Hydrolysis of konjac glucomannan by Trichoderma reesei mannanase and endoglucanases Cel7B and Cel5A for the production of glucomannooligosaccharides

Article abstract of DOI:10.1016/j.carres.2013.02.012

In this paper we describe the enzymatic hydrolysis of konjac glucomannan for the production of glucomannooligosaccharides using purified Trichoderma reesei mannanase, endoglucanases EGI (Tr Cel7b) and EGII (Tr Cel5a). Hydrolysis with each of the three enzymes produced a different pattern of oligosaccharides. Mannanase was the most selective of the three enzymes in the hydrolysis of konjac mannan and over 99% of the formed oligosaccharides had mannose as their reducing end pyranosyl unit. Tr Cel5A hydrolysate shared similarities with mannanase and Tr Cel7B hydrolysates and the enzyme had the lowest substrate specificity of the studied enzymes. The hydrolysate of Tr Cel7B contained a series of oligosaccharides with non-reducing end mannose (M) and reducing end glucose (G) (MG, MMG, MMMG, and MMMMG). These oligosaccharides were isolated from the hydrolysate by size exclusion chromatography in relatively high purity (86-95%) and total yield (23% of substrate). The isolated oligosaccharides were characterized using acid hydrolysis and HPAEC-PAD (carbohydrate composition), HPLC-RI and HPAEC-MS (to determine the DP of purified oligosaccharides), enzymatic hydrolysis (determination of non-reducing end carbohydrate) and NMR (both 1D and 2D, to verify structure and purity of purified compounds). Hydrolysis of konjac mannan with a specific enzyme, such as T. reesei Cel7B or mannanase, followed by fractionation with SEC offers the possibility to produce glucomannooligosaccharides with defined structure. The isolated oligosaccharides can be utilised as analytical standards, for determination of bioactivity of oligosaccharides with defined structure or as substrates for defining substrate specificity of novel carbohydrate hydrolyzing enzymes.

Full text of DOI:10.1016/j.carres.2013.02.012

This document is downloaded from the
Digital Open Access Repository of VTT
Title
Hydrolysis of konjac glucomannan by Trichoderma reesei
mannanase and endoglucanases Cel7B and Cel5A
for the production of glucomannooligosaccharides
Mikkelson, Atte; Maaheimo, Hannu; Hakala, Terhi
Carbohydrate Research. Elsevier. Vol. 372 (2013),
pages 60 - 68
Author(s)
Citation
Date
URL
Rights
2013
http://dx.doi.org/10.1016/j.carres.2013.02.012
This article (post-print version) may be downloaded
for personal use only.
By using VTT Digital Open Access Repository you are
bound by the following Terms & Conditions.
VTT
http://www.vtt.fi
P.O. box 1000
FI-02044 VTT
Finland
I have read and I understand the following statement:
This document is protected by copyright and other
intellectual property rights, and duplication or sale of all or
part of any of this document is not permitted, except
duplication for research use or educational purposes in
electronic or print form. You must obtain permission for
any other use. Electronic or print copies may not be
offered for sale.

Hydrolysis of konjac glucomannan by Trichoderma reesei mannanase and endoglucanases
Cel7B and Cel5A for the production of glucomannooligosaccharides
Atte Mikkelson, Hannu Maaheimo, Terhi K. Hakala1*
VTT Technical research centre of Finland, P.O Box 1000, 02044 VTT, Finland
In this paper we describe the enzymatic hydrolysis of konjac glucomannan for the production
of glucomannooligosaccharides using purified Trichoderma reesei mannanase, endoglucanases
EGI (Tr Cel7b) and EGII (Tr Cel5a). Hydrolysis with each of the three enzymes produced a
different pattern of oligosaccharides. Mannanase was the most selective of the three enzymes in
the hydrolysis of konjac mannan and over 99% of the formed oligosaccharides had mannose as
their reducing end pyranosyl unit. Tr Cel5A hydrolysate shared similarities with mannanase and
Tr Cel7B hydrolysates and the enzyme had the lowest substrate specificity of the studied
enzymes. The hydrolysate of Tr Cel7B contained a series of oligosaccharides with non-reducing
end mannose (M) and reducing end glucose (G) (MG, MMG, MMMG, and MMMMG). These
oligosaccharides were isolated from the hydrolysate by size exclusion chromatography in
relatively high purity (86-95%) and total yield (23% of substrate). The isolated oligosaccharides
were characterized using acid hydrolysis and HPAEC-PAD (carbohydrate composition), HPLC-
RI and HPAEC-MS (to determine the DP of purified oligosaccharides), enzymatic hydrolysis
(determination of non-reducing end carbohydrate) and NMR (both 1D and 2D, to verify structure
and purity of purified compounds). Hydrolysis of konjac mannan with a specific enzyme, such as
T. reesei Cel7B or mannanase, followed by fractionation with SEC offers the possibility to
produce glucomannooligosaccharides with defined structure. The isolated oligosaccharides can
be utilised as analytical standards, for determination of bioactivity of oligosaccharides with
defined structure or as substrates for defining substrate specificity of novel carbohydrate
hydrolyzing enzymes.
Key words: oligosaccharide, Konjac mannan, glucomannan, enzymatic hydrolysis
* Corresponding author: Terhi.Hakala@vtt.fi, tel. +358 20 722 7417, Fax: +358 722 7071

Abbreviations: GM, β-D-Glcp-(14)-D-Manp; MM, β-D-Manp-(14)-D-Manp; MG, β-D-
Manp-(14)-D-Glcp; MMM, β-D-Manp-(14)- β-D-Manp-(14)-D-Manp; MMG, β-D-
Manp-(14)- β-D-Manp-(14)-D-Glcp; GGM, β-D-Glcp-(14)- β-DGlcp-(14)-D-Manp;
GMM, β-D-Glcp-(14)- β-D-Manp-(14)-D-Manp; MGG, β-D-Manp-(14)- β-D-Glcp-
(14)-D-Glcp; MMMG, β-D-Manp-[(14)- β-D-Manp]2-(14)-D-Glcp; MGGM, β-D-Manp-
(14)-[ β-D-Glcp-(14)]2-D-Manp; GGGM, β-DGlcp-(14)-[ β-D-Glcp-(14)]2-D-Manp;
MMMMG, β-D-Manp-[(14) - β-DManp]3-(14)-D-Glcp.

