Báo cáo khoa học: A novel trehalase from Mycobacterium smegmatis ) purification, properties, requirements potx

A novel trehalase from Mycobacterium
smegmatis ) purification, properties, requirements
J. David Carroll
1
, Irena Pastuszak
2
, Vineetha K. Edavana
2
, Yuan T. Pan
2
and Alan D. Elbein
2
1 Department of Microbiology and Immunology, University of Arkansas for Medical Sciences, Little Rock, AR, USA
2 Department of Biochemistry and Molecular Biology, University of Arkansas for Medical Sciences, Little Rock, AR, USA
Trehalose, i.e. a-d-glucopyranosyl-a-d-glucopyranoside,
is a nonreducing disaccharide that is widely distributed
throughout the biological kingdom, being prominent in
prokaryotes and lower eukaryotes, but absent from
mammals [1]. It has a number of important and differ-
ent functions in these various organisms, including:
acting as a reservoir of glucose and ⁄ or energy [2]; ser-
ving as a protectant of proteins and membranes during
various stress conditions [3,4]; having a regulatory role
in the control of glucose metabolism [5]; playing a role
in transcriptional regulation [6]; and serving as an
essential component of various cell wall glycolipids,
Keywords
effect of phosphate; glycosyl hydrolase;
pyrophosphate inhibition; trehalase;
trehalase inhibitors
Correspondence

A. D. Elbein, Department of Biochemistry
and Molecular Biology, University of
Arkansas for Medical Sciences, Little Rock,
AR 72205, USA
Fax: +1 501 686 5189
Tel: +1 501 686 5176
E-mail:
(Received 11 October 2006, revised 18
January 2007, accepted 22 January 2007)
doi:10.1111/j.1742-4658.2007.05715.x
Trehalose is a nonreducing disaccharide of glucose (a,a-1,1-glucosyl-glu-
cose) that is essential for growth and survival of mycobacteria. These
organisms have three different biosynthetic pathways to produce trehalose,
and mutants devoid of all three pathways require exogenous trehalose in
the medium in order to grow. Mycobacterium smegmatis and Mycobacteri-
um tuberculosis also have a trehalase that may be important in controlling
the levels of intracellular trehalose. In this study, we report on the purifica-
tion and characterization of the trehalase from M. smegmatis, and its com-
parison to the trehalase from M. tuberculosis. Although these two enzymes
have over 85% identity throughout their amino acid sequences, and both
show an absolute requirement for inorganic phosphate for activity, the
enzyme from M. smegmatis also requires Mg
2+
for activity, whereas the
M. tuberculosis trehalase does not require Mg
2+
. The requirement for
phosphate is unusual among glycosyl hydrolases, but we could find no evi-
dence for a phosphorolytic cleavage, or for any phosphorylated intermedi-
ates in the reaction. However, as inorganic phosphate appears to bind to,

and also to greatly increase the heat stability of, the trehalase, the function
of the phosphate may involve stabilizing the protein conformation and ⁄ or
initiating protein aggregation. Sodium arsenate was able to substitute to
some extent for the sodium phosphate requirement, whereas inorganic
pyrophosphate and polyphosphates were inhibitory. The purified trehalase
showed a single 71 kDa band on SDS gels, but active enzyme eluted in the
void volume of a Sephracryl S-300 column, suggesting a molecular mass of
about 1500 kDa or a multimer of 20 or more subunits. The trehalase is
highly specific for a,a-trehalose and did not hydrolyze a,b-trelalose or
b,b-trehalose, trehalose dimycolate, or any other a-glucoside or b-glucos-
ide. Attempts to obtain a trehalase-negative mutant of M. smegmatis have
been unsuccessful, although deletions of other trehalose metabolic enzymes
have yielded viable mutants. This suggests that trehalase is an essential
enzyme for these organisms. The enzyme has a pH optimum of 7.1, and is
active in various buffers, as long as inorganic phosphate and Mg
2+
are
present. Glucose was the only product produced by the trehalase in the
presence of either phosphate or arsenate.
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1701
especially in mycobacteria and other related bacteria [7].
In some of these organisms, notably mycobacteria and
corynebacteria, there are three different pathways that
can produce trehalose [8,9], and mutants defective in all
three pathways are unable to grow unless the growth
medium contains or is supplemented with trehalose
[10,11]. Thus, trehalose is essential for mycobacteria
and corynebacteria.
The major enzyme involved in the turnover of treha-
lose, or its conversion to two molecules of glucose, is

trehalase (a,a,1,1-glucosyl hydrolase) [12]. Trehalases
(EC 3.2.1.28) are generally placed in glycoside hydrol-
ase family 65 [13], although Mycobacterium smegmatis
MSMEG4528 and Mycobacterium tuberculosis MT2474
and Rv2402 have been placed in glycoside hydrolase
family 15. This group of enzymes is widely distributed
in the biological world, and trehalases are found in
most organisms that synthesize and ⁄ or utilize trehalose.
In some organisms, such as Saccharomyces cerevisiae,
there are several different trehalases, one of which is
regulated by cAMP and phosphorylation, and another
which is apparently not a regulatory enzyme [14]. On
the other hand, in most bacteria, trehalase does not
appear to undergo post-translational modifications
such as phosphorylation [15], although some of these
enzymes may be transcriptionally regulated.
The trehalase described in this article was purified to
apparent homogeneity from cytoplasmic extracts of
M. smegmatis. This enzyme is unusual among this
group of glycosyl hydrolases [13] in that it has an
almost absolute requirement for inorganic phosphate
and Mg
2+
for activity, although we cannot find any
evidence for a phosphorylated sugar intermediate in
the reaction. The role of the inorganic phosphate and
Mg
2+
may be to stabilize the enzyme conformation,
or to aid in aggregation of the protein to produce act-

ive enzyme. The trehalase may be an essential enzyme
for mycobacteria, as all attempts to isolate mutants
deleted in this protein were unsuccessful. The proper-
ties, amino acid sequence and requirements of this
trehalase are herein described.
Results
Purification of M. smegmatis trehalase
Trehalase was purified about 160-fold from the cytoso-
lic extract of M. smegmatis using the procedures out-
lined in Table 1. The steps in the purification included
ion exchange chromatography on DE-52, hydrophobic
chromatography on phenyl-Sepharose, gel filtration on
Sephacryl S-300, and chromatofocusing. With these
procedures, the trehalase was purified about 160-fold,
with an overall yield of about 0.4%. Figure 1 shows
the protein profiles at each stage of purification, as
demonstrated by SDS ⁄ PAGE. It can be seen in lane 6
that after purification by chromatofocusing, there was
one major band with a molecular mass of about
71 kDa on the SDS gels (Fig. 1). On the other hand,
active trehalase, subjected to gel filtration on Sephacryl
S-300, was eluted from the column in the void volume,
indicating a molecular mass of over 1.5 · 10
6
Da, and
suggesting that the active enzyme is a multimer of 20
Table 1. Purification of trehalase from M. smegmatis.
Purification steps
Protein
(mg)

Activity
(units
a
)
Specific
activity
(unitsÆmg
)1
)
Fold of
purification
1. Extract 3127 709 985 227 1
2. (NH)
2
SO
4
fraction 2329 636 512 273 1.2
3. DE-52,
Phenyl-sepharose
18 13 348 722 3.2
4. Sephacryl S-300 1.9 2818 1452 6.4
5. Chromatofocusing 0.08 2959 36 992 162
a
Units are expressed as nanomoles of glucose released from tre-
halose in 1 min at 37 °C.
Fig. 1. Protein profiles by SDS ⁄ PAGE at various stages in the purifi-
cation of M. smegmatis trehalase. Trehalase was purified as des-
cribed in Table 1, and an aliquot of the enzyme preparation at each
step in the purification was subjected to SDS ⁄ PAGE for 6 h at
30 mA. Proteins were detected by staining with Coomassie Blue.

