Functional divergence of the brain-size regulating gene MCPH1 during primate evolution and the origin of humans

Background One of the key genes that regulate human brain size, MCPH1 has evolved under strong Darwinian positive selection during the evolution of primates. During this evolution, the divergence of MCPH1 protein sequences among primates may have caused functional changes that contribute to brain enlargement. Results To test this hypothesis, we used co-immunoprecipitation and reporter gene assays to examine the activating and repressing effects of MCPH1 on a set of its down-stream genes and then compared the functional outcomes of a series of mutant MCPH1 proteins that carry mutations at the human- and great-ape-specific sites. The results demonstrate that the regulatory effects of human MCPH1 and rhesus macaque MCPH1 are different in three of eight down-stream genes tested (p73, cyclinE1 and p14ARF), suggesting a functional divergence of MCPH1 between human and non-human primates. Further analyses of the mutant MCPH1 proteins indicated that most of the human-specific mutations could change the regulatory effects on the down-stream genes. A similar result was also observed for one of the four great-ape-specific mutations. Conclusions Collectively, we propose that during primate evolution in general and human evolution in particular, the divergence of MCPH1 protein sequences under Darwinian positive selection led to functional modifications, providing a possible molecular mechanism of how MCPH1 contributed to brain enlargement during primate evolution and human origin.

Previous evolutionary analyses of these microcephaly genes showed that four of them, ASPM, CDK5RAP2, CENPJ and MCPH1, evolved rapidly under Darwinian positive selection during the evolution of human and non-human primates [11][12][13][14][15][16]. ASPM also experienced positive selection across anthropoids [14][15][16], while CDK5RAP2 and CENPJ showed accelerated rates of non-synonymous substitutions over the course of primate evolution [11,16]. The signal of positive selection on MCPH1 was observed in the common ancestor of great apes and humans as well as in the human lineage [12], although another study on MCPH1 only detected positive selection in the anthropoids as a whole and no particular acceleration in the human lineage [16]. This rapid evolution suggests these genes may have had a key role in the evolutionary enlargement of the brain, although the link of CENPJ and MCPH1 to the evolution of gross brain size was not confirmed in the association analysis of absolute neonatal brain size among primates [16]. Among the four rapidly evolving microcephaly genes, only ASPM has been experimentally studied to detect the evolutionary consequence of protein sequence changes; mice carrying a truncated ASPM protein were shown to have reductions of both brain and testis size, while the transgenic mouse carrying human ASPM could rescue this phenotype, but did not cause any additional enlargement of the brain [17].
MCPH1 was the first gene identified as being responsible for autosomal recessive primary microcephaly, characterized by significantly reduced brain volume, mental retardation and premature chromosome condensation (PCC) syndrome [2,18]. The MCPH1 gene encodes a 2,508-bp-long coding sequence (CDS) with 14 exons, spanning about 240 kb at 8p23. The MCPH1 protein contains three BRCA1-Carboxyl Terminal (BRCT) domains, including one N terminal BRCT domain and a tandem pair at the C terminus. Numerous studies have implied that the BRCT domains of MCPH1 function as the key component for protein-protein interaction; this seems likely as the interaction of the MCPH1 tandem BRCT domains with proteins like E2F1 and r-H2AX is required for the activation of cell cycle checkpoint, DNA repair and apoptosis [19][20][21][22][23]. Several studies have likewise suggested that MCPH1 may also function as a tumor suppressor [3,23].
Evolutionary studies of MCPH1 have demonstrated a rapid change in protein sequence associated with the brain enlargement during primate evolution and human origin. Interestingly, during two key taxonomic transitions in primates, that is, between lesser apes and great apes, and between great apes and humans, absolute brain volume was greatly enlarged, and MCPH1 might be involved in this process [12]. Additionally, MCPH1 is also highly polymorphic in human populations and still carries the molecular signature of on-going positive selection [12]. Human population studies have reported a sex-specific association between a MCPH1 sequence variant and brain volume [24,25]. These results suggest that the protein sequence changes, especially the humanspecific changes of MCPH1 may have caused the functional changes that explain the genetic basis for the evolution of brain size in primates.
Previously, the MCPH1 protein has been shown to play an essential role during cell cycle and cell apoptosis and it can physically interact with E2F1 to form a complex and bind the promoters of the target genes for regulating their transcriptional activities [19]. Beyond this, MCPH1 alone can also function as a transcriptional regulator, and we previously demonstrated MCPH1 could function as a transcriptional repressor [26]. Together, these regulatory mechanisms allow the experimental testing of the functional changes of MCPH1 during primate evolution.
To detect if the protein sequence divergence of MCPH1 among primates may confer any functional alterations, we selected eight known down-stream genes regulated by E2F1 and MCPH1: p73 [19], p107 [19], p18 [27], p27 [28], p14 ARF [19], Caspase7 [19], CyclinE1 [19] and hTERT [26]. These genes are involved in cell proliferation and apoptosis, critical processes regulating brain development (see Additional file 1: Figure S1). We tested the activating effects (together with E2F1) and the repressing effects (MCPH1 alone) of MCPH1 on these genes' promoter when introducing mutations at the sites containing human-and great-ape-specific amino acid changes. Our results demonstrated that most of the human-specific amino acid substitutions could influence the regulatory effects of MCPH1 on the down-stream genes, and a similar effect was also seen for one of the four great-ape-specific changes, suggesting that the species and lineage specific mutations of MCPH1 are indeed functional and potentially contributed to brain enlargement over the course of primate evolution.

