The b-1,4-endogalactanase A gene from
Aspergillus niger
is specifically induced on arabinose and galacturonic acid and plays
an important role in the degradation of pectic hairy regions
Ronald P. de Vries
1†
, Lucie Par
ˇ
enicova
´
1‡
, Sandra W. A. Hinz
2
, Harry C. M. Kester
1
, Gerrit Beldman
2
,
Jacques A. E. Benen
1
and Jaap Visser
1§
1
Molecular Genetics of Industrial Microorganisms and
2
Food Chemistry, Wageningen University, Wageningen, The Netherlands
The Aspergillus niger b-1,4-endogalactanase encoding gene
(galA) was cloned and characterized. The expression of galA
in A. niger was only detected in the presence of sugar beet
pectin,
D

substituted
D
-galacto-oligosaccharides. GALA was not
active towards
D
-galacto-oligosaccharides that were substi-
tuted with
D
-glucose at the reducing end.
Keywords: Aspergillus niger; b-1,4-endogalactanase; galac-
turonic acid; expression; galactan degradation.
Endogalactanases are involved in the degradation of plant
cell wall polysaccharides, in particular pectin. Two types of
arabinogalactan side chains are present in pectin. Type I
consists of a chain of b-1,4 linked
D
-galactopyranose
linkages, while type II contains a backbone of b-1,3-linked
D
-galactopyranose residues that can be substituted with
b-1,6-linked
D
-galactopyranose residues [1]. Both types can
be substituted with a)1-,3-linked
L
-arabinofuranose chains.
Type I arabinogalactan is degraded by b-1,4-endogalacta-
nase and b-galactosidase. b-1,4-Endogalactanases cleave
within the galactan moiety of type I arabinogalactan,
releasing

was performed using pBluescript SK
+
[18] and pGEM-T
(Promega, Madison, WI, USA). The genomic library of
A. niger has been described previously [19].
Minimal medium and complete medium were descri-
bed before [20]. Liquid cultures were inoculated with
Correspondence to R. P. de Vries, Microbiology, Utrecht University,
Padualaan 8, 3584 CH Utrecht, The Netherlands.
Fax: +31 302513655, Tel.: +31 302533016,
E-mail:
Abbreviations: CREA, carbon catabolite repressor protein; HPAEC,
high performance anion exchange chromatography; GALA, the
A. niger b-1,4-endogalactanase; galA,geneencodingtheA. niger
b-1,4-endogalactanase; Galp, galactopyranose; Glcp, glucopyranose;
LACA, A. niger a-galactosidase; PACC, pH regulatory protein;
TOS, transgalactooligosaccharides.
Present address: Microbiology, Utrecht University, Padualaan 8,
3584 CH Utrecht, The Netherlands.
àPresent address:DipartimentodiBiologia,Universitadeglistudidi
Milano, Via Celoria 26, 20133 Milano, Italy.
§Present address: FGT Consultancy, PO Box 396, 6700 AJ Wagenin-
gen, The Netherlands.
(Received 27 March 2002, revised 1 July 2002,
accepted 21 August 2002)
Eur. J. Biochem. 269, 4985–4993 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03199.x
10
6
sporesÆmL
)1

polypropylene glycol in ethanol was added per litre of
medium as antifoam agent.
Materials
D
-Xylose,
D
-glucose,
D
-fructose,
D
-galactose,
D
-mannose,
and lactose were obtained from Merck (Darmstadt,
Germany).
D
-Glucuronic and
D
-galacturonic acid were
from Fluka (Buchs, Switzerland). Mellibiose, raffinose,
stachyose,
L
-arabinose, gum arabic, gum karaya, locust
bean gum, and beechwood xylan were from Sigma (St. Louis,
Mo.). Potato pectic galactan was from Megazyme Interna-
tional (Bray, Ireland). Taq polymerase was from Gibco BRL
(Breda, The Netherlands). All other standard chemicals
were either obtained from Sigma or Merck. Potato arabi-
nogalactan and onion arabinogalactan were obtained as
described previously (Fractions F44) [21]. Soy arabinoga-

