Intrinsic disorder and coiled-coil formation in prostate
apoptosis response factor 4
David S. Libich
1,
*, Martin Schwalbe
1,
*, Sachin Kate
1
, Hariprasad Venugopal
1
, Jolyon K. Claridge
1
,
Patrick J. B. Edwards
1
, Kaushik Dutta
2
and Steven M. Pascal
1
1 Centre for Structural Biology, Institute of Fundamental Sciences and Department of Physics, Massey University, Palmerston North,
New Zealand
2 New York Structural Biology Centre, NY, USA
Introduction
Prostate apoptosis response factor-4 (Par-4) is an ubi-
quitously expressed and evolutionary conserved protein
that was initially identified as a pro-apoptotic factor in
rat AT-3 androgen-independent prostate cancer cells
exposed to ionomycin [1,2]. The identified pro-apopto-
tic and tumour-suppressive roles of Par-4 are consid-
ered to be its most important cellular functions and,
accordingly, Par-4 is downregulated in various cancers

doi:10.1111/j.1742-4658.2009.07087.x
Prostate apoptosis response factor-4 (Par-4) is an ubiquitously expressed
pro-apoptotic and tumour suppressive protein that can both activate cell-
death mechanisms and inhibit pro-survival factors. Par-4 contains a highly
conserved coiled-coil region that serves as the primary recognition domain
for a large number of binding partners. Par-4 is also tightly regulated by
the aforementioned binding partners and by post-translational modifica-
tions. Biophysical data obtained in the present study indicate that Par-4
primarily comprises an intrinsically disordered protein. Bioinformatic
analysis of the highly conserved Par-4 reveals low sequence complexity and
enrichment in polar and charged amino acids. The high proteolytic suscep-
tibility and an increased hydrodynamic radius are consistent with a largely
extended structure in solution. Spectroscopic measurements using CD and
NMR also reveal characteristic features of intrinsic disorder. Under physio-
logical conditions, the data obtained show that Par-4 self-associates via the
C-terminal domain, forming a coiled-coil. Interruption of self-association
by urea also resulted in loss of secondary structure. These results are
consistent with the stabilization of the coiled-coil motif through an intra-
molecular association.
Abbreviations
CREB, cAMP-responsive element-binding protein; DLS, dynamic light scattering; GST, glutathione S-transferase; HSQC, heteronuclear single
quantum coherence; IDP, intrinsically disordered protein; IPTG, isopropyl thio-b-
D-galactoside; LZ, leucine zipper; NLS, nuclear localization
sequence; Par-4, prostate apoptosis response factor 4; PK, protein kinase; SAC, selective apoptosis of cancer cells.
3710 FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS
neuroblastomas [11], as well as endometrial [12],
pancreatic [13] and gastric [8] cancers.
In addition to its role in cancer, Par-4 is thought to
assist in normal neuronal development by preventing
the hyper-proliferation of nerve tissues, in turn con-

of cytokine signalling box proteins 1, 2 and 4 [37].
Par-4 contains several conserved phosphorylation
sites that are modified by kinases, such as PKA, PKC,
casein kinase II and Akt1, adding a further level of
regulation of the function of Par-4 [38]. Phosphoryla-
tion of an absolutely conserved threonine (rat T155,
human T163 or mouse T156; Fig. 1) by PKA is
required for nuclear translocation [8]. Phosphorylation
of a C-terminal serine residue (rat S249, human or
mouse S231; Fig. 1) by Akt1 effectively inactivates
Par-4 by allowing the chaperone protein 14-3-3 to bind
and sequester it in the cytoplasm, even if it is
phosphorylated on T155 [26].
These multiple interactions coupled with a high
degree of sequence conservation and post-translational
modification suggest that the in vivo role(s) of Par-4
are highly temporally and spatially regulated. Simi-
larly, the ubiquitous expression, post-translational
modifications and a plethora of binding partners are
characteristics common to many intrinsically disor-
dered proteins (IDPs) [39]. In the present study, we
demonstrate that residual structure exists in Par-4
because the measured hydrodynamic radius increased
under denaturing conditions, suggesting that the
ensemble becomes less compact. CD and NMR indi-
cate that Par-4 is primarily intrinsically disordered
under physiological conditions and exists as an ensem-
ble of fast-averaging (on the NMR time-scale) struc-
tures. Furthermore, Par-4 forms a stable coiled-coil
through a self-association event mediated by the C-ter-

(Fig. 1). Furthermore, there is a high degree of
sequence conservation in the C-terminal quarter of Par-
4, which contains primarily a coiled-coil-like sequence
(residues 254–332; Figs 1 and 2A). In particular, a leu-
cine zipper (residues 292–330), which is a subset of the
coiled-coil domain, is almost conserved in all known
Par-4 sequences, suggesting a common functionality
(Figs 1 and 2A). Relatively few Par-4 genes have been
sequenced. It has been suggested that the general pat-
tern of sequence conservation shown in Fig. 1 is likely
to be conserved across other mammalian sequences [1].
D. S. Libich et al. Intrinsic disorder in Par-4
FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS 3711
Based on disembl analysis [41], the majority (> 70%)
of Par-4 is predicted to be disordered. The putative
regions of order in Par-4, as indicated by grey bars in a
disembl plot (Fig. 2B), align with or occur within func-
tionally important regions of Par-4 (Fig. 2A), namely
NLS1, NLS2, SAC and the coiled-coil. Secondary
Fig. 1. Sequence alignment of the prostate apoptosis response factor 4 (Par-4). A BLASTP ⁄ CLUSTALW [102,103] alignment of sequences of
Par-4 from various species: rat (Rattus norvegicus), mouse (Mus musculus), human (Homo sapiens), African clawed frog (Xenopus laevis)
and zebra fish (Danio rero). The amino acids are coloured: red (nonpolar side chains: G, A, V, L, I, M, P, F and W), blue (polar side chains: S,
T, N, Q, Y and C) and green (polar, charged side chains: K, R, H, D and E). Symbols: residues in that column are identical in all sequences
(*); substitutions are conservative (:); and substitutions are semi-conservative (.). The high degree of sequence conservation of Par-4
suggests functional significance and thus resistance to evolutionary pressure. With reference to the numbering of rat Par-4, several seg-
ments are of notable interest: two nuclear localization sequences [NLS1 (20–25) and 2 (137–153)], which are completely conserved among
all known Par-4s, and the SAC domain (137–195), which is defined by being the absolute minimum fragment required for apoptosis and
includes NLS2 [6]. The C-terminal domain (254–332) is a coiled-coil (CC) motif that encompasses a LZ (292–330) as a subset. Two important
phosphorylation sites, T155 and S249, are denoted by red arrows.
Intrinsic disorder in Par-4 D. S. Libich et al.