Identification of lineage-specific MCPH1 amino acid substitutions
In order to identify lineage-specific amino acid substitutions, MCPH1 orthologs of representative primate and other mammalian species were obtained from the NCBI, EMBL and UniProt databases. We used a total of 11 species including 7 primates (human, great apes, lesser ape, Old World monkey and New World monkey) and 4 other representative mammalian species (mouse, rat, dog and cow) ( Figure 1). Using MUSCLE and Clustal W, we aligned the MCPH1 protein sequences (see Additional file 2: Figure S2). During primate evolution, there are two key taxonomic transitions accompanied by brain enlargements (between lesser apes and great apes, and between great apes and humans). We identified nine sites containing substitutions considered as human-specific sites that occur in the human lineage but are relatively conserved in the other species (M96, S101, V310, H314, T377, Y425, L442, R485 and P835; Figure 1, Additional file 3: Table S1). We further selected four sites containing substitutions considered as great-ape-specific since they occurred in the ancestor of Hominidae (I161, E167, A510 and S841; Figure 1, Additional file 3: Table S1). The physicochemical properties of the 13 lineage-specific amino acid substitutions are shown in Additional file 3: Table S1 and the schematic map of the MCPH1 protein domains labeled with the lineage-specific substitutions are shown in Figure 1. All these lineage-specific sites were selected to generate mutant MCPH1 proteins for the reporter gene assays in order to test their functional effects.

Test of protein-protein interaction between MCPH1 and E2F1
Previous studies suggested human MCPH1 (hMCPH1) could interact with E2F1 in vitro and in vivo [19]. To test if the non-human primate MCPH1 can also interact with E2F1, we cloned the rhesus macaque MCPH1 (mMCPH1). The results of the co-immunoprecipitation assay showed that mMPCH1 can also directly interact with E2F1, and no difference was observed for the intensity of the interaction with E2F1 between hMCPH1 and mMCPH1 ( Figure 2A). Given the established MCPH1-E2F1 interaction in human cell lines [19], this proteinprotein interaction mechanism seems likely to have been conserved during primate evolution. We also conducted a cellular co-localization assay for both mMCPH1 and hMCPH1, and the results indicated that both were co-localized with E2F1 ( Figure 2B), consistent with the results of the co-immunoprecipitation assay.
Divergent effects of hMCPH1 and mMCPH1 on transcriptional regulation MCPH1 interacts with E2F1 to enhance its transactivation activity by forming a complex, which binds to the promoters of the E2F1 target genes, including p73, Caspase and RAD51, among others [19]. We, therefore, tested the enhancing effect of MCPH1 using the E2F4B reporter vector commonly used as the positive control in testing the transactivation activity of E2F1 [29]. The expression of the HA-tagged MCPH1 and E2F1 were verified by Western blot using anti-HA monoclonal antibodies ( Figure 2C). Our results showed that hMCPH1 could significantly enhance the transactivation activity of E2F1, but no significant effect was observed for mMCPH1 and gMCPH1 (the gibbon copy of MCPH1) ( Figure 2D). In addition, the difference between E2F1 + hMCPH1 and E2F1 + mMCPH1 is significant (P <0.01, ANOVA test). The same trend was also seen when comparing E2F1 + hMCPH1 and E2F1 + gMCPH1, though not statistically significant ( Figure 2D). Interestingly, the difference between E2F1 + gMCPH1 and E2F1 + mMCPH1 is significant (P <0.05, ANOVA test), suggesting continuum of functional divergence of MCPH1 during primate evolution with a major shift in function between Old World monkeys and apes and a probable further shift during human evolution ( Figure 2D).
To check the conservation of the E2F1 binding sites (CGCGC), we aligned the promoter sequences of the target genes among representative primate species. We found that most of the E2F1 binding sites are highly conserved from humans to marmosets (see Additional file 4: Figure S3), ruling out the possibility of binding bias due to promoter sequence divergence. The E2F1 protein sequences are also highly conserved among primate species (Additional file 5: Figure S4), ruling out the possibility that the different effect of MCPH1 is caused by mutations in E2F1. Collectively,

