A novel role for DYX1C1, a chaperone protein for both Hsp70 and Hsp90, in breast cancer
With three consecutive tetratricopeptide repeat (TPR) motifs at its C-terminus essential for neuronal migration, and a p23 domain at its N-terminus, DYX1C1 was the first gene proposed to have a role in developmental dyslexia. In this study, we attempted to identify the potential interaction of DYX1C1 and heat shock protein, and the role of DYX1C1 in breast cancer.
GST pull-down, a yeast two-hybrid system, RT-PCR, site-directed mutagenesis approach.
Our study initially confirmed DYX1C1, a dyslexia related protein, could interact with Hsp70 and Hsp90 via GST pull-down and a yeast two-hybrid system. And we verified that EEVD, the C-terminal residues of DYX1C1, is responsible for the identified association. Further, DYX1C1 mRNA was significantly overexpressed in malignant breast tumor, linking with the up-regulated expression of Hsp70 and Hsp90.
These results suggest that DYX1C1 is a novel Hsp70 and Hsp90-interacting co-chaperone protein and its expression is associated with malignancy.
KeywordsDYX1C1 Heat shock protein Tetratricopeptide repeats Protein interaction Breast cancer
Over the past couple of decades, our understanding of the role of the tetratricopeptide repeat (TPR) domain has made great progress. A TPR domain is defined by a repeating motif of 34 amino acids that adopt a helix-turn-helix secondary structure (Blatch and Lassle 1999). The concave and convex surfaces at the inner and outer interfaces (D’Andrea and Regan 2003) of this domain facilitate the binding of various ligands. The diversity of co-factors that can associate with this domain, coupled with the biological significance of those protein–protein interactions, have provided an ongoing area of study. TPR-containing proteins have previously been studied for their roles in molecular chaperone complexes, anaphase promoting complexes, transcription repression complexes, and the organization of protein import complexes (Tzamarias and Struhl 1995; Miyata et al. 1997; Moczko et al. 1997). However, these TPR-mediated functions appear to only be the tip of the iceberg. Identifying all of the possible biological functions of TPR-containing proteins and their specific binding partners is an ongoing challenge.
In this study, we specifically look at the human TPR-containing protein, DYX1C1, which contains three consecutive TPR motifs in its C-terminus. Taipale et al. first reported DYX1C1 as a candidate gene for a role in developmental dyslexia (Taipale et al. 2003). Developmental dyslexia is defined as a specific and persistent failure to acquire efficient reading skills despite conventional instruction, adequate intelligence, and socio-cultural opportunities (Chaix et al. 2007). Ramus et al. concluded that developmental dyslexia can be attributed to all neuro-develomental disorders, especially neuronal migration abnormalities (Ramus 2004). When the biological function of DYX1C1 was investigated, DYX1C1 was shown to play a role in the migration of neocortical neurons, and disruption of DYX1C1 leads to neuronal migration disorders similar to those seen in the brains of dyslexics (Wang et al. 2006; Chaix et al. 2007). Currently, the biological functions of DYX1C1 remain only partially identified. Importantly, the molecular mechanism of how DYX1C1 functions still remains unknown and its binding interactions are unreported.
In the present study, we compared expression of DYX1C1 in tumor tissues versus normal tissues. Unexpectedly, we detected expression of DYX1C1 predominately in proliferative tissues. Furthermore, we confirmed that DYX1C1 can interact with heat shock proteins, Hsp70 and Hsp90, via its TPR domain using glutathione S-transferase (GST) pull-down assays and a yeast two-hybrid system, respectively. We hypothesize that the identification of proteins that bind DYX1C1 will provide valuable insight into the role of DYX1C1 in neuronal migration.
Materials and methods
Sequence analysis of DYX1C1
Several online sequence analysis tools were utilized. To search for similar protein sequences in different species, Blast (http://www.ncbi.nlm.nih/Blast/) and Conserved Domain Architecture Retrieval Tool (CDart) (http://www.ncbi.nlm.nih.gov/structure/lexington/lexington.cgi?cmd=rps) were used. To search for conserved domains, Smart (http://smart.embl-heidelberg.de/Smart) was used. To align similar proteins, ClustalW (http://www.ebi.ac.uk/clustalw/) and the ClustalW (1.83) Multiple Sequence Alignments program were used.
