Apoptosis (2009) 14:778787

DOI 10.1007/s10495-009-0349-3

ORIGINAL PAPER

Generation of new TRAIL mutants DR5-A and DR5-B

with improved selectivity to death receptor 5
Marine E. Gasparian Boris V. Chernyak Dmitry A. Dolgikh
Anne V. Yagolovich Ekaterina N. Popova Anna M. Sycheva
Sergey A. Moshkovskii Mikhail P. Kirpichnikov

Published online: 2 May 2009

Springer Science+Business Media, LLC 2009

Abstract TRAIL (tumor necrosis factor (TNF) related

apoptosis-inducing ligand) has been introduced as an
extrinsic pathway inducer of apoptosis that does not have
the toxicities of Fas and TNF. However, the therapeutic
potential of TRAIL is limited because of many primary
tumor cells are resistant to TRAIL. Despite intensive
investigations, little is known in regards to the mechanisms
underlying TRAIL selectivity and efficiency. A major
reason likely lies in the complexity of the interaction of
TRAIL with its five receptors, of which only two DR4 and
DR5 are death receptors. Binding of TRAIL with decoy
receptors DcR1 and DcR2 or soluble receptor osteoprotegerin (OPG) fail to induce apoptosis. Here we describe
design and expression in Escherichia coli of DR5-selective
TRAIL variants DR5-A and DR5-B. The measurements of
dissociation constants of these mutants with all five
receptors show that they practically do not interact with
DR4 and DcR1 and have highly reduced affinity to DcR2
and OPG receptors. These mutants are more effective than
Electronic supplementary material The online version of this
article (doi:10.1007/s10495-009-0349-3) contains supplementary
material, which is available to authorized users.
M. E. Gasparian (&)
D. A. Dolgikh
A. V. Yagolovich

M. P. Kirpichnikov
Laboratory of Protein Engineering, Shemyakin and Ovchinnikov
Institute of Bioorganic Chemistry, RAS, 16/10
Miklukho-Maklaya, 117997 Moscow, Russia
e-mail: marine_gasparian@yahoo.com
B. V. Chernyak
E. N. Popova
A. N. Belozersky Institute of Physico-Chemical Biology,
Moscow State University, 119899 Moscow, Russia
A. M. Sycheva
S. A. Moshkovskii
V. N. Orekhovich Institute of Biomedical Chemistry, RAMS,
119121 Moscow, Russia

123

wild type TRAIL in induction of apoptosis in different

cancer cell lines. In combination with the drugs targeted to
cytoskeleton (taxol, cytochalasin D) the mutants of TRAIL
induced apoptosis in resistant Hela cells overexpressing
Bcl-2. The novel highly selective and effective DR5-A and
DR5-B TRAIL variants will be useful in studies on the role
of different receptors in TRAIL-induced apoptosis in sensitive and resistant cell lines.
Keywords TRAIL
Apoptosis
Death receptors

Decoy receptors

Introduction
Tumor necrosis factor-related apoptosis-inducing ligand
(TRAIL), also called apoptosis-inducing ligand 2 (Apo2L),
triggers programmed cell death in various types of cancer
cells but not in most normal cells [1]. TRAIL is a homotrimeric protein that interacts with five receptors. Binding
with DR4 and DR5 receptors induces death signals to the
intracellular apoptotic machinery [2]. DcR1 and DcR2 are
designated as decoy receptors that compete with DR4 and
DR5 for TRAIL binding and antagonize TRAIL-induced
apoptosis [3]. Osteoprotegerin (OPG) is a soluble TRAILinteracting molecule which may have a more prominent
role in bone and myeloid cell development than in regulating TRAIL-induced apoptosis [4].
Many cancer cell lines express both DR4 and DR5, and
each of these receptors can initiate apoptosis independently
of the other. Recent evidence indicates that both tumor and
normal cells can acquire resistance to TRAIL-induced killing by upregulating either of the two antagonistic receptors,
DcR1 and DcR2 so the four TRAIL receptors are involved in
regulation of TRAIL-induced apoptosis [59]. DcR1 and

Apoptosis (2009) 14:778787

DcR2 are unable to recruit FADD and form active deathinducing signaling complex (DISC). DcR2 inhibits apoptosis by sequestering TRAIL or by forming heteromeric
receptor complex with DR5 that is unable to activate signaling [6, 7]. Inhibition of TRAIL induced apoptosis by
DcR2 critically depends on its association with DR5 via the
NH2-terminal preligand assembly domain overlapping the
first partial cysteine-rich domain of both receptors. Ligand
binding by DcR2 is dispensable for its apoptosis inhibitory
function, thereby excluding the possibility that DcR2 is a
decoy to inhibit apoptosis by binding of TRAIL. It have
been proposed that DcR2 is a regulatory rather than
decoy receptor that inhibits TRAIL-induced apoptosis
signaling through ligand-independent mechanism [6, 8].
To investigate the relative contribution of DR4 and DR5
to ligand-induced apoptosis several receptor-selective
TRAIL variants were generated recently using a novel
approach that enabled phage display of mutated trimeric
proteins. Selective binding to DR4 or DR5 was achieved
with three to six amino acid substitutions [9]. The TRAIL
variants with optimal selectivity and activity were marked
as DR5-8 and DR4-8. The DR4-8 variant showed a
markedly reduced ability to trigger apoptosis, whereas the
DR5-8 variant had minimally decreased or slightly
increased apoptosis-inducing activity in cell lines where
induction of apoptosis was mediated mainly through DR5
receptor. Another DR5-selective TRAIL variant (D269H)
was generated using the automatic design algorithm
FOLD-X. This variant did not induce apoptosis in DR4responsive cell lines but show a significant increase of
biological activity in DR5-responsive cancer cell lines [10].
Here we have generated new DR5-selective variants of
TRAIL (designated as DR5-A and DR5-B) by combining
DR5-8 and D269H variants. DR5-A and DR5-B demonstrated high selectivity to DR5 receptor and practically did
not bind to DR4 or DcR1 receptors. Surface plasmon resonance (SPR) binding experiments demonstrated that
DR5-B variant had also very low affinity to DcR2 and OPG
receptors. DR5-A and DR5-B were significantly more
effective than wild type TRAIL in induction of apoptosis in
different cell lines were the cell death was mediated mainly
through DR5 receptor. The new selective mutants of
TRAIL will serve as a good tool to study the TRAILinduced apoptotic signaling in cell lines with different
expression of death and decoy receptors.

