Inhibition of glutaminyl cyclase alters pyroglutamate
formation in mammalian cells
Holger Cynisa, Stephan Schillinga, Mandy Bodnára, Torsten Hoffmanna, Ulrich Heisera,
Takaomi C. Saidob, Hans-Ulrich Demutha,⁎
aProbiodrug AG, Weinbergweg 22, 06120 Halle/Saale, Germany
bRIKEN Brain Science Institute, Laboratory for Proteolytic Neuroscience, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan
Received 26 May 2006; received in revised form 29 July 2006; accepted 11 August 2006
Available online 16 August 2006
Mammalian cell lines were examined concerning their Glutaminyl Cyclase (QC) activity using a HPLC method. The enzyme activity was
suppressed by a QC specific inhibitor in all homogenates. Aim of the study was to prove whether inhibition of QC modifies the posttranslational
maturation of N-glutamine and N-glutamate peptide substrates. Therefore, the impact of QC-inhibition on amino-terminal pyroglutamate (pGlu)
formation of the modified amyloid peptides Aβ(N3E-42) and Aβ(N3Q-42) was investigated. These amyloid-β peptides were expressed as fusion
proteins with either the pre–pro sequence of TRH, to be released by a prohormone convertase, or as engineered amyloid precursor protein for
subsequent liberation of Aβ(N3Q-42) after β- and γ-secretase cleavage during posttranslational processing. Inhibition of QC leads in both
expression systems to significantly reduced pGlu-formation of differently processed Aβ-peptides. This reveals the importance of QC-activity
during cellular maturation of pGlu-containing peptides. Thus, QC-inhibition should impact bioactivity, stability or even toxicity of pyroglutamyl
peptides preventing glutamine and glutamate cyclization.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Glutaminyl Cyclase; Pyroglutamic acid; Inhibition; Amyloid-β; Alzheimer's disease
The formation of N-terminal pyroglutamic acid (pGlu) is a
posttranslational modification of several hormones such as
gastrin, neurotensin and GnRH . For some of these peptides,
e.g. thyrotropin-releasing hormone (TRH), it has been shown
that this pGlu-modification is crucial for hormonal activity .
Glutaminyl cyclase (QC; EC 126.96.36.199) is a zinc-dependent
metalloenzyme catalyzing the cyclization of amino-terminal
glutamine into pGlu under concomitant liberation of ammonia.
QCs have been identified in a number of animals and plants [3–
6]. Because of its broad substrate specificity, QC has a key
function in posttranslational pyroglutamyl formation of pre-
sumably all pGlu-containing peptides and hormones . QC is
expressed in various tissues of the body with a marked
abundance in different brain regions. The highest expression
was observed in striatum and anterior pituitary . More
detailed studies on the cellular and sub-cellular distribution in
porcine and bovine hypothalamic and pituitary tissue (detection
of QC-immunoreactivity on secretory granules of axonal nerve
endings belonging to the tractus hypothalamo hypophysalis)
revealed the striking evidence, that QC is transported via the
same routes as its substrates, e.g. the hormone precursors of
GnRH and TRH .
Recently, it has been shown that QC is also capable of
converting amino-terminal glutamate to pyroglutamate .
Therefore, QC might also play a role in amyloidotic diseases,
e.g. Alzheimer's Disease (AD), Familial British Dementia
(FBD) or Familial Danish Dementia (FDD), because significant
amounts of the deposited peptides (Aβ, ABri, ADan) are N-
Biochimica et Biophysica Acta 1764 (2006) 1618–1625
Abbreviations: Aβ, amyloid β-peptide; APP, amyloid precursor protein; β-
CTF, β-C-terminal fragment of APP; DMEM, Dulbecco's modified eagle
medium; GnRH, gonadotropin releasing hormone; HRP, horseradish perox-
idase; pGlu, pyroglutamic acid; QC, glutaminyl cyclase; SDS-PAGE, sodium
dodecylsulfate polyacrylamide gel electrophoresis; TMB, tetramethylbenzidine;
mTRH, murine thyrotropin releasing hormone
⁎Corresponding author. Tel.: +49 345 5559900; fax: +49 345 5559901.
E-mail address: Hans-Ulrich.Demuth@probiodrug.de (H.-U. Demuth).
1570-9639/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
terminally modified by a pyroglutamyl residue resulting from
glutamate cyclization [11–13]. These pGlu-containing peptide
species have been suggested to be a potential target for the
development of a treatment strategy due to their pronounced
neurotoxicity, stability and aggregation propensity [12,14–16].
After demonstrating QC-catalyzed N-glutamate peptide cycli-
zation in vitro , the aim of our present study was to show
that QC-inhibitors prevent the pyroglutamate formation in
cultured mammalian cells. Hence, we have screened a number
of established cell lines in culture for QC-activity using a HPLC
method. Having successfully identified several QC-containing
cell lines, we were interested in suppressing QC-activity using a
recently characterized highly potent QC inhibitor P150/03 .
Because of the lack of specific antibodies to analyze the
prevention of pGlu formation at the N-terminus of a variety of
potential QC peptide substrates, we decided to test P150/03
using engineered Aβ(N3E-42) and Aβ(N3Q-42) because of the
availability of specific sandwich ELISAs for Aβ(N3pGlu-42).
2. Materials and methods
2.1. Reverse-transcription PCR
Total RNA was isolated from HEK293 and β-TC 3 cells using the
Nucleospin Kit (Macherey-Nagel) and reversely transcribed by SuperScript II
(Invitrogen). Subsequently, QC was amplified on a 1:12,5 dilution of generated
cDNA product in a 25μl reaction with Herculase Enhanced DNA-Polymerase
(Stratagene). The primer sequences for amplification of QC were: β-TC 3, 5′-
ATATGCATGCATGGCAGGCAGCGAAGACAAGC-3′ (mQC, sense) and 5′-
antisense); HEK293, 5′-CATGGCATGGATTTATTGG-3′ (hQC, sense) and
5′-GACGGTATCAGATGCAGAAC-3′ (hQC, antisense). The PCR products
were purified utilizing the Strataprep PCR Purification Kit (Stratagene) and
confirmed by sequencing.