Distribution of DYX1C1 in normal tissues
Nucleotide sequences of primers used in this study
Primer sequence (5′-3′)
Collection of breast tumor tissue specimens
Forty-one breast tumor tissue specimens were surgically excised from female patients treated in the Department of Breast Surgery, First Affiliated Hospital of Nanjing Medical University. Of the 41 samples, 31 were diagnosed as malignant breast tumors, and 10 were benign. For controls, 10 tissue samples were obtained from the same mastectomy patients (matched pairs), but at a distance of at least 2 cm from the primary tumor. All tissue specimens were snap-frozen in liquid nitrogen and stored at −80°C until used. Forty-one biopsy samples of breast tumor tissues and their associated pathological details were generously provided by First Affiliated Hospital of Nanjing Medical University. Histological differentiation was determined from hematoxylin and eosin (H&E) staining. Assessment of histological grade was according to the method of Elston and Ellis (1991.).
Total RNA extraction, cDNA synthesis, and semi-quantitative RT-PCR
Total RNA was isolated from breast tumor tissues using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. The extracted RNA was reverse transcribed with random primer N6 of the RetroScript kit (Ambion, Austin, TX) according to the manufacturer’s instructions. Hsp70 primers (Table 1) were designed from Genbank entry NM_005345.4 and amplified a 266-bp product. The primers for DYX1C1 and β-actin are listed in Table 1, and the RT-PCR protocol was the same as described for the PCR assays listed above. All RT-PCR experiments were repeated three times. RT-PCR products were separated by electrophoresis and visualized on a 1.5% agarose gel.
Acquisition of gel images and semi-quantitative analysis
Images of the RT-PCR ethidium bromide-stained agarose gels were acquired with a Panasonic CCTV camera (Panasonic System Solution Suzhou, co., Ltd, Suzhou, China) and quantification of the bands was performed by JEDA801 (Jiangsu JEDA Science-Technology Development Co., Ltd). Band intensity was expressed as relative absorbance units. The ratio between the sample RNA to be determined and β-actin was calculated to normalize for initial variations in sample concentration and as a control for reaction efficiency. Mean and standard deviation of all experiments performed were calculated after normalization to β-actin.
Construction of plasmids containing DYX1C1, Hsp70, and Hsp90
Plasmids pGADT7, pGBDT7, pET28a, and pGEX-4T-1 were developed in our lab and include two Sfi-I restriction sites [5′-Sfi-I (A) site (GGCCATTACGGCC); 3′-Sfi-I (B) site (GGCCGAGGCGGCC)] separated by a sufficient number of basepairs to allow efficient cleavage by Sfi-I. The human Hsp70 and Hsp 90 genes were amplified from pEGFP-Hsp70 (Addgene) and pEGFP-Hsp90, respectively, by PCR using oligos containing the Sfi-I (A) and Sfi-I (B) restriction sites. The Hsp70 and Hsp90 open reading frames (ORFs) were cloned into vector, pGADT7 (Clontech), to create pGADT7-Hsp70 and pGADT7-Hsp90 for use in a yeast two-hybrid system. After sequence verification, the Hsp70 insert was subcloned into pGBDT7 and pET28a between the Sfi-I (A) and Sfi-I (B) sites as previously described (Cao et al. 2006), yielding expression constructs BD-bait Hsp70 (pGBDT7-Hsp70) for the yeast two-hybrid assay and His-tagged Hsp70 (pET28a-Hsp70) for both in vitro translation and GST pull-down assays. Hsp70C is a truncated form of Hsp70 that includes the C-terminal residues of Hsp70 (amino acids 449–640) and was generated using primers Hsp70s-F and Hsp70s-R (Table 1). Similarly, Hsp90C is a truncated form of Hsp90 that includes the C-terminal residues of Hsp90 (amino acids 521–774) and was generated using primers Hsp90s-F and Hsp90s-R (Table 1). PCR fragments obtained were digested with Sfi-I and ligated into the Sfi-I sites of pGADT7.
