Reagent and Method for Tumor Detection and Therapy

FIELD OF THE INVENTION

The present invention relates to compositions and methods for cancer diagnosis and therapy, including but not limited to, cancer markers. In particular, the present invention relates to recurrent gene fusions as diagnostic markers and clinical targets for tumors.

BACKGROUND OF THE INVENTION

Cancer, especially malignant cancer, is harmful for human health and brings large economic burdens for the patients. Within the different stages of cancer, the difficulty in curing, curing rate, and life quality are variable. Usually, it is relative easier to treat early stage cancer with less side effects, high curing rates and longer survival time with high life quality. However, at the late stages of cancer, it is much harder to treat cancer, often accompanying with worse side effects and extension of survival time at certain extent. For example, radiotherapy and chemotherapy severely impair patients' life qualities.

Due to the effects of environment, the incidence of cancer is continuously rising. Therefore, it is important to conduct screening and/or early diagnosis of cancer in order to better human health.

Most of the current early screening methods of cancer are dependent on detection of specific cancer biomarkers and checking the immunohistochemistry slides from the patients' tissues. These screening methods are costive, comprehensive at operating, and only specific for certain cancer, causing hard screening of whole body. Sometimes, the improperly biopsies leads to incorrect screening results or diagnosis, and shortness of clinical observation causes false negative screening result.

US20110065113A1 disclosed a method for identifying prostate cancer in a patient comprising: (a) providing a sample from the patient; and (b) detecting the presence or absence in the sample of a gene fusion having a 5' portion from a transcriptional regulatory region of an SLC45A3 gene and a 3' portion from a RAF family gene, wherein detecting the presence in the sample of the gene fusion identifies prostate cancer in the patient. This patent claims that detection of specific fusion gene in patients' samples is a method for tumor diagnosis, suggesting that the presence of fusion gene is an effective tumor biomarker. The disclosures of US20110065113A1 are hereby incorporated by reference in its entirety.

RARS (arginyl-tRNA synthetase) (GI:40068503), is one component of a macromolecular aminoacyl-tRNA synthetase complex during protein synthesis (C LingJ Biol Chem. 2005 Oct 14;280(41):34755-63.). RARS has been shown to be associated with a reduced AIMP1 secretion , probably leading to the generation of the cytokine, EMAP II (A Bottoni, J Cell Physiol. 2007 Aug;212(2):293-7.).

MAD1L1 (Mitotic arrest deficient 1-like protein 1) (GL62243373) is a component of the mitotic spindle-assembly checkpoint that prevents the onset of anaphase until all chromosome are properly aligned at the metaphase plate ( DY Jin, Genomics. 1999 Feb l;55(3):363-4.). It also binds to the TERT promoter and represses telomerase expression(SY Lin, Cell. 2003 Jun 27; 113(7):881-9. ) .

EIF2S3 (Eukaryotic translation initiation factor 2 subunit 3), ( GI: 83656782) functions in the early steps of protein synthesis by forming a ternary complex with GTP and initiator tRNA. This complex binds to a 40S ribosomal subunit, followed by mRNA binding to form a 43S preinitiation complex. Junction of the 60S ribosomal subunit to form the 80S initiation complex is preceded by hydrolysis of the GTP bound to eIF-2 and release of an eIF-2-GDP binary complex in order, for eIF-2, to recycle and catalyze another round of initiation. (SR Kimball, Int J Biochem Cell Biol. 1999 Jan;31(l):25-9.).

TXNDCll (Thioredoxin domain-containing protein 11) (GI: 54633316) may act as a redox regulator involved in DUOX proteins folding. The interaction with DUOX1 and DUOX2 suggest that it belongs to a multi-protein complex constituting the thyroid H(2)0(2) generating system (J Biol Chem. 2005 Jan 28;280(4):3096-103).

SUMMARY OF THE INVENTION

The aim of present invention is to provide a tumor detection method. The aim of present invention is to provide a tumor diagnosis method.

The aim of present invention is to provide a tumor therapy method.

For example, the present invention provides the transcript of RARS-MADILI fusion gene contains exons 1-7 of RARS gene and exon 19 of MAD1L1.

In some embodiments, the present invention provides a tumor detection kit comprising at least one of the following reagents:

(a) oligonucleotide probe comprising a sequence that hybridizes to the DNA of RARS-MADILI fusion gene;

(b) oligonucleotide probe comprising a sequence that hybridizes to the mRNA of RARS-MADILI fusion gene; and

(c) an antibody to the protein encoded by RARS-MADILI fusion gene;

said RARS-MADILI fusion gene comprises a 5' portion of the chimeric genomic DNA from RARS gene and a 3' portion of the chimeric genomic DNA from MAD1L1 gene.

Especially, the junction of the RARS-MADILI fusion gene at least contains the sequence as below: GGTCTTTTATAAGGCCACCAGCCCC (SEQ ID NO: 1), in which the oligonucleotide underlined is the part sequencing from RARS and the rest is from part of sequencing of exon 19 of MAD1L1 gene.

Especially, the cDNA transcribed from RARS-MADILI mRNA at least contains the sequence as below:

CTAACAGTTTCACCTCCTATTGGGGATCTTCAGGTCTTTTATAAGGCCACCAGCCCC TCGGGTTCCAAGATG CAGCTACTGGAGACAGAGTTCTCACAC (SEQ ID NO: 2).

Especially, the protein products of RARS-MADILI should contain at least:

ATSPSGSKMQLLETEFSHTVGELIEVHLRRQDSIPAFLSSLTLELFSRQTVA (SEQ ID NO: 3).

Tumors can be detected by this invention include but not limited to nasopharyngeal carcinoma, head and neck cancer, lung cancer, liver cancer, and gastric cancer.

A method for identifying tumor in a patient comprising:

(a) providing a sample from the patient, and

(b) detecting the presence or absence of RARS-MADILI fusion gene and/or product of RARS-MADILI fusion gene in the sample, where the said RARS-MADILI fusion gene comprising of 5' portion of RARS and a 3' portion of MAD1L1, the presence of said RARS-MADILI fusion gene suggests tumor in the patient.

