¹Department of Biochemistry, University of Leicester, University Road, Leicester LE1 7RH, UK

²Department of Medicine and Therapeutics, University of Aberdeen, Polwarth Building Foresterhill, Aberdeen AB25 2ZD, UK

³Haematology Department, Western Infirmary, Dumbarton Road, Glasgow, UK

Correspondence to: A E Willis, Department of Biochemistry, University of Leicester, University Road, Leicester LE1 7RH, UK

^aCurrent address: Scripps Research Institute, Department of Neurobiology, La Jolla, California, CA 92037 USA

^bCurrent address: Department of Anatomy and Physiology, Medical Science Institute, University of Dundee, Dundee DD1 4HN, UK

^cCurrent address: Department of Biochemistry and Molecular Biology, University of Leeds, Leeds LS2 9JT, UK

The 5' untranslated region of the proto-oncogene c-myc contains an internal ribosome entry segment (IRES) (Nanbru et al., 1997; Stoneley et al., 1998) and thus c-myc protein synthesis can be initiated by a cap-independent as well as a cap-dependent mechanism (Stoneley et al., 2000). In cell lines derived from patients with multiple myeloma (MM) there is aberrant translational regulation of c-myc and this correlates with a C-T mutation in the c-myc-IRES (Paulin et al., 1996). RNA derived from the mutant IRES displays enhanced binding of protein factors (Paulin et al., 1998). Here we show that the same mutation is present in 42% of bone marrow samples obtained from patients with MM, but was not present in any of 21 controls demonstrating a strong correlation between this mutation and the disease. In a tissue culture based assay, the mutant version of the c-myc-IRES was more active in all cell types tested, but showed the greatest activity in a cell line derived from a patient with MM. Our data demonstrate that a single mutation in the c-myc-IRES is sufficient to cause enhanced initiation of translation via internal ribosome entry and represents a novel mechanism of oncogenesis. Oncogene (2000) 19, 4437-4440

c-myc; multiple myeloma; internal ribosome entry segment; translation initiation

Initiation of translation can occur by two different mechanisms; cap-dependent initiation which involves the binding of the eukaryotic initiation factor (eIF) complex 4F to the 7 methyl G cap at the 5' end of the mRNA and ribosomal scanning (Gray and Wickens, 1998; Pain, 1996) or internal ribosome entry. The latter is mediated by structured regions of RNA termed internal ribosome entry segments (IRES), and these have the ability to direct ribosome binding some considerable distance (up to 1000 nts) from the 5' end of the message. IRESs were originally discovered in picornaviral mRNAs, but have now been identified in a number of eukaryotic mRNAs, the protein products of which are involved in the control of cell growth including, FGF-2 (Vagner et al., 1995), VEGF (Miller et al., 1998; Stein et al., 1998), PDGF (Bernstein et al., 1997) and c-myc (Nanbru et al., 1997; Stoneley et al., 1998) and cell death (Coldwell et al., 2000; Holcik et al., 1999). In the case of c-myc we have shown that initiation of protein synthesis can occur by two distinct mechanisms; cap-dependent scanning and IRES mediated (Stoneley et al., 2000). Thus suggesting that control of translation is important in regulating cellular levels of c-myc in addition to the other mechanisms that have been described. These include alterations in rates of transcription of the mRNA, and the stability of both the mRNA and protein (Ross, 1995; Lee et al., 1998; Lüscher and Eisenmann, 1988; Shindo et al., 1993). The number of levels that c-myc expression is controlled at is probably a reflection of the diverse cellular processes in which this proto-oncogene is thought to function e.g. cell proliferation, cell cycle progression and apoptosis (Evan and Littlewood, 1993; Dang, 1999). It is not surprising therefore, that de-regulation of c-myc is associated with a wide range of tumour types and this may occur by amplification or chromosomal rearrangement (Neri et al., 1988; Pelicci et al., 1986; Schwab et al., 1986; Sumegi et al., 1985). However, we have also shown that increased c-myc protein expression can occur from aberrant translational regulation (Paulin et al., 1996; West et al., 1995). For example, in cell lines derived from patients with multiple myeloma (MM, an incurable disease which is characterized by bone marrow plasmacytosis, osteolytic lesions and secretion of a monoclonal immunoglobulin (Niesvizky et al., 1993) there is an up to 20-fold increase in c-myc protein levels that occurs by a translational mechanism (Paulin et al., 1996). This increased c-myc protein expression in MM derived cell lines correlates with a C-T mutation in the region of c-myc DNA that contains the IRES (Paulin et al., 1996). In this study experiments were performed to investigate whether the C-T mutation in the c-myc IRES identified in the cell lines is also present in the bone marrow of MM patients and moreover, if the presence of the mutation alters translation initiation via the IRES.

