Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
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: firstname.lastname@example.org
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 mutations 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 fluorescent protein–Hippi-expressing HeLa cells. In addition, HIPPI interacted in vitro with putative promoter sequences of these genes, containing a similar 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.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Huntingtin interacting protein 1 protein interactor
integrated optical density
‘pseudo’ death effector domain
transcription start site
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 , resulting in increased apoptosis in a specific region of the brain . Among various interacting partners of the protein Htt [3–5], Huntingtin interacting protein 1 (HIP1), identified in the yeast two-hybrid assay  and subsequently characterized as an endocytic adaptor protein with clatharin assembly activity, binds to various cytoskeleton proteins . 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 apoptotic 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 residues) . 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 heterodimer to form. The increased amount of HIP1–HIPPI may in turn lead to the increase in cell death observed in HD . 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 . The bifunctional apoptosis inhibitor, which regulates neuronal apoptosis, also interacts with HIPPI, although the functional relevance of this interaction remains unknown . Very recently, it has been reported that HIPPI interacts with the postsynaptic scaffold protein Homer1c and regulates apoptosis in striatal neurons . 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 .
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 . 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 expression 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 mitochondrial genes ND1 and ND4 and the antiapoptotic gene Bcl-2 are downregulated . Recently, we have also shown that HIPPI can directly interact with the caspase-1 gene upstream 60 bp sequence (− 151 to − 92) in vitro and in vivo. 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.
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, parameters and sequence sets (masked/unmasked) were then assembled and compared, and the redundant motifs were discarded (data not shown). The motif predicted using the methods mentioned was 5′-AAAGA[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 . 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[γP]-labeled dsDNA, AAAGACATG, in the presence of glutathione 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 AAAGACATG 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 fluorescence quenching study (Fig. 1B, panel I). With increasing amounts of dsDNA (AAAGACATG and AAAGAGATG), the fluorescence (λemission = 340 nm, λexcitation = 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, indicating 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 fluorescence 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 presence 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 concentration 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 fluorescence intensities of GST–HIPPI increased, indicating a lesser extent of interactions of GST–HIPPI with AAAGACATG. 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 caspase-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 consensus HIPPI-binding motif AAAGASAHK, i.e. AAAGA[GC]A[ATC][TG], was derived.
Table 1. Summary of binding study with the putative HIPPI-binding motif and its mutants. ND, not determined.
Average Kd (nm)
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 . This has been attributed to the presence of binding sites for other factors within these flanking sequences . As shown above, the 60 bp upstream sequence contains the motif AAAGACATG (− 101 to − 93), and mutation at position − 98 (G to T) abolished the interaction of HIPPI. To check whether this mutation also decreases 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 luciferase gene was driven by the 60 bp region with a mutation at position − 98 (G to T), was decreased (Fig. 3) significantly (P = 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 indicates 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 promoter activity of the 60 bp upstream sequences significantly. 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 substantial promoter activity. This could be due to additional transcription regulator-binding sites within the flanking sequence of the motif. It has been shown that p53 can bind within this region . 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 observed 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 (P = 0.4). Furthermore, there was no significant difference (P = 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 comprising 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 purified 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 interact 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.
As 6X(HN)-pDED of HIPPI does not contain any tryptophan, a 280 nm excitation filter was used, and the fluorescence (characteristics of tyrosine and phenylalanine) 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 (λem = 340 nm; fluorescence intensity changed from 15.62 to 14.42 due to addition of 0.5 µm DNA) or 280 nm (λem = 305 m; fluorescence 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 average 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 analysis, 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.
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)
GFP–N-terminus of HIPPI (fold)
Fold increase: pDED versus N-terminal domain (P-values)
8.9 ± 2.6
54.9 ± 2.1 (6.2-fold)
12.5 ± 3.9
2.4 ± 0.9
32.5 ± 1.8 (13.5-fold)
17.6 ± 1.9 (7.3-fold)
9.2 ± 2.0
23.1 ± 1.3 (2.5-fold)
14.8 ± 4.9
40.4 ± 3.6
90.2 ± 11.1 (2.2-fold)
57.2 ± 10.4
1.5 ± 0.3
10.8 ± 2 (7.2-fold)
5.1 ± 2.2 (3.4-fold)
To test whether the C-terminal pDED of HIPPI could induce apoptosis more efficiently than the N-terminal 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 fragmentation as an indication of apoptosis induction and caspase activation. GFP–pDED-expressing cells exhibited 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 significant (P = 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 (P = 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 (P = 0.047, n = 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 procaspase-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, P = 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 AAAGACATA (− 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 caspase-10 gene, four variant motifs were identified. Among them, AAACAGATG (− 254 to − 262) is present in the positive strand, and the sequences AAAGAAAAG (− 651 to − 643), AAAGAAAAG (− 725 to − 717) and GAAGACATT (− 849 to − 857) are present in the negative strand.
