interacting protein HIP1, with the specific motif present at the putative promoter sequence of the caspase-1, caspase-8 and caspase-10 genes P.. It was observed that the C-terminal ‘pseu
Trang 1interacting protein HIP1, with the specific motif present at the putative promoter sequence of the caspase-1, caspase-8 and caspase-10 genes
P Majumder1, A Choudhury2, M Banerjee1, A Lahiri2and N P Bhattacharyya1
1 Structural Genomics Section, Saha Institute of Nuclear Physics, Bidhan Nagar, Kolkata, India
2 Department of Biophysics, Molecular Biology and Genetics, University of Calcutta, Kolkata, India
It has been known for more than 13 years that
increased CAG repeats beyond position 36 in exon1 of
the Huntingtin (Htt) gene causes Huntington’s disease
[1], resulting in increased apoptosis in a specific region
of the brain [2] Among various interacting partners of
the protein Htt [3–5], Huntingtin interacting protein 1
(HIP1), identified in the yeast two-hybrid assay [6] and
subsequently characterized as an endocytic adaptor
protein with clatharin assembly activity, binds to
various cytoskeleton proteins [7] In the search for interacting partners of HIP1, a novel protein HIPPI (HIP1 protein interactor) has recently been identified HIPPI does not have any known domains except for a
‘pseudo’ death effector domain (pDED) and a myosin-like domain Interaction of HIPPI with HIP1 takes place through the pDED present in both proteins The HIPPI–HIP1 heterodimer recruits procaspase-8, and activates the initiator caspase and its downstream
Keywords
caspase; HIPPI; motif; pDED; transcription
regulation
Correspondence
N P Bhattacharyya, Structural Genomics
Section, Saha Institute of Nuclear Physics,
1 ⁄ AF Bidhan Nagar, Kolkata 700 064, India
Fax: +91 033 2337 4637
Tel: +91 033 2337 5345
E-mail: nitai_sinp@yahoo.com
(Received 11 April 2007, revised 1 June
2007, accepted 5 June 2007)
doi:10.1111/j.1742-4658.2007.05922.x
To investigate the mechanism of increased expression of caspase-1 caused
by exogenous Hippi, observed earlier in HeLa and Neuro2A cells, in this work we identified a specific motif AAAGACATG () 101 to ) 93) at the caspase-1 gene upstream sequence where HIPPI could bind Various muta-tions in this specific sequence compromised the interaction, showing the specificity of the interactions In the luciferase reporter assay, when the reporter gene was driven by caspase-1 gene upstream sequences () 151 to ) 92) with the mutation G to T at position ) 98, luciferase activity was decreased significantly in green fluorescent protein–Hippi-expressing HeLa cells in comparison to that obtained with the wild-type caspase-1 gene
60 bp upstream sequence, indicating the biological significance of such binding It was observed that the C-terminal ‘pseudo’ death effector domain of HIPPI interacted with the 60 bp () 151 to ) 92) upstream sequence of the caspase-1 gene containing the motif We further observed that expression of caspase-8 and caspase-10 was increased in green fluores-cent protein–Hippi-expressing HeLa cells In addition, HIPPI interacted
in vitrowith putative promoter sequences of these genes, containing a sim-ilar motif In summary, we identified a novel function of HIPPI; it binds to specific upstream sequences of the caspase-1, caspase-8 and caspase-10 genes and alters the expression of the genes This result showed the motif-specific interaction of HIPPI with DNA, and indicates that it could act as transcription regulator
Abbreviations
DED, death effector domain; EMSA, electrophoretic mobility shift assay; GFP, green fluorescent protein; GST, glutathione S-transferase;
HD, Huntington’s disease; HIP1, Huntingtin interacting protein 1; HIPPI, Huntingtin interacting protein 1 protein interactor; Htt, Huntingtin; IOD, integrated optical density; pDED, ‘pseudo’ death effector domain; TSS, transcription start site.
Trang 2apoptotic cascades [8,9] It has been shown earlier that
the interaction of HIP1 with normal Htt (a protein with
fewer than 36 Gln) is stronger than that observed with
the mutated Htt (a protein with more than 36 Gln
resi-dues) [10] On the basis of this observation, it has been
proposed that the weaker interaction of HIP1 with
mutated Htt in Huntington’s disease (HD) may
increase the amount of freely available HIP1 and
enhance the propensity for the HIP1–HIPPI
heterodi-mer to form The increased amount of HIP1–HIPPI
may in turn lead to the increase in cell death observed
in HD [8] A role of HIPPI in apoptosis regulation has
also been inferred from other studies Apoptin, a
chicken anemia virus-encoded protein, has been shown
to colocalize with HIPPI in the cytoplasm of normal
cells, whereas in tumor cells the two proteins localize
separately in the nucleus and cytoplasm It has been
proposed that the HIPPI–apoptin interaction may
suppress apoptosis [11] The bifunctional apoptosis
inhibitor, which regulates neuronal apoptosis, also
interacts with HIPPI, although the functional relevance
of this interaction remains unknown [12] Very recently,
it has been reported that HIPPI interacts with the
postsynaptic scaffold protein Homer1c and regulates
apoptosis in striatal neurons [13] All these studies
show that HIPPI, through its interacting partner,
regulates apoptosis Even though the exact function of
HIPPI remains unknown, it has been shown, using
knockout mouse (Hippi–⁄ –), that HIPPI is involved the
Sonic hedgehog signaling pathway [14]
Interactions of several transcription factors with Htt
and alterations of a large number of genes observed in
microarray studies support the hypothesis that the
pathology of HD is mediated through alterations in
transcription [15] In several studies using cellular and
animal models of HD (where the mutated full-length
Htt gene or exon1 are expressed by knockin), the