779

previously described [11]. All variants of TRAIL were

constructed a bacterial expression vector pET-32a encoding human TRAIL amino acids 114281. Amino acid
substitutions in the gene of wild type TRAIL were constructed by oligonucleotide-directed mutagenesis [12] of a
plasmid (pET32a/TRAIL) designed for the intracellular
E. coli expression of TRAIL (114281) under control of
the T7 promoter. Introduction of mutations was confirmed
by DNA sequencing. High level expression of Trx/TRAIL
fusions (*150 mg from 100 ml culture) in E. coli strain
BL-21(DE3) was induced by 0.02 mM IPTG at 28C. The
90% of wild type thioredoxin fused protein were insoluble
and formed inclusion bodies, whereas the fusions of variants D269H, DR5-8, DR4-8, DR5-A, and DR5-B in
6080% have been expressed in soluble form. Introduction
of mutations to TRAIL amino acid sequence resulted in
about three to fivefold increase of protein solubility. The
probable reason is that all introduced mutations lie on the
surface of TRAIL molecule and are more hydrophilic
comparing to wild type amino acid residues. We have
prepared the wild type TRAIL both from inclusion bodies
and from soluble fraction. Neither sedimentation coefficient nor biological activity of these preparations was different. Farther experiments have been done with wild type
TRAIL purified from inclusion bodies because the yield
was sixfold higher in this protocol than in preparation from
cytoplasmic fraction. After solubilization of inclusion
bodies wild type Trx/TRAIL fusion protein was refolded,
purified on Ni-NTA high performance agarose. All mutant
variants were purified from cytoplasmatic protein fraction
without refolding procedure. After cleavage of fusions with
recombinant human enteropeptidase light chain [13] the
target proteins were separated from thioredoxin on Ni-NTA
agarose. The remaining activity of enteropeptidase was
removed by passing the preparations through a STI-agarose
column. All preparations were dialyzed against a buffer
containing 50 mM sodium phosphate (pH 7.5) and
150 mM NaCl, sterilized by filtration, and stored at 4C for
further investigation.
The concentration of TRAIL preparations was determined spectrophotometrically by absorbance at 280 nm
using a molar extinction coefficients calculated from the
known amino acid sequence (25,520 M-1 cm-1 for wild
type and D269H variants; 24,240 M-1 cm-1 for DR5-8,
DR5-A, DR5-B, and 28,650 for DR4-8).
Analytical ultracentrifugation

Materials and methods

Expression and purification of TRAIL variants
Wild type and mutants of TRAIL (114281) expressed in
Escherichia coli BL21 (DE3) strain, and purified as

Analytical ultracentrifugation was carried out with Beckman Model E centrifuge connected to PC. TRAIL solution
(0.5 mg/ml in PBS) was analyzed at 60,000 rpm in standard 0.4 ml cells. Sedimentation coefficient s20,w was calculated using software SEDFIT [14] (available at NIH site

123

780

Apoptosis (2009) 14:778787

Fig. 1 Sedimentation analysis

of purified DR5-A TRAIL
variant. Preparation was
centrifuged at 60,000 rpm in a
standard 0.4 ml cell. Protein
profiles were recorded after 1,
10, 20, 45, 54, 60, 74, and
86 min (curves from the left to
the right) by absorbance at
280 nm. Radius (X scale) rises
from the top to the bottom of the
tube. Shape of obtained curves
and calculated sedimentation
coefficient s20,
w = 3.69 0.04 have proved
almost exclusive globular trimer
content of the preparation

http://www.analyticalcentrifugation.com). Sedimentation
analysis data demonstrated that 98% of recombinant wild
type and mutant TRAIL formed trimers with a relative
molecular mass of 5860 kDa (Fig. 1).
Assay of apoptosis induction
Human acute T leukemia Jurkat, monoblastic leukemia
U937 and Philadelphia chromosome-positive acute myeloid leukemia K562 cell lines were maintained in RPMI
1640 medium supplemented with 10% (v/v) fetal calf
serum, 2 mM glutamine, 100 U/ml streptomycin and
100 U/ml penicillin at 37C and 5% CO2. A concentration
series of the TRAIL variants were made in cell culture
medium. For viability tests, cells were seeded in 96-well
plates (1 9 105 cells/well) and incubated with serial dilutions of TRAIL mutants for 24 h at 37C and 5% CO2.
Subsequently, MTT reagent was added to final concentration 0.5 mg/ml. Cell viability was determined after 3 h of
incubation by measuring the absorption at 570 nm with
Multiscan plate reader. HeLa-Bcl-2 cells were generated
using retrovirus vector [15] in the Institute of Poliomyelitis,
Russian Academy of Medical Sciences and kindly provided
by Professor VI Agol. Human cervical carcinoma HeLa
and HeLa-Bcl-2 cell lines were grown in Dulbeccos
modified Eagles medium (DMEM) supplemented with
10% (v/v) fetal calf serum, 2 mM glutamine, 100 U/ml
streptomycin and 100 U/ml penicillin at 37C and 5% CO2.
In some experiments the inhibitor of protein synthesis