and purified by rescreening at low plaque density. Standard
methods were used for other DNA manipulations, such as
Southern analysis, subcloning, DNA digestions, and lambda
phage and plasmid DNA isolations [23]. Chromosomal
DNA was isolated as previously described [24]. Sequence
analysis was performed on both strands of DNA by using
the Cy5 AutoCycle Sequencing kit (Pharmacia Biotech,
Uppsala, Sweden). The reactions were analyzed with an
ALFred DNA Sequencer (Pharmacia Biotech). Nucleotide
sequences were analyzed with computer programs based on
those of Devereux et al. [25]. RT-PCRs were performed
using the Enhanced Avian RT-PCR kit (Sigma) according
to the suppliers instructions using galA-specific oligonucleo-
tides (5¢-GATGATCTACCCTCTGCTTC-3¢ and 5¢-GTC
ACGGACGGACTGGGT-3¢). Northern analysis was per-
formed as described previously [20]. Per sample, 5 lgof
total RNA was loaded on the gel.
The A. niger galA accession number is AJ305303.
Sequence alignments
Nucleic acid and amino acid sequence alignments were
performed by using the Blast programs [26] at the server
of the National Center for Biotechnology Information
(Bethesda, Md., USA).
Purification of b-1,4-endogalactanase
Culture fluid was collected by filtration through cheesecloth
and diluted twofold with distilled water, after which the pH
was adjusted to 6.5 by the addition of sodium hydroxide.
Proteins were collected by batchwise adsorption to Stream-
line Q XL (Amersham Pharmacia Biotech, Sweden). For
this 40 mL of the matrix was added to the culture fluid and

, areA1/areA
+
[43]
NW290 cspA1, fwnA12, DargB/argB
+
, pyrA6,
prtF28, DpgaA, DpgaB, goxC17
[44]
4986 R. P. de Vries et al.(Eur. J. Biochem. 269) Ó FEBS 2002
Enzyme assays
Transformants producing elevated levels of b-1,4-endoga-
lactanase were selected using AZCL-galactan (Megazyme,
Bray, Ireland) according to the supplier’s instructions.
During the purification endogalactanase activity was moni-
tored using azo-galactan (Megazyme) as the substrate. Two
hundred microlitres of a solution of 1% (w/v) azo-galactan
in water was mixed with 50 lL of a 50-m
M
sodium acetate
buffer (pH 4.2) followed by the addition of 50 lL enzyme
solution. Incubations were performed for 20 min at 40 °C.
Reactions were stopped by the addition of 650 lL ethanol.
The precipitated substrate was removed by centrifugation
and the supernatant was used to measure absorbance at
590 nm. The pH optimum of GALA was determined using
azo-galactan as described above with a pH range of 3.5–7.5.
The specific activity of GALA was determined using
GALACTAZYME tablets (Megazyme) with conditions as
indicated by the supplier and using 0, 10, 25, 50, 100, 200
and 500 lL of the GALA preparation. Incubations were

GALA at 30 °C, pH 4.5 (McIlvain buffer) for 15 min,
30 min, 2 h and 20 h, respectively. The incubation was
stopped by heating the samples for 5 min at 100 °C.
a-L-1,3-Arabinofuranosidase activity in the GALA pre-
paration was measured as described before [27], using 10
and 100 lL of the enzyme preparation (in duplicate) and
incubation times of 1 and 4 h.
Analytical methods
SDS/PAGE. Electrophoresis of proteins was performed
under denaturing conditions on 10% (w/v) gels using the
method of Laemmli [28] in a Mini-V system (Life
Technologies B.V., Breda, The Netherlands).
HPAEC. High performance anion exchange chromato-
graphy (HPAEC) was performed using a SpectraSystem
P4000 (Thermo Separation Products) equipped with a Dio-
nex CarboPac PA-1 (4 · 250 mm) column and a Dionex
ED40 Electrochemical Detector in de Pulsed Amperometric
Detection mode. The samples were analyzed using a linear
gradient of 0–400 m
M
sodium acetate in 100 m
M
sodium
hydroxide for 40 min
D
-galactose,
D
-galactobiose and
D
-galactotetraose were used as standards to identify the

