protein, and total sterols (on a dry weight basis) in the crops of yeasts from high-gravity fermentations with
or without nutritional supplementation under anaerobic or semianaerobic conditions
'4 of the following after semianaerobic
% of the following after anaerobic fermentations: fermentations:
Treatment fermentations:
Glycogen Protein Sterols Glycogen Protein Sterols
Unsupplemented wort 28.4 42.1 0.51 6.7 45.0 0.85
1% yeast extract 25.8 45.6 0.27 7.0 47.9 1.01
supplemented
40 ppm ergosterol- 26.5 36.9 0.40 9.5 43.1 1.09
0.4% Tween 80 (vol/vol)
supplemented
1% yeast extract-40 ppm 10.5 43.5 0.56 7.8 44.6 0.89
ergosterol-0.4%
Tween 80 (vol/vol)
supplemented
642 CASEY, MAGNUS, AND INGLEDEW APPL. ENVIRON. MICROBIOL.

TABLE 2. Q(', values of washed yeasts removed fi-om the anaerobic fermentaction of unsupplemented and supplemented 27% dissolved
solids wort"
0 values" in the pr-esence ot' the lollowing '74 (vol/vol) ethanol
Day Utisupplemented Supplemented
0) 5 1() 15 20 () 5 10 15 2(
0 202 140 108 34 27 194 155 97 35 23
1 303 267 173 93 41 311 244 161 98 64
2 182 140 94 47 33 182 145 105 44 39
3 183 173 117 64 44 181 140 100 55 44
4 171 162 11( 65 31 167 126 84 55 25
5 163 136 '92 48 18 171 125 82 52 34
7 130 104 72 42 29 135 104 76 40 31
9 108 100 72 41 28 ND' ND ND ND ND
13 101 70 44 25 14 ND ND ND ND ND
' Supplemented worts consisted of O.X' yeaist extract-24 ppm ergostei-ol-0.24'i% (vol/vol) Tween 8(). ReduLced levels of supplcments had no negative effects on
the fermentation (6).
Qc,, values are in microliters of CO2 per hoLir per milligram ot' cells (dry weight).
ND, Not determined.

syrup, which is only 70 to 80% fermentable (6). To determine The influence of temperature on these high-gravity wort
the original gravity limit to maltose adjunct worts, worts fermentations was assessed in light of recent reports on the
with up to 35% dissolved solids were fermented (Fig. 2). relationship between it and alcohol tolerance. Supplemented
Although yeast viability remained high in all worts (data not high-gravity corn syrup adjunct worts fermented at 14, 20,
shown), only those with up to 31.9% dissolved solids 25, and 30°C proceeded to a faster completion as the
reached end gravity, producing 16.2% (vol/vol) ethanol. temperature increased (Fig. 3). However, when yeast viabili-
It is important in brewing that a portion of the yeast crop ty was ftollowed to day 5, it was extremely low in the
from the end of one fermentation is able to be used to initiate fermentations at higher temperatures (inset, Fig. 3), despite
a fresh fermentation. Supplemented 28% maltose adjunct the production of similar levels of ethanol in each (ca. 11.4%
worts were repitched five times without any decrease in [vol/vol]). At 14°C, yeast viability remained relatively un-
fermentative ability, with the fermentations being complete changed, even 10 to 14 days after the completion of fermen-

15}
within 5 days in all cases (data not shown). Yeast viability
and cell mass production remained high throughout the five 35
runs, and high-pressure liquid chromatographic analysis of
the beers after runs 1 and 5 indicated normal attenuation of
all the fermentable carbohydrates (<0.2% maltose, <1.4%
maltotriose). Final ethanol concentrations of 14.2, 13.6,
14.3, 13.7, and 14.2% (vol/vol) were reached in runs 1
through 5, respectively. Analysis of glycogen and sterols in
the yeasts from the resulting crops in these fermentations
indicated thait as in the case of single-batch fermentations
(Table 1), the levels of glycogen were ca. 50% lower in the 25
yeasts from the supplemented fermentations, whereas the
levels of the sterols were at least 15% higher.