123

emetine (0.030.3 lg/ml), inhibitor of actin microfilaments

cytochalasin D (1 lg/ml) or inhibitor of microtubules taxol
(10 lM) were added together with TRAIL. The apoptosis
was determined by the number of cells with fragmented
nuclei after staining with Hoechst 33342 (1 lg/ml,
30 min). Condensed and fragmented nuclei were counted
(500 cells per sample) under fluorescent microscope Axiophot (Carl Zeiss).
Surface plasmon resonance measurements (SPR)
of receptor binding
Dissociation constants (KD) for the direct binding of TRAIL
variants to immobilized receptors were determined by SPR
measurements on a ProteOn XPR36 Protein Interaction
Array System (Bio-RAD). The multichannel chip design
allowed monitoring of six sets of interactions simultaneously. All five receptors (TRAILR1(DR4)/Fc, TRAILR2
(DR5)/Fc, TRAILR3(DcR1)/Fc, TRAILR4(DcR2)/Fc, and
Osteoprotegerin/Fc (R&D system) were coupled simultaneously to the one sensor 6 9 6 channel chip surface at a
level of 6,0008,000 resonance units using the amine coupling chemistry. The chip was first activated for 120 s with a
mixture of 1-ethyl-3-(3-dimethyl-aminopropyl) carbodiimide hydrochloride (0.13 M) and N-hydroxysulfosuccinimide
(0.03 M). 2 lg of each receptor in 10 mM sodium acetate
buffer, pH 5.0 was injected across activated channels. The
immobilization reaction was terminated by 1 M ethanolamine, pH 8.5. As a control, the remaining channel was not

Apoptosis (2009) 14:778787

781

coupled with any protein. Sensorgrams were recorded

simultaneously for five concentrations of highly purified
trimeric TRAIL variants ranging from 15 to 250 nM in twofold increments. About 300 ll of wild type TRAIL and
mutant variants were injected at 100 ll/min at 25C by using
PBS (pH 7.4) supplemented with 0.005% Tween-20 as
running and sample buffer. Binding of ligands to the receptors was monitored in real time. Between injections of different TRAIL variants, the receptor sensor surface was
regenerated by using 3 M sodium acetate (pH 5.2) injections.
For determination of KD values different concentrations
range was used for TRAIL variants depending of there
affinities to different receptors (See supplemented data).
Sensorgrams were analyzed by nonlinear regression analysis
according to a 1:1 binding model by ProteOn Manager
Version 2.0.1 software (Bio-Rad) to determine the rates of
dissociation. Dissociation constants KD were determined
from the rate constants as KD = koff/kon at five different
analyzed concentrations. For DR5-B variant binding to
DcR2 and OPG KD values were calculated assuming 1:1
Langmuir binding isotherm and were plotted using a nonlinear curve fit using the same Bio-Rad software.

Results
Generation of new DR5-selective TRAIL mutant
variants
The new DR5-selective TRAIL variants DR5-A and DR5-B
were designed based on previously described variants
DR5-8 [9] generated using a novel phage display technology and D269H [10] generated using the automatic
design algorithm FOLD-X (Table 1). DR5-8 variant demonstrated higher selectivity than D269H but the affinity of
D269H binding to DR5 receptor was several fold higher
than for DR5-8. To improve the selectivity without loosing
affinity to DR5 receptor we have generated DR5-A mutant
by insertion D269H mutation in amino acid sequence of
DR5-8 variant which contained six other amino acid substitutions. Analysis of the original data indicated that D267
residue in TRAIL amino acid sequence did not play a

critical role for DR5 selectivity and substitution D267Q in

DR5-8 provided only modest improvements in DR5
affinity and bioactivity [9]. To clarify the role of this residue we have generated DR5-B mutant variant by changing
back Q 267D in DR5-A as in wild type TRAIL (Table 1).
Determination of dissociation constants for binding
of TRAIL variants to different receptors
Dissociation constants for binding of TRAIL to the
receptors were determined in SPR assays for direct binding
of each TRAIL variant to immobilized receptors. Receptors
were immobilized to the sensor chip (Bio-RAD) with a
high density (6,0008,000 resonance units) to get a sufficient signal for mutant variants. For preliminary estimation
highly purified trimeric TRAIL variants were injected
simultaneously at 250 nM to the 6 9 6 channel sensor chip
with five immobilized receptors (Fig. 2). The chip was
regenerated and sensorgrams were recorded simultaneously
for five concentrations of TRAIL variants ranging from 15
to 250 nM in two-fold increments. KD values were calculated as described in methods.
Our KD values for wt TRAIL (114281) perfectly fits
with the data reported earlier [16]. For example, the
reported KD values for wt TRAIL/Apo2L (91281) binding
to DR4-IgG, DR5-IgG, and DcR2-IgG were 0.8 0.3,
0.9 0.4, and 0.3 0.2 nM, respectively, while our
measurements gave 0.51 0.028, 0.46 0.014, and
0.36 0.009 nM for the same receptors.
Analysis of sensograms demonstrated that binding of
TRAIL variants DR5-A and DR5-B as well as DR5-8 to
DR4 and DcR1 receptors was not detected while KD for
binding of D269H variant to these receptors was increased
only 18.5 and 8 fold, respectively in comparison with wild
type TRAIL (Table 2). At the same time the dissociation
constant of D269H for DR5 receptor was four and ninefold
lower than KD of wild type and DR5-8, respectively. The
insertion of D269H mutation into DR5-8 in DR5-A variant
resulted in 3.5 fold decrease of KD for DR5 without any
loss of selectivity to the other receptors. Affinity of all
these TRAIL variants to DcR2 was only modestly lower
than of wild type but D267Q substitution in DR5-A variant