A tumor therapy kit, including at least one of the following reagents:

(a) reagents that can repress the transcription of theRARS-MADlLl fusion gene comprising 5' portion of RARS and a 3' portion of MAD1L1;

(b) oligonucleotides or modified oligonucleotides to silence the expression of the RARS-MADILI fusion gene comprising 5' portion of RARS and a 3' portion of MAD1L1; and

(c) Chemical that can inactivate the protein of theRARS-MADlLl fusion gene comprising 5' portion of RARS and a 3' portion of MAD ILL

Especially, the reagents repressing the transcripts of 5' portion of RARS and a 3' portion of MAD1L1 are promoter inhibitor, transcription inhibitor, and siRNA.

The transcript of EIF2S3-TXNDC11 fusion gene contains exons 1-2 of EIF2S3 gene and exon 2-12 of TXNDC11. A tumor detection reagent contains at least one of the probes:

1) specifically recognizing the genomic DNAhaving 5' DNA portion of EIF2S3 and 3' DNA portion of TXNDC11

2) specifically detecting a chimeric mRNA transcripts having 5' portion transcribed fromEIF2S3 and 3' portion transcribed from TXNDC11

3) specifically recognizing a chemic protein or peptide having a 5' portion encoded byEIF2S3 and 3' portion encoded by TXNDC11.

Especially, the genomic DNA of EIF2S3-TXNDC11 at least contain part of cadherin-11 and cadherin-5 gene.

Especially, the fusion site of EIF2S3-TXNDC11 gene at least have the following sequencing:

GGATGTTACCAAGTTGACGCCACTTTCACACGAAGTTATCAGCAGACAAGCCACAATTAA CATAGTCGAGC AAAAGATGTGATAATACCAGCAAAGCCACCTGT (SEQ ID NO: 4), wherein the underlined was the portion of EIF2S3 and the rest is part of exon 19 of TXNDC 11.

The mRNA of EIF2S3-TXNDC 11 fusion gene should contain at least: CTAACAGTTTCACCTCCTATTGGGGATCTTCAGGTCTTTTATAAGGCCACCAGCCCCTCG GGTTCCAAGATG CAGCTACTGGAGACAGAGTTCTCACAC (SEQ ID NO: 5).

Tumors detected are nasopharyngeal carcinoma, head and neck cancer, lung cancer, liver cancer, and gastric cancer. A method for identifying tumor in a patient comprising:

(a) providing a sample from the patient, and

(b) detecting the presence or absence of EIF2S3-TXNDC11 fusion gene and/or product of EIF2S3-TXNDC11 fusion gene in the sample, where the said EIF2S3-TXNDC11 fusion gene comprising of 5' portion of EIF2S3 and a 3' portion of TXNDC11, the presence of said EIF2S3-TXNDC11 fusion gene suggests tumor in the patient.

A tumor therapy kit, including at least one of the following reagents:

(a) reagents that can repress the transcription of the EIF2S3 -TXNDC 11 fusion gene comprising 5' portion of EIF2S3 and a 3' portion of TXNDC11;

(b) oligonucleotides or modified oligonucleotides to silence the expression of the EIF2S3-TXNDC11 fusion gene comprising 5' portion of EIF2S3 and a 3' portion of TXNDC11; and

(c) Chemical that can inactivate the protein of the EIF2S3-TXNDC11 fusion gene comprising 5' portion of EIF2S3 and a 3' portion of TXNDC11.

Especially, the reagents repressing the transcripts of 5' portion of EIF2S3 and a 3' portion of TXNDC 11 are promoter inhibitor, transcription inhibitor, and siRNA.

The benefits of present invention:

The gene or protein product of present invention could be used as molecular diagnostic marker and therapeutic targets of many types of tumors, such as nasopharyngeal carcinoma. Present invention will contribute to the early diagnosis and therapy of tumors, increasing the screening efficiency, decreasing the screening cost, enhancing the early diagnostic rate, and improving the patients' survival.

DESCRIPTION OF THE FIGURES

Fig. 1 shows the discovery of RARS-MADILI gene fusion in many kinds of cancer, including (A) Histogram of gene fusion nomination scores in clinically localized nasopharyngeal carcinoma, Head & Neck Squamous Cell Carcinoma, and other types of cancer; (B) Schematic representation of paired-end reads supporting the inter-chromosomal gene fusion between RARS and MAD1L1, resulting in fusion gene RARS-MADILI.

Fig. 2 shows the exon structure of RARS (A) and MAD1L1 (B) normal and fusion transcript RARS-MADILI (C). Fig. 3 a circos plot of the genomic landscape of RARS-MADILI gene fusion discovered by RNA-seq in NPC samples. The outer ring shows RARS and MAD1L1 chromosome ideograms. The RARS-MADILI gene fusion is shown as an arc linking the two genomic loci.

Fig. 4 shows validation of RNA expression of RARS-MADILI gene fusions. (A) qRT-PCR validation of RARS-MADILI gene fusion in C666-1 cells by electrophoresis, and (B) result by sequencing.

Fig. 5 shows validation of protein expression of RARS-MADILI gene fusions. (A) Western blot analysis showing the expression of 34 kDa RARS-MADILI fusion protein in C666-1 and in Hela cells transfected with RARS-MADILI full length fusion construct by N-terminal RARS antibody. (B) Western blot analysis showing the over-expression of 72kDa RARS protein, 90kDa MAD1L1 protein, 34 kDa RARS-MADILI fusion protein in HNE1 cells transfected with empty vector, RARS gene, MAD1L1 gene, and RARS-MADILI full length fusion construct by anti-flag antibody (upper panel) and RARS-MADILI antibody ascites (lower panel). (C) Detection of RARS-MADILI endogenous protein by immune-precipitation of C-terminal MAD1L1 antibody and probed with N-terminal RARS antibody. Rabbit IgG serves as a negative control. 293FT cells transfected with RARS-MADILI was used as a positive control.

Fig. 6 shows RARS-MADILI fusion transcripts by cancer.

Fig. 7 shows genomic organization and FISH validation of RARS and MAD1L1 gene rearrangement. (A) Schematic diagrams in the top panel showing the genomic location of RARS and MAD1L1 genes, respectively. (B-D) FISH validation of RARS-MADILI gene fusion in C666-1 (B), PALM (C), NPC patients (D), respectively.