TABLE 3. Fermentation parameters during the anaerobic 7a

fermentation of the unsupplemented and supplemented 27%/
dissolved solids worts used to determine yeast Q(`,(), and IFl,,
values"
Dissolved solids Yeatst dry wt Yeaist viability
(g/iIt0) ml) (mng/nil) (CFU/ml x 1t) ")
Day
Unsupple- Supple- Unsupple- Supple- Unsupple- Supple- 10 N% A
mented mented mented mented mentedmented
0 27.2 27.7 3.4 3.4 21.4 22.4 N
1 24.0 23.9 6.9 7.5 34.6 35.2 N.
2 19.t) 17.5 9.2 11.4 44.6 54.7
3 17.9 12.6 8.6 13.8 42.1 68.2
4 16.8 10.3 9.( 15.) 45.3 68.0)
5 16.6 9.7 9.9 15.9 50.3 73.9
7 15.5 9.6 9.3 14.8 51.7 73.9
9 14.4 NDD" 9.4 ND 44.7 ND o
0 3 6 9 12 15
13 11.8' ND 8.9 ND 48.6 ND
TIME (days)
Supplemented worts consisted of t).8'4 yeast extr-act-24 ppnm ei-gosterol-
0.24% (vol/vol) Tween 80. FIG. 2. Dissolved solids versus time during the anaerobic fer-
"ND. Not determined. mentation of fully supplemented 27.9% (0). 29.2'% (O), 30.9% (V).
The termentation eventually stuck at 11. 1 g pc- 1()(1 ml ot dissolved solids. 31.9% (0). 33.3% (O) atnd 35%/e (A) dissolved solids worts.
VOL. 48, 1984 HIGH-GRAVITY BREWING 643

highest rates did the fermentations proceed to completion

Temperature Maximum Cell # of % of
(although 17 to 18 days were required), whereas at the four
Coll # 120 hr Maximum lower pitching rates, the fermentations became stuck or
* 14"C 6.2 x107 6.1 x10
7-
98 were only very slowly fermenting (sluggish), even after 3
20"C 6.8x107 5.8x007 weeks. Pitching rate, then, can also influence the rate and
w 85 extent of attenuation, especially in unsupplemented worts.
A 250C 7.2 x107 2.8x107 39 Alternative sources to be used in place of the addition of
* 300C 4.6 x O7 3.0x104 0.1 preformed lipids and yeast extract were explored. It is
E 2C indeed possible to entirely replace the need for preformed
0
0
R
lipids by sparging with oxygen during the first 2 days of
4A
fermentation (4 h per day with 100 ml of oxygen gas per min)
-j
a
(Fig. 4). Cell mass production increased 18% (3.2 mg/ml) in
the oxygenated fermentation over that seen in the fermenta-
0
tion provided with preformed lipids, resulting in a somewhat
a
lower final concentration of ethanol (11.5 versus 12.4% [vol/
I"
w vol]). Final sterol concentrations were 1.14 and 0.71% in the
U,
1

I I yeasts provided with oxygen and preformed lipids, respec-

tively. The timing of the application of the oxygen was found
to be significant, as oxygenation during days 2 through 4 of
fermentation was not as effective as during days 0 through 2.