Table 3. Summary of the presence of similar motifs in caspase-1, caspase-3, caspase-7, caspase-8 and caspase-10 gene upstream sequences.
Allowing one substitution
Allowing one substitution
Allowing one substitution
Allowing two substitutions
Allowing one substitution
Allowing two substitutions
Allowing one substitution
Allowing two substitutions
Allowing two substitutions
Allowing one substitution
Given that the caspase-8 and caspase-10 genes harbor 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 . 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-expressing 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 (P = 0.0004). A similar significant increase in the IOD value of the RT-PCR products for the caspase-10 gene (1.8-fold, P = 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 representative photograph of the RT-PCR products run on agarose gel is shown in Fig. 6B. Equally intense signals for internal control (β-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 µm to 0.05 µm) of the caspase-8 and caspase-10 gene upstream sequences.
Average (n = 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 described previously , and typical cases are shown in Fig. 7B, panel I and panel II, respectively.
In the present work, we have shown that HIPPI interacted 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 comparison with the wild-type 60 bp upstream sequence was observed. The same mutation in the motif also abolished 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 interaction, whereas a change from G to A at the ninth position compromised the interactions (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 AAAGAAAAG (− 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 AAAGACATA (− 808 to 800) at the positive strand (the variant 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 gene 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 determine 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–Hippi cells  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 fluorescence quenching) or in vivo (chromatin immunoprecipitation 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 interact 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 proteins having DNA-binding motifs. However, it contains 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 interaction 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 . DED-containing proteins are known to participate in diverse cellular 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. 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 . We hypothesized that the pDED of HIPPI might have similar DNA-binding ability to that of its distant relative DEDD.Our findings 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 , 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. ]. 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. ]. Recently, transportation of androgen receptor to the nucleus has been reported to be mediated through HIP1 . 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 interaction 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. 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 presence of this motif in the promoters of other genes and regulation by HIPPI is now actively being investigated. Even though the role of HIPPI in HD remains obscure, the protein takes part in the regulation of caspase-1, caspase-8 and caspase-10 gene expression.
HeLa cells were obtained from National Cell Science Center, Pune, India, and routinely grown in MEM medium (HIMEDIA, Mumbai, India) supplemented with 10% fetal bovine serum (Life Technology, Rockville, MD, USA) at 37 °C in 5% CO2 atmosphere under humidified conditions.
Computational analysis to identify the motif
One kilobase upstream sequences for genes (the caspase-1, caspase-3 and caspase-7 genes) whose expressions are increased by exogenous Hippi expression  were retrieved from the ENSEMBL database using the biomart data retrieval tool (http://www.ensembl.org/Multi/martview). Putative cis regulatory elements were searched in these upstream sequences using four widely used motif prediction programs: meme, alignace, bioprospector and mdscan[27–30]. The search was also carried out on a second sequence dataset, which was prepared by masking the promoters of the upregulated genes using repeatmasker (http://repeatmasker.org). The motif prediction process was repeated several times with different parameters and different motif lengths. The results were compiled, and motifs that occur within the 60 bp caspase-1 gene upstream sequence (− 151 to − 92) were selected out using a custom perl script. To test the phylogenetic conservation of the predicted motifs, 1 kb upstream sequences of caspase-1 gene orthologs were downloaded from the DOOP database (http://doop.abc.hu) and were scanned for occurrence of the predicted motifs.
The in silico and experimentally derived motif consensus sequence AAAGASAHK, i.e. AAAGA[GC]A[ATC][TG], was searched for in upstream sequences of the caspase-8 and caspase-10 genes using patmatch.