a4.5-kbEcoRI fragment was cloned into pBluescript
SK
+
, resulting in plasmid pIM3980. Double stranded
sequence was determined for a region of this construct
containing galA and some of the flanking regions,
resulting in the genomic sequence of A. niger galA.The
presence of one putative intron was confirmed by RT-
PCR using total RNA of A. niger mycelium (transferred
to minimal medium containing 15 m
MD
-galacturonic
acid) as a template. The galA gene has a length of
1122 bp, is interrupted by one intron of 72 bp, and
encodes a protein of 350 amino acids. Computer analysis
predicted a eukaryotic signal sequence of 16 amino acids.
The putative mature enzyme has a calculated pI of 3.67, a
calculated molecular mass of 37 053.9 Da and contains
one potential N-glycosylation site. BLAST analysis of the
deduced amino acid sequence of GALA revealed very
high similarity to the endogalactanases of A. tubingensis
(95.7% amino acid sequence identity) and A. aculeatus
(78.9% amino acid sequence identity) and lower similarity
to bacterial endogalactanases (between 21% and 28%
amino acid sequence identity). This is illustrated by a
CLUSTALW analysis [29] of the deduced amino acid
sequences of GALA and b-1,4-endogalactanases of
Aspergillus tubingensis [15], A. aculeatus [12], Bacillus
circulans (Acc. No. P48843), Yersinia pestis [14], Bacillus
subtilis (Acc. No. O07013), Clostridium acetobutylicum

amino acids indicated are those present in at least four sequences. Amino acids present in all sequences are in bold. The eukaryotic signal sequence
of A. niger GALA is depicted in lower case letters. The putative N-glycosylation site of A. niger GALA is in bold, italics, underlined and indicated
above the sequence ("). The two catalytic residues identified in the P. fluorescens endogalactanase [11] are indicated above the alignment (^)andare
in white against a grey background.
4988 R. P. de Vries et al.(Eur. J. Biochem. 269) Ó FEBS 2002
second experiment with identical experimental setup was
performed in which the expression of galA was studied in
two strains, a wild type (N402) and a mutant with a
derepressed phenotype for CREA repression (NW200).
Five of the carbon sources used in this experiment (
D
-xylose,
L
-arabinose,
D
-galactose,
L
-rhamnose,
D
-galacturonic acid,
Fig. 3) have been identified as components of sugar beet
pectin [1], whereas the others served as controls. Expression
of galA in the wild type strain was only detected in
the presence of
D
-galacturonic acid (Fig. 3). However, in the
CREA mutant expression of galA was detected in the
presence of
D
-galacturonic acid and

)1
.
Using galactazyme tablets a specific activity of 12.3 UÆmg
)1
was determined for GALA.
The purified protein had a molecular mass of 48.5 kDa as
determined by SDS/PAGE (Fig. 4). The pH optimum was
between4and4.5.
No activity of GALA could be detected against any of
the transgalactooligosaccharides listed in Materials and
methods.
Hydrolysis of arabinogalactans
The sugar composition of the arabinogalactans from
potato, onion and soy was determined as described [21].
Onion arabinogalactan consists of 99%
D
-galactose and
0.3%
L
-arabinose and is predominantly linear. Potato
arabinogalactan consists of 86%
D
-galactose and 6.6%
L
-arabinose, while soy arabinogalactan consists of 57%
D
-galactose and 38%
L
-arabinose. Methylation analysis
demonstrated that a substantial amount of the

tetramers, these products disappear during prolonged
incubation. A relatively higher amount of monomers
and dimers is released from soy arabinogalactan, which
increased upon prolonged incubation.
MALDI-TOF MS analysis allowed for a more detailed
analysis of the products formed during the incubations,
although the method used cannot detect monomers and
dimers. GALA initially released
D
-galactotriose,
D
-galacto-
tetraose and
D
-galactopentaose from potato and onion
arabinogalactan, while after prolonged incubations pre-
dominantly
D
-galactotriose and
D
-galactotetraose could be
detected (Fig. 6). Using soy arabinogalactan as a substrate,
not only
D
-galactotriose,
D
-galactotetraose and
D
-galacto-
pentaose were detected but also