Table 1 Amino acid changes of TRAIL mutant variants

TRAIL variants

Amino acid changes

References

Wt TRAIL

DR4-8

Y189A

Q193S

N199 V

K201R

Y213 W

S215D

DR5-8

Y189N

R191K

Q193R

H264R

I266L

D267Q

[9]

D269H

D269H

[10]

DR5-A

Y189N

R191K

Q193R

H264R

I266L

D267Q

D269H

Present report

DR5-B

Y189N

R191K

Q193R

H264R

I266L

D269H

Present report

[9]

123

782

Apoptosis (2009) 14:778787

Fig. 2 Sensorgrams of Surface Plasmon Resonance (SPR) of TRAIL

variants binding to the five receptors. Five ligands (wild type TRAIL
and DR5-8, D269HD, DR5-A, DR5-B variants) each at concentration

250 nM were passed simultaneously through five parallel channels of

6 9 6 sensor chip with immobilized DR4, DR5, DcR1, DcR2 and
OPG receptors

Table 2 Dissociation constants of TRAIL variants to five receptors measured by SPR

TRAIL variants

DR5

DR4

KD, 10-9 M DcR1

DcR2

OPG

Wild type

0.51 0.028

0.46 0.014

1.09 0.069

0.36 0.009

0.99 0.017

D269H

0.13 0.003

8.47 0.335

8.69 0.628

0.73 0.014

5.26 0.980

DR5-8

1.18 0.017

NB

NB

3.32 0.048

10.9 0.448

DR5-A

0.33 0.005

NB

NB

4.18 0.082

12.2 0.138

DR5-B

0.71 0.013

NB

NB

101 7.0

221 19a

143 13
DR4-8

NB

3.26 0.010

2.49 0.950

2.10 0.075

21.0 0.140

Standard deviation was calculated by fitting of sensorgrams using ProteOn Manager Version 2.0.1 software (Bio-Rad) as Chi-square criterion
NB no detectable binding
a

KD determined from equilibrium plots (see supplemented data)

(DR5-B variant) dramatically improves selectivity of

TRAIL practically without affecting the affinity to DR5
receptor (Table 2). DR5-B variant demonstrated 280 and
223 fold lower affinity to DcR2 and OPG receptors,
respectively in comparison with wild type TRAIL. In these
cases the rates of dissociation of the receptor-ligand complexes were very high. The reliable value of KD for OPG
was calculated only by measuring of equilibrium binding.
KD for DcR2 was calculated with both kinetics and equilibrium methods that gave the similar results (see Supplemented data and Table 2). These data indicated that D267
position in combination with the substitutions in DR5-A

123

plays an important role in TRAIL-DcR2 or TRAIL-OPG

interactions. Binding of DR5-B variant to DR4 and DcR1
receptors was not detected.
Induction of apoptosis by receptor-selective TRAIL
variants
The biological activity of TRAIL variants was determined
in various cancer cells with different expression of death
and decoy receptors on the cell surface: human cervical
carcinoma HeLa cells expressing DR4 and DR5 in equal
amount and practically non of DcR1 and DcR2 receptors;

Apoptosis (2009) 14:778787

783

human T leukemia Jurkat cells expressing mainly DR5

receptor [7]; monoblastic leukemia U937 cells expressing
mainly DR5 and DcR2 receptors [17, 18] and myeloid
leukemia K562 cells expressing all four receptors [19].
In Hela cells, where apoptosis was mostly mediated
through DR5 receptor, DR-5A and DR5-B variants induced
apoptosis several fold more effectively than wild type
TRAIL (Fig. 3, Table 3). Half-maximal dose of DR5-8
inducing apoptosis in these cells was similar to wild type
but the maximal effect was reduced. In Jurkat cells DR5-A
and DR5-B variants induced apoptosis more effectively
than wild type, DR4-8, DR5-8 or D269H variants (Fig. 3;
Table 3). In contrast, in DR4-responsive K562 cells DR4-8
variant was as effective as wild type TRAIL, when DR5-8,
DR5-A and DR5-B variants demonstrated reduced activity.
In this cell line D269H variant at low concentrations (up to
25 ng/ml) was as effective as DR4-8 and wild type TRAIL

with maximum 60% cell death. This was a saturated percent for wild type and DR4-8 variant when D269H variant
at higher concentrations of ligand (up to 400 ng/ml)
appeared even higher effectivity (up to 85% cell death)
(Fig. 3).
In U937 monoblastic leukemia cell line cytotoxic
activity of DR5-8, DR5-A and DR5-B variants was significantly higher than of wild type TRAIL showing both an
increased maximum response (almost twice) and 1.8, 3.0
and 3.4 decreased EC50 values, respectively. D269H variants activity was similar to the wild type in these cells.
In our previous studies we have demonstrated that
expression of Bcl-2 blocks TRAIL-induced release of
cytochrome c from mitochondria into cytosol and apoptosis
in Hela cells [20]. No one of the DR5-selective TRAIL
variants could overcome this block of apoptosis (Fig. 4).
We have shown earlier that the inhibitor of microtubules,