The requirement for yeast extract could be reduced by
using more malt. For example, instead of using the 11.5%
(22.5% corn adjunct) normal-gravity wort as a base, an all-
malt, normal-gravity wort was used. Corn syrup was used to
adjust the wort to 31%. It was then possible to decrease the
level of yeast extract from 0.8 to 0.2% without decreasing
TIME (days) the rate or extent of fermentation (data not shown). The
FIG. 3. Dissolved solids versus time during the anaerobic fer- additional FAN contained in malt (Table 5) provided suffi-
mentation of 0.8% yeast extract with 24 ppm ergosterol-0.24% (vol/ cient levels of FAN even in the 0.2% yeast-supplemented
vol) Tween 80-supplemented worts at 14, 20, 25. and 30°C. fermentation. Final ethanol concentrations of 9.0, 12.2, 12.4,
and 12.9% (vol/vol) were reached in the unsupplemented,
tation (data not shown). The poor yeast viability seen at the
higher temperatures would certainly preclude its use for 30
repitching in brewing.
The importance of pitching rate in the fermentation of
high-gravity corn syrup adjunct worts was determined. It has
been previously reported that yeast viability and fermenta-
25
tive ability at 0 to 12 h can be improved in unsupplemented
worts by using higher pitching rates (5). When supplemented
29% dissolved solid worts were examined (Table 4), the
fermentations proceeded to completion faster as the pitching
rate increased up to the inoculum of 2.2 107 CFU/ml (the
x 20
fermentation pitched with an inoculum of 3.1 to 107 CFU/ml
fermented at virtually the same rate). All five pitching rates, E~~~~~
however, resulted in fermentation times of 1 week or less.
In unsupplemented high-gravity worts, however (Table 4),
the pitching rate was of more significance. Only at the two 0 .
- ~~~~~~~~~0
TABLE 4. Influence of pitching rate on the fermentation of
unsupplemented and supplemented 29% dissolved solids high-
gravity wort
Pitching rate and days to end fermentation in the following worts":
Unsupplemented Supplemented
Pitching Time Pitching Time
rate (dy)rate (as
(CFU/ml) (days) (CFU/ml) (days)
3.2 x 106 Sluggish at 21 (14%)h 4.6 x 106 7 0
6.8 x 106 Sluggish at 21 (12.5%) 9.0 x 106 6 0 2 4 6 8
1.4 x 107 Stuck at 18 (11.5%) 1.6 x 107 5 TIME (days)
2.6 x 107 Stuck at 18 (11.5%) 2.2 x 5 FIG. 4. Dissolved solids versus during the anaerobic fermenta-
3.7 x 107 17-18 3.1 x 107 5 tion of unsupplemented 30% wort (A). 0.8% yeast extract supple-
4.4 x 107 17-18 mented wort treated with 24 ppm ergosterol-0.24% (vol/vol) Tween
End gravity was 10% dissolved solids. 80 (0), 4-h periods of oxygenation during days 0. 1, and 2 (-), or 4-h
Percent dissolved solids at the day indicated. periods of oxygenation during days 2. 3, and 4 (O).
644 CASEY, MAGNUS, AND INGLEDEW Appi- ENVIRON. MICROBIOL.

TABLE 5. FAN levels during the anaerobic fermentaition of 31% 31% dissolved solids. This occurs at 140C without incremen-
corn syrup adjunct worts (derived from an all-malt wort base) with tal feeding, with high yeast crop viability, and within normal
or without various nutritional supplements brewing time periods (i.e., less than 1 week). The generaliza-
FAN levels (mg/liter) with the following wort treatmcnts: tion, then, that strains of SacchIar,omyvces used in brewing
have only moderate alcohol tolerance (9) compared with
242.8(
ppm veast
extract-24 ppm
0.2( vcast
extract-24 ppm strains used in distilleries (16) does not hold true from a
mented 0.24%/c (vol/vol ergostcrol- cr-gosterol- production viewpoint. Ethanol concentrations of nearly
Trween 80 0.24% k,ol/vol) 0.24's (volv/ol) 16.0% (vol/vol) are considered high, even in the distilling
Tween 80 T\,wcn 80 industry (16).
0) 250 245 526 311 For brewing yeasts to produce these high levels of ethanol
1 181 144 464 20(1 without an excessively long or stuck fermentation, the wort
2 153 113 384 144 must contain two types of nutritional supplements: a nitro-
3 148 96 358 118 gen source (provided in this study by yeast extract) and a
S 146 96 367 113 combination of sterol (provided in this Study by ergosterol.