expression of the caspase-1, caspase-3, caspase-2,
caspase-6 and caspase-7 genes is increased [16,17] How
the expression of these genes is altered is not known
We have previously shown that exogenous
expres-sion of Hippi increases various apoptotic markers In
the course of this study, it was also observed that the
endogenous expression of caspase-1, caspase-3 and
caspase-7 is upregulated in green fluorescent protein
(GFP)–Hippi-expressing cells, whereas the
mitochond-rial genes ND1 and ND4 and the antiapoptotic gene
Bcl-2 are downregulated [9] Recently, we have also
shown that HIPPI can directly interact with the
ca-spase-1 gene upstream 60 bp sequence () 151 to ) 92)
in vitro and in vivo [18] In the present investigation,
we identified and characterized a motif within this
60 bp sequence of the caspase-1 gene where HIPPI
could bind specifically In addition, we observed that a similar motif was present at the putative promoter sequences of the caspase-8 and caspase-10 genes; the expression of these genes was also increased in GFP– Hippi-expressing HeLa cells In vitro experiments showed that HIPPI also interacted with the promoter sequences of these genes
Results
Specific motif at the upstream sequences
of the caspase-1 gene
To search for the specific DNA sequence motif where HIPPI might interact, we analyzed 1 kb upstream regions of the caspase-1, caspase-3 and caspase-7 genes using four different motif prediction algorithms, i.e meme, alignace, bioprospector and mdscan The motifs predicted using the different methods, parame-ters and sequence sets (masked⁄ unmasked) were then assembled and compared, and the redundant motifs were discarded (data not shown) The motif pre-dicted using the methods mentioned was 5¢-AA AGA[CG]A[TA][GT]-3¢ We investigated whether any similar motif was present within the 60 bp stretch of the caspase-1 gene upstream sequence where HIPPI actually interacted [18] It was observed that the motif 5¢-AAAGACATG-3¢ () 101 to ) 93) was present in the positive strand of the caspase-1 gene upstream sequence This motif was conserved in promoters of caspase-1 orthologs from Pan troglodytes (DOOP ID:
83123145,) 245 to ) 253) and Macaca mulatta (DOOP ID: 94252893, ) 245 to ) 253) The motif sequences
of the caspase-3 gene (5¢-AAAGAGATG-3¢, ) 828 to ) 820) and the caspase-7 gene (5¢-AAAGACATA-3¢, ) 245 to ) 253) were present in the positive strand In subsequent studies, we tested whether HIPPI could interact with the 5¢-AAAGACATG-3¢ () 101 to ) 93) motif present in the putative promoter of the caspase-1 gene
Interactions of HIPPI with AAAGACATG and various mutants of this sequence at the caspase-1 gene upstream sequence The specific sequence AAAGACATG identified within the 60 bp upstream sequence was used to test whether HIPPI interacted with this motif The results of a typical electrophoretic mobility shift assay (EMSA) experiment carried out using the above-mentioned sequence and its mutants (mutations at the fourth, fifth and sixth positions) are shown in Fig 1A A mobility shift of the band corresponding to [32P]ATP[cP]-labeled
Trang 3dsDNA, AAAGACATG, in the presence of
gluta-thione S-transferase (GST)–HIPPI (Fig 1A, panel I,
lane 3) indicated interaction of the purified protein
with the motif No shift was observed in the presence
of GST protein only (lane 2)
EMSA with mutants of the 9 bp motif AAAGA
CATG indicated that AAAGAGATG (mutation of
the sixth nucleotide, C to G) interacted with the GST–
HIPPI, as is evident from the mobility shift of the
band corresponding to radiolabeled dsDNA in the
presence of purified protein (Fig 1A, panel II, lanes 3
and 4) However, mutation at the fourth nucleotide
(G to T) and fifth nucleotide (A to C) affected the
interaction In both cases, there was no shift of the
probe, as shown in lane 2 and lane 6, indicating that
GST–HIPPI did not interact with these mutated
motifs
A similar result was also obtained in the
fluores-cence quenching study (Fig 1B, panel I) With
AAAGAGATG), the fluorescence (kemission¼ 340 nm,
kexcitation¼ 295 nm) of GST–HIPPI protein was
reduced and reached a plateau The value of the
dissociation constant, determined from the plateau
region, obtained with AAAGACATG was calculated
to be 1.2 nm (Fig 1C, panel I) A similar result was
obtained with AAAGAGATG, with a dissociation
constant of 0.3 nm (Fig 1C, panel II) A fluorescence
quenching assay with the DNA AAAGACACG (point
mutation at the eighth position T to C of the predicted
motif mentioned above) revealed a decrease in the
intrinsic fluorescence of GST–HIPPI protein,
indica-ting binding of the protein with this mutated motif
(Fig 1B, panel II) The apparent dissociation constant
(Kd) of this binding was 4 nm (Fig 1C, panel III)
However, a similar assay with AAATACATG and
AAAGCCATG did not alter the GST–HIPPI
fluores-cence significantly (Fig 1B, panel I), which further
supported the results of EMSA with the same DNA
sequences, discussed before (Fig 1A, panel II) Further
point mutations at the second (A to G), third (A to
G), seventh (A to C) and ninth (G to A) nucleotides
of the 9 bp motif AAAGACATG and a subsequent
fluorescence quenching study indicated no significant
quenching of fluorescence of GST–HIPPI in the
pres-ence of these mutants This result revealed that GST–
HIPPI did not interact with these mutated sequences
of the 9 bp motif (Fig 1B, panel II)
To explore the nature of the interactions of GST–
HIPPI with AAAGACATG, we increased the
concen-tration of NaCl from 50 mm (normally used in all
binding assays) to 1000 mm As is evident from Fig 2,
with the increasing concentrations of NaCl, the
fluor-escence intensities of GST–HIPPI increased, indicating
a lesser extent of interactions of GST–HIPPI with AA AGACATG This result indicated that the interaction