Bone marrow derived mRNA from patients with MM, premalignant Monoclonal Gammopathy of uncertain significance (MGUS) patients or controls with either non-MM disease or normal individuals, was reverse transcribed. Using oligonucleotides complementary to the P2 promoter of c-myc and to the 3' end of the 5' untranslated region (Figure 1a), the c-myc IRES was amplified by the polymerase chain reaction. Initially T-track sequencing was performed to identify those samples that contained the mutation (Figure 1b). The sequence of this region is CCCTCTC in wild type samples and bands that correspond to two thymidines in this segment of DNA can clearly been seen in the control bone marrow samples (Figure 1b). In the mutant samples the sequence is altered to CTCTCTC, and thus thymidines are now observed. Due to a slight compression in this region, the mutant and wild type sequences migrate to slightly different positions on the gel allowing them to be distinguished between in a mixed population. This is particularly fortuitous since all of the patient samples which scored positive for the mutation contained a mixture of both wild type and mutant sequence, and thus thymidines are observed at this position on the gel (Figure 1b). The difference in intensity in the bands that correspond to the presence of the mutation in the bone marrow samples reflects the fact that there were varying amount of MM plasma cells in this material. All those samples that scored positive were then subjected to full sequencing (Figure 1c). The results from the sequencing analysis are summarised in Table 1. These show that the C-T mutation at position 2756 (numbering as in Watt et al., 1983) was identified in 42% (8/19) MM patients, 20% (2/10) MGUS patients and was undetectable (0/21) in control patients. The difference in allele distribution is chi squared 11.04 (P=0.004) across all three groups and 10.59 (P=0.001) for MM versus the controls. This indicates a strong association of the C-T mutation with MM. Blood samples were then analysed from an additional 20 patients with multiple myeloma and these all scored negative for the presence of the C-T mutation (data not shown) suggesting that the mutation is somatic rather than germ line.

To compare the activity of the mutant and wild type versions of the c-myc IRES, the region of DNA which contains the mutation was cloned into the dicistronic reporter vector pRF (Stoneley et al., 2000). This vector contains the Renilla and Firefly luciferase genes upstream and downstream respectively of a spacer region (Figure 2a) (Stoneley et al., 2000). We have shown previously that insertion of the c-myc IRES into this intercistronic region results in a 50-100-fold increase in the activity of the firefly luciferase by internal ribosome entry (Stoneley et al., 1998). The WT and MT versions of this dicistronic plasmid (pRMF and pRMFmt) were then transfected into a number of cell types. In all cells tested the mutant version of the c-myc IRES was more active than the wild type. Thus the presence of the mutation increased the activity of the luciferase reporter enzyme 1.5-fold in HeLa cells, 1.6-fold in a control SV40 transformed fibroblast line GM637, twofold in a control B-cell line derived from a normal individual, GM03201, whilst in the MM derived cell line, GM2132, the mutant version of the IRES was between 4-6 times more active (Figure 2b). We have shown previously that the efficiency of c-myc IRES-driven translation varies widely between cells lines (Stoneley et al., 2000). However, the enhanced expression of the mutant is not simply a reflection of enhanced IRES expression in GM2132 cells. Thus the relative c-myc-IRES efficiency when expressed as the ratio of Firefly luciferase/Renilla luciferase in MM derived GM2132 cells has a value of 0.75, compared to 4 for HeLa, 1.6 for GM637 and 0.55 for GM03021. The presence of the mutation in the c-myc IRES causes the efficiency of IRES mediated translation in the GM2132 cells to become comparable to that which we observe in HeLa cells where we find that IRESs are most active (Stoneley et al., 2000; Coldwell et al., 2000). Since the major difference in the mutant IRES function was observed in the MM derived cell line, it would suggest that trans-acting factors that are required for internal ribosome entry are more abundant in this cell line. In agreement with this we have shown previously that the MM derived cell line GM2132 expresses an altered repertoire of RNA binding proteins (Paulin et al., 1998). By chemical and enzymatic probing of c-myc IRES RNA in conjunction with computer analysis, we have derived a structural model for the c-myc IRES (LeQuesne et al., submitted). Using this model we predict that the C-T mutation may cause stabilization of a stem loop structure in the 3' end of the IRES (Figure 3). We hypothesise that the increased expression from the C-T form of the c-myc IRES is likely to reside in an enhanced interaction of RNA binding proteins with this structurally altered region.