Lanes: 1, crude extract; 2, ammonium sulfate fractionation; 3,
DE-52 fraction; 4, phenyl-Sepharose fraction; 5, Sephacryl S-300
fraction; 6, preparation after chromatofocusing (the trehalase band
in lane 6 is indicated by arrows); 7, protein standards with masses
of 97 (top band), 66, 45 and 31 kDa.
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1702 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works
or more subunits. The purified enzyme was stable to
storage at ) 20 °C for at least several weeks, but lost
activity upon repeated freezing and thawing. It could
be stored on ice for several weeks with no apparent
loss of activity.
The 71 kDa protein band from lane 6 of the SDS
gels was excised from the gels and subjected to trypsin
digestion and amino acid analysis, with MS being used
to determine the amino acid composition of the various
peptides. The resulting peptide sequences were used to
screen the M. smegmatis mc
2
155 genome sequence
maintained by the Institute for Genomic Research
(TIGR) ( />Page.cgi?org_search ¼ & org ¼ gms.). Using the pro-
gram tblastn, which compares an amino acid query
sequence against a nucleotide sequence database
dynamically translated in all six reading frames, all of
the trehalase-derived peptide sequences aligned with a
single M. smegmatis ORF, MSMEG 4528. This ORF
specifies a 672-residue polypeptide with a calculated
molecular mass of 75.2 kDa. tblastn screening of the
M. tuberculosis H37Rv genome sequence (http://www.

sanger.ac.uk/Projects/M_tuberculosis/) with the amino
acid sequence predicted by MSMEG 4528 identified a
homologous M. tuberculosis ORF, Rv2402. Rv2402
specifies a 642-residue protein, annotated as a ‘con-
served hypothetical protein’ of unknown function. The
respective predicted amino acid sequences of
MSMEG 4528 and Rv2402 are 88% identical, with the
identity distributed evenly throughout the sequence
alignment. The comparison of these two sequences is
presented in Fig. 2. The sequence alignment of the
M. smegmatis trehalase also showed 72% identity
with a hypothetical trehalase from Nocardia farcinica,
63% identity with a proposed trehalase from Frankia,
37% identity with that protein in Burkholderia mallei,
31% identity with Corynebacterium efficiens, 24% iden-
tity with Aspergillus fumigatus , and 28% identity with
Schizosaccharomyces pombe.
Cloning and expression of M. smegmatis
trehalase
The 2019 bp MSMEG 4528 ORF was PCR amplified
from M. smegmatis mc
2
155 genomic DNA using the
oligonucleotide primers pET100 TOPO StreFP 5¢-
CA
CC ATG ATG TGC TGC ATG GTT CTG CAA CA
GA-3¢ and pET100 TOPO treFP 5¢-TGA GCG TCA
CAT CGG GGC GTT-3¢. The pET100 TOPO StreFP
includes the 4 bp sequence ‘
CACC’ (underlined nucleo-

tides) necessary for directional cloning on the 5¢-end.
The bold ‘ATG’ in the FP represents the start codon
and the bold ‘TCA’ in pET100 TOPO StreFP represents
the stop codon of the recombinant ORF. The PCR
product was amplified and ligated with the precut
pET100D-TOPO (Invitrogen), generating the plasmid
pSTRE TOPO. The entire cloned (His)
6
–Stre gene
fusion was sequenced to confirm the fidelity of the
amplification. pSTRE TOPO was transformed into the
Escherichia coli expression strain BL21 star (DE3).
pSTRE TOPO in BL21 star (DE3) was used for further
expression studies.
Properties of the trehalase purified from
M. smegmatis
Effect of time and protein concentration on the activity
and characterization of the product
The conversion of trehalose to glucose was measured
by determining the amount of reducing sugar resulting
from the hydrolysis of trehalose. The formation of
reducing sugar increased in a linear fashion with
increasing time of incubation for at least 1 h, and was
also linear with the amount of enzyme added up to
100 lg of protein (data not shown). These data estab-
lished that all measurements were being made in the
linear range of measurements.
Glucose was the only product identified, both at
early times of incubation and with longer incubations.
Glucose was identified by paper chromatography in

several different solvents that readily separate this
sugar from other monosaccharides, such as mannose
and galactose, and other disaccharides such as maltose,
trehalose and cellobiose. It was also identified by
HPLC on the Dionex Carbohydrate Analyzer, which
also readily separates the various monosaccharides.
The resulting d-glucose was also determined using the
glucose oxidase reagent kit, which is specific for d-glu-
cose. Measurements using glucose oxidase to determine
the amount of glucose released gave very similar values
to those obtained using the reducing sugar test to
measure the amount of glucose.
This trehalase shows an almost absolute requirement
for inorganic phosphate and Mg
2+
for activity (see
below), and therefore it seemed possible that the
enzyme might actually be a phosphorylase, rather than
a glucosyl hydrolase. Therefore, a variety of experi-
ments were done to determine whether any phosphor-
ylated intermediates were produced in this reaction.
Thus, the above assay mixtures were removed after
various times of incubation and were carefully ana-
lyzed for the presence of glucose 1-phosphate or
glucose 6-phosphate. To do this, incubation mixtures
were passed through columns of DE-52 or Dowex-1-
Cl
–
to bind any possible phosphorylated sugars,
and the columns were then eluted with ammonium

J. D. Carroll et al. Novel trehalase from Mycobacterium smegmatis
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1703
Fig. 2. CLUSTALW alignment of predicted amino acid sequences of M. smegmatis trehalase MSMEG 4528 and putative M. tuberculosis
Rv2402. Numbering refers to the individual sequences, rather than to the alignment. Conserved residues and strong and weak conservative
substitutions are indicated by ‘*’, ‘:’ and ‘.’, respectively. Gaps introduced by
CLUSTAL to optimize the alignment are indicated by ‘–’. Polypep-
tide fragments used to identify the M. smegmatis ORF are underlined.
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1704 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works
bicarbonate to remove such phosphorylated sugars.
These eluates were analyzed for the presence of sugar
phosphates. No evidence for the presence of glucose
phosphate could be obtained either in short-time incu-
bations, or in longer incubations.
As phosphorylases can also be assayed in the direc-
tion of synthesis, various incubations were also pre-
pared using either a-glucose 1-phosphate or b-glucose
1-phosphate plus free glucose. In this case, the assay
was set up to measure the formation of trehalose. All
of these assays were also negative for trehalose forma-
tion.
Requirements for enzyme activity
The purified trehalase demonstrated a requirement for
inorganic phosphate, which could also serve as the
buffer in the reaction. Therefore, the effects of a vari-
ety of buffers, all tested at 100 mm and pH 7.0, on the
trehalase activity were determined, in the presence and
absence of added sodium (or potassium) phosphate,
and also in the presence of sodium arsenate rather
than phosphate. The results of these experiments are

presented in Table 2. It can be seen that in the absence
of added potassium (or sodium) phosphate, none of
the other buffers (Hepes, Tris, acetate, citrate, borate,
Mops or Mes) were able to activate the trehalase, as
compared to control incubations with only phosphate
buffer at pH 7.0. However, when 100 mm phosphate
was added to any of these incubations, all of them
(except for those incubations containing citrate or Mes
buffer) gave the same amount of trehalase activity as
the incubation with phosphate buffer alone. The inab-
ility of citrate to act as a favorable buffer may be due
to its strong chelation activity, as it probably competes
favorably with the trehalase for the Mg
2+
also needed
for stimulation. Interestingly, arsenate was able to sub-
stitute for phosphate to some extent with some of
these buffers, but it could not replace phosphate when
either acetate, borate or Tris were used as the buffer.
The studies described here were done with M. smegma-
tis B11. Also shown in Table 2 and discussed in a later
section are comparative results with the trehalase parti-
ally purified from M. tuberculosis H37Rv.
As indicated above, the mycobacterial trehalase
showed an absolute requirement for inorganic phos-
phate (Fig. 3), with optimum activity being observed at
a concentration of 100 mm. Although phosphate could
also serve as the buffer for the reaction, the require-
ment for phosphate was independent of the buffer used,
as indicated in Table 2. Figure 3 demonstrates that