Full-length cDNA of mTRH was isolated from primary cortical neurons
(generously provided by Dr. S. Rossner, Paul Flechsig Institute for Brain
Research, Leipzig, Germany) and subcloned into vector pPCR Script (Strata-
3′ (sense) and 5′-ATATTTACTCCTCCAGAGGTTCCCTG-3′ (antisense). The
pre–pro sequence was amplified applying the Primers: 5′-ATATAAGCTTATG-
CAGGGACCTTGGCTGATG-3′ (sense) and 5′-ATATGCATGCTGTGATC-
CAGGAATCTAAGG-3′ (antisense), subcloned into vector pPCR Script and
confirmed by sequencing.
2.2. Cloning procedure
The cDNA of human APP695 (NLQ) was generated as described elsewhere
. The construct contained Swedish KM595/596NL and London V642I
familial AD mutations, a deletion of amino acids D597, A598, and a point
mutation of E599Q. On basis of NLQ, the sequences of Aβ(N3E-42), and
Aβ(N3Q-42) were amplified and fused to the 3′ end of pre–pro sequence of
mTRH via an artificial SphI cleavage site. The sequence of mature human QC
was also fused to mTRH using the artificial SphI cleavage site. SphI site was
mutated to wild type to possess a lysine–arginine motif for prohormone
convertase cleavage using Primers 5′-TTAGATTCCTGGATCACAAAACGC
[C/G]AATTCCGACATGACTCA-3′ (sense) and 5′-TGAGTCATGTCG-
GAATT[C/G]GCGTTTTGTGATCCAGGAATCTAA-3′ (antisense) for
Aβ(N3E-42) and Aβ(N3Q-42) and 5′-GATTCCTGGATCACAAAACGCCAT-
CATCATCATCATCAT-3′ (sense) and 5′-ATGATGATGATGAT-
GATGGCGTTTTGTGATCCAGGAATC-3′ (antisense) for human QC fusion
protein. Aβ(N3E-42), Aβ(N3Q-42), human QC fusion proteins and cDNA of
NLQ were ligated into the HindIII/NotI-site of vector pcDNA 3.1 (Invitrogen),
confirmed by sequencing and isolated for cell culture purposes using the
EndoFree Plasmid Maxi Kit (Qiagen).
2.3. Cell culture and transfection
Human embryonal kidney cells HEK293, murine monocyte/macrophage
cell line RAW264.7, human glioma U343, murine insulinoma β-TC 3, murine
fibroblast cell line L929, human cervix carcinoma HeLa and human plasma cell
leukaemia cell line L-363 were cultured in appropriate cell culture media in a
humidified atmosphere of 5% CO2at 37 °C. HEK293 cells were transfected
with the APP695 (NLQ) and β-TC 3 cells were transfected with the mTRH-
Aβ(N3Q-42) construct using Lipofectamin2000 (Invitrogen) according to the
manufacturer's manual. Furthermore, β-TC 3 cells were also transfected with
the mTRH-Aβ(N3E-42) construct alone or in combination with the mTRH-QC
construct. Transiently transfected cells were grown over night and afterwards
incubated for 24h with phenolred-free DMEM (Gibco) under serum-free
conditions eitherin presenceor absenceof QC inhibitor1-(3-(1H-Imidazol-1-yl)
propyl)-3-(3,4-dimethoxyphenyl)thiourea (P150/03)  in a concentration of
100μM for β-TC 3 and 10μM for HEK293. The next day, media were collected,
readily mixed with protease inhibitor cocktail (Roche) containing additionally
1mM PefaBloc (Roth) and stored at −80 °C. Cell count of each well was
determined using the CASY cell counter system (Schaerfe System).
Cellswere grown to confluency, pelletized,frozen on dry ice andstored until
assay. Test samples were sonicated and centrifuged in buffer (10mM Tris,
100mM NaCl, 5mM EDTA, 0.5% Triton X-100, 10% Glycerol, pH 7.5) at
16.000×g for 30min and 4 °C. The protein concentration of the resulting
supernatant was determined using the method of Bradford and used as enzyme
source. The QC activity was determined applying an HPLC-assay essentially as
described elsewhere  since continuous assay methods were hampered by
aminopeptidase activities in the crude extracts. The assay is based on conversion
of H-Gln-βNA to pGlu-βNA. The sample consisted of 50μM H-Gln-βNA or
50μM H-Gln-βNA/10μM P150/03 in 25mM MOPS, pH 7.0, 0.1mM N-
Ethylmaleinimide (NEM) and enzyme solution in a final volume of 1ml.
Samples were incubated at 30 °C and constantly shaken at 300rpm in a
thermomixer (Eppendorf). Test samples were removed, and the reaction stopped
by boiling for 5min followed by centrifugation at 16.000×g for 10min. All
HPLC measurements were performed using a RP18 LiChroCART HPLC-
Cartridge and the HPLC system D-7000 (Merck-Hitachi). Briefly, 10μl of the
sample were injected and separated by increasing concentration of solvent A
(acetonitrile containing 0.1% TFA) from 8% to 20% in solvent B (H2O
containing 0.1% TFA). QC activity was quantified from a standard curve of
pGlu-βNA (Bachem) determined under assay conditions.