Fig. 8 shows the representative gel picture of rapid amplification of 5' cDNA ends (RLM-5'RACE) in C666 cells.

Fig. 9 shows rapid amplification of 3' cDNA ends (RLM-3'RACE) in C666 cells. (A) The representative gel picture of

RLM-3' RACE results for C666 cells. (B) The DNA sequencing of RLM-3' RACE results for C666 cells.

Fig. 10 shows fusion fragment between RARS and MAD1L1 of genomic DNA in C666 and PALM cells by long-range

PCR. (A) The schematic depicting the positions of gene specific primers (blue arrows) on RARS part and specific primers (red arrows) on MAD1L1 part of the fusion used for RARS-MADILI. (B) The representative gel picture of

RARS-MADILI fusion fragment results for C666 cells. (C) The schematic depicting the genomic rearrangement of

RARS-MADILI fusion.

Fig. 11 shows the oncogenic potential of the RARS-MADILI fusion. (A) Expression of the RARS-MADILI fusion in nasopharyngeal carcinoma cells (CNE1 and HNE1) leads to increased cellular proliferation. (B) HNE1 stable cells expressing the RARS-MADILI fusion showed increased cell invasion potential by transwell assay. (C) The CNE1 stable cells expressing the RARS-MADILI fusion showed increased cell progression potential under X-ray irradiation. (D) CNE2 stable cells expressing the RARS-MADILI fusion showed increased cell progression potential by colony formation assay.

Fig. 12 shows transformation of BaF3 cells by RARS-MADILI fusion transcript. Fusion constructs RARS-MADILI promoted proliferation in BaF3 cells deprived of IL3.

Fig. 13 shows induction of chromosome instability of Hela cells by RARS-MADILI fusion transcript. (A) Ratio of micronuclei induced by fusion constructs RARS-MADILI and vector control in Hela cells. (B) Multi-spindle poles by fusion construct RARS- in Hela cells.

Fig. 14 shows qPCR confirmation of siRNA knockdown of RARS-MADILI fusion. (A) the schematic depicting the positions of gene specific siRNA on wild-type RARS, wild-type MAD1L1, and RARS-MADILI fusion (B) qPCR confirmation of RARS-MADILI knockdown by siRNA against the fusion junction, wild-type RARS, and wild-type MAD1L1 on C666.

Fig. 15 shows knocking-down of RARS-MADILI fusion reduced the tumor progression and invasive abilities of C666 cells by (A) shRNA and (B) siRNA with soft-agar assay.

Fig. 16 RARS-MADILI induces side population in NPC cells. (A) The expression of RARS-MADILI fusion transcript in C666 SP cells. (B) RARS-MADILI induces side population in CNE2 cells. (C) SP cells from

CNE2-RARS-MAD1L1 stable cell line have increased cell progression potential by colony formation assay. (D) SP cells from CNE2-RARS-MAD1L1 stable cell line have induced invasion capacity by soft-agar assay.

Fig. 17 shows the exon structure of EIF2S3 (A) and TXNDCll (B) normal and fusion transcript EIF2S3- TXNDCll

(C).

Fig. 18 shows validation of expression of EIF2S3- TXNDCll gene fusions. (A) qRT-PCR validation of EIF2S3- TXNDC 11 gene fusion in SUNE2 cells and (B) result by sequencing.

Fig. 19 shows EIF2S3-TXNDC11 fusion transcripts by cancer.

Fig. 20 shows genomic organization and FISH validation of EIF2S3 and TXNDCll gene rearrangement. (A) Schematic diagrams in the top panel showing the genomic location of EIF2S3 and TXNDCll genes, respectively. FISH validation of EIF2S3- TXNDCll gene fusion in (B) CNE2, (C) SUNE2 and HONE1 cells, respectively.

Fig. 21 shows rapid amplification of 5' EIF2S3-TXNDC11 cDNA ends (RLM-5'RACE) in CNE2 cells. (A) The representative gel picture of RLM-5' RACE results for CNE2 cells. (B) The two different versions of DNA sequencing results of RLM-5' RACE in CNE2 cells. (C, D) two versions of 5' EIF2S3-TXNDC 11 cDNA ends after aligning by BLAST analysis.

Fig. 22 shows fusion fragment between EIF2S3 and TXNDCll of genomic DNA in SUNE2 cells by long-range PCR. (A) The schematic depicting the positions of gene specific primers (dark line) on EIF2S3 part and specific primers (dot line) on TXNDCll part of the fusion used for EIF2S3- TXNDCll. (B) The representative gel picture of EIF2S3- TXNDC11 fusion fragment results for SUNE2 cells. (C) The schematic depicting the genomic rearrangement of EIF2S3- TXNDC11 fusion.

DETAILED DESCRIPTION OF THE INVENTION

The inventor of present invention conducted transcriptome sequencing of C666 cells and found the fusion gene of RARS-MADILI. According to the results of massively parallel sequencing, the inventor designed the specific primers forRARS-MADlLl, validating it in tumor tissues, further confirming the transcriptional structure of this fusion gene. The sequencing of the primer are:

RARS-MADILI-F: 5'-CTAACAGTTTCACCTCCTATTGGG-3' (SEQ ID NO: 6);

RARS-MADILI-R: 5'-GTGTGAGAACTCTGTCTCCAGTAGC-3' (SEQ ID NO: 7);

RARS-M AD 1 L 1 -Probe: 5'-fam+TCTTTTATAAGGCCACCAGCCCCTC+tamra-3' (SEQ ID NO: 8);

Amplifying conditions: 50 ^° C 2min; 95 ^° C 2min; 95 ^° C 15sec, 70 ^° C 30sec, 3 cycles; 95 ^° C 15sec, 68 ^° C 30sec, 3 cycles;

95 ^° C 15sec, 66 ^° C 30sec, 3 cycles; 95 ^° C 15sec, 64 ^° C 30sec, total 45cycles; Then statistically analyzing the amplifying results. The result was showed as figure 6.