Hsp70 mutant constructs were assembled using a PCR-based, site-directed mutagenesis approach. The plasmid F2 (pGADT7-GFP-EEVD) from our lab includes the EEVD motif, which are the last four amino acid residues shared by both Hsp70 and Hsp90 at their C-terminus. The plasmid F1 (pGADT7-GFP-MEEVD) contains the C-terminus motif, MEEVD, derived from Hsp90; plasmid F3 (pGADT7-GFP-GPTIEEVD) contains the C-terminal motif, GPTIEEVD, derived from Hsp70; plasmid F7 (pGADT7-MEEVD-GFP) contains the N-terminal motif, MEEVD. In addition, mutants F4 (pGADT7-GFP-MEEVA) and F5 (pGADT7-GFP-GPTIEEVA) contain an amino acid substitution at the extreme C-terminus. Plasmid F6 (pGADT7-GFP) was a negative control. These six mutants, F1–F5 and F7, were generated using primer A (Table 1) as the forward primer, and the reverse primers B, C, D, E, F, or H, respectively (Table 1). Mutant F6 was generated using primer G as the forward primer and primer H as the reverse primer. PCR fragments were digested with Sfi-I and ligated into the Sfi-I sites of the vector pGADT7. The resulting plasmids, F1 (pGADT7-GFP-MEEVD), F2 (pGADT7-GFP-EEVD), F3 (pGADT7-GFP-GPTIEEVD), F4 (pGADT7-GFP-MEEVA), F5 (pGADT7-GFP-GPTIEEVA), F6 (pGADT7-GFP), and F7 (pGADT7-MEEVD-GFP), had their sequences verified and were used in the yeast two-hybrid assays.
Full-length DYX1C1 (amino acids 1–398), and the TPR domain of DYX1C1 (DYX1C1s, comprising amino acids 293–398) were generated using PCR with primers listed in Table 1 and cDNA generated from breast tumor tissues as described above. The Sfi-I fragment of the product was ligated into the Sfi-I (A) and Sfi-I (B) sites of vectors, pGBDT7 and pGEX-4T-1. The resulting plasmids, pGBDT7-F-DYX1C1 and pGBDT7-DYX1C1s, were used in yeast two-hybrid assays, and pGEX-4T-1-GST-F-DYX1C1 and pGEX-4T-1-GST-DYX1C1s were used in GST pull-down assays.
Expression of recombinant Hsp70, DYX1C1, and DYX1C1s for GST pull-down assays
Plasmid pET28a-Hsp70 was transformed into the BL21 (DE3) E. coli strain following the induction of log-phase cultures in YT liquid media with 1.0 mg/ml IPTG for 6 h at 30°C. Bacteria were harvested and purification of Hsp70 for use in the GST pull-down assays was verified by SDS-PAGE. Recombinant GST-DYX1C1 and GST-DYX1C1s proteins were immobilized on MagneGST particles (Promega) from 1 ml of E. coli sample according to the manufacturer’s instructions. Proteins bound to Hsp70 (prey) by the magnetically immobilized GST-DYX1C1 or GST-DYX1C1 s fusion (bait) proteins were visualized on 12% SDS-polyacrylamide gels.
Yeast two-hybrid screening
Binding of DYX1C1 and DYX1C1s to Hsp70 or Hsp90 was investigated using a yeast two-hybrid system with a lacZ reporter. Using the engineered plasmids described above, DYX1C1 and DYX1C1s were expressed as binding or activation domain hybrids and were co-expressed with an Hsp activation or binding domain hybrid in the Y190 yeast strain. Binding interactions were visualized by assaying β-galactosidase levels. After the first round of library screening, the potential interactors were re-transfected into the Y190 strain with the relevant bait or the pGBDT7 vector as negative controls. If re-expression of the selected interactors did not reproduce the phenotypes originally observed during library screening, the interactor was deemed a false positive. β-galactosidase activity was evaluated based on color development in a filter assay. Color development after 12 h was considered a negative interaction, color development in <4 h was labeled a positive interaction, and color development between 4 and 12 h was labeled a weak interaction. We also used this system to determine the specific residues involved in the binding interaction between DYX1C1 and Hsp proteins.
Sequence analysis of DYX1C1
DYX1C1 is overexpressed in breast tumor tissues
Expression level of DYX1C1 in malignant and benign tumor tissues
No. of malignant tumor samples (%)
No. of benign tumor cases (%)
Ten of patients included in this study also were diagnosed with axillary lymph node metastasis. However, no significant correlation between DYX1C1 expression and node metastasis of breast cancer could be identified. Since we were unable to obtain sufficient samples to examine varying grades of breast carcinoma, RT-PCR could not be used to determine whether there was a correlation between DYX1C1 expression and the stage of breast cancer.