8 153 98 376 122
12 170 100 376 122 the most prevalent sterol in Saccharloiwn'ces spp. [34]) with
unsaturiated fatty acid (provided in this study by the oleic
acid fraction of Tween 80). It is important that both supple-
lipids only supplemented, 0.8% yeast extract plus lipids ments be provided simultaneously; the fermentation of high-
supplemented, and 0.2% yeast extract plus lipids supple- gravity worts then proceeds rapidly under both anaerobic (6)
mented worts, respectively. and semianaerobic conditions (Fig. 1).
When the lipids and yeast extract were simultaneously These supplements act by increasing the amount of new
replaced or reduced by oxygenation and increased malt yeast cell mass synthesis over the levels seen in unsupple-
levels, the fermentation of high-gravity maltose adjunct 29% mented fermentations (Table 3). During wort fermentations,
dissolved solids worts proceeded at the same rapid rate (Fig. the rate of sugar conversion into ethanol is considerably
5). higher in growing yeast cells (up to 2.0 g/h per g of yeasts)
than it is in stationary-phase yeast cells (as low as 0.06 g/h
DISCUSSION per g of yeasts) (21, 22). Clearly, it is not realistic to expect
the rapid and complete attenuation of high-gravity worts by
Regular production strains of brewers' yeasts can, with adding only brewing wort syrups to normal-gravity wort
appropriate nutritional supplementation. produce up to bases, ats is the current practice (15). By doing so, lipid and
16.2% (vol/vol) ethanol in the batch fermentation of worts of nitrogen deficiencies will limit the rate and final level of
ethanol production to levels much below the maximum
30
possible, as discussed below.
Lipids that are present in normal-gravity worts in growth-
limiting concentrations (7) must be synthesized by the yeasts
during the first few hours of fermentation while oxygen is
still present (22). However, the decreasing solubility of
25
oxygen in worts with increasing original gravity (3) results in
high-gravity worts which, if left unsupplemented, will cer-
tainly have oxygen as a growth-limiting nutrient. As demon-
strated in this report, in worts with up to 31% dissolved
20 solids, oxygen deficiencies can be overcome by either the
addition of 24 ppm ergosterol with 0.24%/, (vol/vol) Tween 80
as a sour-ce of oleic acid or by periods of oxygenation during
0
the fermentation (Fig. 4 and 5). On an industrial scale, the
latter would likely be a more economical option.
0 l
In low-gravity worts (10 to 12% dissolved solids), a
minimum level of 150 mg of FAN per liter is required to
permit rapid and complete attenuation (24). Virtually all of
15 this nitrogen is utilized within the first 24 h of fermentation
(31). at which point active yeast growth stops (19). In the
0
unsupplemented high-gravity worts used in this study, levels
of FAN ranged from 165 to 250 mg/liter. This was not nearly
enough to support the necessary degree of cell growth
required to rapidly and completely ferment worts as much as
three times more concentrated than normal-gravity worts.
When yeast extract was incorporated into fully supplement-
~ ed worts, considerably more FAN was utilized by the yeasts
than was even available in total in the unsupplemented worts
(days)
TIME (as much as 32(0 mg of FAN per liter). The high levels of
yeast extract required can be drastically reduced by the use
FIG. S. Dissolved solids versus time dui-ing the tnatei-ohic fer- of an all-malt base (Fig. 5).