of functional importance, including the nuclear localization sequences [NLS1 (20–25) and 2 (137–153), coloured green], the region necessary
for SAC (137–195), the coiled-coil C-terminal domain (CC, 254-332, coloured red) and the LZ (292–330, shown with hatching). The rrPar-
4DLZ construct lacks residues 291–332, which is approximately one-half of the coiled-coil and the entire leucine zipper. The rrPar-4SAC con-
struct represents residues 137–195 of Par-4, including NLS2. All three constructs used in the present study have an N-terminal GGS tag, a
remnant from the cleavage of the purification tag, which is omitted here for simplicity. (B)
DISEMBL predicts regions of order ⁄ disorder in pro-
teins using neural networks trained on multiple definitions of disorder [41]. The dashed line in (B) represents a threshold value separating
order and disorder. (C) Secondary structure (a-helix only shown) prediction using
GOR4 [42] and (D) hydrophobic cluster analysis (HCA) [43], a
visually enhanced representation of the primary sequence that highlights clustering of hydrophobic residues using symbols (
,T; ,S;¤,G;
w, P) and colours (red: P and acidic residues D, E, N, Q; blue: basic residues, H, K, R; green: hydrophobic residues, V, L, I, F, W, M, Y;
black: all other residues, G, S, T, C, A). The grey bars indicate the predicted regions of order in (B) and, for comparison, are extended over
(C) and (D).
D. S. Libich et al. Intrinsic disorder in Par-4
FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS 3713
versus a set of folded proteins (black bars). Positive
values indicate a depletion, whereas negative bars indi-
cate an enrichment relative to folded proteins. The pat-
tern of amino acid usage for rrPar-4FL (grey bars) is
in accordance with that generally observed for IDPs
[45,46], namely a depletion of order-promoting amino
acids (L, N, F, Y, I, W, C) and enrichment of dis-
order-promoting residues (S, Q, K, P, E). The amino
acid usage for rrPar-4DLZ and rrPar-4SAC follows a
similar pattern (not shown).
As calculated (i.e. from sequence) and experimen-
tally determined [i.e. from MS, Tricine-PAGE and
dynamic light scattering (DLS)], the molecular weights
for rrPar-4FL, rrPar-4DLZ and rrPar-4SAC are given

a denaturing Tricine-PAGE system (see Experimental
procedures) determined apparent molecular weights of
49.1, 41.5 and 12.4 kDa, respectively. These sizes are
significantly larger (36%, 33% and 77% larger for
rrPar-4FL, rrPar-4DLZ and rrPar-4SAC, respectively)
than the expected MWs determined from the primary
structure or MS (Table 1).
The results of DLS experiments are shown in Table 2
and summarized in Table 1. The measured R
S
for
rrPar-4FL was 189 A
˚
, which is much larger than
expected for a monomeric random coil, suggesting a
polymeric state for rrPar-4FL under these conditions.
Fig. 3. (A) Charge ⁄ hydrophobicity plot of rrPar-4FL (335 residues),
rrPar-4DLZ (293 residues), and rrPar-4SAC (61 residues). The divid-
ing line

R ¼ 2:785

H À 1:151 represents an empirically determined
divisor between intrinsically disordered (high charge, low hydropho-
bicity) and structured (low charge, high hydrophobicity) space.
Proteins such as aprotinin [104], actin [105], ubiquitin [106] and 3C
protease [107] are plotted as examples of classically folded
proteins. (B) Sequence complexity of rrPar-4FL (grey bars) com-
pared with the average amino acid distribution of IDPs (black bars)
relative to the average amino acid distribution of globular proteins.

(calculated) for rrPar-4FL as
either monomeric globular (i.e. folded), molten globule,
pre-molten globule, extended chain or urea-denatured
states are given in Table 2. The experimentally deter-
mined R
S
for rrPar-4DLZ (32.5 A
˚
) and rrPar-4SAC
(20.9 A
˚
) are larger than the expected folded R
S
(25.1
and 14.8 A
˚
, respectively) but still smaller than the cal-
culated random coil R
S
for either protein (Table 2).
This suggests that these constructs exist in an unfolded
yet monomeric form under these conditions. The
volume weighted distributions for rrPar-4FL, rrPar-
4DLZ and rrPar-4SAC are shown in the Supporting
information (Fig. S1A). The relatively broad distribu-
tion of sizes recorded for all three proteins is consistent
with an ensemble of interconverting conformations
rather than one single conformation.
Upon addition of 1 m urea, the measured R
S