Figure 1 presents the discovery of RARS-MA1L1 gene fusion in many kinds of cancer, including: (A) Histogram of gene fusion nomination scores in clinically localized nasopharyngeal carcinoma, Head & Neck Squamous Cell Carcinoma, and other types of cancer. (B) Schematic representation of paired-end reads supporting the inter-chromosomal gene fusion between RARS and MAD1L1, resulting in fusion gene RARS-MADILI.

Fig. 2 shows the exon structure of RARS (A) and MAD1L1 (B) normal and fusion transcript RARS-MADILI(C). Fig. 3 A Circos plot of the genomic landscape of RARS-MADILI gene fusion discovered by RNA-seq in NPC samples. The outer ring shows RARS and MAD1L1 chromosome ideograms. The RARS-MADILI gene fusion is shown as an arc linking the two genomic loci.

To further examine the expression of RARS-MADILI in tumors, RT-PCR combined with Sanger sequencing was applied to detect the fusion gene of RARS-MADILI at mRNA level in C666 (Fig. 4 A). Fig. 4 shows validation of expression of RARS-MADILI gene fusions. ( A) qRT-PCR validation of RARS-MADILI gene fusion in C666-1 cells by electrophoresis and (B) The RARS-MADILI DNA sequencing amplified by above qRT-PCR. The PCR product was further subjected to sequence and the sequencing is:

CTCATCGCTCACCTGCAAGACAAATTTCCAGATTATCTAACAGTTTCACCTCCTATTGGG GATCTTCA GGTCTTTTATAAGGCCACCAGCCCCTCGGGTTCCAAGATGCAGCTACTGGAGACAGAGTT CTCACACACC GTGGGCGAGCTCATCGAGGTGCACCTGCGGCGCCAGGACAGCATCCCTGCCTTCCTCAGC TCGCTCACCCT CG ( The bold characters are from part of RARS gene, the rest is from MAD1L1 gene) (SEQ ID NO: 9).

Fig. 5 shows detection of protein expression of RARS-MADILI gene fusions. The endogenous RARS-MADILI fusion protein in C666 cells was detected with the antibody against N-terminal RARS by Western blotting ( Fig. 5 A) .Using anti-flag antibody, the flag-tagged exogenous proteins (wild-type RARS (72kDa), wild-type MAD1L1 (95kDa), and fusion protein RARS-MADILI (34kDa) were detected by Western blotting (Fig 5B). After pulling-down with the antibody against N-terminal RARS and probing with the antibody against C terminal-MADlLl, the endogenous RARS-MADILI fusion protein in C666 cells was also detected by immune-precipitation (Fig 5C).

Based on these findings, the inventor collected around 500 cases of tumor samples including nasopharyngeal carcinoma, breast cancer, lung cancer, and other epithelial tumor samples and 27cases of normal nasopharyngeal epithelium to detect the fusion gene of RARS-MADILI in tumors by real-time PCR. Fig. 6 shows RARS-MADILI fusion transcripts by cancer. According to Figure 6, RARS-MADILI fusion transcript exists in many types of tumors. The positive detection rate was different among different types of tumors. The positive rate is higher in nasopharyngeal carcinoma, head & neck cancer, and gastric cancer, implying the potential application in tumor diagnosis and targeted therapy. The genomic rearrangement of RARS gene and MAD1L1 gene was detected using Fluorescence in situ hybridization (FISH). According to the position of RARS gene and MAD1L1 gene in chromosomes, the bacterial artificial chromosome (BAC) clones containing either of these two genes were found by searching in Ensemble website. As shown in Fig. 7A, the BAC clone probe containing the single strand of RARS gene is labeled with red fluorescence and the BAC clone probe containing the single strand of MAD1L1 gene is labeled with green fluorescence. Based on the complementary base pairing rules, these probes specifically bound with the single strand nucleotides of genomic DNA in C666 cells (Fig. 7B), PALM cells (Fig. 7C), and nasopharyngeal carcinoma tissue samples from different patients (Fig. 7D), and form the double strand between BAC clone probe and single strand of genomic DNA. In human lymphocytes, the red and the green probes were not merged because RARS gene and MAD1L1 gene are located on different chromosomes. But in tumor samples, if the fusion gene of RARS-MADILI exists, the probe against RARS gene or MAD1L1 gene was located very closely, leading to partially or completely merging as showed in red (Fig. 7 B-D).

5' end of cDNA of RARS-MADILI in C666 was amplified by 5' rapid-amplification of cDNA ends (5'RACE) technology. According to the known sequencing of 3' part of fusion gene-MADlLl, a specific primer called GSP1 primer was designed for reverse transcription: 5'-TTGTGTCTAGGGGAGAAGATT-3' ( SEQ ID NO: 5 ) . After the first strand cDNA synthesis is initiated by reverse transcription, the original mRNA template is destroyed with RNase H, which is specific for RNA:DNA heteroduplex molecules. TdT (Terminal deoxynucleotidyl transferase) and dCTP are used to add homopolymeric tails to the 3' ends of the cDNA. The Abridged Anchor Primer (AAP: 5'- GGCCACGCGTCGACTAGTACGGGGGGGGGG-3' ( SEQ ID NO: 6), provided by Invitrogen5' RACE kit) is able to bind the homopolymeric tails. The specific primers for 5' RARS gene of fusion gene RARS-MADILI are designed: Gsp2: 5 ' - AGGTC AGGCC AAGC AGAGTG-3 ' (SEQ ID NO: 10);

GSP3: 5'-TAGTTGACCTCAGGTGGCCTACA-3' (SEQ ID NO: 11)

and Abridged Universal Amplification Primer (AUAP): 5' -GGCC ACGCGTCGACTAGTAC-3 ' (SEQ ID NO: 12) (provided by Invitrogen5' RACE kit). With AAP, GSP2 and DNA polymerase, the cDNA first strand was used as a template for regular PCR to obtain double strand DNA. In order to increase PCR amplification specificity, PCR products from the primary PCR with the Abridged Anchor Primer were preceded to nest PCR with the AUAP and GSP3 primers. Then after gel electrophoresis and gel extraction of the PCR products, the purified PCR products were inserted into TA clones. After sequencing, the original 5' mRNA of fusion gene RARS-MADILI was obtained. The gel electrophoresis picture was shown in Figure 8. Lane 1 presents the PCR product with GSP1 and AUAP primers in C666 cells, and Lane 2 presents the products with GSP2 and AUAP primers. The arrow pointed to the band of PCR product in lane 2 that is subjected to sequencing. The detailed sequence is:

TTGGGTAACGCCAGGGTTTTCCCAGTCACGACGTTGTAAAACGACGGCCAGTGAATTGTA ATACGACTCAC TATAGGGCGAATTGGGCCCTCTAGATGCATGCTCGAGCGGCCGCCAGTGTGATGGATATC TGCAGAATTCGG CTTGGCCACGCGTCGACTAGTACGGGGGGGGGGCA C7 GGCGAGrGAGACGCrGArGGGAGGArGGACArA CTGGTGTCTGA GTGCTCCGCGCGGCTGCTGCA GCA GGAA GAA GA GATTAAATCTCTGA CTGCTGAAATTGA C CGGTTGAAAAA CTGTGGCTGTTTA GGA GCTTCTCCAAATTTGGA GCA GTTA CAA GAA GAAAATATAAAATTAA A GTATCGA CTGAATATTCTTCGAAA GA GTCTTCA GGCA GAAA GGAA CAAA CCAA CTAAAAATATGATTAA CAT TATTA GCCGCCTA CAA GA GGTCTTTGGTCATGCAATTAA GGCTGCATATCCA GATTTGGAAAATCCTCCTCTG CTA GTGA CA CCAA GTCA GCA GGCCAA GTTTGGGGA CTATCA GTGTAATA GTGCTATGGGTATTTCTCA GATGC TCAAAA CCAA GGAA CA GAAA GTTAA TCCAA GA GAAATTGCTGAAAA CA TTA CCAAA CA CCTCCCA GA CAA TG AATGTATTGAAAAAGTTGAAACTGCTGGTCCTGGTTTTATTAATGTCCACTTAAGAAAGG ATTTTGTATCAGAA CAA TTGA CCA GTCTTCTA GTGAATGGA GTTCAA CTA CCTGCTCTGGGA GA GAATAAAAA GGTTA TA GTTGA CT TTTCCTCCCCTAATATA GCTAAA GA GA TGCA TGTA GGCC A CCTGA GGTCAA CTA AGCCGAATTCC AGC AC ACT GGCGGCCGTTACTAGTGGATCCGAG (SEQ ID NO: 13).

Analyzing with BLAST, in the sequence, the italic characters are aligned with part of human RARS mRNA (NM_002887.3), the underlined characters are aligned with the 5'UTR, and the bold characters are aligned with CDs. First strand cDNA synthesis is initiated at the poly(A) tail of mRNA using the Abridged Anchor Primer(AAP). After first strand cDNA synthesis, the original mRNA template is destroyed with RNase H, which is specific for RNA:DNA heteroduplex molecules. Because there is no satisfied PCR primer in the region of partial RARS gene of fusion gene RARS-MADILI, the inventor designed a GSP1 primer cross the fusion site of RARS gene and MADlLlgene ( GSP1: 5'-GTCTTTTATAAGgccaccagcc-3', the 11 Upper cases is from RARS gene, and the 10 lower cases is from MAD1L1 gene); With AAP, GSP1 and DNA polymerase, the cDNA first strand was used as a template for regular PCR to obtain double strand DNA with a poly(A) tail. As shown in Fig 9, after gel extraction of the PCR products in gel electrophoresis, the purified PCR products are inserted into TA clones, and subjected to sequence. After sequencing, the original 5' mRNA of fusion gene RARS-MADILI was obtained. Fig. 9A shows the representative gel picture of 3' RACE results for C666 cells. CNE2 was used as a negative control. There are two versions of DNA sequencing of 3' RACE results for C666 cells:

version 1:

AAGGAAGGGAAGAAAGCGAAAGGAGCGGGCGCTAGGGCGCTGGCAAGTGTAGCGGTCACG CTGCGCGTA

ACCACCACACCCGCCGCGCTTAATGCGCCGCTACAGGGCGCGTCCATTCGCCATTCA GGCTGCGCAACTGT

TGGGAAGGGCGATCGGTGCGGGCCTCTTCGCTATTACGCCAGCTGGCGAAAGGGGGA TGTGCTGCAAGGC

GATTAAGTTGGGTAACGCCAGGGTTTTCCCAGTCACGACGTTGTAAAACGACGGCCA GTGAATTGTAATAC

GACTCACTATAGGGCGAATTGGGCCCTCTAGATGCATGCTCGAGCGGCCGCCAGTGT GATGGATATCTGCAG

AATTCGGCTTGTCTTTTATAAGGCCACCAGCCCCTCGGGTTCCAAGATGCAGCTACT GGAGACAGAGTTCT

CACACACCGTGGGCGAGCTCATCGAGGTGCACCTGCGGCGCCAGGACAGCATCCCTG CCTTCCTCAGCTC

GCTCACCCTCGAGCTCTTCAGCCGCCAGACCGTGGCGTAGCCTGCAGGCTCGGGGGC ATAGCCGGAGC

CACTCTGCTTGGCCTGACCTGCAGGTCCCCTGCCCCGCCAGCCACAGGCTGGGTGCA CGTCCTGCC

TCTCCAGCCCCACAGGGCAGCAGCATGACTGACAGACACGCTGGGACCTACGTCGGG CTTCCTGCT

GGGGCGGCCAGCACCCTCTCCACGTGCAGACCCCATGCGTCCCGGAGCCTGGTGTGT GGGCGTCG

GCCACCAGCCTGGGTTCCTCACCTTGTGAAATAAAATCTTCTCCCCTAGACAAAAAA AAAACCTATAG

TGAAATCACTAGTGGAACGACGGTAAGCCGAATTCCAGCACACTGGCGGCCGTTACT AGTGGATCCGAGCT

CG(SEQ ID NO: 14)

version2:

CTAGTAACGGCCGCCAGTGTGCTGGAATTCGGCTTGTCTTTTAT L1 GGCCACCAGCCCCrCGGG7 CCA4GAr GTAGCTACTGGTGACAGAGTTCTCACACACCGTGGGCGAGCTCATCGAGGTGCACCTGCG GCGCCAGGACAGC ATCCCTGCCTTCCTCAGCTCGCTCACCCTCGAGCTCTTCAGCCGCCAGACCGTGGCGTAG CCTGCA GGCTCGG GGGCA TA GCCGGA GCCA CTCTGCTTGGCCTGA CCTGCA GGTCCCCTGCCCCGCCA GCCA CA GGCTGGGTGC A CGTCCTGCCTCTCCA GCCCCA CA GGGCA GCAGCA TGA CTGCA GA CA CGCTGGGA CCTATGTCGGGCTTCCT GCTGGGGCGGCCA GCA CCCTCTCCA CGTGCA GA CCCCA TGCGTCCCGGA GCCTGGTGTGTGGGCGTCGGCC A CCA GCCTGGGTTCCTCA CCTTGTGAAATAAAA TCTTCTCCCCTA GAAAAAAAAAAAA CCTATAGTGAAATC A CTAGTGGAACGACGGTAAAGCCGAATTCTGCAGATATCCATCACACTGGCGGCCGCTCGA GCATGCATCTA GAGGGCCCAATTCGCCCTATAGTGAGTCGTATTACAATTCACTGGCCGTCGTTTTACAAC GTCGTGACTGGG AAAACCCTGGCGTTACCCAACTTAATCGCCTTGCAGCACATCCCCCTTTCGCCAGCTGGC GTAATAGCGAA GAGGCCCGCACCGATCGCCCTTCCCAACAGTTGCGCAGCCTGAATGGCGAATGGACGCGC CCTGTAGCGG CGCATTAAGCGCGGCGGGTGTGGTGGTTACGCGCAGCGTGACCGCTACACTTGCCAGCGC CCTAGCGCCCG CTCCTTTCGCTTTCTTCCCTTCCTTTCTCGCCACGTTCGCCGGCTTTCCCCGTCAAGCTC TAAATCG (SEQ ID NO: 15).

Analyzing with BLAST, in the sequence, the italic characters are aligned with part of human MAD1L1 mRNA (NM_003550.2, NM 001013836.1, NM_001013837.1), the underlined characters are aligned with the CDs, and the bold characters are aligned with 3'UTR.

Using the genomic DNA of C666 cells as template, RARS-MADILI fusion gene was amplified by long range PCR at genomic DNA level. As shown in the scheme of Figure 10A, forward primers are designed to target RARS gene as Rl, R2 and R3 primers, the reverse primers are designed to target MAD1L1 gene as M1-M19. The PCR amplified products by primer sets (R1M12, R2M12, R3M12) were subjected to sequence shown in Fig 10 B and IOC. Between the RARS gene (position 167926434) at Chromosome 5 and the MAD1L1 gene (position 1880783) at Chromosome 7, there is an insertion of DNA fragment of HDAC9 intron (position 18268256-18276625 at Chromosome 7).

Primer primer sequencing primer position SEQ ID No. Rl CACCCGGCCTCAACATATATTT chr5: 167923780- 167923801 16

R2 ATCGCTCACCTGCAAGACAAAT chr5: 167924276- 167924297 17

R3 TGAGGAGGGATAATCGCTTGAG chr5:167925136-167925157 18

M12 CATGGGGAAAGTGGGTAAGCTA chr7:1935201-1935222 19

After that, the inventor applied colony formation assay, Transwell experiment, MTT assay to study the function and possible mechanisms of the fusion gene RARS-MADILI, the empty vector was used as a negative control. As shown in Fig. ll,the expression of the RARS-MADILI fusion in nasopharyngeal carcinoma cells (CNE1 and HNEl) leads to increased cellular proliferation (Fig. 11A); HNEl stable cells expressing the RARS-MADILI fusion showed increased cell invasion potential by Transwell assay (Fig. 11B); The CNE1 stable cells expressing the RARS-MADILI fusion showed increased cell invasion potential under X-ray irradiation (Fig. 11C); CNE2 stable cells expressing the RARS-MADILI fusion showed increased cell invasion potential by colony formation assay (Fig. 11D). Based on these results, RARS-MADILI fusion gene induces the colony formation ability, invasion ability, and proliferation of tumor cells, leading to the oncogenesis.

To further detect the transformation ability of RARS-MADILI fusion gene, the inventor infected IL-dependent BaF3 cells with retrovirus encoded with RARS-MADILI fusion gene. As shown in Fig. 12, upon removal of IL3, the overexpression of fusion constructs RARS-MADILI still induces the proliferation of BaF3 cells. Then using the immunostaining with anti-a-tubulin (green) and nuclei (DAPI, blue), the number of micronuclei and multipolar spindle were calculated in the Hela cells with overexpression of RARS-MADILI. The fusion gene RARS-MADlLlincrease the number of micronuclei (Fig 13A and 13B) and induces the presence of multipolar spindles was detected in Hela cells with overexpression of RARS-MADILI (Fig 13C), suggesting that RARS-MADILI induces the genome instability, contributing to the tumor genesis.

The different siRNAs were designed to knock-down the mRNA of RARS gene, MAD1L1 gene, or RARS-MADILI gene and to further explore the function of RARS-MADILI in tumor cells. The scheme of each target sequence by different siRNA was shown in Fig 14A. Real-time analysis showed that siRNA-RARS-05 (siRARS-5) and siRNA-M AD 1 L 1 -03 (siMADlLl-3) knocked-down both RARS-MADILI and wild-type gene (either RARS or MAD1L1), while the rest siRNA only can knock down the related wild-type gene. The colony formation abilities and invasion abilities of C666 cells were reduced after knocking-down by siRARS-5 and siMADlLl-3 (Fig. 15, left panel) and by shRNA (Fig. 15, right panel).

At the end, side population cells (SP cells) are isolated from of C666 by cell sorting and subjected to analyze the endogenous RARS-MADILI expression. As shown in Fig. 16A, the expression level of RARS-MADILI was higher in SP cells than non-SP cells in C666 cell line. Furthermore, the number of SP cells was induced in CNE2 cells after overexpression of RARS-MADILI (Fig. 16B). The SP cells from RARS-MADlLl-expressed CNE2 formed more and larger colonies by colony formation assay (Fig. 16C) and soft-agar assay (Fig. 16D), comparing to vector control, suggesting that SP cells isolated from RARS-MADlLl-expressed CNE2 have more aggressive invasion and proliferation abilities.