BRCT1 BRCT3 BRCT2
Two key taxonomic transitions with brain enlargements  Figure 1 Schematic map of MCPH1 protein domains labeled with lineage-specific amino acid substitutions. Sites containing human specific substitutions are marked with shadows, and the two dramatic brain size enlargement events that coincided with the molecular signatures of Darwinian positive selection during primate evolution are indicated.
these results suggest a functional divergence of MCPH1 between humans and nonhuman primates. We further tested a set of E2F1 target genes (p73, p107, p18, p27, p14 ARF , Caspase7 and CyclinE1). The regulatory network of these genes is shown in Additional file 1: Figure S1. We found that the enhancing effect of hMCPH1 was significantly stronger than mMCPH1 for CyclinE1 and p73 ( Figure 3A, B) and a similar trend was also observed for p14 ARF , though not significantly ( Figure 3C). No significant enhancing effects were detected for the other four genes, p18, p27, p107 and Caspase7 (see Additional file 6: Figure S5). Consistent with the result from the E2F4B assay, these data also suggest a functional divergence between hMCPH1 and mMCPH1 in enhancing the transactivation activity of E2F1.
To further test the functional divergence between hMCPH1 and mMCPH1, we performed another assay to detect the regulatory effect of MCPH1 alone on eight down-stream genes (p73, p107, p18, p27, p14 ARF , Caspase7, CyclinE1 and hTERT). Among the eight genes tested, seven showed a repressing effect for both hMCPH1 and mMCPH1, while only one, Caspase7, showed an activating effect. When comparing the effects between hMCPH1 and mMCPH1, three genes (CyclinE1, p73 and p14 ARF ) showed significant differences (P <0.05) of repressing effect ( Figure 3D-F), the same three which showed enhancing differences between hMCPH1 and mMCPH1 in the E2F1-MCPH1 transactivation assay ( Figure 3A-C). For CyclinE1 and p14 ARF , hMCPH1 showed a stronger repressing effect than mMCPH1, while the opposite effect was observed for p73. We did not detect significant between-species differences for the other four genes with repressing effect (p18, p27, p107 and hTERT) though the trend was the same (see Additional file 7: Figure S6). Caspase 7 was the only gene with an activating effect, and we also observed a significant difference between hMCPH1 and mMCPH1 (P <0.05) (see Additional file 7: Figure S6D). Collectively, the data from the two reporter gene assays suggest a clear functional divergence of MCPH1 between humans and rhesus macaques.