Fig. 3 The biological activity of wild type TRAIL and DR5-selective

variants in HeLa, Jurkat, K562 and U937 cell lines. Viability was
determined as described in Materials and methods 24 h after

addition of TRAIL variants. To increase sensitivity to TRAIL

mediated apoptosis in Hela and U937 cell lines 0.05 and 0.25 lg/
ml emetine, respectively, was added in the growth media

123

784

Apoptosis (2009) 14:778787

Table 3 ED50 values of TRAIL variants induced apoptosis on different cell lines
TRAIL variants HeLa (?0.05 lg/ml emetine) K562

U937 (?0.25 lg/ml emetine)

Jurkat

EC50 (ng/ml) Cell death (%) EC50 (ng/ml) Cell death (%) EC50 (ng/ml) Cell death (%) EC50 (ng/ml) Cell death (%)
Wild type

6.2 0.5

93 6

0.22 0.02

60 4

0.28 0.03

37 3

0.27 0.03

D269H

2.1 0.2

91 6

0.23 0.03a 85 7

0.17 0.02

43 3

0.27 0.03

25 2

DR5-8

6.0 0.6

45 3

0.27 0.02

0.15 0.02

48 4

0.15 0.02

43 3

27 3

22 2

DR5-A

2.2 0.2

86 7

0.73 0.06

34 3

0.07 0.01

56 5

0.09 0.01

43 4

DR5-B

1.5 0.1

78 6

1.11 0.07

35 3

0.07 0.01

53 5

0.08 0.01

50 3

DR4-8

1.6 0.2

25 3

0.23 0.02

62 7

0.71 0.06

26 3

0.26 0.02

14 3

The results are expressed as mean SE of at least three independent experiments

a

Value of EC50 was calculated for concentration of TRAIL up to 25 ng/ml

Fig. 4 Induction of apoptosis in HeLa-Bcl-2 cells by TRAIL variants

in combination with cytochalasin D and taxol. Cells were incubated
with TRAIL variants in the presence of protein synthesis inhibitor
emetine (0.3 lg/ml) for 8 h or in the absence of emetine for 24 h.
Inhibitor of microtubules taxol (10 lM) and inhibitor of actin

microfilaments cytochalasin D (1 lg/ml) were added simultaneously

with TRAIL. Apoptosis was determined by counting cell nuclei with
condensed chromatin after staining with Hoechst 33342 (1 lg/ml).
Data are the mean SD from three independent experiments

taxol and the inhibitor of actin microfilaments, cytochalasin D enhanced efficiency of TRAIL and allow it to
overcome antiapoptotic effect of Bcl-2 [20]. All the DR5selective TRAIL variants did not lose the capacity to break
Bcl-2 block in combination with the inhibitors of cytoskeleton (Fig. 4). It should be mentioned that TRAILinduced apoptosis in HeLa cells was strongly enhanced by
inhibitor of the protein synthesis, emetine. The same effect
in the case of TNF-induced apoptosis was traditionally
ascribed to inhibition of NF-jB-dependent expression of
antiapoptotic proteins. As one can see from Fig. 4, in
combination with all TRAIL variants taxol but not cytochalasin D completely substitute emetine. Moreover, the
effect of taxol was not summated with emetine. The nature
of this effect is currently under investigation.
All our results clearly demonstrated that DR5-A and
DR5-B variants are more specific to DR5 receptor and
demonstrate higher biological activity in comparison to
DR5-8 and D269H TRAIL variants.

Discussion

123

Existence of five receptors creates considerable difficulties

for investigation of TRAIL-induced apoptosis. First,
expression of DR4 and DR5 receptors do not generally
show a direct correlation with sensitivity to apoptosis
stimulation [21]. Second, the molecular mechanisms
involved in TRAIL-induced cell death inhibition by decoy
receptors DcR1 and DcR2 differ in important ways. DcR1
acts merely as a competitor for TRAIL binding, preventing
DR5-associated DISC assembly, while DcR2 impairs DISC
processing and initiator caspase activation [7]. DcR2
interacts with DR5 in the native DISC in a TRAIL
dependent manner and prevents DR4 corecruitment to
DR5. Another study reported that inhibition of apoptosis by
DcR2 does not depend on ligand binding of the receptor.
Rather, inhibition of apoptosis critically depends on the
formation of ligand-independent complexes between DR5
and DcR2 [7]. To facilitate the investigation of TRAIL