Table 2. Effects of various buffers on the activity of trehalase.
Buffer
Trehalase activity (A
540 nm
)
M. smegmatis
B11
M. tuberculosis
H37Rv
Potassium phosphate 1.67 0.84
Hepes 0.03 0.12
+ Potassium phosphate 1.65 0.82
+ Sodium arsenate 0.89 0.69
Tris ⁄ HCl 0.53 0.12
+ Potassium phosphate 1.59 0.27
+ Sodium arsenate 0.45 0.20
Acetate 0.45 0.04
+ Potassium phosphate 1.43 0.86
+ Sodium arsenate 0.13 0.01
Citrate 0.02 0.21
+ Potassium phosphate 0.04 0.90
+ Sodium arsenate 0.03 0.18
Borate 0.03 0.01
+ Sodium phosphate 1.78 0.80
+ Sodium arsenate 0.34 0.06
Mops 0.03 0.20
+ Potassium phosphate 1.49 0.90
+ Sodium arsenate 1.03 0.84
Mes 0.03 0.11
+ Potassium phosphate 0.98 0.87

+ Sodium arsenate 0.70 0.25
Fig. 3. Effect of inorganic phosphate and arsenate on the activity of
the mycobacterial trehalase. Incubations contained 100 m
M Hepes
buffer (pH 7.1), 50 m
M trehalose, 6 mM MgCl
2
, various amounts of
sodium phosphate (r – r) or sodium arsenate (m – m) and 15
units of purified trehalase, all in a final volume of 100 lL. After an
incubation of 6 min at 37 °C, reactions were stopped by heating in
a boiling water bath for 5 min, and the amount of glucose produced
was determined by the reducing sugar test, or by the glucose oxid-
ase assay method. One unit is defined as that amount of trehalase
that produces 1 nmol of glucose in 1 min at 37 °C.
J. D. Carroll et al. Novel trehalase from Mycobacterium smegmatis
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1705
arsenate could substitute for the phosphate requirement
to some extent, although it was not as effective as phos-
phate. However, it did give the same profile as phos-
phate at various arsenate concentrations, suggesting
that it had the same effect on the enzyme. The only
product produced from trehalose in the presence of
arsenate was also identified as glucose by the glucose
oxidase reaction and by HPLC on the Dionex Carbo-
hydrate Analyzer. Although arsenate was somewhat
effective in replacing phosphate as the activator of treh-
alase, in the presence of 100 mm phosphate increasing
concentrations of arsenate inhibited the reaction, and
at equal concentrations of phosphate and arsenate

(100 mm each), the production of glucose was inhibited
by 40%.
The enzyme also showed an absolute requirement
for Mg
2+
, with optimum activity occurring at concen-
trations of 3.5–4 mm (Fig. 4). Mg
2+
could not be
replaced by Ca
2+
,Mn
2+
or Zn
2+
(data not shown).
The pH optimum for trehalase when potassium phos-
phate was used as the buffer, at 6 mm MgCl
2
, was
found to be 7.1 (data not shown). The trehalase activ-
ity dropped sharply at pH values of 6.5 and below, as
well as at pH values of 8.0 and above. Many glycosyl
hydrolases, including a number of trehalases, have pH
optima at about 5.0–5.5. In addition, these enzymes
generally do not show any requirements for inorganic
phosphate, and most are not activated by metal ions
(Table 3).
Substrate specificity and concentration
for optimum activity

Several glucose disaccharides were tested as possible
substrates for the purified trehalase, including maltose
(4-O-a-d-glucopyranosyl-d-glucopyranoside), isomal-
tose (6-O-a-d-glucopyranosyl- d-glucopyranoside), sucrose
(O-b-d-fructofuranosyl-(2 fi 1)-a-d-glucopyranoside),
cellobiose (4-O-b-d-glucopyranosyl-d-glucopyranoside),
p-nitrophenyl-a-d-glucopyranoside, and methyl-a-d-
glucopyranoside. None of these compounds was hydro-
lyzed by the trehalase, even when added to incubation
mixtures at high concentrations (50 mm) (data not
shown). The trehalase also did not hydrolyze a,b-treha-
lose or b,b-trehalose, a,a-trehalose-6,6¢-dibehenate,
trehalulose or nigerose (3-O -a-d -glucopyranosyl-d-
glucopyranoside). Trehalase also was inactive on treha-
lose dimycolate, an important glycolipid found in the
cell wall of M. tuberculosis, and other mycobacteria [7].
In these cases, assays were done by determining the
release of free glucose with the glucose oxidase reagent
kit, and by HPLC on the Dionex Carbohydrate Ana-
lyzer. The enzyme also did not show any activity with
glucose 1-phosphate, glucose 6-phosphate, mannose
1-phosphate or trehalose 6-phosphate, either by glucose
oxidase assay or by HPLC (data not shown).
Several of the above sugars were also tested for
their ability to inhibit the activity of trehalase on
a,a-trehalose. Only methyl-a-d-glucopyranoside
showed any inhibitory effect, giving about 50% inhibi-
tion at a concentration of about 12 mm.
0
.02

.04
.06
.08
01 54
3287
6
C
at
i
c

nocnoe
n
t
r
a
ti
( nom
M)
Activity (A 620)
Fig. 4. Effect of Mg
2+
concentration on the activity of the M. smeg-
matis trehalase. Reaction mixtures were as described in the text,
with 100 m
M sodium phosphate buffer (pH 7.1), 50 mM trehalose,
5 units of purified trehalase, and increasing amounts of MgCl
2
as
indicated. After an incubation of 15 min, reactions were stopped by

heating, and the amount of glucose formed was determined as indi-
cated in Fig. 3.
Table 3. Comparison of trehalases from various organisms.
Organism Molecular mass (kDa) pH optimum K
m
(mM) Cation requirement
Mycobacterium smegmatis  1500 (71 kDa subunit) 7.1 20 Mg
2+
Fusarium oxysporum (acidic) 160 4.6 0.42 None
Fusarium oxysporum (neutral) 100 6.8 8.5 Ca
2+
Aspergillus fumigatus 100 4.0 2.5 –
Neurospora crassa 340 5.0 0.36 –
Saccharomyces cerevisiae 100 7.0 34 –
Escherichia coli 63 6.0 1.9 None
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1706 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works
The enzymatic activity increased with increasing
concentrations of a,a-trehalose in the incubations up
to about 50 mm (Fig. 5). A plot of this data by the
method of Lineweaver and Burk is shown in the inset,
and this plot indicated that the K
m
for trehalose was
about 20 mm.
Table 3 compares some of the properties of the
M. smegmatis trehalase to those of some of the other
trehalases that have been isolated and purified from
various other organisms, and at least partially charac-
terized. Most of these enzymes are from fungi or yeast,