Detection of regulatory changes of MCPH1 for the human-specific sites
In order to test the functional effect of the human specific amino acids changes, we first constructed two mutant hMCPH1s by introducing a backward mutation of changing the human-specific amino acid to an ancestral amino acid (Met96Thr) and introducing a random mutation (Met96Leu) at Site-96 (Figure 1). With the use of the E2F4B transactivation reporter assay, we measured the enhancing effect of the two mutant hMCPH1s. As shown in Figure 4A, both mutant hMCPH1s showed a significantly decreased enhancing effect as compared with the wild-type promoter. (C) The enhancing effect difference between hMCPH1 and mMCPH1 on the p14 ARF promoter. For enhancing effect assay, 0.2 ug target gene promoter construct was co-transfected with 0.2 ug HA tagged E2F1 per/well or 0.2 ug HA tagged E2F1 and 0.2 ug HA tagged hMCPH1 per/ well or 0.2 ug HA tagged E2F1 and 0.2 ug HA tagged mMCPH1 per well in 24-well plates in HEK923T cells, Renilla was used as the internal control. The HA-tagged empty vector was used as the negative control. (D) The repressing effect difference between hMCPH1 and mMCPH1 on the CyclinE1 promoter. (E) The repressing effect difference between hMCPH1 and mMCPH1 on the p73 promoter. (F) The repressing effect difference between hMCPH1 and mMCPH1 on the p14 ARF promoter. For repression assay, 0.2 ug target gene promoter construct was co-transfected with 0.2 ug HA tagged hMCPH1 per/well or 0.2 ug HA tagged mMCPH1 per/well in 24-well plates in HEK923T cells, Renilla was used as the internal control. The HA-tagged empty vector was used as the negative control. All histograms represent the mean ± SD of at least nine data points including both biological and technical replicates (* P <0.05,** P <0.01, *** P <0.001, n.s, not significant).
hMCPH1, similar to the difference between hMCPH1 and mMCPH1, suggesting that the human-specific amino acid change at Site-96 can influence the enhancing effect of MCPH1.
We next generated a series of mutant hMCPH1s carrying backward mutations at the other eight human-specific sites (101, 310, 314, 377, 425, 442, 485 and 835), and tested their enhancing effects. Five of the eight mutant hMCPH1s (101, 310, 377, 425 and 835) showed significantly decreased enhancing effects compared with the wild-type hMCPH1 (P <0.05) ( Figure 4B, C) and the three non-significant sites (314, 442 and 485) showed the same trend. Among the nine human specific sites tested, the majority (6/9) showed a significantly decreased enhancing effect when mutated to the ancestral amino acids, suggesting most of the human-specific MCPH1 amino acid changes are in fact functional.
We also tested the repressing effect on p14 ARF for the human-specific sites. As shown in Figure 4D, E, we detected a significant decrease of the p14 ARF repressing effect for five sites (101, 310, 425, 442 and 485), and the same trend was observed for the other four, though it was not significant. Among the five significant sites, three (101, 310 and 425) overlapped with the six sites showing decreased enhancing effect. Taken together, we detected regulatory changes for most (eight out of nine) human-specific sites, suggesting hMCPH1 has acquired functional modifications during the origin of humans.

Detection of regulatory changes of MCPH1 for the great-ape-specific sites
To detect the functional consequences of MCPH1's four great-ape-specific mutations (Figure 1), we conducted analyses similar to that used for the human-specific sites. The results indicated that for the enhancing assay, although all four great-ape-specific sites showed a decreased level as compared with the wild-type of hMCPH1, none were statistically significant (P >0.05; Figure 5A). In contrast, in the repressing assay, one great-ape-specific site (Site-167) showed a significant decrease of the repressing effect The results of the E2F4B assay testing the changes of enhancing effect. (D-E) The results of the p14 ARF assay testing the repressing effect. All histograms represent the mean ± SD of at least nine data points including both biological and technical replicates (* P <0.05,** P <0.01, *** P <0.001, n.s, not significant).
( Figure 5B), suggesting that this great-ape-specific mutation may have also caused functional modifications of MCPH1 during the evolution of the ancestor of Hominidae.