1. Introduction
Mannans are hemicelluloses that have a role as structural polymers in plant cell walls and as
storage carbohydrate in plant seeds. Mannans can be classified in four subfamilies; linear
mannan, glucomannan, galactomannan and galactoglucomannan. The main structure of linear
mannan consists of mannopyranosyl units bound together with β-D-1,4 linkages.1 The structure
of glucomannan consists of D-mannopyranosyl and D-glucopyranosyl units linked together by β-
D-1,4 –linkages.2 Galactomannans, mainly found in seeds, are substituted in different degrees
with 1,6-linked α-galactopyranosyl units.1 The main hemicellulose present in softwood cell
walls, galactoglucomannan, has a backbone of mannopyranosyl and glucopyranosyl units and is
substituted with α-galactopyranosyl units in varying ratios.3
Glucomannan structure consists of four different β-D-1,4 –linkages: mannose to mannose,
mannose to glucose, glucose to mannose and glucose to glucose. These linkages can be
hydrolysed by endo-1,4- β-mannanases (mannanase, EC 7.2.1.78) and endo- β-glucanases (EG,
endoglucanase, EC 3.2.1.4). Mannanases catalyse the random hydrolysis of β-D-1,4-
mannopyranosyl linkages in the main chain of mannans, glucomannans, and galactomannans.4,5
Endoglucanases belong to the cellulose degrading enzymes and hydrolyse the β-D-1,4 –linkages
in cellulose chain. In addition to cellulose, at least the glucose to mannose and glucose to glucose
linkages in glucomannan are hydrolysable to endoglucanases. Substrate specificity of enzymes
vary depending on their structural features and endoglucanases from certain glycoside hydrolase
(GH) families are able to hydrolyse hemicelluloses such as mannans and xylans as well.6
Enzymes hydrolysing the non-reducing end mannopyranosyl and glucopyranosyl units from
oligosaccharides released from glucomannan are β-mannosidase (EC 3.2.1.25) and β-glucosidase
(EC 3.2.1.21), respectively. The α-galactopyranosyl units can be hydrolysed by α-galactosidase
(EC 3.2.1.22).4
Linear glucomannans from plants are readily available in commercial sources, for example
konjac glucomannan from the tuber of plant Amorphophallus konjac has been studied in detail
and is used as gelling and thickening agent in food and beverages.5 Additionally,
oligosaccharides prepared by enzymatic hydrolysis of konjac glucomannan can be used as
prebiotics as the hydrolysate efficiently enhances the growth of lactobacilli and bifidobacteria.7
The specific nature of enzyme hydrolysis makes enzymes excellent tools for targeted and
controlled hydrolysis of polysaccharides. The advantage of enzymatic hydrolysis over acid
hydrolysis is specificity and thus, although the process is slower, oligosaccharides with desired
degree of polymerization (DP) are obtained without the formation of monosaccharides or
furfural.8 The choice of enzyme will have an effect on the oligosaccharide profile to be obtained
in the hydrolysis of glucomannan. For example hydrolysis of konjac mannan by P. funiculosum
endoglucanase leads to formation of MG, MM, MMG and MMM whereas GM, MM, GGM and
GMM have been detected after hydrolysis by A. niger and Streptomyces sp. mannanase.9 The
substrate specificity can also be exploited in analytical purposes. For example, glucomannan

content of e.g. konjac flour can be determined by enzymatic hydrolysis with mannanase followed
by the alkali treatment to remove acetyl groups and further hydrolysis by β-mannosidase and β-
glucosidase.10 Based on the quantities of different oligosaccharides formed by the enzymatic
hydrolysis it has been concluded that the glucose and mannose units are distributed randomly
along the polysaccharide chain and konjac mannan thus represents a Bernoulli-type chain,9 thus
even features of the polymer structure can be revealed by enzyme hydrolysis.
Trichoderma reesei (syn. Hypocrea jecorina) is an efficient plant biomass degrading fungus,
which is currently one of the main sources of industrial glycoside hydrolases and a commonly
utilized production host of industrial enzymes. The genome of T. reesei was recently sequenced
and the genes encoding carbohydrate-active enzymes were thoroughly investigated. The genome
of T.reesei contains 200 glycoside hydrolase encoding genes, seven and 6 of which encoding
known cellulases and hemicellulases, respectively.11 The aim of the present work was to produce
glucomannooligosaccharides from konjac mannan by enzymatic hydrolysis using selected
purified T. reesei glycoside hydrolases. Although the cellulolytic enzymes of T. reesei have been
widely studied, details on the substrate specificity of the enzymes from this industrially relevant
organism are still needed. In this paper we present detailed analysis of the hydrolysis products of
konjac glucomannan obtained by hydrolysis with two T. reesei endoglucanases belonging to
glycosyl hydrolase (GH) families 5 and 7 (TrCel5A and TrCel7B, formerly known as EGII and
EGI, respectively) and one mannanase belonging to GH family 5.12,13
2. Results
2.1 Enzymatic hydrolysis of konjac mannan
All the three studied enzymes, Tr Cel5A, Tr Cel7B and mannanase from Trichoderma reesei,
hydrolysed konjac mannan efficiently and the amount of released reducing sugars was
approximately 50% of substrate after the 48 hour hydrolysis (Figure 1), which corresponds on
average to total hydrolysis of the substrate into disaccharides. Hydrolysis with mannanase
released reducing sugar ends the most of the three studied enzymes and the hydrolysis was the
fastest in the beginning of the incubation. One of the two studied endoglucanases, Tr Cel7B,
hydrolysed konjac mannan faster, although the overall hydrolysis by Tr Cel5A was higher. At
the end of the 48 h incubation, the hydrolysis by Tr Cel7B and mannanase had plateaued
suggesting that hydrolysis was close to maximal level. The hydrolysis by Tr Cel5A was still
slowly progressing after 48 h, although the hydrolysis was relatively slow.
The mono- and oligosaccarides formed by the hydrolysis by Trichoderma reesei enzymes
were analysed by HPAEC-PAD/MSQ (Figure 2). The yield of each hydrolysis product was
calculated as % of substrate after anhydrocorrection. Monosaccharides were released from the
substrate by all the three enzymes as both glucose (8.2% of substrate by Tr Cel7B and 2.6% by