and have molecular masses of about 100–150 kDa, as
compared to the mycobacterial enzyme, which appears
to be a multimer of 20 or so identical subunits. These
trehalases also vary in terms of pH optima, from aci-
dic trehalases with optima at 4.0–5.6 to neutral treha-
lases with pH optima of about 7, like the
mycobacterial enzyme. In a few cases, such as the
yeast neutral trehalase and the mycobacterial trehalase,
the K
m
for trehalose is quite high (34 and 20 mm,
respectively), but many of the K
m
values are in the low
millimolar range. Finally, only two of the trehalases
have a requirement for a divalent metal ion. The Fusa-
rium oxysporum neutral trehalase requires Ca
2+
, which
is apparently involved in stabilizing the protein, and
the M. smegmatis trehalase requires Mg
2+
, which
appears to be necessary for activity. However, only the
mycobacterial trehalases seem to also require phos-
phate in order to be active.
Effect of various phosphorylated compounds
on trehalase activity
Trehalase activity was inhibited by pyrophosphate and
polyphosphates. Figure 6 shows the effects of increas-

ing concentrations of potassium pyrophosphate on the
hydrolysis of trehalose (formation of glucose) at var-
ious concentrations of phosphate buffer. Thus, with
1mm pyrophosphate, there was little or no effect on
trehalase activity, even at lower concentrations of
phosphate buffer. However, it can be seen that at a
A
B
Fig. 5. Effect of substrate concentration on trehalase activity. Var-
ious amounts of trehalose were added to incubation mixtures con-
taining 100 m
M sodium phosphate buffer (pH 7.1), 6 mM MgCl
2
,
and 15 units of purified trehalase. After an incubation of 10 min at
37 °C, the amount of glucose produced was determined (A). These
data were plotted by the method of Lineweaver and Burke, as
shown in (B).
Fig. 6. Inhibition of trehalase activity by sodium pyrophosphate.
Incubation mixtures contained 100 m
M sodium phosphate buffer
(pH 7.1), 6 m
M MgCl
2
,50mM trehalose, 15 units of purified treh-
alase, and various amounts of sodium pyrophosphate as follows:
(r – r), no pyrophosphate; (h – h), 1 m
M sodium pyrophosphate;
(m – m), 4 m
M sodium pyrophosphate; ( · — · ), 8 mM sodium

pyrophosphate; (* – *), 16 m
M sodium pyrophosphate. After an
incubation of 10 min, reactions were stopped by heating, and the
amount of glucose produced was measured.
J. D. Carroll et al. Novel trehalase from Mycobacterium smegmatis
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1707
pyrophosphate concentration of 8 mm, trehalase was
inhibited by more than 50%, even at phosphate con-
centrations of 100 mm.At16mm pyrophosphate,
there was almost complete inhibition of the enzyme,
even at concentrations of phosphate that were 10-fold
higher.
Polyphosphates were also inhibitors of the trehalase,
and these compounds were better inhibitors than pyro-
phosphate. Four differently sized polyphosphates were
tested, varying in polymerization number from 8 to 60.
All of these polyphosphates were quite similar in inhib-
itory activity, causing 50% inhibition of trehalase
activity when added to incubations at about 100 lg
(approximately 150 nmol for a polymerization number
of 8). These incubations also contained 100 mm potas-
sium phosphate buffer. Other phosphorylated com-
pounds were also inhibitory to varying degrees. ATP
caused about 90% inhibition at a concentration of
20 mm, whereas this concentration of AMP caused
about 50% inhibition. UTP also caused about 90%
inhibition at a concentration of 10 mm. Sodium ortho-
vanadate, an inhibitor of phosphatases, also inhibited
the trehalase by 80% at a concentration of 10 mm, but
no inhibition was observed with sodium fluoride, even

at 40 mm.
Why does the trehalase require inorganic phosphate?
As shown above, the M. smegmatis B11 trehalase
requires inorganic phosphate and Mg
2+
in order to
catalyze trehalose hydrolysis. However, no evidence
for phosphorolytic activity could be demonstrated,
suggesting that phosphate is not directly involved in
the catalytic mechanism. Thus, the question arose as
to whether phosphate might affect the conformation
of the enzyme. In order to determine whether inor-
ganic phosphate bound to the enzyme, [
32
P]inorganic
phosphate was incubated with the purified trehalase
in the presence of Mg
2+
, but without trehalose, for
several minutes on ice, and the mixture was applied
to a Sephacryl S-300 column in the cold. The column
was eluted with buffer, and fractions of 2 mL were
collected and assayed for trehalase activity and for
radioactivity. As seen in Fig. 7, the trehalase emerged
in the void volume (fraction 31 ¼ 62 mL) of a Seph-
acryl S-300 column, and a peak of [
32
P]inorganic
phosphate emerged in the same area of the column
and with the same profile, suggesting that inorganic

phosphate was binding to the protein. Although this
experiment still does not identify a role for inorganic
phosphate in the mechanism of trehalase activity, it
suggests that phosphate may be involved in causing
or expediting the correct conformation of the treh-
alase, or in causing the aggregation of the trehalase
into its active form.
Stabilization of the trehalase by phosphate
and magnesium
One of the possible roles for phosphate in the activity
of the trehalase is to stabilize the enzyme and maintain
it in the stable and perhaps aggregated state. In order
to determine whether phosphate plays a role in stabil-
ity, trehalase was incubated at 50 °C for various times
in sodium phosphate buffer and ⁄ or Mg
2+
, plus either
trehalose or polyethyleneglycol, and the activity of the
enzyme was determined at different times of incuba-
tion. The results of this experiment are shown in
Fig. 8. It can be seen from the upper profile that opti-
mal stability, i.e. very little loss of trehalase activity at
50 °C for 30 min, occurred in the presence of 100 mm
phosphate buffer containing 6 mm MgCl
2
plus 10%
polyethylene glycol. Omitting the polyethylene glycol
or replacing it with NaCl, but still heating the enzyme
in the presence of phosphate buffer and Mg
2+

, resul-
ted in reasonable stability, but significantly less than
obtained with the above incubation with the polyethy-
lene glycol. On the other hand, in the presence of
100 mm phosphate buffer alone or phosphate buffer
with 50 mm trehalose (see profiles 1 and 3), the enzyme
lost most of its activity within 5 min. Incubations with
0
.04
.08
.12
.16
0252
0
3530
4
54
0
5
rFtcaionN umber
Activity (A540),
32
P ( CPM x 10
-3
)
bacde
Fig. 7. Binding of inorganic phosphate to the purified trehalase.
[
32
P]Inorganic phosphate was mixed with purified trehalase and