TPR domain of DYX1C1 binds Hsp70 in vitro
Interaction of DYX1C1 with Hsp70 in vivo
DYX1C1 interacts with Hsp90 via a TPR domain
Using the yeast two-hybrid system, we investigated whether DYX1C1 can interact with Hsp90α, one isoform of Hsp90. Color-based assays revealed that DYX1C1 could interact with both full-length Hsp90α and truncated Hsp90α (Fig. 6b). Since DYX1C1 contains a p23 domain, a domain characterized for binding Hsp90, our results were consistent with previous studies. However, we additionally wanted to know whether TPR domain at the C-terminus of DYX1C1 could mediate this interaction. Therefore, we assessed whether DYX1C1s interacts with Hsp90α or Hsp90α C. We did observe a weaker color change in some of the clones, indicating that the TPR domain of DYX1C1 could mediate weaker interaction with Hsp90α, which might be caused by disruption of p23.
Mutational analysis of the interaction site for Hsp in DYX1C1
Having shown that DYX1C1 is able to associate with both Hsp70 and Hsp90, we generated a series of plasmids coding for different mutants of Hsp70 and Hsp90 to map the heat shock protein (Hsp)-binding domain of DYX1C1 in a yeast two-hybrid system (Fig. 6a). It is well established that TPR co-chaperones recognize the EEVD structural motif that is present in the C-terminus of both Hsp70 and Hsp90. However, the EEVD motif is flanked by different residues in Hsp70 versus Hsp90, GPTIEEVD and MEEVD, respectively (Scheufler et al. 2000). For the yeast two-hybrid screen, these mutant fusion proteins were each co-expressed with a DYX1C1 fusion protein in the Y190 yeast strain and X-gal activity was assayed. Growth on solid SD/-Leu-Trp-His media was detected after 4 days of incubation at 37°C for all pairs. Of the seven pairs tested, two colonies, F2 and F3, produced blue color on the filter within 4 h, demonstrating a strong binding interaction had occurred. With EEVA rather than EEVD at their C-terminus, both F4 and F5 showed no association with the target protein. We did not detect an interaction between DYX1C1 and F1, which share the same last five residues of Hsp90. As a negative control, F7 colonies exhibited no color change. Taken together, our results confirm that EEVD is indispensable for binding interactions between Hsp and DYX1C1, and the octapeptide GPTIEEVD at the C-terminus can also mediate interactions between Hsp70 and DYX1C1. However, our data indicate that MEEVD does not associate with DYX1C1, suggesting there are additional mechanisms by which TPR co-chaperones interact with Hsp90 which deserve further exploration (Fig. 6b).
Heat shock proteins (Hsps) are a well-known family of highly homologous molecular chaperones that are transiently induced under a variety of chemical and physiological stimuli (Marino et al. 2005). Extensive studies have shown the ability of Hsps to maintain cell survival by mediating the renaturation of proteins and participating in anti-apoptotic signaling (Beere 2004; Gething and Sambrook 1992; Nollen and Morimoto 2002; Parsell and Lindquist 1993; Parcellier et al. 2003). Many mechanisms have been proposed to account for the role of Hsps in inhibiting cell apoptosis where, in response to stress, the coordinated activities of Hsps may regulate signaling events within apoptotic pathways to support cell survival. Hsps have been shown to participate in at least six major signaling pathways that involve c-myc, p53, Bcl-2, c-Jun N-terminal kinase (JNK), and cytochrome c (Jolly and Morimoto 2000).