a-helical character in rrPar-4FL is immediately evident
and remains stable up to 65 °C (Fig. 5A,B). By con-
trast, the CD spectra for rrPar-4DLZ (Fig. 5C,D) and
rrPar-4SAC (Fig. 5E,F) show a typical profile of IDPs
with a deep transition at 200 nm [51].
Pairwise overlays of
1
H-
15
N heteronuclear single
quantum coherence (HSQC) spectra for rrPar-4FL,
rrPar-4DLZ and rrPar-4SAC are shown in Fig. 6. The
spectra of all three proteins display the features that
characterize disorder in proteins, namely sharp peaks
and narrow
1
H chemical shift dispersion [51,52].
Chemical shift similarities indicate some structural
similarity between rrPar-4FL and rrPar-4DLZ
(Fig. 6A). Fewer peaks share similar chemical shifts
when comparing rrPar-4FL or rrPar-4DLZ with rrPar-
4SAC (Fig. 6B,C). Thus, the majority of residues in
rrPar-4SAC experience a different local environment
and possibly a different conformation than the SAC
domain in the context of either the rrPar-4FL or
rrPar-4DLZ constructs. Only 160 of the 308 peaks
expected (335 – N-terminal residue – 26 prolyl resi-
dues) for rrPar-4FL and 152 of the 266 expected peaks
(293 – N-terminal residue – 26 prolyl residues) for
rrPar-4DLZ are readily picked. Conversely, 58 peaks

and exposed to trypsin in a 280 : 1 (w ⁄ w) ratio.
D. S. Libich et al. Intrinsic disorder in Par-4
FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS 3715
of the expected 60 (62 – N-terminal residue – one pro-
lyl residue) were readily identifiable for rrPar-4SAC
with only two glycyl residues being unobservable.
To assess the degree of a-helicity in the rrPar-4FL
C-terminus, a CD difference spectrum between rrPar-
4DLZ and rrPar-4FL (25 °C) is shown in Fig. 7A. This
spectrum indicates a well-defined coiled-coil type struc-
ture (Fig. 7A). The two constructs differ in the dele-
tion of the leucine zipper (Fig. 2A); thus, rrPar-4FL
forms a stable coiled-coil under these conditions and
the majority of the a-helical character observed in
rrPar-4FL (Fig. 5A,B) may be attributed to this struc-
ture. The melting temperatures (based on the reduction
of the 222 nm transition in CD spectra) for rrPar-4FL
are 75, 55 and 25 °C when dissolved in native buffer,
native buffer + 1 m urea or native buffer + 6 m urea,
respectively (Fig. 7B). The results of the DLS experi-
ments on rrPar-4FL under the same conditions are
shown in Fig. 7C. As the concentration of urea is
increased from 1 to 6 m, the effective R
S
for rrPar-4FL
is reduced from 189 to 58.5 A
˚
. The latter value is very
close to the predicted R
S

to loss of its transcriptional control and thus lead to
inappropriate survival of damaged or mutated cells [58].
Sequence analysis of Par-4
Bioinformatic analysis of rrPar-4FL reveal characteris-
tic features of IDPs, including high net charge, low
mean hydrophobicity and low sequence complexity
[45,59]. Relative to the amino acid usage observed in
folded proteins, rrPar-4FL, rrPar-4DLZ and rrPar-
4SAC are depleted in order-promoting amino acids
and enriched in disorder promoting residues (Fig. 3B).
The lack of hydrophobic residues inhibits the forma-
tion of a hydrophobic core and thus the formation of
stable tertiary structure (Fig. 2D) [46].
More than 70% of rrPar-4FL is predicted to be dis-
ordered by disembl, providing a strong argument
against the formation of stable global tertiary structure
(Fig. 2B). Hydrophobic cluster analysis is a method of
displaying the primary structure such that the cluster-
ing of hydrophobic residues and thus regions of possi-
ble order become evident [43]. The regions of greatest
hydrophobic clustering in rrPar-4FL correlate well
with the predicted regions of order (Fig. 2B) and with
the secondary structure predictions (Fig. 2C).
Although the majority of rrPar-4FL is predicted to be
disordered, this does not preclude the formation of
short regions of structure or larger but transient sec-
ondary structure elements. Indeed, gor4 predictions of
a-helical structure (Fig. 2C) coincide with the more
ordered regions of rrPar-4FL and fall within the highly
conserved segments of the protein (Fig. 1). Thus,

coiled-coil (Fig. 2) raising the possibility that Par-4
function may be associated with structure stabilization
in these regions.
Intrinsic disorder in proteins is often erroneously
considered to be a featureless random coil, although
proteins do not achieve a completely random confor-
mation even in strongly denaturing conditions [60]. A
more accurate depiction is that IDPs exist as ensembles
of rapidly interchanging conformers that sample vary-
ing regions of secondary structure space [46]. IDPs can
be broadly categorized into three non-exclusive groups
(i.e. a single IDP may fall into more than one
category): random coil, pre-molten globule or molten
globule [61].
Because of the high percentage of rrPar-4FL that is
predicted as disordered, a random coil-like classifica-
tion of the ensemble would appear to be the most
appropriate. Similar to the structural ensemble
described for activator for thyroid hormone and reti-
noid receptors [62], in the absence of interacting part-
ners, rrPar-4FL exists predominantly unfolded in
solution. The kinase-inducible transcriptional-activa-
tion domain of cAMP-responsive element-binding pro-
tein (CREB) has been shown to be an IDP that folds
into an orthogonal a-helix structure upon association
with CREB binding protein [63,64]. The intrinsically
disordered nature along with the CREB binding
protein-induced helical regions could be accurately pre-
dicted from its primary structure [53]. Similarly, the
primarily intrinsically disordered nature and potential