allowed to incubate for 5 min on ice. The mixture was then added
to a 1.6 · 110 cm column of Sephracryl S-300, and the column was
eluted with 20 m
M Tris ⁄ HCl buffer at pH 7.0. Fractions were collec-
ted and assayed for radioactivity (n – n) and for trehalase activity
(r–r). Arrows indicate the position of the void volume (arrow a)
and the molecular mass standards thyroglobulin (669 kDa, arrow b),
b-amylase (200 kDa, arrow c), alcohol dehydrogenase (150 kDa,
arrow d) and bovine serum albumin (66 kDa, arrow e).
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1708 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works
phosphate buffer and polyethylene glycol but without
Mg
2+
gave slightly more stability than phosphate
alone, but much less than incubations with Mg
2+
included. In addition, incubations in 100 mm Hepes
buffer (pH 7.1) or 100 mm Hepes buffer (pH 7.1) plus
6mm MgCl
2
did not provide any additional stability
over that seen in the incubations with the enzyme only
in phosphate buffer. These results indicate that both
phosphate buffer and Mg
2+
are necessary to stabilize
the trehalase against heat denaturation. There is a pre-
cedent for inorganic phosphate affecting the quarter-
nary structure of a protein and causing monomers to

aggregrate into dimers [16].
Inhibitors of the mycobacterial trehalase
Validoxylamine has been reported to be an inhibitor of
various trehalases [17,18]. We tested the effects of this
drug on the trehalase purified from M. smegmatis. Val-
idoxylamine also inhibited this trehalase, and as shown
in Fig. 9, this inhibition was of a competitive nature.
The K
i
for validoxylamine on the trehalase was calcu-
lated to be 5 · 10
)7
m.
Another known trehalase inhibitor is trehazolin
[18,19]. However, trehazolin had no effect on the
mycobacterial trehalase, indicating that this phosphate-
dependent trehalase is different from other trehalases.
The mycobacterial enzyme was also inhibited by the
a-glucosidase inhibitor castanospermine [20], which
caused about 50% inhibition at a concentration of
500 lgÆmL
)1
.
Isolation of recombinant trehalase from E. coli
The E. coli expression strain BL21 was transformed
with the plasmid pSTRE TOPO, as described in
Experimental procedures. The cells were grown in LB
medium containing 100 lgÆmL
)1
ampicillin. Incubation

of the cells with isopropyl thio-b-d-galactoside resulted
in the production of substantial amounts of protein,
but the expressed trehalase protein was associated with
the membrane fraction after centrifugation of the soni-
cated cells and was presumably in inclusion bodies.
This protein could be solubilized by the addition of
Fig. 8. Effect of phosphate and Mg
2+
on the heat stability of the
purified trehalase. Trehalase was incubated for various times at
50 °C under the following conditions: 1 (h – h), 100 m
M sodium
phosphate buffer (pH 7.0); 2 (j – j), 100 m
M sodium phosphate
buffer (pH 7.0) + 10% polyethylene glycol; 3 (n – n), 100 m
M
sodium phosphate buffer (pH 7.0) + 50 mM trehalose; 4 (m – m),
100 m
M sodium phosphate buffer (pH 7.0) + 50 mM treha-
lose + 10% polyethylene glycol; 5 (e – e), 100 m
M sodium phos-
phate buffer (pH 7.0) + 6 m
M MgCl
2
;6(r – r), 100 mM sodium
phosphate buffer (pH 7.0) + 6 m
M MgCl
2
+ 10% polyethylene gly-
col; 7 ( · — · ), 100 m

M sodium phosphate buffer (pH 7.0) + 6 mM
MgCl
2
+ 500 mM NaCl; 8 (– –) 100 mM Hepes buffer (pH 7.0); 9
(* – *), 100 m
M Hepes buffer (pH 7.0) + 6 mM MgCl
2
. An equal ali-
quot of each incubation mixture was removed at the times indica-
ted, and assayed for its ability to catalyze the formation of glucose
from trehalose.
Fig. 9. Inhibition of trehalase by validoxylamine. Incubation mixtures
were prepared as described in the legends to other figures, and
contained 100 m
M sodium phosphate buffer (pH 7.1), 6 mM MgCl
2
,
and 50 m
M trehalose, all in a final volume of 100 lL. Various
amounts of validoxylamine, from 0 to 100 ng, were then added to
these incubations, and the reactions were initiated by adding
10 units of purified trehalase to each assay mixture. Tubes were
incubated for 15 min at 37 °C, and the amount of glucose produced
was determined by the reducing sugar test.
J. D. Carroll et al. Novel trehalase from Mycobacterium smegmatis
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1709
0.5% sarkosyl with 1 mm EDTA to the crude sonicate
before centrifugation. The expressed protein containing
a (His)
6

tag was purified on a nickel column. The elut-
ed fraction showed a single protein with a molecular
mass of about 70 kDa, but we were unable to find any
trehalase activity in the solubilized fraction or in the
purified fraction from the column.
Partial characterization of the trehalase from
M. tuberculosis
The M. tuberculosis trehalase was isolated from cytoso-
lic extracts and partially purified by ammonium sulfate
fractionation and chromatography on a column of
Sephracryl S-300. This partially purified enzyme also
exhibited an absolute requirement for inorganic phos-
phate, as seen in Table 2. However, this enzyme does
not appear to require Mg
2+
for activity, and as shown
in Table 2, this enzyme showed very good activity in
citrate buffer that contained added inorganic phos-
phate. On the other hand, the enzyme purified from
M. smegmatis was not active in citrate even when
phosphate was present. It seems likely that this inhibi-
tion of activity of the M. smegmatis enzyme is due to
the lack of Mg
2+
, as it is probably chelated by the cit-
ric acid. The gene for trehalase in M. tuberculosis has
over 85% identity at the amino acid level to the
M. smegmatis trehalase gene (Fig. 2). Interestingly, the
M. tuberculosis trehalase is not active in Tris buffer,
even when 100 mm phosphate is added (Table 2),

whereas the M. smegmatis trehalase is active in Tris
buffer with added phosphate. This may represent
another difference between these two trehalases. Other
trehalases have been reported to be inhibited by Tris
buffer [21].
Confirmation of M. smegmatis MSMEG 4528
as a trehalase
The identity of MSMEG 4528 as a trehalase was
confirmed by ligating the ORF into the mycobacterial
acetamide-inducible expression vector pSD24 [22]. The
resulting chimeric ORF contained the first codons of the
mycobacterial acetamidase gene amiE fused in frame
with the MSMEG 4528 coding sequence. The entire
p
ami
-MSMEG4528 cassette was excised from pSD24
and inserted into the single-copy integrating shuttle
plasmid pMV306, generating the plasmid p996A661.
This was electroporated into M. smegmatis mc
2
155, and
kanamycin-resistant colonies were recovered. These
transformants contained a stable integrated single
copy of the acetamide-inducible M. smegmatis treh-
alase. A representative transformant and M. smegmatis
containing nonrecombinant pMV306 were cultured with
varying concentrations of acetamide as previously des-
cribed [22]. The cultures were then harvested and
assayed for trehalase activity. The transformant strain,
M. smegmatis p996A661, exhibited an acetamide-

dependent increase in trehalase activity as shown in
Table 4, whereas the strain containing pMV306 was
unaffected by the addition of acetamide.
Discussion
The trehalase described in this report is unusual as far
as glycosyl hydrolases are concerned, as it requires
inorganic phosphate and Mg
2+
for activity. Trehalas-
es and other glycosyl hydrolases in glycoside hydrol-
ase family 65 or glycoside family 15 do not have a
requirement for inorganic phosphate for activity,
unless they are phosphorylases. The reason for this
requirement by the mycobacterial trehalase is still not
known. Several experiments were done to determine
whether a phosphorylated sugar intermediate was
involved in the reaction, and ⁄ or whether the product
of the reaction was actually glucose 1-phosphate or
glucose 6-phosphate. All of these experiments gave
negative results. However, an experiment in which
radioactive inorganic phosphate was incubated briefly
with purified trehalase did suggest that the phosphate
was bound to the enzyme, as active trehalase itself
emerges in the void volume of a Sephracryl S-300 col-
umn, and the mixture of radioactive phosphate and
purified enzyme also emerged in the void volume area,
both being eluted with the same profile. This experi-
ment, and several experiments showing that the treh-
alase was most stable to heating (i.e. up to 30 min at
50 °C) when it was in 100 mm phosphate buffer con-