Apoptosis (2009) 14:778787

induced apoptosis several laboratories have generated DR4

[9, 22, 23] and DR5 selective [9, 10] mutant variants of
TRAIL. All the generated mutants demonstrated reduction
of affinity to decoy receptors and OPG in different extent.
In this study we generated two new DR5 specific TRAIL
variants DR5-A and DR5-B with improved selectivity to
DR5 receptor and high biological activity. In DR5-A variant we simply combined DR5-8 and D269H variants
described before. Using SPR binding measurements we
compare the affinities to all five TRAIL receptors of previously described DR5 selective TRAIL variants DR5-8
and D269H with our generated variants DR5-A and
DR5-B. The dissociation constant of DR5-8 variant to DR5
receptor was ninefold higher in comparison to D269H.
However, the DR5-8 variant was more selective ligand as
far as practically no binding to DR4 and DcR1 receptors
was detected in SPR measurements while D269H variant
has only reduced (18.4 and 8.6 fold, respectively) affinity
to these receptors. In addition D269H variant had also
higher affinity to DcR2 and OPG in comparison with
DR5-8 variant. As it was expected the insertion of D269H
mutation in amino acid sequence of DR5-8 TRAIL variant
(variant DR5-A) resulted in 3.6 fold increase in affinity to
DR5 receptor without significant changes of its selectivity
(Table 2).
Very interesting results were obtained with generated
DR5-B variant where mutation D267Q in DR5-A variant
was changed back to Q267D as in wild type. In phage
display experiments it has been shown that this amino acid
is not critical for DR4 binding, but D267Q substitution in
DR5-8 variant improves the biological activity [9]. Here
we have demonstrated that D267 is a critical amino acid to
reduce affinity to DcR2 and OPG. Dissociation constant of
DR5-B variant to DcR2 was increased 24 and 280 fold in
comparison to DR5-A and wild type TRAIL, respectively.
The affinity to OPG also was reduced 18 folds in DR5-B in
comparison to DR5-A. These SPR binding data indicated
that DR5-B mutant was the most selective to DR5 receptor
among all other variants.
Apoptotic activity of the new TRAIL variants was
analyzed in the cell lines with different selectivity and
sensitivity to TRAIL-induced apoptosis. We have compared our mutants not only with wild type TRAIL but also
with well characterized mutants DR5-8, DR4-8 and D269H
described earlier [9, 10]. Also it was shown that expression
level of DR4 and DR5 in various cell lines did not correlate
with the selectivity of receptor-mediated apoptosis [9, 21,
2426].
Here we have demonstrated that the insertion of D269H
mutation in amino acid sequence of TRAIL DR5-8 variant
improved the efficiency of TRAIL in HeLa, Jurkat and
U937 cells where apoptosis was mostly mediated by DR5
(Fig. 3). The most important was strongly decreased

785

affinity of the new TRAIL variants DR5-A and DR5-B to

DcR2 and OPG. This could be illustrated by induction of
apoptosis in U937 cells expressing mainly DR5 and DcR2
receptors [17, 18]. TRAIL D269H variant had significantly
higher affinity to DR5 than wild type TRAIL but their
proapoptotic efficiency was equally low. It may be suggested that high affinity of D269H variant to DcR2 resulted
in co-recruitment of DcR2 with DR5 within the same DISC
and inhibition of initiator caspase activation [6, 7]. DR5-A
and DR5-B variants have low affinity to DcR2 and this
could be the possible reason why they were more effective
than wild type, DR5-8 or D269H variant in U937 cells. The
detailed analysis of DcR2-DR5 DISC formation induced by
different TRAIL variants is necessary to verify this
suggestion.
In HeLa cells the efficiency of the new TRAIL variants
was similar to D269H mutant. DR4-selective TRAIL variant DR4-8 induced apoptosis only in a small (*25%)
fraction of the cells but the efficiency of receptor-mediated
signaling is remarkably high and similar to that for DR5-B
(EC50 = 1.6 ng/ml and EC50 = 1.5 ng/ml, respectively)
while DR5-B induced apoptosis in 78% of the cells. In
HeLa cells both DR4 and DR5 are expressed and apoptosis
could be mediated by the both receptors. Our data indicated
that TRAIL induced apoptosis through DR4 as effectively
as through DR5 only in a fraction of Hela cells but in the
majority of cell population DR4-dependent pathway was
suppressed.
In Jurkat cells DR5-A and DR5-B variants were more
effective than the other DR5-selective mutants (D269H and
DR5-8). Similarly to HeLa only 26% of Jurkat cells were
killed by DR4-8 variant. Earlier it has been shown that in
Jurcat cells agonistic DR4 mAB induced apoptosis several
folds less effective in comparison to DR5 mAB when cell
surface expression of DR4 receptor was undetectable [27].
In complete agreement with these observations efficiency
of DR4-8 variant in our experiments was 2.5, 4.2, 4.7 and
10 fold lower in comparison to wild type, D269H, DR5-8,
DR5-A and DR5-B variants, respectively (Table 3).
In the case of K562 cells the apoptotic effects of different TRAIL variants were complicated. These cells
express all four TRAIL receptors [19] but in some laboratories K562 cells were found to be resistant to TRAIL
[27, 28]. The authors came to conclusion that in CLL,
K562 and U937 cells there was no clear relationship
between cell surface expression of DR4 or DR5 and sensitivity to TRAIL-induced apoptosis [27, 28]. It was shown
that ionizing radiation sensitized K562 via stimulation of
DR4-mediated apoptosis [29] while inhibitors of HDACi
(histone deacetylase inhibitors) stimulated DR5-mediated
apoptosis [27]. In the other laboratories K562 cells were
quite sensitive to TRAIL [19, 2931] and our data are in
agreement with these observations. The receptor specificity