Studies in animal models and in vitro models have shown that acquired resistance to apoptosis is a trait of most, if not all, types of cancer (Sreedhar and Csermely 2004). It is not surprising, therefore, that Hsp expression is associated with tumor growth. Elevated expression of Hsp70 has been detected in high-grade malignant tumors, and associated with short-term disease-free survival, metastasis, and poor patient prognosis (Kaur and Ralhan 1995; Santarosa et al. 1997). Other members of the Hsp family—including Hsp90a, Hsp90b, and Hsp60—have also been shown to be overexpressed in breast tumors, lung cancer, leukemia, and Hodgkin’s disease (Jameel et al. 1992; Wong and Wispe 1997; Yufu et al. 1992). Strong evidence has shown that higher levels of Hsp transcription in tumor cells is related to the dysfunction of p53, elevated expression of the proto-oncogene HER2 and c-myc, and is crucial to tumorigenesis (Lane et al. 1993). Hsp family members conduct coordinated activities in tumor growth by promoting both autonomous cell proliferation and by inhibiting apoptosis. Hsp90 appears to play an especially important role in both promoting tumor progression by permitting the autonomous state of unstable growth signaling, and by stimulating cell transformation events by facilitating a growing pool of mutant proteins (Calderwood et al. 2006). Unlike Hsp90 that has been shown to be a mediator of proliferation, Hsp70 has been shown to provide protection from programmed cell death (PCD) and senescence. Inactivation or knockdown of Hsp70 can induce a spontaneous PCD event that is rarely observed in normal tissues (Beere and Green 2001). Unfortunately, the mechanistic details of Hsp70 or Hsp90 signaling events in promoting tumor development have not been fully explored.
The pleiotropic role of Hsps is based on one fundamental property, their ability to interact with proteins. Hsp70 and Hsp90 share two domains: a N-terminal ATPase domain and a C-terminal domain for protein–protein interactions. The C-terminal four amino acids, EEVD, endow Hsp70 and Hsp90 with a peptide-binding capacity, which is also required for protection against heat stress (Carrigan et al. 2006). During cancer progression, it is fundamental that Hsp family members cooperate with co-factors. The major co-chaperones of Hsp include BAG-1 (Bcl-2 associated athanogene-1), CHIP (C-terminal Hsp90 interacting protein), Hsp40, hip (Hsc70 interacting protein), hop (Hsp70-Hsp90 organizing protein), and p23. BAG-1 binding to Hsp70 was the first anti-apoptotic protein characterized (Doong et al. 2002; Stuart et al. 1998; Takayama et al. 1995). Hsp40 has been shown to bind unfolded proteins and transfer them to Hsp70 (Hartl and Hayer-Hartl 2002), and contains a J domain that is also maintained by several known oncogenes. Similarly, the family of TPR-containing proteins commonly binds to Hsps. A few TPR-containing proteins have had their binding partners identified and associated with cancer. For example, Hip modulates the Hsp70 chaperones by inhibiting their ATPase activity. By binding to Hsp70 and Hsp90, CHIP inhibits their chaperone activity that is involved in ubiquitin-mediated protein degradation. Hop is an adaptor protein of Hsp70 and Hsp90 which is part of the Hsp70/Hsp90 complex (Frydman and Hohfeld 1997).
In the present study, DYX1C1 is verified to be a co-chaperone of both Hsp70 and Hsp90. The role of DYX1C1 in the developing brain is being studied by a few groups and has been shown to play a role in the migration of neocortical neurons (Wang et al. 2006; Threlkeld et al. 2007). Transfection of N- and C-terminal truncations of DYX1C1 have demonstrated that the C-terminal TPR domain determines the intracellular localization of DYX1C1 to the cytoplasm and the nucleus, although these results were not consistent with data reported by Taipale et al. (2003; Wang et al. 2006). Additional experiments showed that the C-terminus of DYX1C1 is necessary and sufficient for DYX1C1’s function in migration (Wang et al. 2006). Therefore, these data support the significance of the C-terminal TPR domain of DYX1C1 in understanding the function of this gene. The unique domain at the N-terminus of DYX1C1, p23, further specializes the function of DYX1C1 compared to other TPR-containing proteins. P23, the smallest protein in Hsp90 co-chaperone complexes, serves as an important modulator of Hsp90 (Johnson et al. 1994; Johnson and Toft 1994) and stabilizes its ATP-bound form to facilitate the formation of Hsp90 chaperone complexes (Sullivan et al. 2002). However, p23 also participates in a host of cellular process independent of Hsp90 (Freeman et al. 2000; Freeman and Yamamoto 2002). Recent evidence reveals that p23 is degraded by caspases and proteasomes during apoptosis (Mollerup and Berchtold 2005), and is up-regulated following ischemia, in human cancer, and in metastatic tissue (Krebs et al. 2002; Li et al. 2000). Sugt1 is a protein that also contains p23 and TPR domains and has a role in kinetochore assembly in yeast. It is possible that DYX1C1 might have a similar function as Sugt1 (Steensgaard et al. 2004). The combination of functions predicted by both the N-terminal and C-terminal domains of DYX1C1 emphasizes the need for its characterization.