Tr Cel5A) and mannose (7.7% of substrate by mannanase and 1.5% by Tr Cel5A). Thus, the
lowest amount of monosaccharides was released by Tr Cel5A, but it also was the only enzyme of
the three to release significant amount (over 1% of substrate) of both mannose and glucose from
the substrate. Linear manno-oligosaccharides (DP 2-6) were detected in all hydrolysates in total
of 17%, 18% and 4.7% for mannanase, Tr Cel5A and Tr Cel7B, respectively.
In addition to mannose and mannobiose, mannanase hydrolysate contained minor amount of
mannotriose and oligosaccharides with glucose in their non-reducing end and mannose in the
reducing end , namely GM, GGM and GMM (Figure 2B, retention time 52.83, 57.47 and 73.76
min). These oligosaccharides were identified and quantified using in-house standards and
represented 62% of the substrate. The main hydrolysis products of T. reesei mannanase
consisting of only mannose were mannose (7.7% of substrate) and mannobiose (16% of
substrate). Two peaks with DP4 in the mannanase hydrolysate eluting at 78.96 min and 92.48
min (Figure 2B) could not be identified with the analytical standards in use, their yields were
estimated to present 7.6 and 4.5% of substrate, respectively. The concentrations of the
unidentified oligosaccharides in the hydrolysate were estimated based on the standard eluting
closest to them. Altogether 89% of the konjac mannan substrate was hydrolysed into identified
products by Trichoderma reesei mannanase, summing up to 101% yield with the unidentified
peaks, although some minor peaks with DP5-6 in the chromatogram were not included in the
calculations.
The main oligosaccharide products (unidentified peaks 1-4) in the Tr Cel7B hydrolysate
could not be identified with the utilised commercial or in-house standards (Figure 2A, retention
time 16.42, 20.18, 26.62 and 34.93 min). The concentrations of the unidentified oligosaccharides
in the hydrolysate were estimated based on the standard eluting closest to them. The four
unidentified peaks were estimated to represent 48% of the substrate in the hydrolysate of Tr
Cel7B and 32% in the hydrolysate of Tr Cel5A. In addition to the unidentified peaks Tr Cel5A
hydrolysate contained linear manno-oligosaccharides with DP 2-4 (Figure 2C, retention time
24.42, 30.68 and 40.48 min) and oligosaccharides with glucose in their non-reducing end and
mannose in the reducing end (GM and GMM, Figure 2C, retention time 53.75 and 57.71 min).
Thus, the hydrolysate formed by T. reesei Cel5A shared similarities with the Tr Cel7B and
mannanase hydrolysates indicating that the enzyme has lower substrate specificity than the two
other enzymes studied.
2.2 Isolation and identification of glucomannooligosaccharides from Trichoderma reesei
Cel7B and mannanase hydrolysates
In order to isolate and identify the unidentified oligosaccharides present in the Tr Cel7B and
mannanase hydrolysate the freeze-dried hydrolysates were fractionated according to size using
size exclusion chromatography. The oligosaccharide composition of the collected fractions was

analysed by HPAEC-PAD and fractions containing the unidentified oligosaccharides were
combined and freeze-dried. The HPAEC-PAD chromatograms revealed presence of one main
oligosaccharide compound in each combined fractions of Tr Cel7B hydrolysate (Figure 3, each
panel, grey line). The yield of each unidentified oligosaccharide is depicted in Table 1. The
overall yield of isolated oligosaccharides was 23% of the konjac mannan used in the hydrolysis.
The DP of the oligosaccharides in the combined fractions was determined by HPLC-RI using
linear manno-oligosaccharide and cello-oligosaccharide standard mixtures of DP 1-6. In the
chromatograms of each isolated unidentified oligosaccharide only one main peak was observed
although some impurities with were detected (Table 1). The purity of the isolated
oligosaccharides was estimated based on HPLC-RI by comparing the area of the main DP peak
to the total peak area. HPAEC-MS was used to confirm the molecular mass of unidentified
oligosaccharides.
The carbohydrate composition of the unidentified oligosaccharides 1-4 was determined by
acid hydrolysis (Table 1). In order to determine the non-reducing end carbohydrate of the
unidentified oligosaccharides, they were hydrolysed with β-mannosidase and β-glucosidase.
None of the unidentified oligosaccharides were hydrolysable with β-glucosidase (data not
shown), whereas β-mannosidase hydrolysed the four oligosaccharides almost completely into
mannose and glucose (Figure 3, each panel, black line). The carbohydrate composition of the
combined fractions based on enzymatic hydrolysis by β-mannosidase slightly differed from the
composition determined by acid hydrolysis (Table 1). Based on the DP, MSQ detection,
carbohydrate composition and secondary enzymatic hydrolysis by β-mannosidase and β-
glucosidase, the unidentified oligosaccharides were tentatively identified as MG (unidentified 1),
MMG (unidentified 2), MMMG (unidentified 3), and MMMMG (unidentified 4).
The two unidentified peaks in mannanase hydrolysate were eluted into the same fractions by
size exclusion chromatography and could not therefore be separated from each other. Secondary
hydrolysis with β-mannosidase , β-glucosidase, and T.reesei Cel7B was carried out to find out
their structure. Unidentified 6 was hydrolysed into glucose and mannose at ratio 2.7:1 by β-
glucosidase whereas unidentified 5 was not affected suggesting that it has mannose as the non-
reducing sugar unit. Accordingly, unidentified 5 was hydrolysed by β-mannosidase the main
hydrolysis products being GGM (at m/z 511) and mannose (at m/z 187). Interestingly the peak
identified as GMM (at m/z 511) was not completely hydrolysed by β-glucosidase and was not at
all affected by β-mannosidase (Table 2) suggesting that the secondary hydrolysis with β-
glucosidase was incomplete. Both of the unidentified oligosaccharides were hydrolysable by T.
reesei Cel7B, which released MG, GM, MGG, cellobiose, Man and Glc (data not shown). Based
on the response to secondary enzymatic hydrolysis with β-mannosidase , β-glucosidase, and
T.reesei Cel7B the unidentified peaks were tentatively identified as MGGM (unidentified 5) and
GGGM (unidentified 6) (Table 2). HPAEC-MS was used to confirm that the peak mass (m/z)
values were 673 as Li-adducts [M + Li]+ at [M + 7]+ for both unidentified peaks (5 and 6) i.e.
for oligohexoses it means that they are tetrasaccharides (DP4). As the two oligosaccharides were