123

786

of apoptosis was not analyzed in these publications. In our

experiments DR4-selective DR4-8 variant was as effective
as wild type TRAIL, while DR5-selective variants DR5-8,
DR5-A and DR5-B which demonstrated no detectable
binding to DR4 receptor were less effective. These data
indicated that K562 in our laboratory was DR4-responsive
cell line. Induction of apoptosis by DR5-selective variant
D269H in K562 cells was quite unexpected. The KD values
of D269H and DR4-8 variants to DR4 receptor were 18.4
and 7.0 fold higher than for wild type TRAIL, respectively
(Table 2). However, these mutants induced apoptosis as
effectively as wild type TRAIL (Table 3). Probably this
phenomenon is related to the effect of the decoy receptor
DcR1 that inhibits TRAIL-induced cell death by competitive binding of TRAIL [7]. The affinities to DR4 and DcR1
receptors are decreased for D269H (18.4 and 7.9 fold,
respectively) and DR4-8 (7.0 and 2.5 fold, respectively)
variants in parallel. Overriding protective effect of DcR1
these two mutants induced apoptosis in K562 cells with the
same efficiency as wt TRAIL. The reduced affinity of
D269H variant to DcR1 may explain why this mutant at
high concentrations (more than 25 ng/ml) demonstrated
even stronger apoptotic activity than wild type TRAIL
(Fig. 2). We can not exclude also that D269H variant
bypassed internal block of DR5-mediated apoptosis in
K562 due to its very high affinity to DR5 receptor. In this
respect, previous studies have demonstrated an extreme
complexity of the expression and function of TRAIL
receptors in various cell types [24, 32].
Summarizing our results we can assume that generation
of new DR5 specific variants of TRAIL DR5-A and DR5-B
would be useful for studies on the role of different receptors in TRAIL mediated apoptosis in different tumors cells.
High specificity and biological activity of DR5-A and
DR5-B TRAIL variants makes them good candidates for
therapy of resistant tumors.
Acknowledgments The authors would like to thank Natalia
Magretova and Peter Kalmykov from A. N. Belozersky Institute of
Physico-Chemical Biology, Moscow State University for help in
analytical centrifugation. The work was supported by Federal Target
Programme Research and Development in Priority Fields of Scientific and Technologic Complex of Russia for 20072012 (contract
# 02.522.11.2005).

References
1. Ashkenazi A, Pai RC, Fong S, Leung S, Lawrence DA, Marsters
SA et al (1999) Safety and anti-tumor activity of recombinant
soluble Apo2 ligand. J Clin Invest 104:155162. doi:10.1172/
JCI6926
2. LeBlanc HN, Ashkenazi A (2003) Apo2L/TRAIL and its death
and decoy receptors. Cell Death Differ 10:6675. doi:10.1038/
sj.cdd.4401187

123

Apoptosis (2009) 14:778787

3. Ashkenazi A, Dixit VM (1998) Death receptors: signaling and
modulation. Science 281:13051308. doi:10.1126/science.281.
5381.1305
4. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS,
Luthy R et al (1997) Osteoprotegerin: a novel secreted protein
involved in the regulation of bone density. Cell 89:159161.
doi:10.1016/S0092-8674(00)80209-3
5. Bouralexis S, Findlay DM, Atkins GJ, Labrinidis A, Hay S,
Evdokiou A (2003) Progressive resistance of BTK-143 osteosarcoma cells to Apo2L/TRAIL-induced apoptosis is mediated by
acquisition of DcR2/TRAIL-R4 expression: resensitisation with
chemotherapy. Br J Cancer 89:206214. doi:10.1038/sj.bjc.
6601021
6. Clancy L, Mruk K, Archer K, Woelfel M, Mongkolsapaya M,
Screaton G et al (2005) Preligand assembly domain-mediated
ligand independent association between TRAIL receptor 4 (TR4)
and TR2 regulates TRAIL-induced apoptosis. Proc Natl Acad Sci
USA 102:1809918104. doi:10.1073/pnas.0507329102
7. Merino D, Lalaoui N, Morizot A, Schneider P, Solary E, Micheau
O (2006) Differential inhibition of TRAIL-mediated DR5-DISC
formation by decoy receptors 1 and 2. Mol Cell Biol 26:7046
7055. doi:10.1128/MCB.00520-06
8. Merino D, Lalaoui N, Morizot A, Solary E, Micheau O (2007)
TRAIL in cancer therapy: present and future challenges. Expert
Opin Ther Targets 11:12991314. doi:10.1517/14728222.
11.10.1299
9. Kelley FR, Totpal K, Lindstrom SH, Mathieu M, Billeci K, de
Forge L et al (2005) Receptor-selective mutants of Apo2L/
TRAIL reveal a greater contribution of DR5 than DR4 to apoptosis signaling. J Biol Chem 280:22052212. doi:10.1074/
jbc.M410660200
10. van der Sloot AM, Tur V, Szegezdi TE, Mullally MM, Cool RH,
Samali A et al (2006) Designed tumor necrosis factor-related
apoptosis-inducing ligand variants initiating apoptosis exclusively via the DR5 receptor. Proc Natl Acad Sci USA 103:8634
8639. doi:10.1073/pnas.0510187103
11. Gasparian ME, Ostapchenko VG, Yagolovich AV, Tsygannik IN,
Chernyak BV, Dolgikh DA et al (2007) Overexpression and
refolding of thioredoxin/TRAIL fusion from inclusion bodies and
further purification of TRAIL after cleavage by enteropeptidase.
Biotechnol Lett 29:15671573. doi:10.1007/s10529-007-9446-y
12. Kunkel TA (1985) Rapid and efficient site-specific mutagenesis
without phenotypic selection. Proc Natl Acad Sci USA 82:488
492. doi:10.1073/pnas.82.2.488
13. Gasparian ME, Ostapchenko VG, Schulga AA, Dolgikh DA,
Kirpichnikov MP (2003) Expression, purification and characterization of human enteropeptidase catalytic subunit in Escherichia
coli. Protein Expr Purif 31:133139. doi:10.1016/S1046-5928
(03)00159-1
14. Schuck P (2000) Size-distribution analysis of macromolecules by
sedimentation velocity ultracentrifugation and Lamm equation
modeling. Biophys J 78:16061619. doi:10.1016/S0006-3495(00)
76713-0
15. Agol VI, Belov GA, Bienz K, Egger D, Kolesnikova MS, Romanova LI et al (2000) Competing death programs in poliovirusinfected cells: commitment switch in the middle of the infectious
cycle. J Virol 74:55345541. doi:10.1128/JVI.74.12.5534-5541.
2000
16. Hymowitz SG, OConnell MP, Ultsch MH, Hurst A, Totpal K,
Ashkenazi A et al (2000) A unique zinc-binding site revealed by
a high-resolution X-ray structure of homotrimeric Apo2L/
TRAIL. Biochemistry 39:633640. doi:10.1021/bi992242l
17. Wen J, Ramadevi N, Nguyen D, Perkins C, Worthington E,
Bhalla K (2000) Antileukemic drugs increase death receptor 5
levels and enhance Apo-2L-induced apoptosis of human acute
leukemia cells. Blood 96:39003906