taining 6 mm MgCl
2
plus 10% polyethylene glycol,
suggest a role for phosphate as a stabilizer, and per-
haps as an effector of an aggregated and active treh-
alase conformation.
Incubation of the trehalase in other buffers in the
absence of inorganic phosphate or MgCl
2
resulted in
substantial loss of activity in 2 min, at 40 °C or higher.
Thus it seems that at least one function of the phos-
Table 4. MSMEG 4528 codes for trehalase activity.
Acetamide
(m
M)
Trehalase activity (nmol Glc per mg protein)
M. smegmatis (p996A661) M. smegmatis (pMV306)
062 45
1 122 35
5 313 37
20 457 38
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1710 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works
phate ⁄ Mg
2+
mixture may be to stabilize the enzyme.
This stabilization may involve expediting the assembly
of the 71 kDa subunits into a multimer, which appears
to be the active form of this enzyme.

A precedent for the hypothesis that phosphate
expedites aggregation is provided by a study on the
activation and dimer formation of rat renal phosphate-
dependent glutaminase [16]. In that case, inorganic
phosphate activated the glutaminase and converted it
from a monomer to a dimer. These workers found a
strong correlation between dimer formation and acti-
vation of activity, and concluded that glutaminase is
only active as a dimer, or a larger aggregate. The
results shown here suggest that the trehalase is also
activated and aggregated by phosphate in a similar
manner.
Interestingly, the trehalase partially characterized
from M. tuberculosis also requires inorganic phosphate
for activity. However, with the partially purified
enzyme, we could not demonstrate any requirement
for Mg
2+
. It could be that the metal ion is bound to
the trehalase in this case, but the addition of EDTA to
these reaction mixtures did not inhibit the activity, and
nor did the addition of citrate buffer to the assay mix-
tures. We will have to purify this enzyme to homogen-
eity in order to determine whether it does have a
requirement for Mg
2+
. The gene for the trehalase from
M. tuberculosis shows over 85% identity at the amino
acid level to the gene from M. smegmatis, and both
proteins appear to have similar activities. There is also

over 85% sequence identity to the hypothetical protein
from M. avium. However, the aggregated form of the
tuberculosis trehalase emerges from a Sephacryl S-300
column later than the enzyme from M. smegmatis, sug-
gesting that it is not as highly aggregated. Again, it
will be necessary to obtain pure protein before we can
verify these initial observations. Although the myco-
bacterial trehalases have only moderate similarity (less
that 35%) to the many other microbial trehalases (i.e.
from E. coli or Sa. cerevisiae or various fungi), they do
show over 70% identity to an unidentified protein
found in Nocardia sp.
We were able to clone the M. smegmatis gene for
the trehalase, and to transfect E. coli with this gene.
The transfected cells did produce a protein that migra-
ted as a 71 kDa protein on SDS gels. However, this
protein was not active, and all attempts to produce
active trehalase have been unsuccessful. We also made
numerous attempts to knock out the gene for this
trehalase in M. smegmatis. Despite the fact that we
have been able to successfully knock out the genes for
the trehalose phosphate synthase, the trehalose phos-
phate phosphatase, the trehalose synthase and the
maltooligosyl trehalose synthase, this technology has
not been successful with the trehalase. We have been
unable to construct a trehalase mutant. These results
may suggest that the trehalase is an essential enzyme
for the growth or survival of this organism.
The predicted size difference between M. smegmatis
MSMEG 4528 and M. tuberculosis RV2402 is due to

30 N-terminal amino acid residues in MSMEG 4528.
The 31st residue, a valine specified by a GTG codon,
corresponds to the N-terminal methionine residue of
Rv2402. It is possible that Val31 represents the actual
translation start of the M. smegmatis trehalase. In this
event, the protein would have a predicted molecular
mass of 72 kDa rather than the 75.2 kDa of the pre-
dicted product of the full-length ORF. The smaller
value is more consistent with the observed molecular
mass of 71 kDa.
The trehalase described here is highly specific for
a,a-1,1-trehalose, and will not hydrolyze any of the
other trehalose isomers (i.e. a,b-trehalose or b,b-treha-
lose), or a variety of a-orb-linked disaccharides
(maltose, isomaltose, cellobiose, sucrose) or other gly-
cosides (a-methylglucoside). As indicated above, the
enzyme is unusual in its requirement for inorganic
phosphate, and it has not been possible to obtain a
mutant that is lacking this enzyme, suggesting that it
may be essential. However, we do not know the role
of this enzyme in the homeostasis of mycobacteria.
M. smegmatis, unlike M. tuberculosis, can utilize treha-
lose as a sole carbon source (data not shown). We spe-
culate that trehalase plays a role in trehalose
utilization, although this does not explain the presence
of trehalase in M. tuberculosis. It is possible that
M. tuberculosis does not contain a specific trehalose
uptake pathway, and that the trehalase is redundant.
As the M. tuberculosis genome is significantly smaller
than that of M. smegmatis (4.4 Mbp versus 6.98 Mbp),

it is possible that M. tuberculosis has dispensed with
genetic information not relevant to its specialized
intracellular lifestyle.
The inhibition of the M. smegmatis trehalase by
polyphosphates suggests an intriguing possibility for
regulation of trehalose metabolism. In many cells,
polyphosphates appear to protect cells from adverse
conditions such as heat or desiccation [23], but the
mechanism of this protection is not understood at this
time. Polyphosphates are present in various mycobac-
teria, and play an important role in metabolism and
cell function [24]. Trehalose has also been shown to
protect many different types of living cells from a vari-
ety of adverse conditions [25], and also to protect iso-
lated proteins and membranes from heat, freezing or
desiccation [26]. Is it possible that the protective effect
J. D. Carroll et al. Novel trehalase from Mycobacterium smegmatis
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1711
of polyphosphates is due to inhibition of the mycobac-
terial trehalase, which might then result in an increase
in the intracellular content of free trehalose.
Experimental procedures
Bacterial strains and culture conditions
Mycobacterium smegmatis mc
2
155 was provided by WR
Jacobs Jr, Albert Einstein College of Medicine, New York.
M. smegmatis B11 is a spontaneous, streptomycin-resistant
derivative of mc
2

155. M. tuberculosis H37Rv was generously
supplied by K Eisenach, University of Arkansas for Medical
Sciences. The E. coli strains TOP10 and BL21Star (DE3)
(Invitrogen, Carlsbad, CA) were used for cloning and
expression studies, respectively. E. coli strains were cultured
in LB broth and on LB agar supplemented with 100 lgÆmL
)1
ampicillin, 20 lgÆmL
)1
kanamycin or 10 l gÆmL
)1
tetracyc-
line, individually or in combination when applicable.
M. smegmatis was cultured in Trypticase Soy broth and
maintained on slants of Trypticase Soy agar. In some experi-
ments, this organism, as well as M. tuberculosis H37Rv, was
also cultured in Middlebrook 7H9 broth supplemented with
10% (v ⁄ v) oleic acid ⁄ albumin ⁄ dextrose complex. For some
studies, M. smegmatis was also grown in a mineral salts
medium containing the following: glutamic acid (1% w ⁄ v);
citric acid (0.2%); KH
2
PO
4
(0.05%); MgSO
4
.7H
2
O (0.05%);
ferric ammonium citrate (0.005%); ZnSO