Discussion
MCPH1 has experienced strong Darwinian positive selection during primate evolution, but there are no data showing functional divergence. Here we demonstrated evidence of functional divergence of MCPH1 between humans and nonhuman primates. Most of the human-specific amino acid changes could alter the regulatory effects of MCPH1 on the transcription of the down-stream genes, and a similar effect was observed for one of the four great-ape-specific amino acid changes. Accordingly, our data support the hypothesis that selection on MCPH1 has resulted in functional divergence at the protein level, which potentially contributes to changes in the development and evolution of brain size. Absolute brain size has increased in parallel across primate evolution. Along the two branches we focused on in this analysis, absolute brain volume increased from 70 to 152 ml to 230 to 565 ml during the transition between lesser apes and great apes [30,31], and from 230 to 565 ml to 1,129 to 1,685 ml during the transition between great apes and humans [30,31]. Previous studies indicated that there were accelerated amino acid substitutions during both the origin of Hominidae's ancestor and of our own species, paralleling the two brain enlargements [12], suggesting that the amino acid substitutions of MCPH1 were probably adaptive and may have contributed to the brain expansion during primate evolution. In addition, the gradient change of MCPH1's transcription regulation from macaque to gibbon, and to humans ( Figure 2D) seems to imply a continuum of functional divergence rather than a number of discrete shifts, which calls for further functional tests in extensive primate lineages.
Interestingly, all the human-and great-ape-specific mutations are located in the non-BRCT domains (Figure 1). Since the three BRCT domains of MCPH1 are critical for protein-protein interaction, the amino acid changes during primate evolution seems not to have caused drastic functional alteration, but rather a modification of the existing function.
Inferring the exact functional alterations of the human-specific and great-ape-specific mutations is difficult. Previous studies have shown that the middle domain (residues 367 to 485) of MCPH1, where the four human-specific mutations are located, is the binding domain by Condensin II for homologous recombination repair [32,33], an important mechanism for cell cycle checkpoints and genome integrity. Concurrently, all four human-specific sites located in this middle domain showed altered regulatory effects when mutated into ancestral amino acids, suggesting that the human-specific mutations may have changed the binding property with Condensin II. Additionally, all four human-specific mutations caused changes in physicochemical properties of amino acids (see Additional file 3: Table S1).
We also found that for the regulatory changes of the down-stream genes, almost half (three out of eight) of the tested genes (p73, CyclinE1 and p14 ARF ) had significant differences between humans and rhesus macaques, either in the enhancing assay with MCPH1-E2F1 or in the repressing assay with MCPH1 alone (Figure 3), indicating a functional divergence between humans and non-human primates. The protein p73 is involved in both cell cycle regulation and induction of apoptosis [34]. E2F1 is an important regulator of p73, especially during brain development [35]. CyclinE1, meanwhile, is involved in cell cycle and is a key target gene of E2F1 [36] and has been shown to take part in the determination of the number of neurons during mouse corticogenesis by regulating the G1 mode of cell division [37]. p14 ARF is an alternate reading frame product of CDKN2A involved in cell cycle regulation that is also involved in self-renewal of neural stem cells and neural development [38][39][40][41]. Additionally, human population studies have reported that the MCPH1 sequence variants were associated with brain volume in a sex-specific manner [24,25]. Recently, it was also reported that MCPH1 might have contributed to the evolution of sexual dimorphism in brain mass across anthropoid primates [42]. In fact, two of the downstream genes regulated by MCPH1, p73 and cyclinE1, were reported to be associated with sex dimorphism during germ line development [43,44], suggesting that the regulation of MCPH1 on brain development may differ between males and females. Taken together, the strengthened transactivation effect of human MCPH1 on these down-stream genes may contribute to the greatly enlarged neuro-progenitor pool in the human brain during neurogenesis, which is in line with recent studies that suggest MCPH1's functional role in neuro-progenitor cells through the Chk1-Cdc25-Cdk1 pathway [45,46].
Conversely, when MCPH1 acts alone as a transcription repressor, there were also differences between humans and rhesus macaques on the repressing effect of the down-stream genes (p73, CyclinE1 and p14 ARF ), implying that a homeostasis of gene expression regulation by MCPH1 is required during neurogenesis.
Although we observed functional divergence of MCPH1 due to its protein sequence changes during primate evolution and human origin, it should be stressed that we did not establish a direct link between the adaptive changes of MCPH1 and the ever-increasing brain size in primates. As shown in the MCPH1 knock-out mice analysis, the truncated MCPH1 not only caused a reduction in brain size, but also resulted in a reduction of testis size [45], suggesting that MCPH1 may also play a role during testis development. Accordingly, we cannot rule out the possibility that the adaptive evolution of MCPH1 in primates may be caused by selection on other phenotypes, though the current evidences mostly favor enlargement of the brain.
Initially proposed by King and Wilson [47], the importance of cis-regulatory changes in human evolution has recently been tested and confirmed [48]. However, our functional data of MCPH1 suggests that protein sequence changes may also have significant phenotypic effects. Hence, the evolution of an important trait like brain function may require genetic alterations at multiple regulatory levels.

Conclusions
We demonstrated the existence of functional alterations caused by the lineage-specific mutations of MCPH1 during the evolution of primates, especially during the origin of humans. The functional changes of MCPH1 are likely executed by regulating several key down-stream genes.

Ethical statement
The research protocol of this study was approved by the internal review board of Kunming Institute of Zoology, Chinese Academy of Sciences (Approval ID: SYDW-2012011).

Cell culture
The HEK293T cell line was obtained from ATCC. Cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) (Gibco, Rockville, MD, USA) with 10% fetal bovine serum (Hyclone, Logan, UT, USA) at 37°C in a humidified atmosphere containing 5% CO 2 .