HEK293 cells were transfected with APP (NLQ) or with vector alone in 6-
well plates and lysed 1 day post transfection by adding 100μl 5× SDS sample
buffer directly to each well. Cells were collected by scraping using rubber
policeman, diluted by addition of 200μl PBS and sonicated for 15s. Afterwards,
the sample was heated to 100 °C for 5min, cooled on ice and centrifuged at
13.000rpm for 5min. 10μl, 20μl, 40μl and 60μl of the sample were loaded onto
a SDS-PAGE gel (8%) and separated. Proteins were transferred to a
nitrocellulose membrane (Roth), which was blocked using 3% (w/v) dry milk
in TBS-T (20mM Tris/HCl; pH 7.5; 500mM NaCl, 0.05% (v/v) Tween20). APP
was detected by consecutive incubation of the blot membrane with a polyclonal
APP antibody (Cell Signaling) and an anti-rabbit antibody conjugated with
horseradish peroxidase (Cell Signaling) in TBS-Tcontaining 5% (w/v) dry milk
and visualized using the SuperSignal West Pico System (Pierce) according to the
2.6. Enzyme-linked immunosorbent assay (ELISA)
Aβ(42) concentration in cell culture supernatants was determined using
specific sandwich ELISAs detecting either total Aβ(x-42) or the N-terminally
pyroglutamated variant Aβ(N3pGlu-42) (IBL-Hamburg). Briefly, 100μl of the
sample or adequate standard were added to 96-well microtiter-plates coated with
an antibody specific for Aβ(42) followed by incubation over night at 4 °C. The
1619H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625
next day, plates were washed and incubated with antibodies detecting either
Aβ(x-42) or Aβ(N3pGlu-42). Both antibodies were conjugated with horse-
radish peroxidase (HRP). After a final washing step, bound enzyme activity was
measured using a TMB peroxidase substrate in a colorimetric reaction. The
absorbance at 450nm was determined using a Sunrise plate reader (Tecan).
3.1. Determination of QC activity in cell lines
With emphasis on QC activity, we have analyzed mamma-
lian cell lines representing different derivations in order to select
an optimal line for expression of QC substrate precursors
mTRH-Aβ(N3E-42), mTRH-Aβ(N3Q-42) or APP (NLQ).
Murine QC was recently isolated from mouse insulinoma cell
line β-TC 3 and detected in mouse monocyte/macrophage cell
line RAW264.7 [19,20]. In addition, the human cell lines HeLa,
HEK293, U343 and L-363 as well as the murine cell line L929
were analyzed in the present study applying an HPLC method
for determination of QC activity. Incubation of cell supernatant
with Q-βNA substrate led to a linear rise in pGlu-βNA product
over a time period of 60min, as exemplarily shown for HEK293
and β-TC 3 (Fig. 1). In both cases, the activity could be
completely blocked by addition of QC-specific inhibitor P150/
03 at a concentration of 10μM. Furthermore, the enzymatic
activity was also abolished by boiling the cell supernatant for
5min before applying it to the assay (not shown), suggesting
that conversion of Q-βNA to pGlu-βNA in cell supernatant is
exclusively enzyme-catalyzed. Among the examined cell lines,
β-TC 3 contained the highest QC activity (16.3±3.0nmol/h/
mg), followed by L-363 (11.5±0.2nmol/h/mg), HeLa (9.5±
0.4nmol/h/mg), U343 (9.2±1.3nmol/h/mg), L929 (8.1±
0.9nmol/h/mg), RAW264.7 (6.5±1.9nmol/h/mg) and
HEK293 (3.5±0.9nmol/h/mg) (Fig. 2). The activity determina-
tions show that QC is present in a broad range of cells varying
from neuronal-derived cells (U343, HEK293) to blood cells
(RAW264.7, L-363) and even cells of peripheral tissues (HeLa,
β-TC 3, L929). For further expression studies, we have chosen
β-TC 3 due to its high QC content and HEK293 since this cell
line is a well-established model for expression and amyloido-
genic processing of APP [21,22].
3.2. Verification of QC expression in β-TC 3 and HEK293
To substantiate the findings obtained by the HPLC-assay, we
have analyzed the QC expression in β-TC 3 and HEK293 using
RT-PCR. These investigations were performed to prove that QC
activity arises from expression of murine or human QC mRNA
and not from putative isoenzymes . Therefore, total-RNA
was isolated from both cell lines for cDNA generation. RT-PCR
yielded products for both cell lines (Fig. 3), and the QC cDNA
sequence was confirmed by sequencing (not shown). We have
also found QC transcripts in RAW264.7, L-929 and L-363 (not
shown) whereas HeLa and U343 were not further characterized.
Although these RT-PCR results do not rule out putative
isoenzymes, the synopsis of the RT-PCR and QC activity
Fig. 1. Time dependent formation of pGlu-βNA in HEK293 and β-TC 3. Q-
βNA (50μM) was incubated with cell extract in presence (HEK293—filled
triangles; β-TC 3—filled squares) or absence (HEK293—open triangles; β-TC
3—open squares) of 10μM P150/03. Assay was carried out in 25mM MOPS,
pH 7.0, containing 0.1mM N-Ethylmaleinimide (NEM) at 30 °C. The inset
shows the chemical structure of P150/03. For data evaluation, only the phase of
linear product formation was used. Therefore, the time point at 60min for β-TC
3 was omitted.
Fig. 2. Distribution of QC activity among various cell lines. Cell lines and
determined QC activity were as follows: β-TC 3 (16.3±3.0nmol/h/mg), L-363
(11.5±0.2nmol/h/mg), HeLa(9.5±0.4nmol/h/mg), U343(9.2±1.3nmol/h/mg),
L929 (8.1±0.9nmol/h/mg), RAW264.7 (6.5±1.9nmol/h/mg), HEK293 (3.5±
Fig. 3. Analysis of QC expression by RT-PCR. (A) Detection of QC transcripts
in HEK293. Lanes: 1, DNA standard; 2, amplified PCR product of human QC;
3, negative control; (B) Detection of QC transcripts in β-TC 3. Lanes: 1, DNA
standard; 2 amplified PCR product of murine QC; 3, negative control.
1620H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625
determinations suggests that enzymatic activity originates
exclusively from QC, which is further substantiated by the
complete inhibition of the cyclization activity in both cell
extracts. Thus, β-TC 3 and HEK293 are suitable for expression
of the QC substrate precursors and to evaluate the effect of QC
inhibitors in these cell lines.