of same size, they could not be separated by size exclusion from each other and their identity was
not further confirmed by NMR.
2.3 Identification of glucomannooligosaccharides by NMR spectroscopy
13
Finally the identification of the unknown oligosaccharides 1-4 was confirmed by 1H and C
1
NMR spectroscopy. The structural reporter group regions of the H NMR spectra of the four
samples in D2O are shown in figure 4. All samples are P2 purified samples and named after main
identified compound. In the spectrum of unidentified 1 (Figure 4, A) two doublets were observed
at 5.229 and 4.673 ppm. The coupling constants (3.8 Hz and 8.0 Hz, respectively) confirm that
these are the α and β H1 signals of the reducing end glucose.9 In addition, a signal was observed
at 4.763 ppm, a typical position for a β1-4 linked mannose,9 confirming the structure as Manβ1-
4Glc. The 1D 1H spectrum also reveals typical H2 signals of Man and β-Glc at 4.074 and 3.856
ppm, respectively.15 Although these structural reporter group signals already confirm the
1
structure, full assignment of the H and 13C NMR signals was carried out from DQFCOSY,
TOCSY and HSQC spectra, and are presented in tables 3 and 4, respectively. The 13C NMR data
reveals clear glycosylation shifts of the C4 signals of the both anomers of the glucose, further
confirming the position of linkage between Man and Glc. The 13C chemical shifts were similar to
the ones published earlier for the same disaccharide structure.16
Structural reporter group signals similar to the ones observed above for MG were found also
in the 1H spectrum of unidentified 2 (Figure 4, B), but now there were two signals at the H1 and
H2 regions of mannose, confirming the structure as MMG. The same trend continued in the
spectra of unidentified 3 (Figure 4, C) showing a new signal at 4.774 ppm, between the typical
anomeric signals of terminal and glucose-bound mannoses already present in the spectrum of
MMG.9,15 This confirms the structure as MMMG. In addition, a small signal (4.760 ppm) and a
group of doublets (about 4.51 ppm, probably mid-chain Glc)9 with small intensity were observed,
arising from the small fraction (about 10%) of other oligosaccharide structures present in the
1
sample as was already revealed from the chromatography results. The H spectrum of the last
sample, unidentified 4 (Figure 4, D), showed H1 and H2 signals of a fourth mannose, resonating
at similar chemical shifts as the middle-chain mannose signals in MMMG (4.77 ppm).15 This
confirms the structure as MMMMG. The weak anomeric β -glucose signals9 from the small
fraction of other oligosaccharide structures were observed also in this sample.
Full assignments of the 1H and 13C NMR signals (Tables 3 and 4, respectively) were obtained
from the 2D experiments and they confirmed the assignments of the structural reporter groups
1
presented above. Two clearly different H spin systems were observed for the mannoses in the
TOCSY spectra (spectra not shown, see Table 3), corresponding to the terminal and mid-chain
1
mannoses. Although the H chemical shifts of the anomeric protons of glucose and mannose

bound mid-chain mannoses were slightly different, no differences could be observed for the
other proton or carbon signals. However, in the HSQC spectra of MMMG and especially
1
MMMMG, the H4-C4 peak of the mid-chain mannoses was exceptionally broad in the H
dimension and, in the case of MMMMG, also unsymmetrical. This suggests that there is a small
1
difference in the H chemical shift of these overlapping signals, but the resolution of the spectra
13
was not high enough to split the peak. In all cases, the C chemical shifts of C4s of the mid-
chain mannoses and the reducing end glucose showed a clear glycosylation shift of about 10 ppm,
as illustrated for MMG in figure 5. This confirms that C4 is the glycosylation site in each linkage.
The 13C chemical shifts of MMG were similar to the ones published for the same structure.16 The
1
2D experiments also confirmed that the H signals at about 4.51 ppm arise from -Glc moieties
having a 1H spin system very similar to the glucoses in the main oligosaccharides. This suggests
that these are oligosaccharides with a non-reducing glucose substituted at position 4.9
2.4 Analysis of hydrolysis specificity of Trichoderma reesei Cel7B, Cel5A and
mannanase
In order to make conclusions of the enzyme specificity, the observed oligosaccharides were
analysed based on their reducing and non-reducing end pyranosyl units (Table 5). The most
selective hydrolysis pattern was observed with T .reesei mannanase, which formed over 99% of
oligosaccharides with reducing end mannose. As expected, the hydrolysis products of the two
endoglucanases differed from those of the mannanase. T. reesei Cel7B clearly preferred
glucopyranosyl units in its -1 subsite of the active site, as 84% of the identified oligosaccharides
possessed glucose in their reducing end (Table 5). As much as 91% of the formed
oligosaccharides had mannose in their non-reducing end, which is explained by the preference of
glucopyranosyl in the -1 subsite and the ability of the enzyme to degrade the glucose to mannose
bond. Unexpectedly some linear manno-oligosaccharides (~5% of substrate) were present in the
T.reesei Cel7B hydrolysate.
The observed pattern of non-reducing and reducing end pyranosyl units in the Tr Cel5A
hydrolysate shows lower specificity of the enzyme in comparison to mannanase and Tr Cel7B.
As the used Tr Cel5A preparate is obtained from a culture filtrate of a wild type Trichoderma
reesei and not from a Man- strain, it needs to be noted that the produced linear
mannooligosaccharides are different from that of purified mannanase. Mannanase released
mannose (7.7% of substrate) and mannobiose (16%) as main linear manno-oligosaccharides
whereas Tr Cel5A released only little mannose (1.5%) and the main products were mannobiose
(5.2%), mannotriose (8.3%) and mannotetraose (3.8%), which suggests that the observed man-
man hydrolysis is truly carried out by Tr Cel5A rather than mannanase impurity in the enzyme
preparate.