mentation of maltose ad'juc ts. The worts were either- prepar-ed
for an (Ill-malt base anti supplemented wivth 0).2"1 yeast extraict -anti It is apparent. therefore, that a frequently mentioned
oxygenated during dayvs (). 1. and 2 ( 1()( ml of 0, gas per min) (40) oi- advantage of high-gravity brewing, namely. the increased
prepared fro.i the solids vesse and supplemented with f.er/- yield of ethanol resulting from the nonproportional increase
yeast extract with 24 ppm ergosterol-0.24'/r (vol/vol) Tween 8)) (U). in cell mass production with the increase in wort gravity (30),
VOL. 48, 1984 HIGH-GRAVITY BREWING 645

is in fact an important disadvantage in worts above the creased accumulation of intracellular ethanol at higher tem-
current commercial limit of 16 to 18% dissolved solids. peratures (26). Therefore, decreased fermentation times
Although the use of higher levels of non-nitrogenous and arising from higher temperatures of fermentation are at the
cheaper adjuncts is possible in worts up to this limit, above price of poor yeast crop viability, which is something that
these levels, nutrient deficiencies will prematurely terminate would eliminate their use in brewing. In any case, the
growth and therefore lead to significantly prolonged and in increased production of esters and higher alcohols at higher
some cases even stuck fermentations. In the past, such temperatures would preclude the use of temperatures above
problems have been attributed to the toxic influence of 14°C for the brewing industry (10). However, in the gasohol
ethanol (9, 37) when in fact growth-limiting levels of nitrogen or distilling industries, in which end yeast crop viability is
and oxygen have not permitted sufficient new yeast cell mass not as critical, the combination of nutritional supplementa-
synthesis. Indeed, it appears that Saccharomyces carlsber- tion and temperatures up to 30°C may be more practical.
gensis can be repitched repeatedly in 30% dissolved solids Yeast crop cellular composition (especially glycogen and
wort, producing up to 14.2% (vol/vol) ethanol each time, sterols) were found to differ considerably between anaerobic
without any increase in the time required for fermentation or and semianaerobic conditions and to depend on the nutri-
decrease in the extent of attenuation. In unsupplemented tional supplement used. The level of protein fell consistently
worts as low as 15.3% dissolved solids (with 40% adjunct), within the 40 to 50% range, but the amounts of cellular
deteriorations in the rate of attenuation, the end gravity glycogen and sterols fluctuated greatly. Glycogen levels
reached, and the final ethanol concentration obtained have were all less than 10% in yeasts under semianaerobic condi-
been observed, even though the first generation gave normal tions, regardless of whether they were unsupplemented or
and complete attenuation (18). The claim that ethanol con- supplemented, whereas up to nearly 30% concentrations
centrations as low as 8 to 9% (vol/vol) are sufficiently high to were seen in the anaerobic fermentations (Table 1). Howev-
kill off a large enough portion of the yeast crop to prevent its er, the level in the fully supplemented anaerobic yeasts,
use for repitching (35) is therefore not accurate. which gave the fastest fermentation time (6), was 10.5%; this
Supplementation was shown to have no influence on the was the only crop under anaerobic conditions to contain less
fermentative tolerance of the yeasts to ethanol (Table 2). than 26% glycogen.
This possibility was explored because sake yeasts only The level of glycogen in brewers' yeasts is an important
produce the 20 to 23% (vol/vol) levels of ethanol if the consideration, as it is the sole source of metabolic energy for
proteolipid component of Aspergillus oryzae is present (17). lipid synthesis and hexose transport in the first few hours of
In other strains of Saccharomyces, exposure to 0.5 to 1.5 M fermentation (32). Because of this, levels decline during the
ethanol causes an increase in the proportion of monounsa- first 24 h of fermentation but then rise and peak at values as
turated fatty acyl residues in cellular phospholipids (4). Also high as 40% at the end of the growth phase, before gradually
enhanced are tolerances to the effects of ethanol on genera- declining during the stationary phase (19, 32). Pitching rates
tion time, viability, and solute uptake when enrichment with resulting in the presence of 160 to 200 mg of glycogen per
a C18:2 unsaturated fatty acid rather than a C18:1 unsaturated liter are suggested to ensure enough glycogen for the synthe-
fatty acid was carried out (35). The early rise and then sis of adequate lipid levels (32). Although this might suggest
continued decline in fermentative power values has been a potential problem for the repitching of yeasts from the
attributed to an increase in the cellular levels of hexokinase supplemented fermentations, this was not found to be the
early in the fermentation, followed by gradual irreversible case, as added lipids can compensate for low glycogen levels
inhibition and denaturation of hexokinase activity by the (Fig. 3).