of GST–HIPPI with AAAGACATG was electrostatic
in nature, although other possibilities cannot be ruled out
The above results showed that purified GST–HIPPI interacted with the 9 bp motif AAAGACATG present
at the upstream sequence () 101 to ) 93) of the cas-pase-1 gene, and that mutation at the sixth and eighth positions of the motif did not affect this binding, as is evident from the significant quenching of GST–HIPPI protein fluorescence observed with the respective sequences (Fig 1B, panels I and II) A summary of the results is shown in Table 1 From the experimental studies described above with the various mutant motifs and their interactions in vitro with HIPPI, the consen-sus HIPPI-binding motif AAAGASAHK, i.e AAAG A[GC]A[ATC][TG], was derived
Reduction of the promoter activity of the 60 bp ()151 to ) 92) caspase-1 gene upstream sequence
by mutation at position) 98 (G to T) to the specific motif AAAGACATG ()101 to ) 93) in GFP–Hippi-expressing cells
We have earlier shown that the 717 bp () 700 to + 17) and 60 bp () 151 to ) 92) sequences can act as the promoter in the luciferase reporter assay in HeLa
as well as in Neuro2A cells It has been shown that the luciferase activity of pGL3 when driven by the 717 bp caspase-1 gene upstream sequence is higher than that obtained with the 60 bp-driven construct [18] This has been attributed to the presence of binding sites for other factors within these flanking sequences [19] As shown above, the 60 bp upstream sequence contains the motif AAAGACATG () 101 to ) 93), and muta-tion at posimuta-tion ) 98 (G to T) abolished the interaction
of HIPPI To check whether this mutation also decrea-ses the expression of the reporter gene driven by this mutated 60 bp caspase-1 gene upstream sequence
in vivo, we carried out the luciferase assay after cloning both the wild-type 60 bp sequence and the mutated
60 bp sequence in pGL3 The luciferase activity, seen
in GFP–Hippi-expressing HeLa cells when the lucif-erase gene was driven by the 60 bp region with a mutation at position ) 98 (G to T), was decreased (Fig 3) significantly (P¼ 0.01) in comparison with that obtained with the wild-type 60 bp sequence The result of this experiment is shown in Fig 3, and indi-cates that mutation of the specific site of the binding motif at the putative promoter sequence of the caspase-1 gene, where HIPPI can bind, decreased the
Trang 4promoter activity of the 60 bp upstream sequences
sig-nificantly As shown above, the interaction of HIPPI
with the mutated 9 bp motif (G to T at the fourth
position of the motif) was abolished, whereas the
60 bp sequence with the mutated motif exhibited
sub-stantial promoter activity This could be due to
addi-tional transcription regulator-binding sites within the flanking sequence of the motif It has been shown that p53 can bind within this region [19] The luciferase activity of the pGL3 driven by the mutated 60 bp caspase-1 gene upstream sequence in GFP–Hippi-expressing HeLa cells was similar (3.0 ± 1.1) to that
A
B
C
shifted
1 2 3
1 2 3 4 5 6 band
Probe
1.00E+008
AAAGAGATG
AAGGACATG AAAGACCTG
AAAGACATG
K d =1.2 nM
AAAGAGATG
K d =0.29 nM AAAGACACG
K d =4 nM
AAAGACACG AAAGACATA AGAGACATG
AAAGACATG AAAGCCATG AAATACATG
8.00E+007 6.00E+007 4.00E+007 2.00E+007 0.00E+000
14 16
12 10 8
6 4 0.00 0.01 0.02 0.03 0.04 0.05 0.00 0.02 0.04 0.06
[DNA] µM [DNA] µM
1.50E-008 1.40E-008
1.28E-008 1.26E-008
0.105 0.100 0.095 0.090 0.085 0.080
1.20E-008 1.10E-008 1.00E-008
F 340
F 340
0 25 50 75 100
1/[DNA] µM
20 25 30 35 40 45 50
1/[DNA] µM
0 11 22 33 44 55
1/[DNA] µM
0.08
Probe (9 bp) AAAGACATG
II I
I
shifted band
Fig 1 In vitro binding assay of putative HIPPI-binding motif AAAGACATG and its mutant (A) EMSA of binding of the end-labeled 9 bp motif AAAGACATG and two of its mutants with purified GST–HIPPI protein Panel I Lane1: probe (200 n M of [ 32 P]ATP[cP]-labeled 9 bp motif AAAGACATG) only Lane 2: probe + 4 l M GST protein Lane 3: probe + 1.7 l M GST–HIPPI Panel II Typical results of similar analysis with the same 9 bp DNA with mutation at the fourth, fifth or sixth base and GST–HIPPI protein are shown Lane 1: 200 n M [32P]ATP[cP]-labeled AAATACATG (probe only) Lane 2: 200 n M same probe + 3.8 l M GST–HIPPI protein Lane 3: band corresponding to 200 n M [ 32 P]ATP[cP]-labeled AAAGAGATG (probe only) Lane4: 200 n M probe + 3.8 l M GST–HIPPI protein Lane 5: AAAGCCATG (probe only) Lane 6: result obtained with 200 n M same probe + 3.8 l M GST–HIPPI protein (B) Quenching of intrinsic fluorescence of GST–HIPPI (0.8 l M ) at 340 nm (kexc¼ 295 nm) in the presence of the 9 bp putative motif sequence and six of its mutants Panel I Fluorescence quenching of GST–HIPPI protein due to addition of AAAGACATG (red line) and its mutants: AAAGAGATG (black line), AAAGCCATG (green line), AAATACATG (blue line) Inset: Curves representing wavelength scan of each point of quenching experiment with the 9 bp motif sequence AAAGACATG Fluor-escence intensities were measured in a Fluoromax 3 spectrofluoremeter Panel II Study of any quenching of intrinsic fluorFluor-escence of GST– HIPPI due to addition of point-mutated sequences of the 9 bp potent HIPPI-binding motif, namely: AAGGACATG (black line), AGAGACATG (red line), AAAGACCTG (green line), AAAGACACG (blue line) and AAAGACATA (sky blue line) Positions of point mutations are indicated by underlines (C) Linear plot of 1 ⁄ DF versus 1 ⁄ c, where as DF is the change in fluorescence with respect to the intrinsic fluorescence of GST– HIPPI due to addition of DNA with concentration c (l M ) Kdvalues were calculated from such plots In panel I, DF is of GST–HIPPI versus c
of 9 bp motif sequence AAAGACATG In panel II and panel III, DF is of GST–HIPPI versus c of mutated products of 9 bp sequence: AAAGA GATG and AAAGACACG, respectively.
