3.2. Comparison of Microbial Growth on Lupin Protein Isolate Solutions
The growing parameters consisting of CFU, pH and glucose for all eight microorganisms investigated are shown in
Table 3. In addition, growth curves of four of these microorganisms were selected in
Figure 1 in order to highlight the temporally different transitions into the exponential phase.
The results showed that all microorganisms were able to grow in LPI solution. The minimum increase in CFU/mL (ΔE
CFU) was recorded for
L. reuteri with 1.36×10
7 CFU/mL and the maximum for
S. xylosus with 6.01×10
8 CFU/mL. The results of the pH curve showed that
L. amylolyticus and
L. helveticus appear to have the best metabolism and were adapted most rapidly to the lupine solution. The pH curves showed a direct and constant pH acidification over 24 h for
L. helveticus (exemplarily shown in
Figure 1a) and
L. amylolyticus (similar to
Figure 1a). After 24 h of fermentation, a change into the stationary phase could not be observed for both microorganisms.
S. xylosus (exemplarily shown in
Figure 1b) and
L. delbrueckii (similar to
Figure 1b) showed a lag phase and a transition into the log phase after approximately 8 h.
S. xylosus reached the stationary phase after 14 h.
L. delbrueckii did not show a clear transition into the stationary phase after 24 h.
L. sakei subsp.
carnosus (exemplarily shown in
Figure 1c) and
L. reuteri (similar to
Figure 1c) changed from the lag phase to the log phase after 10 h and reached the stationary phase after 18 h. The largest lag phases were recorded for
L. parabuchneri (exemplarily shown in
Figure 1d) and
L. brevis (similar to
Figure 1d). Both microorganisms reached the log phase after 14 h and changed to the stationary phase after 20 h. The results of glucose concentrations showed that the added carbon source of 5 g/kg glucose was metabolized by all microorganisms. After 24 h fermentation, residues of glucose were present in all fermented samples. The degradation of glucose (ΔE
Glucose) ranged from 3.4 g/kg with remaining 1.7 g/kg glucose after 24 h for
L. amylolyticus to 4.8 g/kg with remaining 0.2 g/kg glucose after 24 h for
L. reuteri, respectively. Fritsch et al. [
21] and Lampart-Szczapa, Konieczny, Nogala-Kałucka, Walczak, Kossowska and Malinowska [
9] showed similar results and confirmed the suitability of lupine flour and lupine protein, respectively, for lactic fermentation.
3.3. Sensory Anaylsis
In the retronasal sensory evaluation, the following six aroma qualities and corresponding references were selected by the panelists for the description of LPI and fermented LPI: cheesy (butanoic acid); popcorn-like, roasty (2-acetylpyrazine); earthy, moldy, beetroot-like (geosmin); pea-like, green bell pepper-like (2-isopropyl-3-methoxypyrazine); cooked potato-like (3-(methylthio-)propanal); and oatmeal-like, fatty (oatmeal).
Comparative aroma profile analyses (
Figure 2) of LPI,
L. brevis,
L. amylolyticus and
S. xylosus were emphasized to highlight the increase of aroma perception in
cheesy and
roasty,
popcorn-like as well as the reduction of the mean aroma perceptions in comparison to LPI, across the panel. The primary aroma attributes in the LPI were
pea-like,
green bell pepper-like, and
oatmeal-like,
fatty with median values of 4.5 and 3.0, respectively. Otherwise, the aroma impression was evaluated with low intensities for
earthy,
moldy,
beetroot-like;
cooked potato-like, and
popcorn-like,
roasty with median values of 2.5, 2.5, and 1.0, respectively, while the attribute
cheesy was imperceptible (median value of 0). The dominant aroma attributes in the LPI samples obtained after fermentation with
L. reuteri were described as
cheesy (median of 4.0) and
oatmeal-like,
fatty (median value of 3.0). The aroma perception was otherwise evaluated with low intensities for
earthy,
moldy,
beetroot-like;
pea-like,
green bell pepper-like, and
cooked potato-like with values of 2.0 and
popcorn-like,
roasty with a value of 1.0. The aroma perception of the samples obtained after
L. brevis fermentation was evaluated with a maximum intensity of 3.0 for
oatmeal-like,
fatty. Followed by low intensities of
earthy,
moldy,
beetroot-like;
cooked potato-like, and
popcorn-like,
roasty with values of 2.0 and
cheesy and
pea-like,
green bell pepper-like with values of 1.0. Fermentation of LPI with
L. amylolyticus was described with a dominant aroma impression of
popcorn-like,
roasty with a value of 5.0, followed by
oatmeal-like,
fatty with a value of 3.5. Low intensities were judged for
pea-like,
green bell pepper-like with a value of 2.0,
cooked potato-like with a value of 1.5, and
earthy,
moldy,
beetroot-like with 1.0. Attribute
cheesy was not perceptible (median value of 0). The aroma profile of the
L. parabuchneri fermented samples showed the aroma impressions of
pea-like,
green bell pepper-like;
cheesy; and
oatmeal-like,
fatty with intensities of 4.0, 3.0, and 3.0 respectively. Less dominant were the attributes
popcorn-like,
roasty and
cooked potato-like, both with median values of 2.0, and
earthy,
moldy,
beetroot-like with an intensity median value of 1.5.