3.3. Expression of mTRH-Aβ(N3Q-42) in β-TC 3
For the expression of mTRH-Aβ(N3Q-42), the pre–pro
sequence of mTRH was cloned, containing a signal peptide with
an adjacent propeptide terminating at arginine76. The sequence
was fused N-terminally to the Aβ(N3Q-42) sequence. The
selection of murine insulinoma cell line β-TC 3 for transfection
was based on its high QC content, reliable transfection rates,
and characterization as hormone-secreting cell line. According
to the well-established steps of releasing TRH from its precursor
protein, it was anticipated that mTRH-Aβ(N3Q-42) is directed
to the Endoplasmatic Reticulum via the mTRH signal peptide
and that Aβ(N3Q-42) is liberated within the secretory pathway
by prohormone convertases . β-TC 3 was transfected with
the construct mTRH-Aβ(N3Q-42) or with the vector alone and
the amount of Aβ(42) and Aβ(N3pGlu-42) was determined
using commercially available ELISA kits. In this regard, the
Aβ(x-42) ELISA recognizes all N-terminally truncated iso-
forms of the Aβ-peptide, but the Aβ(N3pGlu-42) ELISA is
specific for formation of N-terminal pGlu. The expression of
mTRH-Aβ(N3Q-42) leads to significantly elevated Aβ(42)
levels in converted media of β-TC 3 incubated in absence of
P150/03 (49.7±8.9pg/ml/1.5×106cells) (P<0.05; Student's t-
test) or with media containing P150/03 (100μM) (47.7±3.2pg/
ml/1.5×106cells) (P<0.01; Student's t-test) compared to
vector-transfected controls without inhibitor (32.7±1.1pg/ml/
1.5×106cells) or with inhibitor (22.9±1.8pg/ml/1.5×106
cells) (Fig. 4A). Thus, the incubation of Aβ-transfected cells
with P150/03 did not change the levels of Aβ(42) within 24h
of the assay, suggesting addition of P150/03 does not affect
Aβ(N3Q-42) expression or secretion. The significant amounts
of Aβ(42) detected in vector-transfected controls were always
observed for β-TC 3, but not for HEK293 (Fig. 4A). The N-
terminal glutamine of secreted Aβ N3Q-42 is converted to
pyroglutamic acid (63.5±16.4pg/ml/1.5×106cells) as deter-
mined by pGlu ELISA, implying a processing of the fusion
protein within the secretory pathway. Moreover, the incuba-
tion of Aβ(N3Q-42) expressing β-TC 3 cells with P150/03
leads to significantly reduced pGlu formation at the N-
terminus of N3Q-42 (17.5±4.3pg/ml/1.5×106cells) (P<0.01;
Student's t-test) indicating that the QC inhibitor P150/03 is
capable of preventing pGlu formation within this model
system (Fig. 4B).
3.4. Expression of APP (NLQ) in HEK293
To examine whether QC inhibition leads also to reduced
pGlu formation at the N-terminus of Aβ(N3Q-42) after the
amyloidogenic proteolytic processing, we expressed APP
(NLQ) in HEK293 cells. HEK293 were identified to belong
to the neuronal lineage even though they are based on primary
cultures of human embryonic kidney cells and they comprise a
suitable cell system for expression of APP [21,22]. After
transfection of HEK293 with the construct APP (NLQ), the
protein is abundantly detectable in cell lysates showing two
bands at 100kDa corresponding to mature and immature forms
of APP695 after 30s development time. The vector-transfected
controls did not show a signal for APP in consequence of the
short development time (Fig. 5). However, native APP in
HEK293 can be visualized by longer development times (not
shown). In analogy to the direct expression model, the extent of
N-terminal cyclization of liberated Aβ(N3Q-42) in presence or
in absence of P150/03 was determined. Thereby, it became
evident that HEK293 cells appear to be more sensitive to P150/
03 treatment than β-TC 3, since Aβ(42) was significantly
reduced after inhibitor administration at a concentration of
100μM, indicating a toxic effect, although no toxicity was
observed on β-TC 3 and HEK293 up to a concentration of
1mM during routine determination of cytotoxicity. Due to the
unspecific effect, the concentration of P150/03 was reduced by
one order of magnitude to 10μM, which corresponds to the
200fold of the Kivalue . At this concentration, a putative
cytotoxic effect (or a total Aβ(42) modulating effect) was not
Fig. 4. (A) Determination of Aβ(x-42) secreted by transfected β-TC 3. Cells
transfected with construct mTRH-Aβ(N3Q-42) or with vector alone were
incubated either with (100μM) or without QC specific inhibitor P150/03. Aβ
concentration measured by Aβ(42) specific ELISAwas normalized to cell count
(pg/ml/1.5× 106cells). (*, Student's t-test, P<0.05, n=3; ***, Student's t-test,
P<0.001, n=3). (B) Determination of Aβ(N3pGlu-42) secreted by transfected
β-TC 3. Cells were incubated with (100μM) or without P150/03. Extend of
cyclizationwas measuredby Aβ(N3pGlu-42)specific ELISA andnormalized to
cell count (pg/ml/1.5×106cells). (**, Student's t-test, P<0.01, n=3).
1621H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625
observed as suggested by the Aβ(42) concentration in the cell
culture supernatant of HEK293 incubated with (173.6±29.6pg/
ml/1×106cells) or without (152.9±39pg/ml/1mio 106cells)
inhibitor (Fig. 6A). However, by applying 10μM of P150/03 to
HEK293 cells expressing APP (NLQ) we significantly reduced
the amount of Aβ(N3pGlu-42) by 50% (122.8± 13.3pg/ml/
1×106cells) compared to the untreated control (245.5±44.4pg/
ml/1×106cells) (P<0.001; Student's t-test) (Fig. 6B).