Fusion gene EIF2S3-TXNDC 11

The inventor conducted transcriptome sequencing of SUNE2 cells and found the fusion gene of EIF2S3-TXNDC11. The structure of EIF2S3-TXNDC11 is shown in Fig. 17. According to the results of massively parallel sequencing, the inventor designed the specific primers for EIF2S3-TXNDC11, validating it in SUNE2 and CNE2 cells (nasopharyngeal carcinoma cell lines) by PCR and gel electrophoresis (Fig. 18A), and subjected with sequencing (Fig. 18B), further confirming the transcriptional structure of this fusion gene. The amplified sequence of EIF2S3-TXNDC11 is:

CACGAAGTTATCAGCAGACAAGCCACAATTAACATAGTCGAGCAAAAGATGTGATAA TACCAGCAAAGC

ACGTTCGACGGGATTCAGAGGTGGTACTGCTCTTCTTCTATGCCCCT (SEQ ID NO: 20) (the bold characters are from part of EIF2S3, the rest are from part of TXNDC 11 ) .

The PCR primers are:

EIF2S3-TXNDC 11 -F: 5 ' -GGATGTTACC AAGTTGACGCC-3 ' (SEQ ID NO: 21);

EIF2S3-TXNDC 11 -R: 5'-ACAGGTGGCTTTGCTGGTATTA-3' (SEQ ID NO: 22);

EIF2S3-TXNDCll-Probe: 5'- fam+CAGACAAGCCACAATTAACATAGTCGAGCA+tamra -3' (SEQ ID NO: 23) ; Additional 100 NPC samples were analyzed with the expression of EIF2S3-TXNDC11 by real-time PCR. Amplifying conditions: 50 ^° C 2min; 95 ^° C 2min; 95 ^° C 15sec, 70 ^° C 30sec, 3 cycles; 95 ^° C 15sec, 68 ^° C 30sec, 3 cycles; 95 ^° C 15sec, 66 ^° C 30sec, 3 cycles; 95 ^° C 15sec, 64 ^° C 30sec, total 45 cycles; Then statistically analyzing the amplifying results. Only two cases were examined the presence of EIF2S3-TXNDC11. The result was showed as figure 19.

The genomic rearrangement of EIF2S3 gene and TXNDC 11 gene was detected using Fluorescence in situ hybridization (FISH). According to the position of EIF2S3 gene and TXNDC 11 gene in chromosomes, the bacterial artificial chromosome (BAC) clones containing any of these two genes were found by searching in Ensemble website. As shown in Fig. 20A, the BAC clone probe containing the single strand of EIF2S3 gene is labeled with red fluorescence and the BAC clone probe containing the single strand of TXNDC 11 gene is labeled with green fluorescence. Based on the complementary base pairing rules, these probes specifically bound with the single strand nucleotides of genomic DNA in CNE2 cells (Fig. 20B), SUNE2 cells, and HONE1 cells (Fig. 20C), and form the double strand between BAC clone probe and single strand of genomic DNA. In human lymphocytes, the red and the green probes were not merged because EIF2S3 gene and TXNDC 11 gene are located on different chromosomes. But in tumor samples, if the fusion gene of EIF2S3-TXNDC11 exists, the probe against EIF2S3 gene or TXNDC 11 gene was located very closely, leading to partially or completely merger as showed in red (Fig. 20B-C).

5' end of cDNA of EIF2S3-TXNDC 11 in C666 was amplified by 5' rapid-amplification of cDNA ends (5'RACE) technology. According to the known sequencing of 3' part of fusion gene- TXNDC 11 (human TXNDC 11 mRNA sequence NW_004078085.1), a specific primer called GSP1 primer was designed for reverse transcription: 5'-GCTTTGCTGGTATTATCACATCTTT-3' (SEQ ID NO: 24). After the first strand cDNA synthesis is initiated by reverse transcription, the original mRNA template is destroyed with RNase H, which is specific for RNA:DNA heteroduplex molecules. TdT (Terminal deoxynucleotidyl transferase) and dCTP are used to add homopolymeric tails to the 3' ends of the cDNA. The Abridged Anchor Primer (AAP: 5'- GGCCACGCGTCGACTAGTACGGGGGGGGGG-3' (SEQ ID NO: 25), provided by Invitrogen5' RACE kit) is able to bind the homopolymeric tails. The specific primers for 5' EIF2S3 gene of fusion gene EIF2S3-TXNDC11 are designed: EIF2S3-TXNDC 11 -GSP2: 5'-GTGGCTTGTCTGCTGATAACT-3' (SEQ ID NO: 26) ( from human EIF2S3 mRNA sequencing NM_001415.3 ) ;

EIF2S3-TXNDC 11 -GSP3: 5'-AGTGGCGTCAACTTGGTAA-3' (SEQ ID NO: 27) ( from human EIF2S3 mRNA sequencing NM_001415.3 ) ;

And Abridged Universal Amplification Primer (AUAP): 5' -GGCCACGCGTCGACTAGTAC-3' (provided by Invitrogen 5' RACE kit). With AAP, GSP2 and DNA polymerase, the cDNA first strand was used as a template for regular PCR to obtain double strand DNA. In order to increase PCR amplification specificity, PCR products from the primary PCR with the Abridged Anchor Primer were preceded to nest PCR with the AUAP and GSP3 primers. Then after gel electrophoresis (Fig. 21 A) and gel extraction of the PCR products, the purified 150 bp PCR products were inserted into TA clones and then the TA colonies were digested to confirm the correct insertion (Fig. 21B). After sequencing, the original 5' UTR sequence of mRNA of fusion gene EIF2S3-TXNDC11 was obtained. There are two version of 5'UTR in CNE2. In the sequence of 5'UTR in CNE2, the italic characters are mRNA sequence, the underlined characters are 5'UTR sequences, the bold characters are CDs sequence, the vector is PCR2.1, inserted sequence of version 1 has additional 11 bp at 5'UTR (Fig. 21C) comparing to NCBI Reference Sequence (NM_001415.3), while inserted sequence of version 2 is short of 8 bp at 5'UTR (Fig. 21D) comparing to NCBI Reference Sequence (NM_001415.3)). The detailed sequencing are as below:

version 1: TCAGCTCATTTTTAACCATTAGGCCGAAATCGGCAAAATCCCTTATAAATCAAAAGAATA GACCGAGATAGG GTTGAGTGTTTTCCAGTTTGGAACAAGAGTCCACTATTAAAGAACGTGGACTCCAACGTC AAAGGGCGAA