4
(0.002%); and a
carbohydrate source (glucose or trehalose, 1–4% w ⁄ v). The
pH of the medium was adjusted to 7 with KOH before auto-
claving.
Reagents and materials
Trehalose, trehalose 6-phosphate, maltose, sucrose, a- and
b-glucose 1-phosphate, mannose 1-phosphate, other sugars
and sugar phosphates, DEAE-cellulose, amino-hexose-agar-
ose, phenyl-Sepharose, various other chromatographic res-
ins, glucose oxidase ⁄ peroxidase assay kits, molecular mass
markers for gel filtration, buffers and other biochemicals
were all obtained from Sigma Chemical Co., St Louis, MO.
Bio-Rad protein reagent, DE-52, hydroxyapatite and all
electrophoresis reagents were obtained from Bio-Rad (Her-
cules, CA). Trypticase Soy broth was obtained from Becton
Dickinson (Franklin Lakes, NJ), and LB broth was
obtained from Fisher Scientific Co. (Pittsburgh, PA) Seph-
acryl S-300 and Sephacryl S-200 were obtained from
Amersham Pharmacia Biotech Inc. (Piscataway, NJ).
[
3
H]Trehalose was prepared by incubating UDP-[
3
H]glucose
plus glucose 6-phosphate with the purified mycobacterial
trehalose phosphate synthase as described previously [27].
The radioactive trehalose phosphate was isolated by ion
exchange chromatography, and was then treated with puri-
fied trehalose phosphate phosphatase [28] to obtain free

radioactive trehalose. [
32
P]Inorganic phosphate
(10 mCiÆmL
)1
) was purchased from American Radiolabeled
Chemicals, Inc. (St Louis, MO).
Assay of trehalase activity
The enzymatic activity of the trehalase was measured by
following the release of glucose, either by the reducing
sugar method [29], or with the glucose oxidase test kit.
Assay mixtures contained the following components in a
final volume of 100 lL: 100 mm phosphate buffer (pH 7.1),
6mm MgCl
2
,50mm trehalose, and various amounts of
trehalase. Incubations were at 37 °C for various times, and
incubations were stopped by heating in a boiling water
bath. When other sugars were tested as substrates, they
were added at various concentrations up to 50 mm. Other
additions to the reaction mixtures were made as described
in Results.
As the trehalase required inorganic phosphate for enzy-
matic activity, it was important to show that the observed
activity was due to a glucosyl hydrolase rather than to a
phosphorylase. Thus, incubation mixtures were prepared
with either 5 mm a-glucose 1-phosphate or b-glucose
1-phosphate plus 5 mm glucose in Hepes buffer at pH 7.0
or pH 6.0. Following an incubation of 15–60 min, reaction
mixtures were passed through a column of DE-52 (HCO

3
–
)
to bind any possible phosphorylated sugars. The neutral
fractions from these columns were assayed on the Dionex
(Bannockburn, IL) Carbohydrate Analyzer for the presence
of trehalose. Some incubations were also prepared with
50 mm trehalose and 50 mm [
32
P]inorganic phosphate, plus
5mm MgCl
2
. After incubation, these reaction mixtures
were applied to cellulose thin layer plates, which were sub-
jected to ascending chromatography in methanol ⁄ formic
acid ⁄ H
2
O [80 : 15 : 5]. In this solvent, inorganic phosphate
migrates near the solvent front, whereas sugar phosphates
migrate at an R
f
of about 0.5.
Purification of trehalase
Mycobacterium smegmatis was grown in 2 L Erlenmeyer
flasks containing 1 L of Trypticase Soy broth. Cells were
harvested by centrifugation (5500 g, Beckman J-25 centri-
fuge/JLA 16.250 rotor), washed with NaCl ⁄ P
i
, and stored as
a paste in aluminum foil at ) 20 °C until used. One hundred

grams of cell paste was suspended in 500 mL of ice-cold
20 mm sodium phosphate buffer (pH 7.5), and cells were
disrupted by sonic oscillation. Cell walls and membranes
were removed by high-speed centrifugation (100 000 g,
Beckman L7-65 centrifuge/50.2 Ti rotor), and the superna-
tant liquid was designated the crude extract. All purification
steps were done at 4 °C unless otherwise specified. The
supernatant liquid was placed in an ice bath on a magnetic
stirrer, and solid ammonium sulfate was added slowly to
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1712 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works
30% saturation. The resulting precipitate was removed by
centrifugation (12 000 g, Beckman J-25 centrifuge/JA-20
rotor) and discarded. The supernatant liquid was then
brought to 60% saturation by the slow and continual addi-
tion of solid ammonium sulfate. The resulting precipitate
was isolated by centrifugation (12 000 g, Beckman J-25 cen-
trifuge/JA-20 rotor), dissolved in 20 mL of 20 mm sodium
phosphate buffer (pH 7.5), and dialyzed overnight against
the same buffer.
The dialyzed enzyme was applied to a column of DE-52
that had been equilibrated with 20 mm phosphate buffer
(pH 7.5). The column was washed well with 20 mm phos-
phate buffer, and the wash, which did contain some trehalase
activity, was discarded. Trehalase was eluted from the
column with a 0–0.5 m linear gradient of KCl in 20 mm phos-
phate buffer. Fractions were collected, and every other frac-
tion was assayed for trehalase activity. This trehalase activity
was purified to apparent homogeneity as described below.
Active fractions eluted from the DE-52 column were pooled,

adjusted to 1 m KCl, and applied to a column of phenyl-
Sepharose that had been equilibrated with 1 m KCl in phos-
phate buffer. The enzyme was eluted with a reverse linear
gradient of 1–0 m KCl in phosphate buffer. Active fractions
were pooled and concentrated using an Amicon (Millipore,
Billarica, MA) filtration apparatus.
The concentrated enzyme was then applied to a Seph-
acryl S-300 column, and the column was washed with
20 mm phosphate buffer (pH 7.5). Active fractions were
pooled and concentrated using the Amicon filtration appar-
atus. This fraction was then subjected to chromatofocusing
as follows. The concentrated enzyme was applied to a
column of Polybuffer exchange (PBE94) that had been
equilibrated with 25 mm histidine ⁄ HCl buffer (pH 6.0). The
enzyme was eluted from the column with Polybuff-
er 74 ⁄ HCl (pH 4.0). The enzyme emerged from the column
between pH 4.4 and 4.1. This fraction showed one major
protein band of about 71 kDa, which was sequenced and
identified in the M. smegmatis genome.
Analysis of sugars
Sugars were identified and quantified using high-perform-
ance anion-exchange chromatography on the Dionex Car-
bohydrate Analyzer. Samples were injected into a
CarboPacPA-1 column equilibrated with a mixture of water
(98 parts) and 400 nm NaOH (two parts). Sugars were elut-
ed with an increasing gradient of NaOH, and were detected
by pulsed amperometry as recommended by the manufac-
turer (Dionex, Technical Note 20).
Other methods
Protein was measured with the Bio-Rad protein reagent,