3. Discussion
The hydrolysis efficiency of the three enzymes was evaluated based on the analysis of
released reducing sugars during a 48-h hydrolysis. All the three studied Trichoderma reesei
enzymes hydrolysed konjac mannan efficiently, reducing sugars over 50% of the substrate
(Figure 1). The most efficient enzyme in hydrolysis of konjac mannan was, in terms of both
hydrolysis reaction velocity and overall hydrolysis yield, was mannanase. Based on the shape of
the hydrolysis curve, the hydrolysis by EGI and mannanase had reached maximal level after 48 h.
However, the hydrolysis by EGII was still slowly progressing after 48 h and further hydrolysis of
the detected oligosaccharides might thus occur.
Although Trichoderma reesei mannanase was very efficient in the hydrolysis of konjac
mannan, it also was selective at its -1 subsite as almost all the detected oligosaccharides had a
non-reducing end mannose unit. Similar degradation pattern of konjac mannan has previously
been observed by mannanases from Aspergillus niger9,14 and Streptomyces sp..9 Hydrolysis of
galactoglucomannan present in pine kraft pulp by T. reesei mannanase revealed that it is able to
hydrolyse β-D-1,4 –pyranosyl linkage between two mannose units and between mannose and
glucose but not between glucose and mannose leading to the formation of e.g. GM
disaccharides.15 The observed cleavage pattern by T. reesei mannanase belonging to GH family 5
is in accordance with the conclusions made on the substrate specificity of GH family 5 and GH
family 26 mannanases.18 Based on site-directed mutagenesis of different mannanases it has been
concluded that the GH 5 mannanases can accommodate glucose in their -2 and +1 subsites but
not in -1 subsite of the active site.18 This property leads to the formation of oligosaccharides with
either glucose or mannose in non-reducing end and exclusively mannose in reducing end as
hydrolysis products.
Endoglucanases in general are able to hydrolyze the linkage β-D-1,4 –linkage between two
glucose units and additionally all endoglucanases from T.reesei are known to be able to cleave
the β-D-1,4 –linkage between a glucopyranosyl and mannopyranosyl unit in glucomannan.4 T.
reesei Cel7B is known to hydrolyse xylan, but not homopolymeric mannans.19 Based on the
work by Vlasenko et al. 2010,6 this hydrolysis pattern is typical for GH family 7 endoglucanases
as none of the seven studied GH family 7 endoglucanases hydrolyse linear mannan or
galactomannan, although some of them are relatively efficient in their activity towards xylans.
Unexpectedly some linear manno-oligocaccharides were present in the T. reesei Cel7B
hydrolysate. The Mn molecular weight of the utilised konjac mannan substrate is relatively low,
13 000 g mol-1 (~ DP 70),20 and the linear manno-oligosaccharides may have been released from
konjac mannan chains possessing mannose in the reducing end. The observed hydrolysis product
pattern of T. reesei Cel7B resembles that obtained previously with Trichoderma viride EG IV
(GH family 7),17 i.e. release of glucose and oligosaccharides with reducing end glucose and non-
reducing end mannose, even though linear manno-oligosaccharides were not observed in T.
viride EG IV hydrolysate.

Similar hydrolysis products to T.reesei Cel5A were observed in the konjac mannan
hydrolysate of Penicillium funiculosum endoglucanase.9 Identification of dimeric and trimeric
oligosaccharides from konjac mannan hydrolysed by Penicillium funiculosum β-endoglucanase
revealed that MG, MM, MMG and MMM were the dominant di- and trisaccharides present in
the hydrolysate. Glucose and mannose are present in the reducing end of the oligosaccharides in
1:1 ratio and the non-reducing end ratio presents the Man:Glc ratio of the original konjac
mannan substrate. Trichoderma reesei Cel5A belongs to GH family 5 and as described by
Vlasenko et al. 20106, several GH family 5 endoglucanases are able to hydrolyse mannans. In
addition, unlike other T. reesei endoglucanases, Cel5A (Cel 5a) (previously known as EGIII) has
been found to possess mannanase activity.21 The determined Km values of the enzyme were
similar towards both Locust bean gum galactomannan and carboxymethylcellulose, but the kcat
was tenfold higher to carboxymethylcellulose than galactomannan.
4. Conclusions
In this paper we present a detailed analysis of glucomannooligosaccharides obtained by
hydrolysis of konjac glucomannan with two T. reesei endoglucanases (Tr Cel5A and Tr Cel7B,
formerly known as EGII and EGI, respectively) and one mannanase (GH family 5). By
enzymatic hydrolysis of konjac mannan DP2-6 oligosaccharides with varying sequence were
produced, isolated and identified by the combination of NMR and secondary enzyme hydrolysis.
The analysis of the hydrolysis products provides knowledge on the substrate specificity of the
hydrolytic enzymes of one of the most studied plant biomass degrading fungi. Based on our
results Tr Cel5A hydrolyses the cleavage of all the glycosidic bonds present in glucomannan
main chain, whereas Tr Cel7B and T.reesei mannanase have higher substrate specificity,
especially in their -1 subsite of the active site. One of the challenges in studying enzymatic
hydrolysis of mannans with varying mannose:glucose:galactose ratio is the lack of commercial
standard samples for analysis of the hydrolysis products. The combination of specific enzymatic
hydrolysis and fractionation with size exclusion gives the possibility to produce oligosaccharides
with defined structure to be used as analytical standards, for determination of bioactivity of
oligosaccharides with defined structure or as substrates for defining specificity of novel
carbohydrate hydrolyzing enzymes.
5. Experimental
5.1 Materials
Enzyme treatments were carried out with mannanase, endoglucanase I (Tr Cel7a, GH family
7) and endoglucanase II (Tr Cel5A, GH family 5) purified from Trichoderma reesei culture