In conclusion we have identified a mutation in the c-myc IRES in the plasma cells isolated from bone marrow in 42% of MM patients samples and have shown in a tissue culture based assay that in a MM derived cell line, this mutation increases the expression of a downstream reporter by up to sixfold. Clearly this is not the only mechanism to increase c-myc expression since 58% of samples do not contain this mutation. However, we have found some additional mutations in the c-myc IRES in bone marrow material obtained from these patients and those currently under investigation. Several examples exist in viral systems which show that single mutations are capable of altering IRES function, for example a single substitution in the FMDV-IRES has been shown to increase the degree of internal ribosome entry up to fivefold (Martinez-Salas et al., 1993). However, this is the first example of a mutation in a eukaryotic IRES leading to increased translation initiation by internal ribosome entry. This represents a novel mechanism of c-myc de-regulation and has profound implications in tumorigenesis.

Acknowledgements

This work was funded by grants from the Leukaemia Research Fund (SA Chappell, M de Schoolmeester and M Helfrich), Medical Research Council (FEM Paulin; JPC LeQuesne, held a MRC studentship) and the Cancer Research Campaign (M Stoneley) and the BBSRC (advanced fellowship to AE Willis).

Auscbel RM. (1987). In: Current Protocols in Molecular Biology. Greene J. (ed.). Wiley-Interscience: New York., pp. 11-9.17.

Bernstein J, Sella O, Le S and Elroy-Stein O. (1997). J. Biol. Chem. 272, 9356-9362. MEDLINE

Choczynski P and Sacchi N. (1987). Analyt. Biochem. 162, 156-159. MEDLINE

Coldwell M, Mitchell S, Stoneley M, MacFarlane M and Willis A. (2000). Oncogene 19, 899-907. MEDLINE

Dang C. (1999). Mol. Cell Biol. 19, 1-11. MEDLINE

Evan GI and Littlewood TD. (1993). Curr. Opin. Genet. Dev. 3, 44-49. MEDLINE

Gray N and Wickens M. (1998). Annu. Rev. Cell Dev. Biol. 14, 399-458. MEDLINE

Holcik M, Lefebvre C, Ye C, Chow T and Korneluk R. (1999). Nature Cell Biol. 1, 190-192. Article MEDLINE

Lee CH, Leeds P and Ross J. (1998). J. Biol. Chem. 273, 25261-25271. MEDLINE

Lüscher B and Eisenmann RN. (1988). Mol. Cell. Biol. 8, 2504-2512. MEDLINE

Martinez-Salas E, Saiz J-C, Davila M, Belsham GJ and Domingo E. (1993). J. Virol. 67, 3748-3755. MEDLINE

Miller D, Dibbens J, Damert A, Risau W, Vadas M and Goodall G. (1998). FEBS Letts. 434, 417-420.

Nanbru C, Lafon I, Audiger S, Gensac G, Vagner S, Huez G and Prats A-C. (1997). J. Biol. Chem. 272, 32061-32066. MEDLINE

Neri A, Barriga F, Knowles DM, Magrath IT and Dalla-Favera R. (1988). Proc. Natl., Acad. Sci. USA 85, 2748-2752.

Niesvizky R, Siegel D and Michaeli J. (1993). Blood Rev. 7, 24-33. MEDLINE

Pain VM. (1996). Eur. J. Biochem. 236, 747-771. MEDLINE

Paulin FEM, Chappell SA and Willis AE. (1998). Nucleic Acids Res. 26, 3097-3103. MEDLINE

Paulin FEM, West MJ, Sullivan NF, Whitney RL, Lyne L and Willis AE. (1996). Oncogene 13, 505-513. MEDLINE

Pelicci P, Knowles DM, Magrath IT and Dalla-Favera R. (1986). Proc. Natl. Acad. Sci. USA 2984-2988.

Ross J. (1995). Microbiol. Rev. 59, 16-95.