Trang 5observed in HeLa cells (without any detectable HIPPI
expression) when the luciferase gene was driven by the
60 bp wild-type sequence (1.2 ± 1.3) The difference
was not statistically significant (P¼ 0.4) Furthermore,
there was no significant difference (P¼ 0.6) between
the luciferase activities in HeLa cells expressing pGL3
driven either by the 60 bp wild-type sequence
(1.2 ± 1.3) or the 60 bp mutated (1.7 ± 0.9) sequence
Thus, these luciferase activities could be due to the
presence of promoter-binding site(s) within the 60 bp
caspase-1 gene upstream sequence other than for
HIPPI This indicated that, due to point mutation
at position 98 (G to T), HIPPI could not bind to the
caspase-1 gene upstream to transcribe the downstream
gene; this was manifested by about a two-fold decrease
in luciferase acitivity Thus, the results of promoter
assay experiments further confirmed the in vitro result
that mutation of the motif abolished the binding of
HIPPI to the specific sequence of the caspase-1 gene
upstream sequence, and the increased expression of
caspase-1 in GFP–Hippi-expressing cells was due to interaction of HIPPI with this motif
Interaction of pDED of HIPPI with upstream sequences of the caspase-1 gene
To check which portion of HIPPI was responsible for this interaction, a cDNA portion corresponding to the two termini of HIPPI, i.e the N-terminal portion com-prising amino acid residues 10–334 (NCBI protein ID NP_060480) and the C-terminal pDED region (amino acids 335–429), were cloned and expressed in bacteria, and the proteins were purified Interactions of the puri-fied 6X(HN)-pDED and the N-terminal domains of HIPPI [also tagged with 6X(HN)] were studied in vitro
by EMSA and fluorescence quenching The results revealed that the 6X(HN)-pDED domain of HIPPI interacted with the 60 bp upstream sequence of the caspase-1 gene (Fig 4A, panel II, lanes 1 and 3) In contrast, the N-terminal region of HIPPI did not inter-act with the upstream sequence of the caspase-1 gene (Fig 4A, panel I, lane 3) This result showed that the C-terminal end containing the pDED domain of HIPPI could interact with the upstream sequence of the caspase-1 gene
16000000
GST-Hippi 25µg + AAAGACATG 0.05 µM
R2=0.92422 14000000
12000000
10000000
8000000
6000000
4000000
2000000
0
Fig 2 (A) Sigmoidal curve (R 2 ¼ 0.924) showing gradual increase
in fluorescence intensity of GST–HIPPI (0.8 l M ) at 340 nm (k exc ¼
295 nm) at saturation level of binding with the 9 bp motif
AAAGA-CATG with increasing salt (NaCl) concentrations ranging from
50 m M to 1 M
Table 1 Summary of binding study with the putative HIPPI-binding
motif and its mutants ND, not determined.
Fluorescence quenching Result Average Kd(n M )
8
p=0.01
7 6 5
4 3 2 1 0
Fig 3 The caspase-1 gene upstream 60 bp and a point mutation (G to T) incorporated in the same sequence at position ) 98, cloned
in the pGL3 enhancer plasmid (60 M ) and transfected (4 lg each) in GFP–Hippi-expressing cells Cells expressing GFP–Hippi were mon-itored by the presence of GFP under a fluorescence microscope About 80–90% of cells expressed GFP after 20 h of transfection with GFP–Hippi Transfected cells are denoted as 60_Hi and
60 M _Hi, respectively The corresponding average fold increase (n ¼ 3) in luciferase activities compared to control (cell expressing only pGL3 without any insert) are given in bar diagrams P-values
of significance are mentioned above the bar diagram in each of the cases studied.