filtrate as described by Pere et al. (1995)22 and Rättö et al. (1993).23 Mannanase and
endoglucanase activities were determined as described by Pere et al. (1995)22 and the activities
were expressed as katals. One nanokatal (nkat) of enzyme catalyses the release of 1 nmol of
reducing sugars from the substrate polymer (locust bean gum for mannanase and HEC for
endoglucanase) in one second. Protein content of the preparates was determined with BIORAD
protein assay. For the secondary hydrolysis of the oligosaccharides commercial β-mannosidase
from Cellulomonas fimi (Megazyme, Ireland) was used. Commercial enzyme product Novozyme
188 was used as a source of Aspergillus niger β-glucosidase.
Low viscosity native konjac glucomannan (Megazyme, Ireland) was used as substrate in the
enzyme hydrolysis. The carbohydrate composition of the substrate was determined after acid
hydrolysis with 4% H2SO4 at 120°C for 1 h by HPAEC (Dionex ICS-3000) with pulse
amperometric detection using CarboPac PA1 column. The determined Man:Glc ratio of the
substrate was 1.64:1 and it contained traces of galactose (0.9% of dw) and arabinose (0.1% of
dry weight). The used substrate has been characterized in detail by Chua et al. (2012)20 and
based on their analysis the Mw of the substrate is 5.5 x 104g/mol and the degree of acetylation
1.8%.
5.2 Enzymatic hydrolysis of konjac mannan
Enzyme hydrolysis was performed in 1% substrate concentration in 50-mM ammonium
acetate buffer (pH 5), 45°C for 48 hours with mixing. Enzyme dosage was 2 mg of protein/g of
substrate. After hydrolysis the solutions were boiled for 10 min to inactivate the enzyme. In
order to isolate formed oligosaccharides the hydrolysis was performed to 4 g batches of konjac
mannan.
The enzyme hydrolysis was followed by analysing the increase of reducing sugars.The
concentration of reducing sugars was determined on a 96-well plate using a colorimetric DNS
(dinitrosalicylic acid) method described by Bernfeld (1955).24 The colored nitro ammonium
compound resulting from reaction of DNS with reducing sugars was quantified by reading the
absorbance at 540 nm using a Multiscan MS spectrophotometer. Distilled water was used as a
blank and 0.1-0.5 mg/ml D-(+)-glucose was used as standard.
5.3 Fractionation of konjac mannan hydrolysates by size exclusion chromatography
The hydrolysate was fractionated by size exclusion using preparative BioGel P2 (BioRad)
column as described by Tenkanen et al. (1997).15 Freeze-dried hydrolysates were dissolved into
distillated water for fractionation by gel filtration. Samples ca. 5 g were diluted using 50 ml of
water to gain ca. 10% (w/v) of end concentration. Whole 50 ml of diluted sample was injected
into preparative column of BioGel P2. P2 was eluted with water using 10 ml/min flow, totally
200 fractions were collected using collection time of 3 min for each fraction. The formed
fractions were analysed by HPAEC-PAD and samples containing the target oligosaccharides

were combined. The oligosaccharides in the combined samples were characterized using acid
hydrolysis and HPAEC-PAD (carbohydrate composition), HPLC-RI (to determine the DP of
purified oligosaccharides), enzymatic hydrolysis (determination of non-reducing end
carbohydrate) and NMR (to verify structure and purity of purified compounds).
5.4 Liquid chromatographic analysis
Oligosaccharides present in the hydrolysate were analysed with HPAEC (Dionex ICS-3000 )
with pulse amperometric and MSQ detection using CarboPAC 200 column with a configuration
described by Bruggink et al. (2005).25 The system was equilibrated with 3 mM NaOH. After
injection of a 25µl sample 3 mM NaOH was run through the column for 43 minutes. Two linear
gradients first from 3 mM to 100 mM NaOH during 37 minutes and then from 100 mM NaOH to
100 mM NaOH + 150 mM Na-acetate during 30 minutes were then created. Finally the column
was washed with 100 mM NaOH + 300 mM Na-acetate and then with 300 mM NaOH. The flow
rate was 0.3 ml min-1. The formed oligosaccharides were identified based on elution of
commercial standards (linear manno-oligosaccharides from MegaZyme, linear cello-
oligosaccarides from Serva, Seikagu, and Merck) and elution of in-house standards (GM, GGM,
GMM) described in Tenkanen et al. (1997).15 Yield of each oligosaccharide was calculated as
percentage from the substrate after anhydrocorrection with coefficient calculated from the
equation: mhydrolysis product ÷ (mhydrolysis product + mH2O). HPAEC-MS was used to verify and to
identify unidentified compound as Li –adducts [M + Li]+ at [M + 7]+ in the positive mode.
HPLC-RI was also utilized to identify the purity of purified compound and to determine the
DP of those. Used equipment was HPLC-RI (PerkinElmer, Flexar) system equipped with
Phenomenex RSO-oligosaccharide 4 % Ag+, 200x10 mm -column. Running conditions were
isocratic 0.3 ml/min of milliQ -water. The DP range and elution order was calibrated by using
linear manno- and cellooligosaccharide standard mixtures of DP 1-6.
5.5 NMR Spectroscopy
For NMR experiments the oligosaccharide samples were dissolved in 600 µl of D2O (99.9%
D, Aldrich) containing 0.05 wt. % 3-(trimethylsilyl)propionic-2,2,3,3-d4 acid as an internal
standard. All NMR experiments were carried out at 8ºC on a 600 MHz Bruker Avance III NMR
spectrometer equipped with a Q-CPN cryoprobe. The residual water signal was suppressed by
presaturation using the noesy1d pulse sequence. DQFCOSY, TOCSY and 13C multiplicity edited
HSQC 2D spectra were recorded with standard pulse sequnces. In TOCSY the mixing time was
1
80 ms and in HSQC the JHC coupling constant was estimated to 150 Hz. The chemical shifts
were referenced to internal TSP.