Our current study focuses on the function of the TPR domain of DYX1C1 in mediating interactions with Hsp70 and Hsp90 to provide some additional insight into the role of this protein. Sequence alignments of DYX1C1 showed that human DYX1C1exhibits a high degree of genetic homology among different species. DYX1C1 is also predominately overexpressed in malignant cancer tissues compared to paired adjacent normal tissue and benign tumor tissues. Protein interaction assays in vivo and in vitro verify that the TPR domain of DYX1C1 interacts with the C-terminal EEVD motif contained in both Hsp70 and Hsp90. Our study indicates that DYX1C1 might not only function in neural migration, but may also have an important role in regulating cell apoptosis and cell survival in association with Hsp70 and Hsp90.
It is well established that the chemokine CXCL12 plays a role in the proliferative phenotype of benign and malignant proliferative diseases of the breast and the prostate (Cabioglu et al. 2005). Recently, CXCL12 transcription was shown to regulate DYX1C1 in human prostate epithelia cells as modulated by ELK-1, a transcriptional signal of the early growth response-1 (EGR1) gene that facilitates tumor cell growth, angiogenesis, and cell survival (Begley et al. 2007; Adamson and Mercola 2002). Data that characterizes that interactions between CXCL12, ELK-1, and DYX1C1 strengthen our conclusion that DYX1C1 plays a role in cell proliferation and tumor progression. Latest research has suggested the similar conclusion that alternative transcripts of DYX1C1 might be used as a biomarker to detect specific cancer (Kim et al. 2009).
Given the molecular heterogeneity observed for cancer cells, Hsps represent a key factor in the modulation of co-chaperone proteins to endow cancer cells with a capacity for acquired resistance to apoptosis. Thus, the study of Hsp co-chaperones represents an important area of study which may identify valuable tumor markers and possible therapeutic targets. Recently, a number of Hsp90 inhibitors have entered phase 1 and 2 clinical trials and have a promising future in clinical application (Neckers and Neckers 2002; Neckers et al. 1999). Similarly, DYX1C1, the Hsp70 and Hsp90 co-chaperone, also has the potential to be applied in clinical diagnosis and treatment with additional study.
In summary, this study identified DYX1C1 as a co-chaperone of Hsp70 and Hsp90 and its overexpression in breast malignant tumor. We detected DYX1C1 overexpression in malignant breast tumor tissue compared to benign breast tumor tissue, and also detected a significant difference in expression of DYX1C1 in cancer tissue versus adjacent normal tissue. Moreover, this study confirmed the hypothesis that DYX1C1 can interact with Hsp70 and Hsp90 via its TPR domains using GST pull-down assays and yeast two-hybrid system, respectively. We also confirmed that the C-terminal residues of EEVD in Hsp70 and Hsp90 are the residues that specifically interact with DYX1C1. Based on our data, we hypothesize that the TPR domains of DYX1C1 mediate interactions between DYX1C1 and Hsp70 and Hsp90 to promote cell proliferation. Additional details of DYX1C1’s protein–protein interactions will continue to improve our understanding of the role of DYX1C1 in neuronal migration and other important, conserved roles that DYX1C1 may have.
This work was supported by China national 973 funds (No. G1999055901 and No. 2009CB941701), Program for Changjiang Scholars and Innovative Research Team in University and construction of medical key discipline and talent fostering strategy. Besides, we thank generous technical assistance from Professor Sha Jiahao.
- Miyata Y, Chambraud B, Radanyi C, et al. (1997) Phosphorylation of the immunosuppressant FK506-binding protein FKBP52 by casein kinase II: regulation of HSP90-binding activity of FKBP52. Proceedings of the National Academy of Sciences 94, 14500-5Google Scholar
- Taipale M, Kaminen N, Nopola-Hemmi J et al (2003) A candidate gene for developmental dyslexia encodes a nuclear tetratricopeptide repeat domain protein dynamically regulated in brain. Proceedings of the National Academy of Sciences 100, 11553-8Google Scholar