preparation was assayed for this activity. However, no
a-
L
-1,3-arabinofuranosidase activity could be detected in
the preparation (data not shown).
This paper demonstrates the important role of GALA in
the degradation of both linear and
L
-arabinose-substituted
galactan side chains of pectin. This is in agreement with a
previous study in which the synergy of enzymes degrading
the pectin side chains was studied [34]. GALA had a positive
effect on the activity of the other enzymes involved in the
degradation of these side chains.
DISCUSSION
A. niger galA is highly similar to the b 1,4-endogalactanase
encoding genes from A. tubingensis and A. aculeatus.The
similarity of the galA genesishigherbetweenthetwo
biseriate species, A. niger and A. tubingensis,thanbetween
the monoseriate A. aculeatus galA gene and either of the
other two Aspergillus genes. A similar observation has
previously been reported for pectin lyase encoding genes of
these species [35]. Significant similarity was also detected to
a number of bacterial b-1,4-endogalactanases. The highest
similarity was detected in the region around the first
catalytic residue identified for the P. fluorescens b-1,4-
endogalactanase [11]. The similarity around the second
catalytic residue is lower.
The determined molecular mass for A. niger GALA
(48.5 kDa) is higher than the molecular mass based on the

-galactose residues but no b-1,4-linked galactan chains.
Therefore it only requires an exo-acting galactose releasing
enzyme (LACA). Previously it was shown that lacA is under
the control of the xylanolytic activator protein [32,41]. The
increase in expression levels of galA in the CREA mutant,
indicates that at least one of the detected putative CREA
binding sites in the promoter of galA is functional.
Hydrolysis of arabinogalactans with a different degree of
L
-arabinose substitution by GALA results in different
products. Arabinogalactans with a low degree of
L
-arabi-
nose substitution (potato, onion) as substrates result in the
liberation of monomers, dimers, trimers and tetramers of
D
-galactose. Using arabinogalactans with a high degree of
L
-arabinose substitution (soy), higher oligomers appear
initially, but these disappear if the incubation is continued.
Another difference is that only with the latter substrate
arabinose-substituted galacto-oligosaccharides are detected.
These disappear in time, suggesting the presence of an
arabinose-releasing activity in the GALA preparation.
However, no a-1-,3-
L
-arabinofuranosidase activity could
be detected using PNP-a-
L
-arabinofuranoside as a sub-

[10,12,36,39]. However, none of these papers reported the
release of
L
-arabinose substituted galacto-oligosaccharides.
ACKNOWLEDGEMENTS
The authors thank Simon Flitter for the identification and isolation
of the galA containing phage clones and Matthew Illsley for the
analysis of a-
L
-1,5-arabinofuranosidase activity in the endogalactanase
preparation.
REFERENCES
1. de Vries, R.P. & Visser, J. (2001) Aspergillus enzymes involved in
degradation of plant cell wall polysaccharides. Microb.Mol.Biol.
Rev. 65, 497–522.
2. Labavitch, J.M., Freeman, L.E. & Albersheim, P. (1976) Puri-
fication and characterization of a b-1,4-galactanase which
degrades a structural component of the primary cell walls of
dicots. J. Biol. Chem. 251, 5904–5910.
3. Emi, S., Fukumoto, J. & Yamamoto, T. (1971) Studies of hemi-
cellulolytic enzymes of Bacillus subtilis.PartI.Purification,crys-
tallization and some properties of arabinogalactanase. Agric. Biol.
Chem. 35, 1891–1898.
4. Emi, S. & Yamamoto, T. (1972) Purification and properties of
several galactanases of Bacillus subtilis var. amylosacchariticus.
Agric. Biol. Chem. 36, 1945–1954.
5. Tsumura, K., Hashimoto, Y., Akiba, T. & Hirokoshi, K. (1991)
Purification and properties of galactanases from alkalophilic
Bacillus sp. S-2 and S-39. Agric. Biol. Chem. 55, 1265–1271.
6. Yamamoto, T. & Emi, S. (1988) Arabinogalactanase of Bacillus