Trang 6As 6X(HN)-pDED of HIPPI does not contain any
tryptophan, a 280 nm excitation filter was used, and
the fluorescence (characteristics of tyrosine and
phe-nylalanine) was measured at 305 nm A decrease in
the fluorescence intensity of 6X(HN)-pDED due to
the addition of the 60 bp region of the caspase-1 gene
upstream sequence was observed Addition of the
caspase-1 gene upstream 60 bp sequence could also quench the intrinsic fluorescence of 6X(HN)-pDED of HIPPI from 6.997 to 1.802 (Fig 4B, panel I) with an apparent binding constant (Kd) 0.34 nm; a double reciprocal plot is shown in Fig 4B, panel II However, addition of the upstream sequences of the caspase-1 gene (717 bp) to the N-terminal domain (without the pDED domain) of HIPPI did not decrease the fluorescence intensities determined by exciting either at 295 nm (kem¼ 340 nm; fluorescence intensity changed from 15.62 to 14.42 due to addition
of 0.5 lm DNA) or 280 nm (kem¼ 305 m; fluores-cence intensity changed from 8.77 to 7.59) This result also showed that pDED of HIPPI actually interacted with the caspase-1 gene upstream sequences We recently observed that pDED of HIPPI could also interact in vivo with the caspase-1 gene upstream sequence (data not shown)
Increase in caspase-1 gene expression and induction of apoptosis by C-terminal pDED
of HIPPI The role played by pDED of HIPPI in alteration of caspase-1 gene expression in HeLa cells was monitored
by western blot analysis using antibody to caspase-1 (Fig 5, middle panel) The band intensities were measured using image master vds software The aver-age integrated optical density (IOD) of three different experiments is shown in Table 2 The results indicated that caspase-1 expression, as detected by western blot
A
B
shifted
labeled
band
Casp1
ups 717 bp
shifted
labeled
band
Casp1ups
II I
C1ups 717bp (HIPPI-pDED)
C1ups 60bp (6XHN-HIPPI_pDED)
Kd=0.34n M
C1ups 717bp (HIPPI-Nterm)
C1ups 60bp (HIPPI-pDED)
9
7
5
4
2
0
0.00 0.01 0.02
0.03 0.04
0.18 0.19 0.20 0.21
0.17
0 20 40 60 80100 120 140
60 bp
Fig 4 (A) In vitro binding assay of pDED of HIPPI with the
ca-spase-1 gene upstream sequences In panel I, typical results of
EMSA with the 717 bp caspase-1 probe and 6X(HN)-pDED of Hippi
and 6X(HN)-tagged N-terminal domain of Hippi protein are shown.
Lane 1 shows the result with the probe only (400 n M ) Lanes 2 and
3 show results obtained when 400 n M probe was allowed to
inter-act with 2.7 l M 6X(HN)-pDED of HIPPI and 5.2 l M 6X(HN)-tagged
N-terminal domain of HIPPI, respectively In panel II, 500 n M 60 bp
caspase-1 gene upstream sequence ( ) 151 to ) 92) labeled with
[32P]dCTP[aP] was used as probe Lanes 1 and 2 show results
obtained with probe only and probe + 4.8 l M BSA (nonspecific
pro-tein), respectively Lane 3 shows the result when 500 n M probe
was allowed to react with 1.6 l M 6X(HN)-pDED of HIPPI, and lane
4 shows the result obtained when 500 n M labeled 60 bp probe was
allowed to react with 1.6 l M 6X(HN)-pDED protein in the presence
of 500 n M unlabeled 60 bp caspase-1 gene upstream sequence (B)
Panel I, quenching of intrinsic fluorescence of 6x(HN)-tagged pDED
of HIPPI (pDED-HIPPI, 2 l M ) at 305 nm in presence of the 717 bp
(squares) and 60 bp (triangles) upstream sequences of the
ca-spase-1 gene (denoted as C1ups 717 bp and C1ups 60 bp,
respect-ively) and any change in intrinsic fluorescence of 6x(HN)-tagged
N-terminal domain of HIPPI (HIPPI-Nterm, 1 l M ) at 305 nm in the
presence of 717 bp upstream sequence of the caspase-1 gene,
denoted by circles Panel II shows a linear plot of 1 ⁄ DF versus 1 ⁄ c
where, DF represents the decrease in intrinsic fluorescence of
6x(HN)-tagged HIPPI-pDED protein in the presence of the 60 bp
upstream sequence of the caspase-1 gene; c, concentration (l M ).
The apparent binding constant (K d ) was calculated from this plot
and is given with the graph. Fig 5 Role of exogenous pDED of HIPPI in alteration of caspase-8
and caspase-1 activation in HeLa cells Western blot analysis using antibodies to caspase-8 (upper panel) and caspase-1 (middle panel) with total protein isolated from HeLa cells (H), HeLa cells expres-sing GFP-tagged N-terminus of HIPPI containing Myosin-like domain (HiN), and that expressing GFP-tagged pDED of HIPPI (HiD) The upper bands of the upper panel correspond to proca-spase-8 (57 kDa), and the lower bands represent the 12 kDa activa-ted caspase-8; the upper bands of the middle panel correspond to procaspase-1 (45 kDa), and the lower bands correspond to the
20 kDa activated caspase-1 The lowermost panel shows the level
of b-actin (14.4 kDa).