using BSA as the standard. The molecular mass of the treh-
alase was estimated by gel filtration on Sephacryl S-300.
Molecular mass standards included thyroglobulin
(669 kDa), b-amylase (200 kDa), alcohol dehydrogenase
(150 kDa) and BSA (66 kDa). SDS ⁄ PAGE was performed
according to Laemmli in 10% polyacrylamide gel [30]. The
gels were stained with 0.25% Coomassie Blue in 10% acetic
acid and 50% methanol. DNA manipulations were per-
formed using standard techniques. Requisite enzymes were
supplied by New England Biolabs (Beverly, MA). DNA
sequence analysis was performed in-house.
Sequence analysis
Purified trehalase (Fig. 1, lane 6) was eluted from the gel
and subjected to trypsin digestion and amino acid analysis
using Q-TOF MS to determine amino acid compositions of
the various peptides. The data from these peptides was used
to locate the ORF coding for trehalase in the M. smegmatis
genome. ORFs were identified by blastp alignment with
predicted amino acid sequences on GenBank. Multiple
amino acid alignments were performed using the clustalw
alignment program at a website maintained by the Euro-
pean Bioinformatics Institute (EMBL-EBI; http://www.
ebi.ac.uk/clustalw/). Basic sequence analysis, including iden-
tification of restriction sites, translations, and DNA
sequence alignment, were performed using the gene-jockey
program (Biosoft, Cambridge, UK).
Acknowledgements
J. D. Carroll and Y. T. Pan are supported by a
Research Grant (RG-9639-N) from the American
Lung Association.

References
1 Elbein AD (1974) The metabolism of a,a-trehalose. Adv
Carbohyd Chem Biochem 30, 227–256.
2 Sussman AS & Lingappa BT (1959) Role of trehalose in
ascospores of Neurospora tetrasperma. Science 130,
1343–1344.
3 Arguelles JC (2000) Physiological roles of trehalose in
bacteria and yeast: a comparative analysis. Arch Micro-
biol 174, 217–224.
4 Crowe JH, Crowe LM & Chapman D (1984) Preserva-
tion of membranes in anhydrobiotic organisms: the role
of trehalose. Science 223, 701–703.
5 Thevelein JM & Hohmann S (1995) Trehalose synthase:
guard to the gate of glycolysis on yeast. Trends Biochem
Sci 20, 3–10.
6 Burklen L, Schock F & Dahl MK (1998) Analysis of
the interaction between the Bacillus trehalose repressor
TreR and the tre operator. Mol Gen Genet 260, 48–55.
7 Brennan PJ & Nikaido H (1995) The envelope of myco-
bacteria. Annu Rev Biochem 64, 29–63.
J. D. Carroll et al. Novel trehalase from Mycobacterium smegmatis
FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works 1713
8 DeSmet KAL, Weston A, Brown IN, Young DB &
Robertson BD (2000) Three pathways for trehalose
biosynthesis in mycobacteria. Microbiology 146,
199–208.
9 Elbein AD, Pan YT, Pastuszak I & Carroll JD (2003)
New insights on trehalose: a multifunctional molecule.
Glycobiology 13, 17R–27R.
10 Murphy HN, Stewart GR, Mischenko VV, Apt AS,

Harris R, McAlister MS, Driscoll PC, Young DB &
Robertson BD (2005) The OtsAB pathway is essential
for trehalose biosynthesis in Mycobacterium tuberculosis.
J Biol Chem 280, 14524–14529.
11 Woodruff PJ, Carlson BL, Siridechadilok B, Pratt MR,
Senaratne RH, Mougous JD, Riley LW, Williams SJ &
Bertozzi CR (2004) Trehalose is required for growth of
Mycobacterium smegmatis. J Biol Chem 279, 28835–
28843.
12 Thevelein JM (1984) Regulation of trehalose mobiliza-
tion in fungi. Microbiol Rev 48, 42–59.
13 Davies G & Henrissat B (1995) Structure and mechan-
isms of glycosyl hydrolases. Structure 3, 853–859.
14 Nwaka S & Holzer H (1998) Molecular biology of tre-
halose and trehalases in the yeast, Saccharomyces cerevi-
siae. Prog Nucleic Acids Res Mol Biol 58, 197–237.
15 Uhland K, Mondigler M, Spiess C, Prinz W &
Ehrmann M (2000) Determinants of translocation
and folding of TreF, a trehalase of Escherichia coli.
J Biol Chem 275, 23439–23445.
16 Godfrey S, Kuhlenschmidt T & Curthoys NP (1977)
Correlation between activation and dimer formation of
rat renal phosphate-dependent glutaminase. J Biol Chem
252, 1927–1931.
17 Kameda K, Asano N, Yamaguchi T & Matsui K (1986)
Validoxylamines as trehalase inhibitors. J Antibiotics 40,
563–565.
18 Asano N (2003) Glycosidase inhibitors: update
and perspectives on practical use. Glycobiology 13,
93R–104R.

19 Ando V, Satake H, Ito K, Sato A, Nakajima M,
Takahashi S, Haruyama H, Ohkuma Y, Kinoshita T &
Enokita R (1991) Trehazolin, a new trehalase inhibitor.
J Antibiotics 44, 1165–1168.
20 Pan YT, Hori H, Saul R, Sanford BA, Molyneux RJ &
Elbein AD (1983) Castanospermine inhibits the
processing of the oligosaccharide portion of
the influenza viral hemagglutinin. Biochemistry 22 ,
3975–3983.
21 Terra WR, Ferreira C & de Bianchi AG (1978) Physical
properties and Tris inhibition of an insect trehalase and
a thermodynamic approach to the nature of its active
site. Biochim Biophys Acta 524, 131–141.
22 Daugelat S, Kowall J, Mattow J, Bumann D, Winter R,
Hurwitz R & Kaufmann SHE (2003) The RD1 proteins
of Mycobacterium tuberculosis: expression in Mycobac-
terium smegmatis and biochemical characterization.
Microbes Infect 5, 1082–1095.
23 Kornberg A, Rao NN & Ault-Riche D (1999) Inorganic
polyphosphates: a molecule with many functions. Annu
Rev Biochem 69, 89–125.
24 Szymona O (1974) Studies on inorganic polyphosphate
metabolism in
Mycobacterium phlei. II. Direct
utilization of inorganic polyphosphates for synthesis
of nucleic acids in M. phlei. Acta Microbiol Pol A 6,
301–310.
25 Singer MA & Lindquist S (1998) Thermotolerance in
Saccharomyces cerevisiae: the yin and yang of trehalose.
TIB Tech 16, 460–468.

26 Crowe JH, Hoekstra FA & Crowe LM (1992) Anhydro-
biosis. Annu Rev Physiol 54, 579–599.
27 Pan YT, Carroll JD & Elbein AD (2002) Trehalose-
phosphate synthase of Mycobacterium smegmatis. Clon-
ing, expression and properties of recombinant enzyme.
Eur J Biochem 269, 6091–6100.
28 Klutts S, Pastuszak I, Korath Edavana V, Thampi P,
Pan YT, Abraham EC, Carroll JD & Elbein AD (2003)
Purification, cloning, expression and properties of myco-
bacterial trehalose-phosphate phosphatase. J Biol Chem
278, 2093–2100.
29 Nelson N (1944) A photometric adaptation of the Som-
ogyi method for the determination of glucose. J Biol
Chem 153, 375–380.
30 Laemmli UK (1970) Cleavage of structural proteins
during the assembly of the head of bacteriophage T4.
Nature 227, 680–685.
Novel trehalase from Mycobacterium smegmatis J. D. Carroll et al.
1714 FEBS Journal 274 (2007) 1701–1714 ª 2007 FEBS No claim to original US government works