M Tris, pH 7.0, 20 mM NaCl) only (filled
circles), buffer + 1
M urea (open diamonds) and buffer + 6 M urea
(open triangles). (C) Volume distribution of DLS measurements of
rrPar-4FL showing the apparent hydrodynamic radius of the parti-
cles: buffer (10 m
M Tris, pH 7.0, 20 mM NaCl) only (white bars),
buffer + 1
M urea (grey bars) and buffer + 6 M urea (hatched bars).
The reduction of the apparent R
S
upon increasing urea concentra-
tion suggests the disruption of a polymeric complex.
Intrinsic disorder in Par-4 D. S. Libich et al.
3718 FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS
[45,50,66]. Although rrPar-4FL, rrPar-4DLZ and BSA
contain an approximately equal percentage of trypsin
cut sites, BSA is digested at a much slower rate
(Fig. 4). This implies that a significant portion of the
conformational ensemble of the rrPar-4 proteins are
more exposed to the solvent than BSA and largely lack
protection by folded and stable tertiary structure.
The hydrodynamic radius of Par-4 is larger than
that predicted by sequence analysis
The observable Stokes radius of a protein increases in
proportion to its degree of ‘unfoldedness’; thus, an
IDP will have an observable R
S
larger than a folded
globular protein of the same MW [48,67]. Some exam-

rrPar4DLZ and rrPar-4SAC (see Fig. S1) is consistent
with this type of conformational exchange. Interest-
ingly, although the addition of 1 m urea causes a sub-
tle but significant increase in R
S
for rrPar4DLZ and
rrPar-4SAC (Table 2), the width of the distributions
are largely unaltered. Together, these observations
suggest that 1 m urea can disrupt some folding
elements and bring the conformation of the ensemble
closer to random coil, but conformational exchange
continues.
Secondary structure of Par-4 assessed by CD and
NMR
The CD spectra for rrPar-4DLZ and rrPar-4SAC are
exemplary of IDPs with a deep transition at 200 nm
and a minor transition at 222 nm (Fig. 5) [51]. The
CD spectra of IDPs are often complicated by minor
contributions from secondary structure elements, such
as alpha or poly-proline type II helices [70]. Decon-
volution of the 25 °C spectra estimates 32%, 17% and
18% of combined regular and distorted a-helix for
rrPar-4FL, rrPar-4DLZ and rrPar-4SAC, respectively
[71]. All three constructs remain relatively stable
throughout the heating cycle because the 5–65 °C
traces exhibit similar features. Thus, in addition to the
coiled-coil region of rrPar-4FL, other regions of these
proteins may transiently populate a-helical or other
secondary structures. Interestingly, although the over-
all temperature-induced changes are minor, an isodich-

structs. Possible reasons for this feature include poor
chemical shift dispersion and intermediate exchange
[74]. A detailed examination of the dynamics of the
visible regions of the proteins (dependent on assign-
ments) may help to elucidate the time scales of motion
involved and thus more definitive statements could
then be made about particular residues or regions of
rrPar-4LZ and rrPar-4DLZ [75].
The spectrum of rrPar-4SAC (Fig. 6) is much more
complete than those recorded for rrPar-4FL and
rrPar-4DLZ. Nonetheless, a similar degree of disorder
is suggested by the peak shape and chemical shift dis-
persion. The size of the rrPar-4SAC (7 kDa) relative
to that of the other constructs (> 30 kDa) is likely to
be a contributing factor in the observance of these res-
onances because fewer residues equates to less chance
of spectral overlap and a lower likelihood of slow to
intermediate exchange.
D. S. Libich et al. Intrinsic disorder in Par-4
FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS 3719
Evidence for self-association of Par-4 mediated
by a coiled-coil
The putative coiled-coil region of Par-4 (residues 254–
332; Fig. 1) is the site of recognition and association
for the majority of known binding partners [76]: the
deletion of the leucine zipper renders Par-4 incapable
of binding to proteins such as Wilms’ tumour 1,
aPKCf, p62, death-associated protein-like ⁄ zipper
interacting kinase, Akt1, E2F1 and b-site amyloid
precursor protein cleaving enzyme 1 [17,31,76,77].

than expected if the ensemble consisted of coil-like
monomers. The constructs lacking the coiled-coil
region (rrPar-4DLZ and rrPar-4SAC) did not have
appreciably large R
S
relative to the estimated random
coil values (Table 2).
The reduction of the R
S
from 189 A
˚
to 78 A
˚
for
rrPar-4LZ in 1 m urea is likely to be a result of the
disruption of a noncovalent interaction. Furthermore,
the observed R
S
in 6 m urea (58 A
˚
) is close to the cal-
culated random coil value for rrPar-4FL (Fig. 7C).
Thus, as the concentration of urea is decreased, a poly-
meric state of rrPar-4FL forms in association with
increased stabilization of the coiled-coil. Melt curves
constructed from the reduction in intensity of the
222 nm CD transition demonstrate that the rrPar-4FL
complex is much less stable in increasing concentra-
tions of urea, and show that the melting temperature
for the rrPar-4FL complex increases from 25 °Cin6m