Acknowledgements
Ulla Vornamo and Nina Vihersola are thanked for skilful technical assistance. The research
leading to these results has received funding from the European Community’s Seventh
Framework Programme FP7/2007-2013 under grant agreement n° CP-IP 228589-2 AFORE.
References
1. Moreira, L.R.S.; Filho, E.X.F. Appl. Microbiol. Biotechnol., 2008, 79, 165-178.
2. Katsuraya, K.; Okuyama, K.; Hatanaka, K.; Oshima, R.; Sato, T.; Matsuzaki, K.; Carbohydr.
Polym., 2003, 53, 183-189.
3. Willför, S.; Sjöholm, R.; Laine, C.; Roslund, M.; Hemming, J.; Holmbom, B. Carbohydr.
Polym., 2003, 52, 175–187.
4. Biely, P.; Tenkanen, M. Enzymology of hemicellulose degradation. In Trichoderma and
Gliogladium. Harman, G.E., Kubicek, C.P., Eds.; Taylor & Francis Ltd, London & Bristol,
1998; Vol. 2, pp 25-47.
5. van Zyl, W.; Rose, S.H.; Trollope, K.; Görgens, J.F. Process Biochem., 2010, 45, 1203-1213.
6. Vlasenko, E.; Schülein, M.; Cherry, J.; Xu, F. Biores. Technol., 2010, 101, 2405-2411.
7. Al-Ghazzewi, F.H.; Khanna, S.; Tester, R.F.; Piggott, J. J. Sci. Food Agric., 2007, 87, 1758-
1766.
8. Akipnar, O.; Erdogan, K.; Bakir, U.; Yilmaz L. LWT – Food Sci Technol., 2010, 43, 119-125.
9. Cescutti, P.; Campa, C.; Delben, F.; Rizzo, R. Carbohydr. Res., 2002, 337, 2505-2511.
10. McCleary, B.V.; Charnock, S.J.; Rossiter, P.C.; O´Shea, M.F.; Power, A.M., Lloyd, R.M. J.
sci. Food Agric., 2006, 86, 1648-1661.
11. Martinez, D.; Berka, R.M.;Henrissat, B.;Saloheimo, M.; Arvas, M.; Baker, S.E.; Chapman,J.;
Kubicek, C.P.; Han, C.S.; Ho, I.; Larrondo, L.F.; de Leon, A.L.; Magnuson, J.K.; Merino, S.;
Misra, M.; Nelson, B.; Putnam, N.; Robbertse, R.; Salamov, A.A.; Schmoll, M.; Terry, A.;
Thayer, N.; Westerholm-Parvinen, A.; Schoch, C.L.; Yao, J.; Barabote, R.; Nelson, M.A.;
Detter, C.; Bruce, B.; Kuske, C.R.; Xie, G.; Richardson, P.; Rokhsar, D.S.;Lucas, S.M.;
Rubin, E.M.; Dunn-Coleman, N.; Ward, M.; Brettin, T.S. Nature Biotech., 2008, 26, 553–
560.
12. Saloheimo, M., Nakari-Setälä, T., Tenkanen, M. and Penttilä, M. Eur. J. Biochem. 1997, 249,
584–591.
13. Stålbrand, H.;Saloheimo, A.;Vehmaanperä, J.; Henrissat, B.; Penttilä,. M. Appl. Environ.
Microbiol., 1995, 1090-1097.

14. Takahashi, R.; Kusakabe, I.; Kusama, S.; Sakurai, Y.; Murakami, K.; Maekawa, A.; Suzuki,
T. Agric. Biol. Chem., 1984, 48, 2943-2950.
15. Tenkanen, M.; Makkonen, M.; Perttula, M.; Viikari, L.; Teleman, A. J. Biotechnol., 1997, 57,
191-204.
16. Bock, K.; Pedersen C.; Pedersen, H. Adv. Carboh. Chem. Biochem., 1984, 42, 93-225
17. Albrecht, S.; van Muiswinkel, G.C.J.; Xu, J.; Schols, H.A.; Voragen, A.G.J.; Gruppen, H. J.
Agric. Food Chem., 2011, 59, 12658-12666.
18. Tailford, L., Ducros, V. M.-A.; Flint, J.E.; Roberts, S.M.; Morland, C.; Zechel, D.L.; Smith,
N.; Bjornvad, M.E.; Borchert, T.V.; Wilson, K.S.; Davies, G.J.; Gilbert, H.J. Biochemistry,
2009, 48, 7009-7018.
19. Bailey, M.; Siika-aho, M.; Valkeajärvi, A.; Penttilä, M. Biotechnol. Appl. Biochem. 1993, 17,
65-76.
20. Chua, M.; Chan, K.; Hocking, T.J.; Williams, P.A.; Perry, C.J.; Baldwin, T.C. Carbohydr.
Polym., 2012, 87, 2202-2210.
21. Macarrón, R.; Acebal, C.; Castillón, M.P.; Clyessens, M. Biotechnol. Lett., 1996, 18, 599-
602.
22. Pere, J.; Siika-aho, M.; Buchert, J.; Viikari, L. Tappi J., 1995, 78, 71-78.
23. Rättö, M.; Poutanen, K. Biotechnol. Lett., 1988, 10, 661-664.
24. Bernfeld, P. Amylases, α and β. In: Colowick, S. P. and Kaplan, N. O. (eds.). Methods in
enzymology, vol 1. New York, NJ, USA: Academic Press. 1955. pp. 149-158.
25. Bruggink, C.; Maurer, R.; Herrmann, H.; Cavalli, S.; Hoefler, F. J. Chromatogr. A, 2005,
1085, 104 -109.

Figure 1. Enzymatic hydrolysis of konjak mannan by Trichoderma reesei endoglucanases (Tr
Cel7B and TrCel5A) and mannanase (man) was followed by release of reducing sugars using
dinitrosalicylic acid method (DNS). The data represents average and standard deviation of three
parallel reducing sugar analyses.

Figure 2. HPAEC –PAD chromatograms of konjac mannan hydrolysed by Trichoderma reesei
endoglucanase Cel7B (A), mannanase (B) and endoglucanase Cel5A (C). The degree of
polymerization (DP) of the unidentified peaks were determined using MSQ –detection.

Figure 3. HPAEC-PAD chromatogram of fractions isolated by gel filtration from Trichoderma
reesei Cel7B hydrolysate of Konjac mannan A) Unidentified oligosaccharide 1(MG), B)
Unidentified oligosaccharide 2 (MMG), C) Unidentified oligosaccharide 3 (MMMG), D)
Unidentified oligosaccharide 4 (MMMMG). The chromatograms on top in each panel are the
fractions as such, the chromatogram below are the fractions after secondary hydrolysis with β-
mannosidase from Cellulomonas fimi.

1
Figure 4. Structural reporter group regions of the 600 MHz H NMR spectra of A) MG
(unidentified 1), B) MMG (unidentified 2), C) MMMG (unidentified 3), D) MMMMG
(unidentified 4). The most important structural reporter groups observed in these spectra are H1
of the reducing end α and β glucose (doublets at 5.23 and 4.67 ppm, respectively), H1 of the
mannoses (between 4.7 and 4.8 ppm) and H2 of the mannoses (around 4.1 ppm). The letters b-e
refer to the different mannose units of the saccharides, starting from the reducing end. The
spectra were recorded at 8ºC in D2O and the chemical shifts were referenced to internal TSP (0
ppm).