Cloning of fluorescent MCPH1 plasmids
The full length cDNAs of human MCPH1 and macaque MCPH1 were PCR amplified, and the PCR products were digested with Age I and EcoR I and cloned into frame for N-terminal fusions into the cFUGW plasmids. The final constructs were confirmed by sequencing (ABI-3130 automatic sequencer).

Transient transfection and luciferase reporter assays
All transfections were carried out in triplicates in the 24-well plates (Corning, Corning, NY, USA). About 2 × 10 5 cells were seeded for 24 h prior to transfection. Briefly, equal numbers of cells were plated in the 24-well and 6-well plates and grown to 80% confluence. The indicated amounts of vectors were mixed in OPTI-MEM medium (Gibco) with Lipofectamine 2000 (Invitrogen). The solution was incubated for about 30 minutes at room temperature and then placed on the cultured cells. After four to six hours, the medium was changed into DMEM (Gibco) with 10% fetal bovine serum (Hyclone). For luciferase assay, cells were grown in the 24-well plates and transfected with the indicated amounts of vectors, including pTK-Renilla as an internal control, and Lipofectamine 2000 (Invitrogen) was used. Luciferase activity was assayed 28 to 32 h after transfection. The luciferase activity in cell extracts was determined by the Dual-luciferase Reporter Assay System (Promega, Madison, WI, USA) according to the manufacturer's protocol. The relative light units were measured using a luminometer.
The wide range of relative luciferase activity seen in different panels was likely due to the different amount of cells used at each biological replicate, which did not influence the measurement of relative activity. At least three technical replicates were conducted for each experiment. To avoid transfection efficiency bias, we also performed at least three biological replicates for the luciferase assay. The promoter constructs of p73, p107, p18, p27, p14 ARF , Caspase7 and CyclinE1 were kindly provided by Dr. Wuhan Xiao from Institute of Hydrobiology, Chinese Academy of Sciences, and these constructs were published before [26,49,50].
To test the co-localization of MCPH1 and E2F1, the GFP tagged MCPH1 expression vector was co-transfected with the RFP tagged E2F1 expression vector into the HEK293T cells. After 24 to 48 h, the cells were checked under fluorescent microscopy.

Generating MCPH1 mutants
The human MCPH1 gene copy was used to generate mutants carrying mutations at the sites with humanspecific and great-ape-specific mutations. A total of 13 sites were tested, including the 9 sites containing human-specific mutations and the 4 sites containing great-ape-specific mutations. The mutant MCPH1s were prepared using the QuickChange II XL site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) and specific oligodeoxynucleotide primer sets. The intended mutations were confirmed by sequencing. The oligodeoxynucleotide primers used for generating the mutant MCPH1s are shown in Additional file 8: Table S2 and Additional file 9: Table S3.

Co-immunoprecipitation
Flag-E2F1 and HA-MCPH1 were co-transfected into the HEK293T cells in the six-well plates by Lipofectamine 2000 with the total amount of 16 ug DNA. After 36 to 38 h of transfection, cells were lysed with 400 ul lysis buffer (50 ml Tris-HCl, pH = 7.4; 150 mM NaCl; 1 mM EDTA; 1% Triton-100; 1 Mm Na 3 VO 4 ) containing a cocktail of protease inhibitors (Sigma Chemical St. Louis, MO, USA). Cell lyses were incubated with HA agarose beads (Sigma Chemical) overnight at 4°C with lysis buffer, and boiled in 2 × protein loading buffer. Blots were blocked in 0.05% Tris-buffered saline (TBS), 20% Tween and 5% non-fat milk followed by incubation with the indicated primary (anti-Flag (M2) from Sigma or anti-HA from Covance, Princeton, NJ, USA) and secondary antibody (anti-mouse from KPL, Inc. Maryland, Washington, D.C, USA) in this buffer. Immunoreactivity was detected with an enhanced chemiluminescence system (Pierce Protein Biology, Rockford, IL) with colored markers (Fermentas, Pittsburgh PA, USA) as the molecular size standard.

Western blotting
Proteins from the HEK293T cells were homogenized in RIPA lysis buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1 mM EDTA; 1% Triton-100; 1 mM Na 3 VO 4 ) containing a cocktail of protease inhibitor (Sigma Chemical). Extracted proteins (15 to 20 μg) were separated by SDSpolyacrylamide gel electrophoresis and electrophoreticly transferred to a membrane incubated with anti-HA monoclonal antibody (Covance). Immunoreactivity was detected with an enhanced chemiluminescence system (Pierce Protein Biology) with colored markers (Fermentas) as the molecular size standard.