in the context of rrPar-4FL, the C-terminus forms a
coiled-coil at neutral pH. This suggests that a region
N-terminal to the leucine zipper domain provides an
electrostatic surface to counter the negative charges in
the leucine zipper domain or otherwise helps to stabi-
lize the coiled-coil. Using a series of deletion mutants,
Gao et al. [35] demonstrated that an N-terminal region
of Par-4 is able to interact with the coiled-coil region.
This may be a required intramolecular interaction for
stabilization of the coiled-coil during self-association
or with other binding partners at neutral pH. Alterna-
tively, interactions with other parts of the protein,
including the N-terminal region of the coiled-coil (i.e.
N-terminal to the leucine zipper), may comprise a req-
uisite trigger sequence for coiled-coil formation [87].
Although the CD spectrum of rrPar-4FL showed
a-helical character (Fig. 5A), there is no obvious sign
of helix formation in the corresponding HSQC spec-
trum and, as noted, the total number of peaks
observed is only approximately 150. Recently, Liew
et al. [88] demonstrated that the signals observed in a
1
H-
15
N HSQC of a glutathione S-transferase (GST)
fusion peptide arose almost exclusively from the target
protein and not GST. They argued that because GST
forms a 52 kDa dimer, the signals arising from GST
would be broadened beyond observable limits and,
thus, the remaining resonances are from the flexible

bind N-terminally [36,37]. Indeed, Par-4 has been dem-
onstrated to mediate the ternary complex between
aPKCf and p62 [27].
The advantage of disorder in dynamic processes
Intrinsic disorder may impart several advantages to
Par-4 in its role as a pro-apoptotic factor. Because dis-
ordered regions are solvent exposed, they are easily
accessible for post-translational modifications, such as
phosphorylation, ubiquitination or Ubl-conjugation,
etc., which enables precise control of function, localiza-
tion and turnover rate [39,89]. Notably for Par-4, the
phosphorylation of T155 (Fig. 1) is required for
nuclear translocation and subsequent initiation of
apoptosis [8]. The extreme proteolytic sensitivity of
IDPs offers an additional layer of cellular control via
rapid, controlled turnover [90]. Disordered regions also
confer an increased structural plasticity and, conse-
quently, IDPs are able to bind multiple targets with
high specificity yet in a readily reversible manor. The
‘fly-casting’ mechanism has been proposed to describe
how disordered segments bind their targets with
low affinity and fast association ⁄ dissociation rates
[45,51,69,91]. Two extensively studied proteins, p53
and high mobility group protein A, have been shown
to interact with multiple binding partners primarily
through disorder containing regions [56]. Similarly,
because of intrinsic disorder, Par-4 could contain mul-
tiple specific binding sites such that it binds to different
partners simultaneously, as discussed in the case of
aPKCf ⁄ p62. An extended conformation also has the

classified as disordered when plotted in charge-hydro-
phobicity space. disembl predicts that the majority of
Par-4 (> 70%) is disordered, yet ordered segments
align well with predicted secondary structure elements
(a-helix) and regions of hydrophobic clusters. Limited
proteolysis and DLS experiments demonstrate that
rrPar-4FL is primarily extended in solution, exhibiting
high susceptibility to trypsin and a large hydrodynamic
radius. Furthermore, CD and NMR experiments
revealed characteristic spectral features of intrinsic dis-
order. Taken together, these data demonstrate that
rrPar-4FL does not maintain a stabilized global
tertiary structure, but does not preclude the possible
formation of transient and ⁄ or local structure.
D. S. Libich et al. Intrinsic disorder in Par-4
FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS 3721
Although primarily disordered, rrPar-4FL is able to
self-associate via the C-terminus forming a stable
coiled-coil in this region. Self-association behaviour
was not observed for any of the other constructs used
in the present study, each of which lacked the C-termi-
nus. Although previous experiments with the leucine
zipper domain of Par-4 showed that self-association
required acidic pH, the same requirement was not
observed in the present study, possibly because of the
influence of other regions of Par-4 providing charge-
mediated or other forms of stabilization. Intrinsic dis-
order imparts many advantages to a multifunctional
protein such as Par-4. Protein–protein and protein–
ligand interactions can be highly specific yet readily

included between thioredoxin and the N-terminus of the
target. An rTEV-protease cleavage site was included for
removal of the thioredoxin and hexa-histidine tags. The
fusion tag was PCR amplified using pET32a (Merck Bio-
sciences) as a template and subsequently inserted into the
NdeI ⁄ BamHI restriction sites of pET23a. The primers were:
forward 5¢-CTGGCATATGAGCGATAAAATTATTCAC-
3¢ and reverse 5¢-CCGGGGATCCCTGAAAATACAGG
TTTTCGGTCGTTGGGATATCGTAATCGTGATGGTG
ATGGTGATGCATATG-3¢.
The rrPar-4FL construct (residues 1–332, rat sequence
numbering) was prepared by PCR amplification using the
four primers 5¢-CAGGGATCCATGGCGACCGGCG
GCTATCGGAG-3¢,5¢-CTTGGCGGCTGGATCTCCGCC
GCTCGAAC-3¢,5¢-GTTCGAGCGGCGGAGATCCAGCC
GCCAAG-3¢ and 5¢-CAGGTCGACTTACCTTGTCAGC
TGCCCAACAAC-3¢ to remove an internal BamHI site on
the racine Par-4 cDNA. The PCR product was then cloned
into the BamHI ⁄ SalI sites of pCFE-TrxH-TEV. The rrPar-
4DLZ (residues 1–290) construct lacking the leucine
zipper was PCR amplified with the primers 5¢-CAG
GGATCCATGGCGACCGGCGGCTATCGGAG-3¢ and
5¢-CCGGAAGCTTTTATTCTTCTTTATCTTGCATCAG-
3¢ using the full-length construct as a template. The PCR
product was then cloned into the BamHI ⁄ HindIII sites of
pCFE-TrxH-TEV. The rrPar-4SAC (residues 137–195) con-
struct representing the SAC domain was cloned in the same
manner as rrPar-4DLZ using the primers 5¢-GAGGAT
CCAGGAAAGGCAAAGGGCAGATCG-3¢ and 5¢-GCA
AGCTTTTATGCTTCATTCTGGATGGTG-3¢.