Trang 7analysis, was increased in GFP–pDED-expressing cells
by 4.4 ± 0.7-fold as compared to that in parental
HeLa cells However, this increase in the N-terminal
part of HIPPI-expressing cells was only 1.4-fold The
increase in caspase-1 expression was again 6.2 ±
1.1-fold in pDED HIPPI-expressing cells as compared to
the N-terminal part of HIPPI-expressing cells
To test whether the C-terminal pDED of HIPPI
could induce apoptosis more efficiently than the
N-ter-minal domain in our system, these two domains cloned
in pEGFP C1 vectors were transfected into HeLa cells
After 32 h, when 80–90% of cells were expressing
GFP-tagged protein, we determined the nuclear
frag-mentation as an indication of apoptosis induction and
caspase activation GFP–pDED-expressing cells
exhib-ited nuclear fragmentation in 32.5 ± 1.8% of the total
cell population, whereas this value in the GFP-tagged
N-terminal domain of HIPPI-expressing cells was only
17.6 ± 1.9% This difference was statistically
signifi-cant (P¼ 0.0006) Thus, GFP–pDED of HIPPI was
more effective in inducing apoptosis in HeLa cells
Fluorometric determination of caspase-1 activity by a
commercially available kit indicated that, in GFP–
pDED of HIPPI-expressing cells, caspase-1 activity
was 1.6-fold higher (P¼ 0.02) than that observed in
the GFP–N-terminal domain of HIPPI-expressing
HeLa cells Fluorometric determination of caspase-8
activity indicated that in HeLa cells expressing GFP–
pDED of HIPPI, caspase-8 activation was also 1.6-fold
higher in comparison to that obtained in the
GFP–N-terminal domain of HIPPI-expressing cells This value
was also statistically significant (P¼ 0.047, n ¼ 3)
Activation of caspase-1 and caspase-8 in
GFP–pDED-expressing cells was further supported by western blot
analysis (Fig 5) using total protein isolated from HeLa
cells expressing pDED and the N-terminal domain of
HIPPI It is evident from Fig 5 that ectopic pDED
expression in HeLa cells induced cleavage of proca-spase-8 (Fig 5, upper panel) and procaspase-1 (Fig 5, middle panel) proteins more efficiently as compared to that of the N-terminal domain of HIPPI A similar higher activation (2.1-fold, P¼ 0.03) of caspase-3 was observed in GFP–pDED of HIPPI-expressing cells in comparison to that observed in GFP-–N-terminal HIPPI-expressing HeLa cells These results are shown
in Table 2
Presence of the motif and the putative promoter sequences of caspase-8 and caspase-10 increased expression of the genes in GFP–Hippi-expressing HeLa cells
The derived motif AAAGASAHK, i.e AAAGA[GC] A[ATC][TG], was used to search for the presence of the motif at the 1000 bp upstream sequences of the caspase-8 and caspase-10 genes using motiflocator (http://www.esat.kuleuven.ac.be/dna/BioI/Software html) These genes are supposed to be involved in HD The results for similar motifs identified in the caspase-1 (for reference) caspase-3, caspase-7, caspase-8 and caspase-10 genes are shown in the Table 3 In Table 3, the start and end positions of the motifs are indicated
by distance from the transcription start site (TSS) The position upstream of the TSS of any gene is denoted
by ‘–’ followed by the distance from the TSS In the caspase-8 gene, the putative upstream sequence motifs AAAGAGAAC () 955 to ) 963) in the positive strand, and AAAGAAAAG () 418 to ) 410) and AA AGACATA () 800 to ) 808) in the negative strand, were observed (variants are underlined) As shown above, mutation at the last base, G to A, abolished the interaction of HIPPI, so the last motif would not interact with HIPPI The other two motifs might be the target of HIPPI In the upstream sequence of the
Table 2 Comparison of apoptosis induction and alteration in caspase-1 gene expression in GFP-tagged pDED of HIPPI and N-terminal domain of HIPPI-expressing HeLa cells.
GFP–pDED
of HIPPI (fold) P-values
GFP–N-terminus
of HIPPI (fold) P-values
Fold increase: pDED versus N-terminal domain (P-values) Caspase-1
expression
8.9 ± 2.6 54.9 ± 2.1
(6.2-fold)
(0.0005) Nuclear
fragmentation
2.4 ± 0.9 32.5 ± 1.8
(13.5-fold)
(7.3-fold)
0.0006 Caspase-8
activation
9.2 ± 2.0 23.1 ± 1.3
(2.5-fold)
(0.047) Caspase-1
activation
40.4 ± 3.6 90.2 ± 11.1
(2.2-fold)
(0.02) Caspase-3
activation
1.5 ± 0.3 10.8 ± 2
(7.2-fold)
(3.4-fold)
(0.03)
Trang 8caspase-10 gene, four variant motifs were identified.
Among them, AAACAGATG () 254 to ) 262) is
present in the positive strand, and the sequences AA
AGAAAAG () 651 to ) 643), AAAGAAAAG () 725
to) 717) and GAAGACATT () 849 to ) 857) are
pre-sent in the negative strand
Given that the caspase-8 and caspase-10 genes
har-bor similar motifs as that in the caspase-1 gene and
increase caspase-1 expression, we first tested the
expression of the caspase-8 and caspase-10 genes in
GFP–Hippi-expressing cells by the semiquantitative
RT-PCR described previously [18] The numbers of
PCR cycles and the amount of total RNA were chosen
so that the yield of RT-PCR products was in the linear range The IOD value of the RT-PCR product obtained with RNA isolated from GFP–Hippi-expres-sing HeLa cells was increased 2.5-fold in comparison
to the value obtained when RNA from the HeLa cells was used This increase was statistically significant (P¼ 0.0004) A similar significant increase in the IOD value of the RT-PCR products for the caspase-10 gene (1.8-fold, P¼ 0.0002) was detected A bar diagram showing the mean IOD of bands corresponding to caspase-8 and caspase-10 gene-specific products run on 1.5% agarose gel is shown in Fig 6A) A representa-tive photograph of the RT-PCR products run on
Table 3 Summary of the presence of similar motifs in caspase-1, caspase-3, caspase-7, caspase-8 and caspase-10 gene upstream sequences.