Motif sequences and methods for making dsDNA and labeling
Motif sequence AAAGACATG (designated as P1) in the upstream region of the caspase-1 gene (− 101 to − 93) and its complementary sequence CATGTCTTT (P1C) were synthesized (IDT, Coralville, IA, USA). In addition, oligonucleotides, namely AGAGACATG (P2), AAGGACATG (P3), AAATACATG (P4), AAAGCCATG (P5), AAAGAGATG (P6), AAAGACCTG (P7), AAAGACACG (P8), and AAAGACATA (P9), and their complementary sequences CATGTCTCT (P2C), CATGTCCTT (P3C), CATGTATTT (P4C), CATGGCTTT (P5C), CATCTCTTT (P6C), CAGGTCTTT (P7C), CGTGTCTTT (P8C), and TATGTCTTT (P9C), were also synthesized chemically. The underlined sequences are changes from the original motif observed at the caspase-1 gene upstream sequence (P1).
Each single-stranded oligonucleotide was mixed with reverse oligonucleotide in a 1 : 1 molar ratio in sterile water; the mixture was then heated to 95 °C for 15 min and slowly cooled to 4 °C to allow perfect annealing. The kinase reaction was carried out with these double-stranded oligoneucleotides using polynucleotide kinase in the presence of 1 × polynucleotide kinase buffer and [32P]ATP[γP] (BRIT, Hyderabad, India). The radiolabeled oligonucleotides were used for EMSA as described previously .
Cloning of the Hippi pDED domain and N-terminal domain of Hippi in the bacterial expression plasmid pProTET and mammalian vector pEGFP
The pDED and N-terminal region of HIPPI encoded by the cDNA (gi|19923513) region (1003–1278 and 30–1003), respectively, were amplified by PCR using specific primer sets, namely, Hi_pDEDF (5′-ACGCGTCGACGTCGGAAATGGAGGAGTGACGG-3′), Hi_pDEDR (5′-CGGGATCCCGTTAATAAAAGCCTGTTGCTGGTT-3′), Hi_Nterm F (5′-ACGCGTCGACGTCATGACTGCTGCTCTGGCCGT-3′), and Hi_Nterm R (5′-CGGGATCCCGCTGCTGGTATCGCTCCTTTG-3′), and cloned in pPROTET and pEGFP plasmids using methods essentially described previously . pPROTET Clones were transformed in Escherichia coli strain BL21 Pro, and protein expression was induced by incubating with anhydrotetracycline (90 ng·mL−1) for 5 h. Proteins were isolated from cells by a freeze–fracture method, and purified by Ni2+–nitrilotriacetic acid affinity chromatography. Finally, the sizes of the pDED and N-terminal domain of HIPPI with the tag 6X(HN) were determined by 12.5% SDS/PAGE. The sizes were 13 kDa for pDED, and 40 kDa for the N-terminus of HIPPI. pEGFP clones were transfected in HeLa cells as described previously .
Different concentrations of GST–HIPPI or 6X(HN)-pDED of HIPPI were added to the probe 9 bp motif sequences, and mutated forms (allowed to bind with GST–HIPPI) or caspase-1 gene upstream sequences [allowed to bind with 6X(HN)-pDED of HIPPI] in binding buffer (1×: Hepes, 12.5 mm; EDTA, 0.5 mm; dithiothreitol, 0.25 mm; KCl, 37.5 mm; glycerol, 5%; MgCl2, 2.5 mm) containing 50 ng·µL−1 poly(dI:dC) were incubated at room temperature for 40 min. At the end of incubation, products were loaded on 5% polyacrylamide gel and the gel was run at 200 V for 4.5 h at 4 °C. The gel was then dried at 80 °C for about 45 min. The dried gel was exposed to X-ray film (Kodak, Mumbai, India) overnight at − 80 °C. After the film had been developed, positions of bands on the film were indicative of the positions of the probe. Cold 60 bp caspase-1 gene upstream sequence (without [32P]dCTP[αP] incorporation) in excess was used as competitive inhibitor of the reaction between the labeled 60 bp caspase-1 gene upstream sequence and 6X(HN)-pDED of HIPPI to determine the specificity of the interaction.