L. sakei subsp.
carnosus fermentation was described with
popcorn-like,
roasty as main aroma impression with an intensity of 4.0, followed by
oatmeal-like,
fatty with an intensity of 3.0. The attributes
earthy,
moldy,
beetroot-like;
pea-like,
green bell pepper-like, and
cooked potato-like were described equally less intensely with values of 2.0. The attribute
cheesy was imperceptible in the samples obtained after fermentation with
L. sakei subsp.
carnosus. Samples fermented with
S. xylosus exhibited a
cheesy intensity of 5.0, followed by
oatmeal-like,
fatty with a value of 3.0. The other attributes were less intense with
pea-like,
green bell pepper-like (2.5),
earthy,
moldy,
beetroot-like (2.0),
popcorn-like,
roasty (2.0), and
cooked potato-like (1.0). Similar to
L. amylolyticus and
L. sakei subsp.
Carnosus fermentation, the attribute
popcorn-like,
roasty was described as a dominant aroma impression for the
L. helveticus fermented samples with an intensity of 4.0. The aroma profile of
L. helveticus was otherwise described with aroma impressions
pea-like,
green bell pepper-like with a value of 3.0. The attributes
oatmeal-like,
fatty;
cooked potato-like;
earthy,
moldy,
beetroot-like, and
cheesy were rated less intensively with values of 2.5, 2.0, 1.0, and 0.5 respectively. The main aroma impression of
L. delbrueckii was assessed as
oatmeal-like,
fatty with an intensity of 4.0. The aroma impressions
popcorn-like,
roasty;
earthy,
moldy,
beetroot-like;
pea-like,
green bell pepper-like; and
cooked potato-like were rated with an intensity of 2.0. The attribute
cheesy could not be observed by the panelists.
The aroma profile showed, with the exception of the
L. parabuchneri fermented samples, that the aroma perception
pea-like,
green bell pepper-like decreased in intensity due to the fermentation. The maximum decreases were determined for the samples obtained after
L. brevis fermentation with an intensity of 1.0 compared to unfermented LPI with a value of 4.5. Furthermore, fermentation increased the intensity of the aroma impressions
popcorn-like,
roasty and
cheesy. In the samples fermented with
L. amylolyticus,
L. sakei subsp.
carnosus and
L. helveticus the aroma impressions
popcorn-like,
roasty increased from 1.0 for unfermented LPI to 5.0, 4.0 and 4.0 respectively.
S. xylosus and
L. reuteri showed an increase in intensity of the attribute
cheesy from 0 for unfermented LPI to 5.5 and 4.0 respectively. Several authors confirm for lupin, soy and pea protein, respectively, the significant modification of aroma profile by fermentation [
6,
7,
8,
10]. The authors described the reduction of
n-hexanal content, which contributes most to the
green and
beany off-flavor of pea, by fermentation with lactic acid extract in pea protein extract and soy, respectively [
6,
7,
8]. Further studies showed that the fermentation of soy protein isolate with lactic acid bacteria significantly reduced the aroma impression of
beany [
10]. A statement about the reduction of
n-hexanal content due fermentation cannot be obtained in this study. A reduced perception of the
pea-like,
green bell pepper-like aroma may also be caused by masking effects. The total aroma intensities of all samples differed only slightly with median values of 5.5 for the samples fermented with
L. amylolyticus, 5.0 for LPI,
L. parabuchneri,
S. xylosus,
L. reuteri, and
L. sakei subsp.
carnosus fermented samples, 4.5 for
L. helveticus and
L. delbrueckii, and 4.0 for
L. brevis fermented samples.
The taste impressions
bitter and
salty were analyzed by the panel in the sensory evaluation and displayed in
Figure 3. The
bitter intensity of unfermented LPI was described with a mean value of 2.3. All fermented samples did not differ significantly (
p < 0.05) in bitterness compared to LPI, with the exception of
L. sakei subsp.
carnosus (1.0). However, the trend of the mean values of the fermented samples showed slightly lower intensities of bitterness compared to LPI. The intensity of
salty was described for LPI with a mean value of 1.9. In comparison, no significant differences (
p < 0.05) in the intensity of the fermented samples were found. However, the samples obtained after fermentation with
L. amylolyticus,
S. xylosus, and
L helveticus tended to be more salty.
The hedonic evaluation was performed for a first indication of the prevalence of the samples and was not performed according to ISO standards. The evaluation (
Figure 4) of the panel resulted in a rating of 4.2 for unfermented LPI. The sample, which was most popular with the panelists (6.4), was fermented with
L. sakei subsp.
carnosus, followed by the samples fermented with
L. helveticus (5.5) and
L. amylolyticus (5.4). The most unpopular samples were the ones fermented with
S. xylosus and
L. reuteri with values of 2.8 and 3.0, respectively. The results showed that both saltiness and bitterness do not have a considerable effect on the acceptance of the samples. The acceptance seems to be influenced by the differences in the aroma profile. It was found that the samples with the aroma attributes popcorn like were rated as popular by all subjects. The samples with the maximum evaluation had the attribute
popcorn-like,
roasty in their aroma profile as dominant aroma impression and also the highest intensity in this attribute compared to LPI and all other fermented samples. Meanwhile, the aroma impression
cheesy dominated in
S. xylosus and
L. reuteri fermented samples—the samples with the minimal hedonic rating. In addition, both of these samples had the highest intensity of this attribute compared to LPI and the other samples.