To balance the reduced concentration of the inhibitor,
HEK293 were incubated with the inhibitor 3h prior to the
assay. Again, no significant differences were detected in Aβ(42)
concentration between samples incubated with (202.9±44.9pg/
ml/1×106cells) or without (192.9±28.6pg/ml/1×106cells)
inhibitor (Fig. 6A). The amount of detectable N3pGlu-42,
however, was reduced to 31% (153.9±25.2pg/ml/1×106cells)
compared with the untreated control (485.9±67pg/ml/1×
106cells) (P<0.001; Student's t-test) (Fig. 6B). Surprisingly,
for every Aβ(N3pGlu-42) determination a higher Aβ content
was observed than determined with Aβ(x-42) specific ELISA.
This might be a result of different detection antibodies in both
ELISAs displaying different specificities.
3.5. Co-expression of mTRH-Aβ(N3E-42) and mTRH-QC in
The N-terminal pGlu modification in the disease-related
proteins Aβ, ABri and ADan arises from the cyclization of
glutamate and not from glutamine. To address the question,
whether QC is also capable of cyclizing glutamate under
physiological conditions in cell culture, the construct mTRH-
Aβ(N3E-42) possessing a N-terminal glutamate after proces-
sing was expressed in β-TC 3. In accordance with the findings
after expression of mTRH-Aβ(N3Q-42), the transfection of
mTRH-Aβ(N3E-42) led to elevated amounts of Aβ(42) in the
cell culture media (65.6±4.6pg/ml/1.5×106cells) but the
amount of Aβ(N3pGlu-42) remained below the limit of
quantification (0.13±0.18pg/ml/1.5×106cells). However, the
QC-catalyzed cyclization of glutamate in vitro is slow [10,33].
Therefore, it was assumed that the QC activity in β-TC 3 might
be not sufficient to produce significant Aβ(N3pGlu-42) levels
within 24h, i.e. during the time of analysis. Consequently, we
co-transfected the constructs mTRH-Aβ(N3E-42) and mTRH-
QC to increase the QC activity in β-TC 3. This approach led to
elevated amounts of Aβ(42) (42.0±1.7pg/ml/1.5×106cells)
and, interestingly, to significant levels of Aβ(N3pGlu-42) (2.5±
0.6pg/ml/1.5×106cells) (P<0.01; Student's t-test) (Fig. 7A).
The over-expression of QC was monitored by activity
determination in cell culture media (not shown). In consistence
with the findings obtained by expression of mTRH-Aβ(N3Q-
42) in β-TC 3, the application of P150/03 in a concentration of
100μM did not change the amount of Aβ(42) detected after co-
transfection of mTRH-Aβ(N3E-42) and mTRH-QC (49.7±
5.9pg/ml/1.5×106cells vs. 48.8±2.9pg/ml/1.5×106cells).
Fig. 6. (A) Analysis of Aβ(x-42) secreted by HEK293. Cells were transfected
with APP (NLQ) or with vector alone in absence (closed bars) or presence
(open bars) of 10μM P150/03. Inhibitor was added to HEK293 either together
with assay medium (N3Q-42, no preincubation) or cells were preincubated with
inhibitor 3h prior to the assay (N3Q-42-PI). Aβ was determined by Aβ(x-42)
specific ELISA and normalized to cell count (pg/ml/1×106cells), (n=6). (B)
Analysis of Aβ(N3pGlu-42) secreted by transfected HEK293. Cells were
incubated in absence (closed bars) or presence (open bars) of 10μM P150/03.
Inhibitor was added to HEK293 either together with assay medium (N3Q-42,
no preincubation) or cells were preincubated with inhibitor 3h prior to the
assay (N3Q-42-PI). Aβ was determined by Aβ(N3pGlu-42) specific ELISA
and normalized to cell count (pg/ml/1×106cells), (***; Student's t-test,
Fig. 5. Western Blot for APP (NLQ) detection in HEK293. Lanes: kDa, Protein
Standard Dual Color (Bio-Rad); 1, vector 10μl; 2, vector 20μl; 3, vector 40μl;
4, vector 60μl; 5, APP(NLQ) 10μl; 6, APP(NLQ) 20μl; 7, APP(NLQ) 40μl; 8,
1622H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625
However, the amount of Aβ(N3pE-42) was significantly
reduced in the cell culture medium supplemented with P150/
03 (0.98±1.12pg/ml/1.5×106cells) compared to medium
without inhibitor (6.0±2.9pg/ml/1.5×106cells) (P<0.05;
Student's t-test) (Fig. 7B).
It is important to underline that within this study the
determined Aβ(42) concentrations did not significantly differ in
samples incubated with or without the inhibitor, but the extent
of pGlu formation at the N-terminus of Aβ peptides was always
significantly reduced. Thus, the QC-specific inhibitor P150/03
not only modulates the cyclization of glutamine but also
decreases the QC-catalyzed cyclization of glutamate at the N-
terminus of artificial Aβ peptides.
For several bioactive peptides, a N-terminal pyroglutamyl
residue is described, which is formed post-translationally from a
glutamine precursor. This pGlu formation is catalyzed by QC
and pGlu-containing peptides and proteins are conserved in
invertebrates like the marine snail Aplysia californica and the
Brazilian armed spider Phoneutria nigriventer as well as in
vertebrates like Mus musculus, Bos taurus and Homo sapiens
[4,20,25–27]. Two major functions are attributed to the pGlu
residue. It stabilizes proteins by preventing N-terminal
degradation by aminopeptidases and thus prolongs their
biological half-life. On the other hand, for some mammalian
peptides like TRH the pGlu modification was shown to be
essential for biologic activity . Therefore, blockage of pGlu
formation at the N-terminus of selected peptides seems to be a
promising approach for modulating their activity throughout the
body either by their metabolic destabilization or by their
functional inactivation. Previous publications have shown that
QC is expressed in various tissues, with a marked abundance in
brain, e.g. cerebral cortex, hypothalamus, hippocampus and
striatum, in peripheral glands like thyroid and thymus and in
human B-lymphocytes [3,8,23]. The current study demonstrates
that QC activity is readily detectable in mammalian cell lines
from very different origins (Fig. 2). This ubiquitous occurrence
within the cell lines tested was a surprising finding because it
was presumed that QC is co-expressed with an adequate
substrate within the secretory pathway. In fact, for almost all of
those cell lines no QC substrate is described. Nevertheless,
further screening of mammalian cell lines did not reveal one
single cell line possessing no QC-activity (unpublished results).