buildup of intracellular ethanol levels with time (27). Like glycogen, sterol levels varied between anaerobic and
As the ethanol concentration in this Warburg assay in- semianaerobic conditions. Levels in the semianaerobically
creased up to 20% (vol/vol), so did the degree of inhibition of fermented yeasts ranged from 0.85 to 1.09%, depending on
glycolytic activity. The relationship between the activity the type of supplementation (Table 1), compared with levels
remaining (relative to the 0% [vol/voli control) and the of only 0.27 to 0.56% in the anaerobically grown yeasts.
concentration of ethanol was found to be linear. However, Clearly, oxygen is being introduced during the course of the
when the IF50 values of these plots were calculated, little semianaerobic fermentations due to the higher levels of
difference between the two sets of yeasts was seen at any cellular sterols in these yeasts. Brewers' yeasts normally
point throughout the fermentation. The values fluctuated have ca. 0.1% total sterols (dry weight) at pitching but reach
within the range of 10.1 to 13.6% (vol/vol) ethanol, with the 1.0% or greater within the first several hours, utilizing wort
highest IF50 values in fact being observed in the yeasts from oxygen for de novo synthesis (22). The 1.0% level is a
the unsupplemented fermentation. Improvements in the sup- maximum ceiling which is subsequently diluted down to a
plemented fermentations on a per milligram of yeasts (dry 0.1% lower limit by passage of sterols to daughter cells (22).
weight) basis were not due to supplementation increasing the Based on the results described above, it is clear that
tolerance of brewers' yeast fermentative ability to ethanol. cellular sterol levels alone are not the growth-limiting factor
The enhanced death of yeasts in the high-gravity wort in yeasts from unsupplemented high-gravity wort fermenta-
fermentations with increasing temperature (Fig. 3) was not tions. In none of the yeast crops from these fermentations
an unexpected result. Ethanol and temperature effects on was the lower limit of 0.1% sterols observed. In fact, the
yeasts are known to be closely related as (i) ethanol becomes lowest level reached was 0.48%, but this level was reached
increasingly toxic to growth and viability at higher tempera- at day 6 offermentation and remained unchanged up until the
tures (25, 26), (ii) ethanol decreases the optimum tempera- end of fermentation (day 12). It was not, therefore, the level
ture of fermentation (14), (iii) ethanol decreases the optimal of cellular sterols which caused the cessation in new yeast
and maximum temperatures for growth (36), and (iv) higher cell mass production in the unsupplemented fermentation,
temperatures of fermentation result in decreased final etha- but other growth-limiting nutrients. In the yeasts from
nol concentrations in the mash (28). supplemented fermentations, the addition of the lipids in the
The explanation for the increased inhibitory effects of supplement elevated even further the levels of sterols away
ethanol at higher temperatures has been attributed to in- from the growth-limiting level of 0.1%.
646 CASEY, MAGNUS, AND INGLEDEW Appt-. ENVIRON. MICROBIOL.

It remains to be determined whether the 16 to 17% (vol/ 13. Giudici, P., and M. Guerzoni. 1982. Sterol content as a character
vol) levels of ethanol reported here represent the true upper for selecting yeast strains in enology. Vitis 21:5-14.
limit for brewers' yeasts. As sake yeast can only produce the 14. Gray, W. D. 1941. Studies on the alcohol tolerance of yeasts. J.
20 to 23% (vol/vol) levels of ethanol by the sequential Bacteriol. 42:561-574.
addition of substrate over a period of weeks (17), it may well 15. Hackstaff, B. W. 1978. Various aspects of high gravity brewing.