Schwab M, Klempnauer K-H, Alitalo K, Varmus H and Bishop M. (1986). Mol. Cell. Biol. 6, 2752-2755. MEDLINE

Shindo H, Tani E, Matsumuto T, Hashimoto T and Furuyama J. (1993). Acta Neuropathologica 86, 345-352. MEDLINE

Stein I, Itin A, Einat P, Skaliter R, Grossman Z and Keshet E. (1998). Mol. Cell. Biol. 18, 3112-3119. MEDLINE

Stoneley M, Paulin FEM, LeQuesne JPC, Chappell SA and Willis AE. (1998). Oncogene 16, 423-428. MEDLINE

Stoneley M, Subkhankulova T, Le Quesne J, Coldwell M, Jopling C, Belsham G and Willis A. (2000). Nucleic Acids Res. 28, 687-694. MEDLINE

Sumegi J, Hedberg T, Bjorkholm M, Godal T, Mellstedt H, Nilsson MG, Perlman C and Klein G. (1985). Int. J. Cancer 36, 367-371. MEDLINE

Vagner S, Gensac M-C, Maret A, Bayard F, Amalric F, Prats H and Prats A-C. (1995). Mol. Cell. Biol. 15, 35-44. MEDLINE

Watt R, Nishikaura K, Sorrentino J, ar-Rushdi A, Croce CM and Rovera G. (1983). Proc. Natl. Acad. Sci. USA 80, 6307-6311. MEDLINE

West MJ, Sullivan NF and Willis AE. (1995). Oncogene 13, 2515-2524.

Figure 1 C-myc IRES derived from patients with MM contains a C-T mutation at 2756. (a) Schematic diagram showing position of the PCR primers and the location of the mutation (numbering as in Watt et al., 1983). (b) Section of a polyacrylamide gel showing direct T-track sequencing of DNA samples derived from bone marrow of MM or control patients. Patient samples that scored positive for the C-T mutation at 2756 are marked*. (c) Section of a gel showing an example of the full DNA sequence pattern obtained from patient samples that contain the mutation and a control sample. Total cellular RNA was prepared according to the method of Chomczynski and Sacchi (1987). After the removal of contaminating DNA the samples were incubated for 1 h with Moloney Murine Leukaemia Virus reverse transcriptase in a mixture containing 0.5 mM dNTPs, 10 mM DTT, 1 l random hexamer primers (100 pmoles/l) and 1 unit RNasin (Promega). Two l of these samples were then used in a polymerase chain reaction. The Primers FP1 (5'-GCCGGATCCGGCCCTTTATAATGCGAC-3') and FP2 (5'-GTGGAATTCTGGTTTTCCACTACCCGAAA-3') were used to amplify c-myc exon 1 (2289-2881). PCR reactions contained 10 l 10´PCR Buffer (Advanced Biosystems), 10 l MgCl₂ (25 mM), 1 l dNTPs (10 mM of each of dATP, dCTP, dTTP, dGTP), 1 unit Taq DNA polymerase (Advanced Biosystems). The PCR reactions were carried out in a Techne DNA Thermal Cycler at 94°C for 3 min followed by 37 cycles of (94°C for 2 min, 63°C for 3 min, 72°C for 2 min) and 72°C for 10 min

Figure 2 The mutant version of the IRES is most active in MM derived cell lines. (a) Schematic diagram of dicistronic construct used to determine the expression of the c-myc IRES. The corresponding region of DNA which contained the mutation was sub-cloned into the discistronic plasmid vector pRF, which encodes two luciferase genes, Renilla luciferase upstream and Firefly luciferase downstream. (b). Activities of the wild type and mutant version of the c-myc IRES in a range of cell lines. All cells were obtained from NIGMS. Cells were grown in RPMI (GM2132, GM03201) or DMEM (HeLa, GM637) containing 10% foetal calf serum in a humidified atmosphere containing 5% CO₂. Cells were transfected with pRMF (that contains the wild type IRES; Stoneley et al., 1998) or pRMFmt and the -galactosidase encoding construct pcDNA3. 1/HisB/Lacz (Invitrogen) using either calcium phosphate DNA co-precipitation (Auscbel, 1987) or by electroporation (GM2132 and GM03201) using a BioRad Gene pulser. Cells were harvested after 48 h and luciferase expression was determined using the dual-luciferase assay system (Promega). -galactosidase expression was determined using a Galactolight plus system (Tropix). Both activities were measured in an Opticomp-1 Luminometer (MGM instruments). Variations in transfection efficiency were corrected by normalizing luciferase activity to -galactosidase activity. All assays were performed in triplicate on three independent occasions

Figure 3 The increased activity of the c-myc IRES is likely to be caused by a structural change. We have derived a secondary structure model for the c-myc IRES (LeQuesne et al., submitted) and using this model we predict that the C-T mutation would result in the insertion of an extra stem loop region in the 3' end of the UTR

Table 1 Summary of patient samples that were examined for the presence of the mutation