Seven years ago, researchers also found a ubiquitous distribu-
tion of QC activity in all tissues they examined and they have
speculated that tissue-specific forms of QC might exist because
of obvious discrepancies between QC expression studies using
Northern Blot and QC activity determination . Clearly, it is
reasonable to take the existence of isoenzymes of QC into
consideration. However, P150/03, an imidazole-derivative,
completely blocked QC activity in cell extracts of the utilized
cell lines HEK293 and β-TC 3 suggesting either a very high
homology of QC and the putative isoenzyme regarding their
active sites or simply the absence of such an isoenzyme. In
addition, the RT-PCR revealed transcripts of QC in HEK293, β-
TC 3 and in the other cell lines (not shown) indicating that the
QC activity origins substantially from QC and not from a
putative isoenzyme. Finally, it has been described that putative
isoforms of QC are not inhibited by imidazole . Moreover, it
should also be noted that these cell lines represent immortalized
tumor cells and it cannot be excluded that during transformation
of a cancer cell QC activity is upregulated.
Having successfully identified a number of cell lines
possessing QC activity, the effort was to demonstrate potency
of a QC-specific inhibitor in cell culture. Because of the lack of
antibodies or ELISAs specific for the N-terminal pGlu
Fig. 7. (A) Determination of Aβ(x-42) (closed bars) and Aβ(N3pGlu-42) (open
bars) secreted by β-TC 3. Cells were transfected with construct mTRH-
Aβ(N3E-42) in absence or presence of construct mTRH-QC. Aβ concentration
measured by Aβ(42) and Aβ(N3pGlu-42) specific ELISA was normalized to
cell count (pg/ml/1.5×106cells). (**, Student's t-test, P<0.01, n=3). (B)
Determination of Aβ(x-42) (closed bars) and Aβ(N3pGlu-42) (open bars)
secreted by β-TC 3. Cells were transfected with constructs mTRH-Aβ(N3E-42)
and mTRH-QC and incubated either with (100μM) or without P150/03. Aβ
concentrationsweredeterminedby Aβ(42)andAβ(N3pGlu-42) specificELISA
and normalized to cell count (pg/ml/1.5×106cells). (*, Student's t-test,
1623H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625
modification of nearly all QC substrates, an expression model
was used, where the QC substrate Aβ(N3Q-42) was expressed
in β-TC 3 as fusion protein with pre–pro sequence of mTRH to
mimic pGlu hormone maturation. This model is based on the
generation of a native QC substrate from a prohormone
precursor after prohormone convertase cleavage, its post-
translational pGlu modification and the prevention of pGlu
formation by a QC-specific inhibitor within the secretory
pathway. We also expressed a mutant form of APP (NLQ) in
HEK293, which resembles, after its secretase liberation from
the larger amyloid precursor protein, the subsequent pGlu
formation of the amyloidogenic species Aβ(N3pGlu-42). In
contrast to the mTRH-Aβ(N3Q-42) construct which is
processed within the secretory pathway, i.e. under conditions
where QC is abundant, the liberation of Aβ from APP has been
described within the endosomal/lysosomal as well as within the
secretory pathway [28–30]. This in addition raises the
interesting question, whether QC inhibition is capable of
suppressing pGlu-Aβ formation despite of the potentially
different sub-cellular localization of QC and Aβ in both
experimental approaches. Finally, the expression of mTRH-
Aβ(N3E-42) in combination with mTRH-QC was to prove the
involvement of QC in glutamate cyclization in mammalian
cells. As expected, after expression of mTRH-Aβ(N3Q-42) and
mTRH-Aβ(N3E-42) in β-TC 3 and APP (NLQ) in HEK293,
Aβ(42) was liberated from both different precursors by
endogenous proteolysis. Aβ(42) levels did not differ between
experiments in presence or absence of P150/03, indicating that
the inhibitor exhibits no unspecific effects on the APP-
processing machinery of the cell in its concentration range
studied. But strikingly, the pGlu formation at the N-terminus of
Aβ(N3Q-42) was significantly suppressed in presence of the
inhibitor suggesting a specific inhibition of this post-transla-
tional modification reaction in both model systems. This is of
particular interest in the case of APP processing. The generation
of pGlu-Aβ indicates that QC and the β-secretase cleavage
products, i.e. Aβ and β-CTF, are co-localized in the secretory
pathway at least in our cell-based model system. The result
supports our hypothesis of potentially intraneuronal induction
of neurodegeneration by toxic pGlu-peptide species. In fact,
intracellular toxic oligomer formation as perhaps causal, early
event becomes more and more accepted [29,30].
N-terminal pyroglutamated Aβ-peptides are a major hall-
mark of Alzheimer's disease amyloid plaques [11,12,31], are
more prone to aggregation [15,34], and more toxic to neurons
and astrocytes than full length Aβ(42) . The precursor of
pGlu in disease-related peptides in amyloidotic diseases is
glutamate instead of glutamine. Recently, we could show that
QC is responsible for pGlu formation from glutamate at the N-
terminus of Aβ peptides in vitro . This study demonstrates
for the first time that QC is also capable of converting glutamate
into pGlu in living cultured cells. Based on our results we
cannot exactly predict in which cellular compartment the pGlu-
Aβ peptide derived from the APP-precursor is formed, but the
TRH-Aβ-precursor is very likely processed within the secretory
pathway. However, in all experimental settings, P150/03
suppresses pGlu-formation from a glutamine and a glutamate
precursor. Therefore, it is tempting to conclude, that QC
inhibition has an impact on pGlu-formation independent from
the pathway, in which a certain QC substrate is posttranslation-
ally generated by convertases and/or secretases. Thus, alteration
of pGlu formation using highly potent QC inhibitors comprises
an interesting new tool with broader applicability.