Fig 6 (A) Bar diagrams showing mean of IOD of bands obtained by RT-PCR at caspase-8 and caspase-10 loci, along with error bars calcula-ted on the basis of three independent experiments using mRNA isolacalcula-ted from GFP–Hippi-expressing (gray bar) and parental (white bars) HeLa cells Levels of significance (P-values) are shown on the top of the bars (B) Picture of RT-PCR products run on 1.5% agarose gel and stained with ethidium bromide, representing PCR amplification with RNA isolated from HeLa cells (lane 1:H), GFP–Hippi-expressing HeLa cells (lane 2:Hi), and reaction carried out where no RNA was added (lane 3:ve) The uppermost panel shows the PCR reaction carried out with caspase-8 (164 bp)-specific primers; the middle panel represents PCR amplification using primers specific for caspase-10 (178 bp) genes; and the lowermost panel shows bands (315 bp) corresponding to b-actin (loading control).
Trang 9agarose gel is shown in Fig 6B Equally intense signals
for internal control (b-actin gene-specific primers) were
obtained in all the cases (Fig 6B, lowermost panel)
Fluorescence quenching assay to measure the
interactions of GST–HIPPI with the caspase-8 and
caspase-10 gene upstream sequences
Expression of caspase-8 and caspase-10 increased in
GFP–Hippi-expressing cells, as described above, and
DNA sequences similar to the putative HIPPI-binding
motif were present within the 1000 bp upstream
sequences of the caspase-8 and caspase-10 genes
(Table 3) To check interactions of GST–HIPPI with
these upstream regions containing the motifs, the
caspase-8 gene upstream 710 bp () 991 to ) 282) and
caspase-10 gene upstream 768 bp () 914 to ) 147)
regions were PCR-amplified The results, shown in
Fig 7A, indicated quenching of GST–HIPPI intrinsic
fluorescence at 340 nm, due to addition of increasing
concentrations (0.001 lm to 0.05 lm) of the caspase-8
and caspase-10 gene upstream sequences
Average (n¼ 2) Kd values for binding of purified
GST–HIPPI with the caspase-8 gene (0.32 ± 0.13 nm)
and the caspase-10 gene (11 ± 3.8 nm) upstream
sequences were calculated from reciprocal plots as
des-cribed previously [18], and typical cases are shown in
Fig 7B, panel I and panel II, respectively
Discussion
In the present work, we have shown that HIPPI
inter-acted specifically with the motif AAAGACATG () 101
to) 93) present in the upstream region of the caspase-1
gene in vitro Decreased expression of the reporter gene
luciferase when driven by the 60 bp caspase-1 upstream
sequence () 151 to ) 92) containing this motif with a
mutation at position 98 position (G to T) in
compar-ison with the wild-type 60 bp upstream sequence was
observed The same mutation in the motif also
abol-ished the interactions of HIPPI in vitro In addition, we
observed that HIPPI could interact with the putative
promoter sequences of the caspase-8 and caspase-10
genes Expression of caspase-8 and caspase-10, as
detected by semiquantitative RT-PCR, was also
increased in the GFP–Hippi-expressing HeLa cells
The motif derived from bioinformatics analysis and
interaction studies of HIPPI with various mutants of
AAAGACATG present in the caspase-1 gene was
AAAGASAHK, i.e AAAGA[GC]A[ATC][TG] For
the caspase-8 gene, there are three variations in the
motif from that of the motif in the caspase-1 gene
upstream region Our experimental data suggest that
the change of C at the sixth position to G, and of T to
C at the eighth position, did not decrease the interac-tion, whereas a change from G to A at the ninth posi-tion compromised the interacposi-tions (Fig 1B, panel I and panel II) The change at the ninth position of the motif in the caspase-8 gene upstream sequence is G to
C As the caspase-8 gene 710 bp upstream sequence () 991 to ) 282) was shown to interact with HIPPI, we speculated that HIPPI interacted with the motif AAAGAGAAC () 963 to ) 955) We could not exclude the possibility that the other motif AAAG AAAAG () 418 to ) 410) present in the negative strand of the caspase-8 gene promoter interacted with HIPPI Further experiments are necessary to establish this Interaction of HIPPI with the motif AAAGA CATA () 808 to 800) at the positive strand (the vari-ant substitution G to A is underlined) of the caspase-8 gene was not possible, as we showed above that this particular motif did not interact with HIPPI (Fig 1B) The putative HIPPI-binding motifs at the caspase-10
11
A
B
Caspase10 ups + GST-Hippi
Casp8ups (GST-HIPPI)
K d =15 n M
Caspase8 ups + GST-Hippi
10 9 8 7 6
F340 5 4 3 2
0.140 0.145
0.135 0.130 0.125
F 0.15 0.20 0.25
0.10 0.05 0.00
0.00 0.01 0.02
[DNA] µ M
0.03 0.04 0.05
0 10 20
1/[DNA] µ M
1/[DNA] µ M
30 40 50
Fig 7 (A) In vitro binding study of GST–HIPPI with caspase-8 gene ( ) 991 to ) 282) and caspase-10 gene () 914 to ) 147) upstream sequences The upper line (squares) represents a gradual decrease
in the intrinsic fluorescence of GST–HIPPI protein (0.8 M ) at
340 nm (k ex ¼ 295 nm) due to addition of the caspase-10 gene upstream sequence The lower line in the graph (triangles) repre-sents a similar alteration in fluorescence intensity of GST–HIPPI when the caspase-8 gene upstream sequence was added gradually
to it (B) Linear plot of 1 ⁄ DF versus 1 ⁄ c, where DF represents change in intrinsic fluorescence of GST–HIPPI due to addition of upstream sequences from the caspase-8 (panel I) and caspase-10 (panel II) genes, and c represents final concentration (l M ) of DNA allowed to bind with the protein.