Master Brew. Assoc. Am. Tech. Quart. 15:1-7.
be possible with the sequential addition of adjunct to supple- 16. Harrison, J. J., and J. C. Graham. 1970. Yeasts in distillery
mented brewers' wort to produce beers of up to 20 to 23% practice, p. 283-348. In A. H. Rose and J. S. Harrison (ed.).
(vol/vol) ethanol. Should this be possible, it would suggest The yeasts, vol. 3. Academic Press. Inc., New York.
the need for a complete reevaluation of the definition of 17. Hayashida, S., D. Feng, and M. Hongo. 1974. Function of the
alcohol tolerance in the various species of Sacwharoinyc es. high concentration alcohol-producing factor. Agric. Biol. Chem.
In addition, it is shown that the self-imposed wort gravity 38:2001-2006.
limit of 16 to 18% for high-gravity brewing in industry should 18. Hsu, W. P., A. Vogt, and L. Bernstein. 1980. Yeast nutrients and
not be ascribed to sluggish and stuck fermentations, yeast beer quality. M.B.A.A. Tech. Quart. 17:85-88.
death, or yeast intolerance to alcohol. It would appear that 19. Ingledew, W. M. 1975. Utilization of wort carbohydrates and
nitrogen by Saccharoinv'ces carlsbergensis. Master Brew. As-
brewers and alcohol manufacturers could easily consider soc. Am. Tech. Quart. 12:146-150.
production of worts with higher gravities and enjoy larger 20. Jones, M., and J. Pierce. 1964. Some factors influencing the
economies of labor, capital, and energy. individual amino acid composition of wort. Proc. Am. Soc.
Brew. Chem. 22:130-136.
ACKNOWLEDGMENTS 21. Kirsop, B. H. 1971. Yeast metabolism and sugar utilization.
We thank Molson Breweries of Canada Ltd. and the Natural Brew. Guardian 100:56-58.
Sciences and Engineering Research Council of Canada for research 22. Kirsop, B. H. 1982. Developments in beer fermentation. Top.
funds that partially supported this project. G. P. Casey is the Enzyme Ferment. Biotechnol. 6:79-131.
recipient of an National Sciences and Engineering Research Council 23. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.
(of Canada) postgraduate scholarship. 1951. Protein measurement with the Folin phenol reagent. J.
We thank Molson (Saskatchewan) Brewery Ltd. and Labatts Biol. Chem. 193:265-275.
(Saskatchewan) Brewery Ltd. for providing yeast slurry and brew- 24. Meilgaard, M. C. 1976. Wort composition: with special refer-
ing supplies. ence to the use of adjuncts. Master Brew. Assoc. Am. Tech.
Quart. 13:78-90.
25. Nagodawithana, T. W., C. Castellano, and K. H. Steinkraus.
LITERATURE CITED 1974. Effect of dissolved oxygen. temperature. initial cell count.
1. Andreasen, A. A., and J. B. Stier. 1953. Anaerobic nutrition of and sugar concentration on the viability of Saccharoinv (ces
Saccharotnvces cerevisiae. 1. Ergosterol requirements for (cerevisiae in rapid fermentations. Appl. Microbiol. 28:383-391.
growth in a defined medium. J. Cell. Comp. Physiol. 41:23-36. 26. Navarro, J. M., and G. Durand. 1978. Alcoholic fermentation:
2. Andreasen, A. A., and J. B. Stier. 1954. Anaerobic nutrition of effect of temperature on ethanol accumulation within yeast
Saccharomtncces cerev'isiae. II. Unsaturated fatty acid require- cells. Ann. Microbiol. (Inst. Pasteur) 129B:215-224.
ment for growth in a defined medium. J. Cell. Comp. Physiol. 27. Navarro, J. M., and J. D. Finck. 1982. SaccharomYces iwaruml
43:271-281. hexokinase behavior during alcoholic fermentation. Cell. Mol.