pooled fractions containing the rrPar-4 fusion proteins were
dialysed against lysis buffer. The purification tags (thiore-
doxin and hexa-histidine) were cleaved from the rrPar-4
proteins with rTEV at room temperature and passed again
over the Ni-nitrilotriacetic acid column. The cleavage leaves
a three residue (Gly-Gly-Ser) remnant at the N-terminus of
all the rrPar-4 constructs. The eluted fractions were subse-
quently dialysed against 10 mm Tris (pH 7.4) and 20 mm
NaCl.
Ion-exchange chromatography was used as a final purifi-
cation step for rrPar-4DLZ and rrPar-4SAC. The constructs
were purified on SP-sepharose column (GE Healthcare)
using a linear gradient of 0–100% high salt buffer over
Intrinsic disorder in Par-4 D. S. Libich et al.
3722 FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS
20 min. The low salt buffer contained 20 mm NaPO
4
(pH
6) and 50 mm NaCl; the high salt buffer contained 20 mm
Tris (pH 7.5) and 1 m NaCl. Fractions containing the
protein of interest were pooled and dialysed against 10 mm
Tris (pH 7.0), 20 mm NaCl and were concentrated by
centrifugation using a Vivaspin 20 device (Vivascience AG,
Hannover, Germany). The rrPar-4FL construct was further
purified by RP-HPLC using a Delta Pak C18-300 A
˚
,
300 · 3.9 mm column, (Waters Corporation, Milford, MA,
USA) with a linear gradient of 20–45% acetonitrile
containing 0.08% trifluoroacetic acid. The rrPar-4FL frac-

intensities of the Tricine-PAGE gel band representing the
undigested band by densitometry using the Gel Doc Imager
and Quantity One software package (Bio-Rad, Hercules,
CA, USA).
DLS
The apparent Stokes radii of the rrPar-4 constructs were
analysed using a Zetasizer Nano ZS (Malvern Instruments,
Malvern, UK). Sample concentrations were 0.3 mgÆmL
)1
in
native buffer (10 mm Tris, pH 7.0, 20 mm NaCl) or native
buffer plus 1 or 6 m urea. DLS data were obtained at
25 °C using a low-volume disposable 1 cm pathlength plas-
tic cuvette (Sarstedt, Nu
¨
rnbrecht, Germany) and five
successive scans were collected and averaged for each pro-
tein sample. Samples were prepared 1 day in advance and
maintained overnight at 4 °C to allow any bubbles to dissi-
pate and were then allowed to equilibrate to 25 °C before
measurements were made. The diffusion coefficients were
extracted from the correlation curve and the hydrodynamic
radius was calculated using the Stokes–Einstein equation.
The highest peak of the resulting histogram recorded for
each sample was taken as the mean diameter for that
particular sample.
CD
Spectra were recorded on a Chirascan CD spectropolarime-
ter (Applied Photophysics, Leatherhead, UK) equipped with
a recirculating water bath. Samples were at a concentration

was uniformly
15
N labelled at a concentration of 0.34 mm
in 10 mm Tris (pH 7.0), 20 mM NaCl.
1
H-
15
N HSQC spectra were recorded with the settings:
rrPar-4FL: 200 transients, 2048 · 128 points (F
2
· F
1
) and
spectral widths of 8389.2 and 2128.9 Hz for F
2
and F
1
,
respectively; rrPar-4DLZ: 20 transients, 2048 · 128 points
(F
2
· F
1
) and spectral widths of 8389.2 and 2128.9 Hz for
F
2
and F
1
, respectively; rrPar-4SAC: 24 transients,
2048 · 256 points (F

NM, Weinstein MH, Strange R & Rangnekar VM
(1997) Immunohistochemical analysis of the proapop-
totic protein Par-4 in normal rat tissues. Cell Growth
Differ 8, 881–890.
2 Sells SF, Wood DP Jr, Joshi-Barve SS, Muthukumar
S, Jacob RJ, Crist SA, Humphreys S & Rangnekar
VM (1994) Commonality of the gene programs
induced by effectors of apoptosis in androgen-depen-
dent and -independent prostate cells. Cell Growth
Differ 5, 457–466.
3 Ranganathan P & Rangnekar VM (2005) Regulation
of cancer cell survival by Par-4. Ann N Y Acad Sci
1059, 76–85.
4 Chakraborty M, Qiu SG, Vasudevan KM & Rangne-
kar VM (2001) Par-4 drives trafficking and activation
of Fas and Fasl to induce prostate cancer cell apopto-
sis and tumor regression. Cancer Res 61, 7255–7263.
5 Camandola S & Mattson MP (2000) Pro-apoptotic
action of PAR-4 involves inhibition of NF-kappaB
activity and suppression of BCL-2 expression. J Neuro-
sci Res 61, 134–139.
6 El-Guendy N, Zhao Y, Gurumurthy S, Burikhanov R
& Rangnekar VM (2003) Identification of a unique
core domain of par-4 sufficient for selective apoptosis
induction in cancer cells. Mol Cell Biol 23, 5516–5525.
7 Johnstone RW, See RH, Sells SF, Wang J, Muthuk-
kumar S, Englert C, Haber DA, Licht JD, Sugrue SP,
Roberts T et al. (1996) A novel repressor, par-4, mod-
ulates transcription and growth suppression functions
of the Wilms’ tumor suppressor WT1. Mol Cell Biol