In the present study, pGlu formation in cell culture
experiments was never completely blocked (Figs. 4B, 6B) by
administration of P150/03, which was in contrast to the total
inhibition of QC activity observed in cell extracts. The detection
of remaining Aβ(N3pGlu-42) despite inhibitor administration is
not surprising because P150/03 is an imidazole derivative 
and the acidic environment of secretory granules leads to the
protonation of a ring nitrogen and thus, partially hinders the
interaction with the active site metal of the target enzyme .
This assumption is substantiated by the decline of the Kivalue
of P150/03 from 60nM at pH 8.0 to 200nM at pH 6.0.
Moreover, a significant amount of spontaneous cyclization must
also to be taken into consideration. Incubation of N-glutamine
peptides in absence of enzyme (100mM MES, pH 7.0, 30 °C)
leads to a relatively high amount (approximately 20% within
24h) of spontaneous glutamine cyclization similar to the
observations here in cell culture (F. Seifert et al., manuscript
in preparation) (Figs. 4B, 6B).
In conclusion, our study verifies QC as a potential target in
Alzheimer's disease. Preclinical investigations will target the
formation of toxic pGlu-Aβ species by inhibitors of QC in
different animal models. However, by administration of QC
inhibitors not only the formation of toxic pGlu-Aβ peptides
from N-glutamate substrates will be influenced but rather all
native N-glutamine substrates. Thus, due to the relatively fast
spontaneous cyclization of the N-glutamine peptides, partial
and reversible blockage of the pGlu-formation should be
beneficial, if long-term application is considered. Further
investigations will concentrate also on other native QC
substrates with relevance in certain diseases, e.g. the contribu-
tion of gastrin to colorectal cancer .
This work was supported by grants of the BMBF (grant
#0313185). The authors gratefully thank Dr. Steffen Rossner
(Paul Flechsig Institute for Brain Research, Leipzig, Germany)
for providing cultures of murine primary cortical neurons. We
thank Dr. Ingo Schulz and Mirko Buchholz for helpful
discussions, Nadine Schreier for routine cytotoxicity tests and
Anett Stephan for technical assistance. The help of Prof. Dr.
Robert C. Bateman Jr. and Jan Eggert for critical reading of the
manuscript is also gratefully acknowledged.
 A.C. Awade, P. Cleuziat, T. Gonzales, J. Robert-Baudouy, Pyrrolidone
carboxyl peptidase (Pcp): an enzyme that removes pyroglutamic acid
(pGlu) from pGlu-peptides and pGlu-proteins, Proteins 20 (1994) 34–51.
 G.N. Abraham, D.N. Podell, Pyroglutamic acid. Non-metabolic formation,
function in proteins and peptides, and characteristics of the enzymes
effecting its removal, Mol. Cell. Biochem. 38 Spec. No. (1981) 181–190.
1624H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625
 W.H. Busby Jr., G.E. Quackenbush, J. Humm, W.W. Youngblood, J.S. Download full-text
Kizer, An enzyme(s) that converts glutaminyl-peptides into pyroglutamyl-
peptides. Presence in pituitary, brain, adrenal medulla, and lymphocytes,
J. Biol. Chem. 262 (1987) 8532–8536.
 W.H. Fischer, J. Spiess, Identification of a mammalian glutaminyl cyclase
converting glutaminyl into pyroglutamyl peptides, Proc. Natl. Acad. Sci.
U. S. A. 84 (1987) 3628–3632.
 M. Messer, Enzymatic cyclization of L-glutamine and L-glutaminyl
peptides, Nature 197 (1963) 1299.
 S. Schilling, A.J. Niestroj, J.U. Rahfeld, T. Hoffmann, M. Wermann, K.
Zunkel, C. Wasternack, H.U. Demuth, Identification of human glutaminyl
cyclase as a metalloenzyme. Potent inhibition by imidazole derivatives and
heterocyclic chelators, J. Biol. Chem. 278 (2003) 49773–49779.
 S. Schilling, S. Manhart, T. Hoffmann, H.H. Ludwig, C. Wasternack, H.U.
Demuth, Substrate specificity of glutaminyl cyclases from plants and
animals, Biol. Chem. 384 (2003) 1583–1592.
 T. Pohl, M. Zimmer, K. Mugele,J. Spiess, Primary structure and functional
expression of a glutaminyl cyclase, Proc. Natl. Acad. Sci. U. S. A. 88
 T.M. Bockers, M.R. Kreutz, T. Pohl, Glutaminyl-cyclase expression in the
bovine/porcine hypothalamus and pituitary, J. Neuroendocrinol. 7 (1995)
 S. Schilling, T. Hoffmann, S. Manhart, M. Hoffmann, H.U. Demuth,
Glutaminyl cyclases unfold glutamyl cyclase activity under mild acid
conditions, FEBS Lett. 563 (2004) 191–196.
 T.C. Saido, T. Iwatsubo, D.M. Mann, H. Shimada, Y. Ihara, S. Kawashima,
Dominant and differential deposition of distinct beta-amyloid peptide
species. A beta N3(pE), in senile plaques, Neuron 14 (1995) 457–466.
 Y. Harigaya, T.C. Saido, C.B. Eckman, C.M. Prada, M. Shoji, S.G.
Younkin, Amyloid beta protein starting pyroglutamate at position 3 is a
major component of the amyloid deposits in the Alzheimer's disease brain,
Biochem. Biophys. Res. Commun. 276 (2000) 422–427.
 J. Ghiso, T. Revesz, J. Holton, A. Rostagno, T. Lashley, H. Houlden, G.
Gibb, B. Anderton, T. Bek, M. Bojsen-Moller, N. Wood, R. Vidal, H.
Braendgaard, G. Plant, B. Frangione, Chromosome 13 dementia
syndromes as models of neurodegeneration, Amyloid 8 (2001) 277–284.
 T.C. Saido, Alzheimer's disease as proteolytic disorders: anabolism and
catabolism of beta-amyloid, Neurobiol. Aging 19 (1998) S69–S75.