3. Baker, C., and S. Morton. 1977. Oxygen levels in air-saturated Biol. 28:85-89.
worts. J. Inst. Brew. 83:348-349. 28. Ough, C. S. 1966. Fermentation rates of grape juice. 11. Effects
4. Beaven, M. J., C. Charpentier, and A. H. Rose. 1982. Production of initial 'Brix, pH and fermentation temperature. Am. J. Enol.
and tolerance of ethanol in relation to phospholipid fatty-acyl Vitic. 17:20-26.
composition in Saccharomyces ucreivisiae NCYC 431. J. Gen. 29. Owades, J. L. 1981. The role of osmotic pressure in high and low
Microbiol. 128:1447-1455. gravity fermentations. Master Brew. Assoc. Am. Tech. Quart.
5. Casey, G. P., and W. M. Ingledew. 1983. High gravity brewing: 18:163-165.
influence of pitching rate and wort gravity on early yeast 30. Palmer, A. K., and H. Rennie. 1974. Ester control in high gravity
viability. J. Am. Soc. Brew. Chem. 41:148-153. brewing. J. Inst. Brew. 80:447-454.
6. Casey, G. P., C. A. Magnus, and W. M. Ingledew. 1983. High 31. Patel, G. B., and W. M. Ingledew. 1973. Trends in wort
gravity brewing: nutrient enhanced production of high concen- carbohydrate utilization. Appl. Microbiol. 26:349-353.
trations of ethanol by brewing yeast. Biotech. Lett. 5:429-434. 32. Quain, D. E., and R. S. Tubb. 1982. The importance of glycogen
7. David, M. H., and B. H. Kirsop. 1972. The varied response of in brewing yeasts. Master Brew. Assoc. Am. Tech. Quart.
brewing yeasts to oxygen and sterol treatment. J. Am. Soc. 19:29-33.
Brew. Chem. 30:14-16. 33. Quain, D. E., and R. S. Tubb. 1983. A rapid and simple method
8. Davis, B. D. 1947. The estimation of small amounts of fatty acid for the determination of glycogen in yeast. J. Inst. Brew. 89:38-
in the presence of polyoxyethylene sorbitan partial fatty acid 40.
esters (''Tween") and serum proteins. Arch. Biochem. 15:351- 34. Rattray, J. B. M., A Schibeci, and D. K. Kidby. 1975. Lipids of
358. yeasts. Bacteriol. Rev. 39:197-231.
9. Day, A., E. Anderson, and P. A. Martin. 1975. Ethanol tolerance 35. Rose, A. H. 1980. Recent research on industrially important
of brewing yeasts. Proceeding of the 15th Congress of European strains of SacchiaromYvces cerevisiae. Soc. Appl. Bacteriol.
Brewing Convention, p. 377-391. 9:103-121.
10. Engan, S., and 0. Aubert. 1977. Relations between fermentation 36. Van Uden, N., and H. Duarte. 1981. Effects of ethanol on the
temperature and the formation of some flavour components. temperature profile of Saccharomvces cerevisiae. Z. Allg. Mik-
Proceedings of the 16th Congress of European Brewing Conven- robiol. 21:743-750.
tion, p. 591-609. Schweizer Baruerei-Rundschau, Zurich. 37. White, F. H. 1978. Ethanol tolerance of brewing yeast. Proceed-
11. European Brewing Convention. 1975. Free alpha amino nitrogen ings of the 15th Convention of the Society of Institute Brewing
in worts and beer. 3rd ed., p. 61-62. (Australia and New Zealand), p. 133-146.
12. European Brewing Convention Analytica Microbiologica. 1981. 38. Whitear, A. L., and D. Crabb. 1977. High gravity brewing-
Part 11. J. Inst. Brew. 87:303-321. concepts and economics. The Brewer 63:60-63.