13
Figure 5. 150 MHz C NMR DEPT(135) spectrum of the trisaccharide MMG at 8ºC. The clear
glycosylation shifts of a4, a4 and b4 13C signals compared to the non-glycosylated c4 confirm
the sites of the glycosidic bonds. As illustrated, the residues are named by letters a-c, starting
from the reducing end glucose. The assignments are based on 2D COSY, TOCSY and HSQC
spectra and the exact chemical shifts are given in table 4 with reference to internal TSP, 0 ppm..

Table 1 Identification of oligosaccharides isolated by size exclusion chromatography from
Trichoderma reesei Cel7B konjac mannan hydrolysate
Proposed
sequence
DPa,b MSQ
detection ratio
Man:Glc Man:Glc
Yield Yield
Estimated
purity (%)
ratio
(enzyme
hydrol.)
(mg)
(% of
subst.)
(SIM
ions
(acid
hydrol.)
m/z)b
Unidentified 1
Unidentified 2
Unidentified 3
MG
2
3
4
5
349
0.9:1
1.6:1
2.2:1
2.5:1
0.9:1
1.7:1
3.1:1
5.1:1
180
174
238
323
4.5
4.4
6.0
8.1
96
95
86
81
MMG
511
MMMG
673
Unidentified 4 MMMMG
421c, 836
aHPLC-RI: DP was measured with Phenomenex RSO-oligosaccharide column which DP range
was calibrated using linear manno- and cellooligosaccharides having degree of polymerization of
1-6.
bMSQ detection: SIM (single ion monitoring) mas per charge values for sugars can be detected
as Li-adducts [M + Li]+ at [M + 7]+ in the positive mode. For oligohexoses m/z values were
checked using linear manno- and cello-oligosaccharides (DP 1-6)
cThe adduct of saccharide and two lithium ions has two overall charges, therefore the total mass
has to be divided by a factor 2 in order to obtain the mass per charge ratio (m/z).
Table 2 Secondary hydrolysis products obtained from unidentified oligosaccharides 5 and 6
enriched from T. reesei mannanase konjac mannan hydrolysate by size exclusion
chromatography into fractions 120 – 128.
Unident. 5 Unident. 6 GMM (%) GGM (%) Man2 (%)
Man (%)
Glc (%)
(%)
(%)
42
29
28
30
23
-
35
-
-
-
-
-
-
Fraction 120-128
β-mannos.a
-
31
-
5
2
β-glucos.b
43
3
19
Secondary hydrolysis performed by aCellulomonas fimi β-mannosidase and bAspergillus niger β-
glucosidase

Table 3 1H chemical shifts of the glucomannooligosaccharides produced by enzymatic
hydrolysis of konjac glucomannan recorded at 600 MHz.
Compound Residuea Chemical shiftb
H1
H2
H3
H4
H5
H6
H6’
MG
5.228
4.673
4.763
3.588
3.291
4.071
3.861
3.679
3.664
3.680
3.692
3.585
3.931
3.586
3.433
3.822
3.878
3.950
3.767
3.738
3.735
-Glca
-Glca
-Manb
MMG
5.229
4.673
4.785
4.751
3.588
3.291
4.136
4.077
3.864
3.683
3.818
3.665
3.684
3.70
3.843
3.561
3.935
3.590
3.553
3.452
3.828
3.882
3.926
3.962
3.773
3.740
3.769
3.730
-Glca
-Glca
-Manb
-Manc
MMMG
5.229
4.672
4.784
4.774
4.746
3.585
3.298
3.866
3.684
3.684
3.70
3.934
3.590
3.828
3.880
3.770
3.739
-Glca
-Glca
-Manb
-Manc
-Man
4.138
4.074
3.827
3.664
3.841
3.566
3.564
3.446
3.928
3.960
3.768
3.731
MMMMG
5.229
4.672
4.783
4.770
4.757
3.587
3.291
3.865
3.684
3.684
3.70
3.936
3.591
3.828
3.882
3.771
3.740
-Glca
-Glca
-Manb
-Manc-d
-Mane
4.138
4.074
3.826
3.666
3.843
3.564
3.565
3.451
3.930
3.963
3.765
3.731
a The residues are named by letters a-e, starting from the reducing end glucose.
bIn ppm relative to internal TSP at 0 ppm (D2O, 8ºC)

Table 4. 13C chemical shifts of the glucomannooligosaccharides produced by enzymatic
hydrolysis of konjac glucomannan recorded at 600 MHz.
Compound
Residuea Chemical shiftb
C1
C2
C3
C4
C5
C6
Man-Glc
94.64
98.58
102.85 73.40
73.90
76.52
74.14
77.06
75.61
81.51
81.45
69.45
72.77
77.46
79.29
62.90
63.02
63.75
-Glca
-Glca
-Manb
Man2-Glc
Man3-Glc
Man4-Glc
94.65
98.63
102.80 72.78
103.05 73.31
73.88
76.56
74.10
77.00
74.34
75.58
81.47
81.36
79.45
69.49
72.77
77.53
77.88
79.28
62.85
63.00
63.32
63.85
-Glca
-Glca
-Manb
-Manc
94.64
98.62
102.87 72.76
103.00 73.30
73.92
76.58
74.09
77.00
74.31
75.57
81.48
81.35
79.40
69.47
72.75
77.57
77.87
79.28
62.85
62.95
63.27
63.82
-Glca
-Glca
-Manb,c
-Mand
94.65
98.61
102.84 72.78
103.00 73.30
73.89
76.54
74.09
77.00
74.32
75.57
81.43
81.32
79.35
69.47
72.77
77.5
77.85
79.28
62.85
62.94
63.26
63.83
-Glca
-Glca
-Manb-d
-Mane
a The residues are named by letters a-e, starting from the reducing end glucose.
bIn ppm relative to internal TSP at 0 ppm (D2O, 8ºC)

Table 5 Share of reducing and non-reducing end gluco- and mannopyranosyl units in the konjac
mannan hydrolysates of T.reesei Cel7B, Cel5A and mannanase
Tr Cel7B
Tr Cel5A Tr mannanase
hydrolysate hydrolysate hydrolysate
(%)a
83.3
16.7
12.5
87.5
(%)a
49.8
50.2
29.1
70.9
(%)a
Reducing end Glc
1.0
Reducing end Man
Non-reducing end Glc
Non-reducing end Man
99.0
71.2
28.8
Sum of identified
hydrolysis productb (%)
67.4
75.9
100.9
aPercentage of identified hydrolysis products
bSum of identified hydrolysis products was calculated using anhydrocorrection