RM (2008) Critical role of prostate apoptosis response-
4 in determining the sensitivity of pancreatic cancer
cells to small-molecule inhibitor-induced apoptosis.
Mol Cancer Ther 7, 2884–2893.
14 Bieberich E, MacKinnon S, Silva J, Noggle S &
Condie BG (2003) Regulation of cell death in mitotic
neural progenitor cells by asymmetric distribution of
prostate apoptosis response 4 (PAR-4) and
simultaneous elevation of endogenous ceramide. J Cell
Biol 162, 469–479.
15 Blaschke AJ, Staley K & Chun J (1996) Widespread
programmed cell death in proliferative and postmitotic
regions of the fetal cerebral cortex. Development 122,
1165–1174.
16 Loo DT, Copani A, Pike CJ, Whittemore ER,
Walencewicz AJ & Cotman CW (1993) Apoptosis is
induced by beta-amyloid in cultured central
nervous system neurons. Proc Natl Acad Sci USA 90,
7951–7955.
17 Xie J & Guo Q (2005) PAR-4 is involved in regulation
of beta-secretase cleavage of the Alzheimer amyloid
precursor protein. J Biol Chem 280, 13824–13832.
18 Duan W, Rangnekar VM & Mattson MP (1999)
Prostate apoptosis response-4 production in synaptic
compartments following apoptotic and excitotoxic
insults: evidence for a pivotal role in mitochondrial
dysfunction and neuronal degeneration. J Neurochem
72, 2312–2322.
19 Zeitlin S, Liu JP, Chapman DL, Papaioannou VE &
Efstratiadis A (1995) Increased apoptosis and early

814–830.
26 Goswami A, Burikhanov R, de Thonel A, Fujita N,
Goswami M, Zhao Y, Eriksson JE, Tsuruo T &
Rangnekar VM (2005) Binding and phosphorylation of
par-4 by akt is essential for cancer cell survival. Mol
Cell 20, 33–44.
27 Chang S, Kim JH & Shin J (2002) p62 forms a ternary
complex with PKCzeta and PAR-4 and antagonizes
PAR-4-induced PKCzeta inhibition. FEBS Lett 510,
57–61.
28 Page G, Kogel D, Rangnekar V & Scheidtmann KH
(1999) Interaction partners of Dlk ⁄ ZIP kinase:
co-expression of Dlk ⁄ ZIP kinase and Par-4 results in
cytoplasmic retention and apoptosis. Oncogene 18,
7265–7273.
29 Roussigne M, Cayrol C, Clouaire T, Amalric F &
Girard JP (2003) THAP1 is a nuclear proapoptotic
factor that links prostate-apoptosis-response-4 (Par-4)
to PML nuclear bodies. Oncogene 22, 2432–2442.
30 Boosen M, Vetterkind S, Koplin A, Illenberger S &
Preuss U (2005) Par-4-mediated recruitment of Amida
to the actin cytoskeleton leads to the induction of
apoptosis. Exp Cell Res 311, 177–191.
31 Lu C, Chen JQ, Zhou GP, Wu SH, Guan YF & Yuan
CS (2008) Multimolecular complex of Par-4 and E2F1
binding to Smac promoter contributes to glutamate-
induced apoptosis in human- bone mesenchymal stem
cells. Nucleic Acids Res 36, 5021–5032.
32 Park SK, Nguyen MD, Fischer A, Luke MP, Affarel
B, Dieffenbach PB, Tseng HC, Shi Y & Tsai LH

introducing the D2 concept. Annu Rev Biophys 37,
215–246.
40 Oldfield CJ, Cheng Y, Cortese MS, Romero P,
Uversky VN & Dunker AK (2005) Coupled folding
and binding with alpha-helix-forming molecular recog-
nition elements. Biochemistry 44, 12454–12470.
41 Linding R, Jensen LJ, Diella F, Bork P, Gibson TJ &
Russell RB (2003) Protein disorder prediction: implica-
tions for structural proteomics. Structure 11, 1453–
1459.
42 Garnier J, Gibrat JF & Robson B (1996) GOR method
for predicting protein secondary structure from amino
acid sequence. Comput Methods Macromol Seq Anal
266, 540–553.
43 Callebaut I, Labesse G, Durand P, Poupon A, Canard
L, Chomilier J, Henrissat B & Mornon JP (1997)
Deciphering protein sequence information through
hydrophobic cluster analysis (HCA): current status and
perspectives. Cell Mol Life Sci 53, 621–645.
44 Uversky VN, Gillespie JR & Fink AL (2000) Why are
‘natively unfolded’ proteins unstructured under
physiologic conditions? Proteins 41, 415–427.
45 Dunker AK, Lawson JD, Brown CJ, Williams RM,
Romero P, Oh JS, Oldfield CJ, Campen AM, Ratliff
CM, Hipps KW et al. (2001) Intrinsically disordered
protein. J Mol Graph Model 19, 26–59.
46 Tompa P (2002) Intrinsically unstructured proteins.
Trends Biochem Sci 27, 527–533.
47 Uversky VN (2002) What does it mean to be natively
unfolded? Eur J Biochem 269, 2–12.

Inform 11, 161–171.
56 Iakoucheva LM, Brown CJ, Lawson JD, Obradovic Z
& Dunker AK (2002) Intrinsic disorder in cell-signaling
and cancer-associated proteins. J Mol Biol 323,
573–584.
57 Hanahan D & Weinberg RA (2000) The hallmarks of
cancer. Cell 100, 57–70.
58 Hartwell LH & Kastan MB (1994) Cell cycle control
and cancer. Science 266, 1821–1828.
59 Romero P, Obradovic Z, Li X, Garner EC, Brown CJ
& Dunker AK (2001) Sequence complexity of
disordered protein. Proteins 42, 38–48.
60 Shortle D (1996) The denatured state (the other half of
the folding equation) and its role in protein stability.
FASEB J 10, 27–34.
61 Uversky VN & Fink AL (2004) Conformational
constraints for amyloid fibrillation: the importance of
being unfolded. Biochim Biophys Acta 1698, 131–153.
62 Demarest SJ, Martinez-Yamout M, Chung J, Chen H,
Xu W, Dyson HJ, Evans RM & Wright PE (2002)
Mutual synergistic folding in recruitment of CBP ⁄ p300
by p160 nuclear receptor coactivators. Nature 415,
549–553.
63 Richards JP, Bachinger HP, Goodman RH & Bren-
nan RG (1996) Analysis of the structural properties
of cAMP-responsive element-binding protein (CREB)
and phosphorylated CREB. J Biol Chem 271, 13716–
13723.
64 Radhakrishnan I, Perez-Alvarado GC, Parker D,
Dyson HJ, Montminy MR & Wright PE (1997)