teriol. 183, 4823–4838.
14. Parkhill, J., Wren, B.W., Thomson, N.R., Titball, R.W., Holden,
M.T.G.,Prentice,M.B.,Sebaihia,M.,James,K.D.,Churcher,C.,
Mungall, K.L., Baker, S., Basham, D., Bentley, S.D., Brooks, K.,
Cerdeno-Tarraga, A.M., Chillingworth, T., Cronin, A., Davies,
R.M., Davis, P., Dougan, G., Feltwell, T., Hamlin, N., Holroyd,
S.,Jagels,K.,Leather,S.,Karlyshev,A.V.,Moule,S.,Oyston,
P.C.F., Quail, M., Rutherford, K., Simmonds, M., Skelton, J.,
Stevens, K., Whitehead, S. & Barrell, B.G. (2001) Genome
sequence of Yersinia pestis, the causative agent of plague. Nature
413, 523–527.
15. Vlugt-Bergmans, C.J.B. & van Ooijen, A.J.J. (1999) Expression
cloning in Kluyveromyces lactis. Biotechniques 13, 87–92.
16. Henrissat, B. & Bairoch, A. (1993) New families in the classifica-
tion of glycosidases based on amino acid sequence similarities.
Biochem. J. 293, 781–788.
17. Sorenson, S.O., Pauly, M., Bush, M., Skjot, M., McCann, M.C.,
Borkhardt, B. & Ulskov, P. (2000) Pectin engineering: Modifica-
tion of potato pectin by in vivo expression of an endo-1,4-beta-
D
-galactanase. Proc. Natl Acad. Sci. USA 97, 7639–7644.
18. Short, J.M., Fernandez, J.M., Sorge, J.A. & Huse, W.D. (1988)
Lambda ZAP: a bacteriophage expression vector with in vivo
excision properties. Nucleic Acids Res. 16, 7583–7600.
19. Harmsen, J.A.M., Kusters-van Someren, M.A. & Visser, J. (1990)
Cloning and expression of a second Aspergillus niger pectin lyase
gene (pelA): indications of a pectin lyase gene family in A. niger.
Curr. Genet. 18, 161–166.
20. de Vries, R.P. & Visser, J. (1999) Regulation of the feruloyl
esterase (faeA)genefromAspergillus niger. Appl. Environ.

(1993) Specific binding sites in the alcR and alcA promoters of the
ethanol regulon for the CreA repressor mediating carbon cata-
bolite repression in Aspergillus nidulans. Mol. Microbiol. 7,847–
857.
4992 R. P. de Vries et al.(Eur. J. Biochem. 269) Ó FEBS 2002
31. Orejas, M., Espeso, E.A., Tilburn, J., Sarkar, S., Arst, H.N.J. &
Penalva, M.A. (1995) Activation of the Aspergillus PacC tran-
scription factor in response to alkaline ambient pH requires pro-
teolysis of the carboxy-terminal moiety. Genes & Development. 9,
1622–1632.
32. de Vries, R.P., van den Broeck, H.C., Dekkers, E., Manzanares,
P., de Graaff, L.H. & Visser, J. (1999) Differential expression
of three a-galactosidase genes and a single b-galactosidase
gene from Apergillus niger. Appl. Environ. Microbiol. 65, 2453–
2460.
33. Kusters-van Someren, M., Flipphi, M., de Graaff, L.H., van den
Broeck, H., Kester, H., Hinnen, A. & Visser, J. (1992) Char-
acterisation of the Aspergillus niger pelB gene: structure and regu-
lation of expression. Mol. General Genet. 234, 113–120.
34. de Vries, R.P., Kester, H.C.M., Poulsen, C.H., Benen, J.A.E. &
Visser, J. (2000) Synergy between accessory enzymes from
Aspergillus in the degradation of plant cell wall polysaccharides.
Carbohydr. Res. 327, 401–410.
35. Par
ˇ
enicova
´
, L., Benen, J.A.E., Samson, R.A. & Visser, J. (1997)
Evaluation of RFLP analysis of the classification of selected black
Aspergilli. Mycol. Res. 101, 810–814.

master strains for assignment of genes to six linkage groups in
Aspergillus niger. Curr. Genet. 14, 437–443.
43. de Vries, R.P., Visser, J. & de Graaff, L.H. (1999) CreA modulates
the XlnR induced expression on xylose of Aspergillus niger genes
involved in xylan degradation. Res. Microbiol. 150, 281–285.
44. Warren, M.E., Kester, H., Benen, J., Colangelo, J., Visser, J.,
Bergmann, C. & Orlando, R. (2002) Studies on the glycosylation
of wild type and mutant forms of Aspergillus niger pectin methy-
lesterase. Carbohydr. Res. 337, 803–812.
45. Melchers,W.J.G.,Verweij,P.E.,vandenHurk,P.,vanBelkum,
A., de Pauw, B.E., Hoogkamp-Korstanje, A.A. & Meis, J.F.G.M.
(1994) General primer-mediated PCR for detection of Aspergillus
species. J. Clin. Microbiol. 32, 1710–1717.
Ó FEBS 2002 A. niger b-1,4-endogalactanase (Eur. J. Biochem. 269) 4993