Apoptosis (2009) 14:778787

18. Phillips TA, Ni J, Pan G, Ruben SM, Wei Y, Pace JL et al (1999)
TRAIL (Apo-2L) and TRAIL receptors in human placentas:
Implications for immune privilege. J Immunol 162:60536059
19. Mirandola P, Gobbi G, Ponti C, Sponzilli I, Cocco L, Vitale L
(2006) PKCe control protection against TRAIL in erythroid
progenitors. Blood 107:508513. doi:10.1182/blood-2005-072676
20. Gasparian ME, Domnina LV, Ivanova OY, Izyumov DS,
Lomakin AY, Popova EN et al (2008) Cytoskeleton inhibitors
combined with TRAIL induce apoptosis in HeLa Carcinoma cells
overexpressing antiapoptotic protein Bcl-2. Biochemistry (Mosc)
73:358362
21. Ashkenazi A (2002) Targeting death and decoy receptors of the
tumour-necrosis factor superfamily. Nat Rev Cancer 2:420430.
doi:10.1038/nrc821
22. MacFarlane M, Kohlhaas SL, Sutcliffe MJ, Dyer MJS, Cohen
GM (2005) TRAIL receptor-selective mutants signal to apoptosis
via TRAIL-R1 in primary lymphoid malignancies. Cancer Res
65:1126511270. doi:10.1158/0008-5472.CAN-05-2801
23. Tur V, van der Sloot AM, Reis CR, Szegezdi E, Cool RH, Samali
A et al (2008) DR4-selective tumor necrosis factor-related
apoptosis-inducing ligand (TRAIL) variants obtained by structure-based design. J Biol Chem 283:2056020568. doi:10.1074/
jbc.M800457200
24. Ashkenazi A, Dixit VM (1999) Apoptosis control by death and
decoy receptors. Curr Opin Cell Biol 11:255260. doi:10.1016/
S0955-0674(99)80034-9
25. Griffith TS, Rauch ST, Smolak PJ, Waugh JY, Boiani N, Lynch
DH et al (1999) Functional analysis of TRAIL receptors using
monoclonal antibodies. J Immunol 162:25972605

787
26. Ashkenazi A, Herbst RS (2008) To kill a tumor cell: the potential
of proapoptotic receptor agonists. J Clin Invest 118:19791990.
doi:10.1172/JCI34359
27. MacFarlane M, Inoue S, Kohlhaas SL, Majid A, Harper N,
Kennedy DBJ et al (2005) Chronic lymphocytic leukemic cells
exhibit apoptotic signaling via TRAIL-R1. Cell Death Differ
12:773782. doi:10.1038/sj.cdd.4401649
28. Inoue S, MacFarlane M, Harper N, Wheat LMC, Dyer MJS,
Cohen GM (2004) Histone deacetylase inhibitors potentiate TNFrelated apoptosis-inducing ligand (TRAIL)-induced apoptosis in
lymphoid malignancies. Cell Death Differ 11(Suppl 2):193206.
doi:10.1038/sj.cdd.4401535
29. Pietro R, Secchiero P, Rana R, Gibellini D, Visani G, Bemis K
et al (2001) Ionizing radiation sensitizes erythroleukemic cells
but not normal erythroblasts to tumor necrosis factor-related
apoptosis-inducing ligand (TRAIL)-mediated cytotoxicity by
selective up-regulation of TRAIL-R1. Blood 97:25962603.
doi:10.1182/blood.V97.9.2596
30. Perez-Cruz I, Carcamo JM, David W, Golde DW (2007) Caspase8 dependent trail-induced apoptosis in cancer cell lines is inhibited by vitamin C and catalase. Apoptosis 12:225234.
doi:10.1007/s10495-006-0475-0
31. Gregorini A, Tomasetti M, Cinti C, Colomba D, Colomba S
(2006) CD38 expression enhances sensitivity of lymphoma T and
B cell lines to biochemical and receptor-mediated apoptosis. Cell
Biol Int 30:727732. doi:10.1016/j.cellbi.2006.05.004
32. Kim K, Fisher MJ, Xu SQ, El-Deiry WS (2000) Molecular
determinants of response to TRAIL in killing of normal and
cancer cells. Clin Cancer Res 6:335346

123