Acids Res 32, W668–W673.
72 Dyson HJ & Wright PE (2001) Nuclear magnetic
resonance methods for elucidation of structure and
dynamics in disordered states. Methods Enzymol 339,
258–270.
73 Dyson HJ & Wright PE (2004) Unfolded proteins and
protein folding studied by NMR. Chem Rev 104,
3607–3622.
74 Cavanagh J (2007) Protein NMR Spectroscopy: Princi-
ples and Practice, 2nd edn. Elsevier Academic Press,
Amsterdam, Boston, MA.
75 Dyson HJ & Wright PE (2002) Insights into the
structure and dynamics of unfolded proteins from
nuclear magnetic resonance. Adv Protein Chem 62,
311–340.
76 Zhao Y & Rangnekar VM (2008) Apoptosis and
tumor resistance conferred by Par-4. Cancer Biol Ther
7, 1867–1874.
77 Gurumurthy S & Rangnekar VM (2004) Par-4 induc-
ible apoptosis in prostate cancer cells. J Cell Biochem
91, 504–512.
78 Harbury PB, Zhang T, Kim PS & Alber T (1993) A
switch between two-, three-, and four-stranded coiled
coils in GCN4 leucine zipper mutants. Science 262,
1401–1407.
79 Landschulz WH, Johnson PF & McKnight SL (1988)
The leucine zipper: a hypothetical structure common to
a new class of DNA binding proteins. Science 240,
1759–1764.
Intrinsic disorder in Par-4 D. S. Libich et al.

single point mutations. Protein Sci 12, 257–265.
87 Steinmetz MO, Jelesarov I, Matousek WM, Honnap-
pa S, Jahnke W, Missimer JH, Frank S, Alexandrescu
AT & Kammerer RA (2007) Molecular basis of
coiled-coil formation. Proc Natl Acad Sci USA 104,
7062–7067.
88 Liew CK, Gamsjaeger R, Mansfield RE & Mackay JP
(2008) NMR spectroscopy as a tool for the rapid
assessment of the conformation of GST-fusion pro-
teins. Protein Sci 17, 1630–1635.
89 Xie H, Vucetic S, Iakoucheva LM, Oldfield CJ, Dunker
AK, Obradovic Z & Uversky VN (2007) Functional
anthology of intrinsic disorder. 3. Ligands, post-transla-
tional modifications, and diseases associated with
intrinsically disordered proteins. J Proteome Res 6,
1917–1932.
90 Wright PE & Dyson HJ (1999) Intrinsically unstruc-
tured proteins: re-assessing the protein structure-func-
tion paradigm. J Mol Biol 293, 321–331.
91 Shoemaker BA, Portman JJ & Wolynes PG (2000)
Speeding molecular recognition by using the folding
funnel: the fly-casting mechanism. Proc Natl Acad Sci
USA 97, 8868–8873.
92 Meszaros B, Tompa P, Simon I & Dosztanyi Z (2007)
Molecular principles of the interactions of disordered
proteins. J Mol Biol 372, 549–561.
93 Adzhubei AA & Sternberg MJ (1993) Left-handed
polyproline II helices commonly occur in globular
proteins. J Mol Biol 229, 472–493.
94 Williamson MP (1994) The structure and function of

15
N chemical shift referencing in biomolecular
NMR. J Biomol NMR 6, 135–140.
102 Altschul SF, Madden TL, Schaffer AA, Zhang J,
Zhang Z, Miller W & Lipman DJ (1997) Gapped
BLAST and PSI-BLAST: a new generation of protein
database search programs. Nucleic Acids Res 25, 3389–
3402.
103 Thompson JD, Higgins DG & Gibson TJ (1994)
CLUSTAL W: improving the sensitivity of progressive
multiple sequence alignment through sequence weight-
ing, position-specific gap penalties and weight matrix
choice. Nucleic Acids Res 22, 4673–4680.
104 Hamiaux C, Perez J, Prange T, Veesler S, Ries-Kautt
M & Vachette P (2000) The BPTI decamer observed in
acidic pH crystal forms pre-exists as a stable species in
solution. J Mol Biol 297, 697–712.
105 Kabsch W, Mannherz HG, Suck D, Pai EF & Holmes
KC (1990) Atomic structure of the actin:DNase I
complex. Nature 347, 37–44.
106 Vijay-Kumar S, Bugg CE & Cook WJ (1987) Structure
of ubiquitin refined at 1.8 A resolution. J Mol Biol
194, 531–544.
107 Matthews DA, Dragovich PS, Webber SE, Fuhrman
SA, Patick AK, Zalman LS, Hendrickson TF, Love
RA, Prins TJ, Marakovits JT et al. (1999) Structure-
assisted design of mechanism-based irreversible
D. S. Libich et al. Intrinsic disorder in Par-4
FEBS Journal 276 (2009) 3710–3728 ª 2009 The Authors Journal compilation ª 2009 FEBS 3727
inhibitors of human rhinovirus 3C protease with