CGTAAAGCACTAAATCGGAACCCTAAAGGGAGCCCCCGATTTAGAGCTTGACGGGGA AAGCCGGCGAACG

TGGCGAGAAAGGAAGGGAAGAAAGCGAAAGGAGCGGGCGCTAGGGCGCTGGCAAGTG TAGCGGTCACG

CTGCGCGTAACCACCACACCCGCCGCGCTTAATGCGCCGCTACAGGGCGCGTCCATT CGCCATTCAGGCTG

CGCAACTGTTGGGAAGGGCGATCGGTGCGGGCCTCTTCGCTATTACGCCAGCTGGCG AAAGGGGGATGTG

CTGCAAGGCGATTAAGTTGGGTAACGCCAGGGTTTTCCCAGTCACGACGTTGTAAAA CGACGGCCAGTGA

ATTGTAATACGACTCACTATAGGGCGAATTGGGCCCTCTAGATGCATGCTCGAGCGG CCGCCAGTGTGATGG

ATATCTGCAGAATTCGGCTTGGCCACGCGTCGACTAGTACGGGGGGGGGGGAGCCGG GGrGA77TCC7TCCr

CTTTTGGCAACATGGCGGGCGGA GAA GCTGGA GTGA CTCTA GGGCA GCCGCATCTTTCGCGTCA GGATCTCA

CCA CCTTGGA TGTTA CCA A GTTGA CGCCA CTTTCA CA CGAA GTTATCA GCA GA CAA GCCA CATACCGTC AGC

ACCACGCATAGCCGAATTCCAGCACACTGGCGGCCGTTACTAGTgGATCCg(SEQ ID NO: 28) and version 2:

CCTTTAGCAGCCCTTGCGCCCTGAATTTGTAAAATTCGCGTTAAATTTTGTTAAATC AGCTCATTTTTTAACC

AATAGGCCGAAATCGGCAAAATCCCTTATAAATCAAAAGAATAGACCGAGATAGGGT TGAGTGTTGTTCCA

GTTTGGAACAAGAGTCCACTATTAAAGAACGTGGACTCCAACGTCAAAGGGCGAAAA ACCGTCTATCAGG

GCGATGGCCCACTACGTGAACCATCACCCTAATCAAGTTTTTTGGGGTCGAGGTGCC GTAAAGCACTAAAT

CGGAACCCTAAAGGGAGCCCCCGATTTAGAGCTTGACGGGGAAAGCCGGCGAACGTG GCGAGAAAGGAA

GGGAAGAAAGCGAAAGGAGCGGGCGCTAGGGCGCTGGCAAGTGTAGCGGTCACGCTG CGCGTAACCACC

ACACCCGCCGCGCTTAATGCGCCGCTACAGGGCGCGTCCATTCGCCATTCAGGCTGC GCAACTGTTGGGAA

GGGCGATCGGTGCGGGCCTCTTCGCTATTACGCCAGCTGGCGAAAGGGGGATGTGCT GCAAGGCGATTAAG

TTGGGTAACGCCAGGGTTTTCCCAGTCACGACGTTGTAAAACGACGGCCAGTGAATT GTAATACGACTCAC

TATAGGGCGAATTGGGCCCTCTAGATGCATGCTCGAGCGGCCGCCAGTGTGATGGAT ATCTGCAGAATTCGG

CTTGGCCACGCGTCGACTAGTACGGGGGGGGGGGGGCrC7 7 GGCAA CArGGCGGGCGGA GAA GCTGGA G

TGA CTCTA GGGCA GCCGCA TCTTTCGCGTCA GGAA CTCA CCA CCTTGGATGTTA CCA A GTTGA CGCCA CTTT

CA CA CGAA GTTATCA GCA GA CAA GCCA CAAGCCGAATTCC AGC AC ACTGGCGGCCGTTACTAGTGGATCCG

AGCTCG(SEQ ID NO: 29)

Using the genomic DNA of SUNE2 cells as template, EIF2S3-TXNDC11 fusion gene was amplified by long range PCR with LA Taq® Hot Start Version DNA polymerase (TAKARA provided) at genomic DNA level. As shown in the scheme of Figure 22A, forward primers are designed to target the second exon of EIF2S3 gene as El primer 5'-TTGACGCCACTTTCACACGAAG-3' (SEQ ID NO: 30) (from human EIF2S3 mRNA sequence NM_001415.3), the reverse primers are designed to target the second exon of TXNDC11 gene as Tl primer 5' -CCTCTGAATCCCGTCGAACGTA-3' (SEQ ID NO: 31) (from human TXNDC11 mRNA sequence NW_004078085.1). The amplifying condition is 93 ;©33m in , ^° C68L0sec ;&fMmche cycle, adding 10 seconds to the extension time of every consecutive cycle till the 40th cycle, at last extend additional 6 min.

The PCR amplified products by primer set (El Tl) were sub cloned into TA vector (Fig. 22B) and subjected to sequence shown in Fig. 22C. Between the EIF2S3 gene (UCSC human genome GRCh37/hgl9 position 24073786) at Chromosome X and the TXNDCllgene (UCSC human genome GRCh37/hgl9 position 11830973) at Chromosome 16, there is an insertion of 6 pieces of DNA fragments: fragment 24073786-24073908 (EIF2S3 gene) of Chromosome X; fragment 66018003-66023649 of Chromosome 16; fragment 80169640-80170488 of Chromosome 16; fragment 11831302-11831864 of Chromosome 16; fragment 11830580-11830885 of Chromosome 16; fragment 11830016-11830973of Chromosome 16 (TXNDC11 gene).

Above all, RARS-MADILI and EIF2S3-TXNDC11 fusion genes play important roles in tumor genesis and development of cancers, and might be useful in early diagnosis and targeted therapy as molecular diagnostic markers and therapeutic targets.