 W. He, C.J. Barrow, The A beta 3-pyroglutamyl and 11-pyroglutamyl
peptides found in senile plaque have greater beta-sheet forming and
aggregation propensities in vitro than full-length A beta, Biochemistry 38
 C. Russo, E. Violani, S. Salis, V. Venezia, V. Dolcini, G. Damonte, U.
Benatti, C. D'Arrigo, E. Patrone, P. Carlo, G. Schettini, Pyroglutamate-
modified amyloid beta-peptides-AbetaN3(pE)-strongly affect cultured
neuron and astrocyte survival, J. Neurochem. 82 (2002) 1480–1489.
 M. Buchholz, U. Heiser, S. Schilling, A.J. Niestroj, K. Zunkel, H.U.
Demuth, The first potent inhibitors for human glutaminyl cyclase:
synthesis and structure–activity relationship, J. Med. Chem. 49 (2006)
 K. Shirotani, S. Tsubuki, H.J. Lee, K. Maruyama, T.C. Saido, Generation
of amyloid beta peptide with pyroglutamate at position 3 in primary
cortical neurons, Neurosci. Lett. 327 (2002) 25–28.
 T. Chikuma, K. Taguchi, M. Yamaguchi, H. Hojo, T. Kato, Improved
determination of bovine glutaminyl cyclase activity using precolumn
derivatization and reversed-phase high-performance liquid chromatogra-
phy with ultraviolet detection, J. Chromatogr., B Anal.. Technol. Biomed.
Life Sci. 806 (2004) 113–118.
 S. Schilling, H. Cynis, A. von Bohlen, T. Hoffmann, M. Wermann, U.
Heiser, M. Buchholz, K. Zunkel, H.U. Demuth, Isolation, catalytic
properties, and competitive inhibitors of the zinc-dependent murine
glutaminyl cyclase, Biochemistry 44 (2005) 13415–13424.
 G. Shaw, S. Morse, M. Ararat, F.L. Graham, Preferential transformation of
human neuronal cells by human adenoviruses and the origin of HEK 293
cells, FASEB J. 16 (2002) 869–871.
 C. Haass, M.G. Schlossmacher, A.Y. Hung, C. Vigo-Pelfrey, A. Mellon,
B.L. Ostaszewski, I. Lieberburg, E.H. Koo, D. Schenk, D.B. Teplow,
Amyloid beta-peptide is produced by cultured cells during normal
metabolism, Nature 359 (1992) 322–325.
 P.A. Sykes, S.J. Watson, J.S. Temple, R.C. Bateman Jr., Evidence for
tissue-specific forms of glutaminyl cyclase, FEBS Lett. 455 (1999)
 P. Schaner, R.B. Todd, N.G. Seidah, E.A. Nillni, Processing of
prothyrotropin-releasing hormone by the family of prohormone con-
vertases, J. Biol. Chem. 272 (1997) 19958–19968.
 R.W. Garden, T.P. Moroz, J.M. Gleeson, P.D. Floyd, L. Li, S.S. Rubakhin,
J.V. Sweedler, Formation of N-pyroglutamyl peptides from N-Glu and N-
Gln precursors in Aplysia neurons, J. Neurochem. 72 (1999) 676–681.
 A.M. Pimenta, B. Rates, C. Bloch Jr., P.C. Gomes, M.M. Santoro, M.E.
de Lima, M. Richardson, M.N. Cordeiro, Electrospray ionization
quadrupole time-of-flight and matrix-assisted laser desorption/ionization
tandem time-of-flight mass spectrometric analyses to solve micro-
heterogeneity in post-translationally modified peptides from Phoneutria
nigriventer (Aranea, Ctenidae) venom, Rapid Commun. Mass Spectrom.
19 (2005) 31–37.
 I. Song, C.Z. Chuang, R.C. Bateman Jr., Molecular cloning, sequence
analysis and expression of human pituitary glutaminyl cyclase, J. Mol.
Endocrinol. 13 (1994) 77–86.
 C. Haass, C.A. Lemere, A. Capell, M. Citron, P. Seubert, D. Schenk, L.
Lannfelt, D.J. Selkoe, The Swedish mutation causes early-onset
Alzheimer's disease by beta-secretase cleavage within the secretory
pathway, Nat. Med. 1 (1995) 1291–1296.
 K. Takeda, W. Araki, T. Tabira, Enhanced generation of intracellular
Abeta42 amyloid peptide by mutation of presenilins PS1 and PS2, Eur. J.
Neurosci. 19 (2004) 258–264.
 A.M. Cataldo, S. Petanceska, N.B. Terio, C.M. Peterhoff, R. Durham, M.
Mercken, P.D. Mehta, J. Buxbaum, V. Haroutunian, R.A. Nixon, Abeta
localization in abnormal endosomes: association with earliest Abeta
elevations in AD and Down syndrome, Neurobiol. Aging 25 (2004)
 L. Miravalle, M. Calero, M. Takao, A.E. Roher, B. Ghetti, R. Vidal,
Amino-terminally truncated Abeta peptide species are the main component
of cotton wool plaques, Biochemistry 44 (2005) 10810–10821.
 A.M. Smith, S.A. Watson, Review article: gastrin and colorectal cancer,
Aliment. Pharmacol. Ther. 14 (2000) 1231–1247.
 K.F. Huang, Y.L. Liu, W.J. Cheng, T.P. Ko, A.H. Wang, Crystal
structures of human glutaminyl cyclase, an enzyme responsible for pro-
tein N-terminal pyroglutamate formation, Proc. Natl. Acad. Sci. U. S. A.
102 (37) (2005) 13117–13122.
 S. Schiling, T. Lauber, M. Schaupp, S. Manhart, E. Scheel, G. Bohm,
H.U. Demuth, On the seeding and oligomerization of pGlu-amyloid
peptides (in vitro), Biochemistry (2006) (in press).
1625H. Cynis et al. / Biochimica et Biophysica Acta 1764 (2006) 1618–1625