Trang 10gene upstream sequences are AAACAGATG () 254 to
) 262) in the positive strand, and AAAGAAAAG
() 651 to 643), AAAGAAAAG () 725 to ) 717) and
GAAGACATT () 849 to ) 857) in the negative
strand We were unable to exclude any of the motifs
as the target of HIPPI, as we observed that HIPPI
interacted with the putative promoter 768 bp () 914
to ) 147) sequence of the caspase-10 gene (Fig 7B,
panel II) Further experiments are necessary to
deter-mine the specific sequences where HIPPI could interact
at the upstream sequence of this gene
It is interesting to note that even though the initial
motif search revealed that the upstream sequence of the
caspase-7 gene contains the AAAGACATA sequence
present in duplicate within the) 355 to ) 347 and ) 315
to 307 regions, our experiments with this motif revealed
that HIPPI did not interact with it (Fig 1B, panel II)
Thus, the increase in caspase-7 expression in GFP–
Hippicells [9] might not be due to the direct interaction
of HIPPI with the promoter sequence This was further
supported by the observations that purified HIPPI did
not interact with the caspase-7 gene upstream 592 bp
sequence () 1080 to 489) in vitro (by EMSA and
fluor-escence quenching) or in vivo (chromatin
immunopre-cipitation assay using antibody to HIPPI) (data not
shown) We also failed to detect any interactions of
purified HIPPI with the caspase-3 gene upstream 652 bp
sequence () 997 to ) 346) (data not shown), even
though the exact motifs with which HIPPI could
inter-act were present in the negative strand (Table 3) of the
gene The reason behind this still remains obscure;
whether strand bias or the neighboring nucleotides
prevented the interaction remains to be determined
HIPPI does not have any similarity with known
pro-teins having DNA-binding motifs However, it
con-tains the pDED at the C-terminus (amino acid
residues 335–426) and a myosin-like domain The
pDED of HIPPI shows only 34.9% similarity and
21% identity to other known death effector domains
(DEDs), and 39.2% similarity and 26.7% identity with
the pDED of HIP1 It has been shown that the
inter-action of HIPPI with HIP1 is mediated through the
pDED present at HIP1 The pDED differs from its
conformational neighbor DED by the presence of
charged residues at the interacting helices, as opposed
to the hydrophobic ones in the later [8]
DED-contain-ing proteins are known to participate in diverse
cellu-lar functions, including apoptosis through receptor
signaling [20,21] Other DED-containing proteins, such
as DEDD, are known to bind DNA and inhibit RNA
polymerase I activity in vivo [22] Direct evidence that
the DED-containing proteins DEDD and FLAME-3
interact with the transcription factor TFIIIC102 and
thus regulate the transcription of the target genes has been also provided [23] We hypothesized that the pDED of HIPPI might have similar DNA-binding ability to that of its distant relative DEDD.Our find-ings that the purified C-terminal pDED of HIPPI was able to interact with the caspase-1 gene upstream sequence (Fig 4A,B), similar to what was observed with full-length HIPPI [18], and that the exogenous expression of cDNA corresponding to the pDED of HIPPI alone was sufficient to increase caspase-1 expression and apoptosis (Fig 5, Table 2) showed that the C-terminal pDED of HIPPI contributed to the increased expression of caspase-1 and apoptosis HIPPI generally resides in the cytoplasm How this cytoplasmic protein is transported to the nucleus remains unknown In an earlier study, we showed that exogenously expressed Hippi in HeLa cells can be detected in the nuclear fraction [Fig 2(b) in Majumder
et al [9]] It can be seen that there was no detectable endogenous expression of Hippi in HeLa cells, whereas the expression of HIP1 was detected in HeLa cells [Fig 2(c), III, in Majumder et al [9]] Recently, trans-portation of androgen receptor to the nucleus has been reported to be mediated through HIP1 [24] On the basis of this observation, we hypothesized that HIP1 might play similar role in the transport of HIPPI into the nucleus We are presently testing this hypothesis The role of HIPPI in HD, if any, remains unknown Even though the caspase-8 and caspase-10 genes have been implicated in poly-Q-mediated toxicity [25,26], it
is not known whether the expression of these genes is altered The role of HIPPI in the increased expression
of caspase-1 observed in various models and HD patients has to be established We speculated that excess available HIP1 in HD, due to weaker interac-tion of HIP1 with the mutated Htt allele, leads to the formation of more HIPPI–HIP1 heterodimer, which in turn increases caspase-1 expression We were able to immunoprecipitate the caspase-1 gene promoter by antibody to HIPPI after crosslinking DNA protein
in vivo [18] However, we failed to do the same thing with antibody to HIP1 (data not shown), indicating that HIP1 does not directly interact with the putative promoter of the caspase-1 gene Thus, the role of HIP–HIPPI heterodimer formation in the increased expression of caspase-1, caspase-8 and caspase-10 is not clear We speculated that it might be necessary for transporting HIPPI into the nucleus
In summary, together with our earlier observations [9,18], we observed in the present work that HIPPI interacted with the specific motif present in the putative promoters of the caspase-1, caspase-8 and caspase-10 genes and altered the expression of these genes The