Fluorimetric quenching study
The purified GST–HIPPI protein was diluted (final concentration varied from 0.8 µm to 4 µm) in reaction buffer (100 mm Tris/HCl, pH 8.0, 50 mm NaCl). The concentration of DNA was increased gradually (3 nm to 200 nm) to the fixed amount of the protein, and the fluorescence intensities were measured at 340 nm [305 nm for 6X(HN)-pDED of HIPPI], exciting at 295 nm [280 nm for 6X(HN)-pDED of HIPPI protein] in a Hitachi 4010 Spectrofluorimeter (Hitachi, Tokyo, Japan) or Spex FluoroMax 3 Spectrofluorimeter (Edison, NJ, USA). Changes in fluorescence intensities (ΔF) due to addition of upstream sequences of the caspase-1 gene were calculated. From double reciprocal plot of ΔF and concentrations of DNA in the ranges where the decrease in fluorescence intensities reached saturation, apparent dissociation constants (Kd) for each DNA–protein binding reaction were calculated following the methods described by Sing & Rao .
Semiquantitative RT-PCR with gene-specific primers
Methods for RNA isolation, first-strand DNA synthesis, etc. have been published previously . After isolation, RNA was treated with RNase-free DNase (Sigma Chemicals, St Louis, USA) to remove possible genomic DNA contaminants. The first strand of cDNA was synthesized as described before, and used to study the expression of caspase-8 and caspase-10 by PCR. The caspase-8 gene-specific primers were: forward, 5′-AAGCAAACCTCGGGGATACT-3′; reverse, 5′-GGGGCTTGATCTCAAAATGA-3′. The caspase-10 gene-specific primers were: forward, 5′-GACGCCTTGATGCTTTCTTC-3′; reverse, 5′-ATGAAGGCGTTAACCACAGG-3′. PCR conditions for these two genes were similar, except for the annealing temperature. The common PCR conditions used for each of these loci were: initial denaturation at 94 °C for 1 min, followed by 35 cycles each containing three steps, denaturation at 94 °C for 15 s, annealing at 50 °C (for caspase-8) or at 60 °C (for caspase-10) for 30 s, and extension at 72 °C for 1 min and finally extension for 10 min at 72 °C.
The caspase-1 gene upstream 60 bp DNA (− 151 to − 92) was cloned in pGL3 as described previously . A mutation (G to T) at position − 98 of the 60 bp (− 151 to − 92) upstream sequence of the caspase-1 gene was introduced using specific primers (forward, 5′-CCTGATGCAGGCTACAGTTCT-3′; and reverse, 5′-GCATATGCATGTATTTATTTTTCTTC-3′) and standard procedures. The specific mutation (G to T) was confirmed by sequencing. HeLa cells were transfected with GFP–Hippi. Twenty-four hours after transfection, more than 90% of the transfected cells were expressing GFP-tagged protein (as visualized under a fluorescence microscope) . The 60 bp or mutated 60 bp sequences (G to T at position − 98) of the caspase-1 gene upstream sequence cloned in pGL3 vector and control pGL3 plasmid (without any insert) were transfected separately in HeLa cells expressing GFP–Hippi, and the cells were grown in the presence of geniticin (marker present at the pEGFP plasmid). Control pGl3 and 60 bp mutated 60 bp sequences of the caspase-1 gene upstream sequences cloned in pGL3 plasmid were also transfected into the parental HeLa cells. All these transfections were carried out using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA), following the procedure provided by the manufacturer. After 48 h of transfection of pGL3, cells were harvested and lysed, and luciferase substrate (Promega, Madison, WI, USA) was added. Luciferase activity was measured in a Sirius tube luminometer (Berthold Detection Systems, Pforzheim, Germany). In the same experiments, the transfection efficiencies of pGL3-containing upstream sequences of the caspase-1 gene in both control and GFP–Hippi-transfected cells were monitored by cotransfecting the β-galactosidase gene containing vector pSV-β-galactosidase (Promega) and measuring the β-galactosidase activity along with luciferase activity. Appropriate correction was made for equal transfection, using results obtained with the β-galactosidase activity. The transfection efficiency of the Hippi construct cloned in pEGFP plasmid was monitored by assaying GFP–Hippi expression in those cells as described previously , and appropriate correction was incorporated on the basis of this.
Detection of nuclear fragmentation, and caspase-1, caspase-3 and caspase-8 activation
Nuclear fragmentation, and caspase-1, caspase-3 and caspase-8 activation, were detected using methods described previously .
Western blot analysis
The methods of total protein isolation from exponentially growing cells and western blot analysis using antibodies to caspase-1 and caspase-8 were similar to those published previously .