The initiation of cap-dependent translation involves ˜13 tightly controlled protein factors (reviewed in (Jackson et al., 2010)). Among these, eIF4E binds the mRNA cap structure and interacts with a scaffold (eIF4G) and the eIF4A RNA helicase (a DEAD box protein also known as DDX2). During initiation these and other factors form the eIF4F complex and together with the 40S ribosomal unit proceed to a transcript's 5′UTR for a translation start site. The eIF4A RNA helicase is directly involved in scanning and recent studies have defined co-factors and the molecular mechanics of its helicase activity (Marintchev, 2009, 2013; Parsyan et al., 2011; Svitkin, 2001). However, the precise mRNA features that necessitate the eIF4A helicase action are not known.
The activation of protein translation contributes to malignant transformation. For example, activation of the RAS, ERK, and AKT signaling pathways stimulates cap-dependent translation (reviewed in (Blagden and Willis, 2011; D'Ambrogio et al., 2013; Guertin and Sabatini, 2007). Moreover, the rate limiting eIF4E translation factor is expressed at high levels in many cancers and can transform rodent fibroblasts and promote tumor development in vivo (Lazaris-Karatzas et al., 1990; Ruggero et al., 2004; Wendel et al., 2004). Accordingly, cap-dependent translation is an emerging target for cancer therapies (see recent review by (Blagden and Willis, 2011). Notably, three distinct natural compounds target the eIF4A helicase and these are silvestrol isolated from plants in the Malaysian rainforest (Cencic, 2009), pateamine A found in marine sponges off the coast of New Zealand (Northcote et al., 1991), and hippuristanol which is produced by pacific corals (Li et al., 2009b). These compounds show promising preclinical activity against different cancers (Bordeleau et al., 2005; Bordeleau et al., 2006; Cencic et al., 2007; Schatz et al., 2011; Tsumuraya et al., 2011a). Other strategies to inhibit translation include rapamycin and mTORC1 kinase inhibitors (Hsieh et al., 2012; Thoreen et al., 2009), inhibitors of the eIF4E kinase MNK1/2 (Furic et al., 2010; Ueda et al., 2004; Wendel et al., 2007), a peptide (4EGI-1) that interferes with the eIF4E-eIF4G interaction (Moerke et al., 2007), and the anti-viral ribavirin that may bind eIF4E directly (Kentsis et al., 2004; Yan et al., 2005).
The recently developed transcriptome-scale ribosome footprinting technology greatly facilitates the study of protein translation. Briefly, the technology is based on the identification of ribosome-protected RNA fragments in relation to total transcript levels using deep sequencing (Ingolia et al., 2009). The technology has been applied to explore translational effects in various biological contexts, and perhaps the most relevant to this study are reports of the translational effects of mTORC1 inhibition on mRNAs harboring TOP- and TOP-like sequences (Hsieh et al., 2012; Thoreen et al., 2012).
In one embodiment, a method is provided for identifying an agent capable of modulating cap-dependent mRNA translation. The method comprises comparing translation efficiency in the presence and absence of the agent in an in-vitro or in-vivo translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs. eIF4A refers to eIF4A1 or eIF4A2, and RNA helicases include, but are not limited to, eIF4A1, eIF4A2, DHX9 or DHX36. The modulation of translation in the presence of the agent indicates the agent as capable of modulating cap-dependent mRNA translation. In one embodiment, modulating is decreasing, suppressing or inhibiting cap-dependent mRNA translation. In one embodiment, the agent stabilizes the binding of eIF4A to the eIF4A-dependent translation-controlling motif of the mRNA. In one embodiment, the eIF4A-mRNA complex stabilizing motif of the mRNA is located in the 5′ UTR.
In one embodiment, the eIF4A-dependent translation-controlling motif comprises a G-quadruplex structure. In one embodiment, the G-quadruplex structure comprises a (GGC/A)4 motif. In one embodiment, the (GGC/A)4 motif comprises GGCGGCGGCGGC (SEQ ID NO:1). In one embodiment, the eIF4A-dependent translation-controlling motif comprises a sequence selected from SEQ ID NO:4, SEQ ID NO:5, SEQ ID NO:6, SEQ ID NO:7, SEQ ID NO:8, SEQ ID NO:9 or SEQ ID NO:10. In one embodiment, the eIF4A-dependent translation-controlling motif comprises a sequence selected from among SEQ ID NO:10 to SEQ ID NO:62. In one embodiment, the eIF4A-dependent translation-controlling motif is at least one sequence selected from SEQ ID NO:1 or from among SEQ ID NO:4 to SEQ ID NO:62.
In one embodiment of the methods described herein, the mRNA encodes a transcription factor. In one embodiment, the mRNA encodes an oncogene. In other embodiments, the mRNA encodes NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2. In other embodiments, the mRNA is from a gene selected from Table 3A. In other embodiments, the mRNA is from a gene selected from Table 3B. In other embodiments, the mRNA is from a gene selected from Table 3C.
In one embodiment of the method, the agent suppresses the growth of cancer cells in vitro or in vivo. In one embodiment, the agent interferes with eIF4A activity. In one embodiment, the agent increases eIF4A activity. In one embodiment, the agent inhibits eIF4A helicase activity. In one embodiment, the agent increases eIF4A helicase activity. In one embodiment, the agent promotes the stabilizing the binding of eIF4A with an eIF4A-dependent translation-controlling motif. In one embodiment, the agent does not trigger feedback activation of Akt.
In one embodiment, the modulation of translation in the foregoing method is measured by a fluorescence reporter assay. In one embodiment, the assay comprises renilla luciferase expression.
In one embodiment, a method is provided for identifying an agent that modulates eIF4A activity, the method comprising comparing translation efficiency in the presence and absence of the agent in an in-vitro or in-vivo translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs, wherein the increase or decrease in translation efficiency in the presence of the agent indicates the agent as capable of increasing or decreasing eIF4A activity.
In one embodiment, a method is provided for identifying an agent that inhibits eIF4A activity, the method comprising comparing translation efficiency in the presence and absence of the agent in an in-vitro or in-vivo translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs, wherein a decrease in translation efficiency in the presence of the agent indicates the agent as capable of inhibiting eIF4A activity.
In one embodiment, a method is provided for determining whether an mRNA sequence comprises at least one eIF4A-dependent translation-controlling motif, the method comprising comparing translation efficiency in the presence and absence of an agent that inhibits eIF4A activity in an in-vivo translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs, wherein a decrease in translation efficiency in the presence of the agent indicates the mRNA sequence possesses at least one eIF4A-dependent translation-controlling motif.
In one embodiment, a method is provided for determining whether a cancer or tumor is susceptible to an agent that inhibits eIF4A activity, the method comprising identifying the presence of at least one eIF4A-dependent translation-controlling motif in mRNA from the cancer or tumor, wherein the presence of the at least one eIF4A-dependent translation-controlling motif indicates susceptibility of the cancer or tumor to the agent. In one embodiment, the level of expression of MYC is not predictive of the susceptibility of a cancer or tumor to an agent that inhibits eIF4A activity.
In one embodiment, methods are provided for 1) measuring the effect of known RNA helicases such as eIF4A, DHX9 or DHX36 on G-quadruplex unwinding; 2) investigating the effect of other cofactors/inhibitors required for eIF4A activity; 3) a screening method to identify other proteins that can unwind G-quadruplexes; or 4) identifying and establishing the effect of small molecules that stabilize the G-quadruplex structure, by utilizing a fluorescence resonance energy transfer (FRET)-based assay utilizing an oligonucleotide comprising a G-quadruplex labeled with a fluorophore at the 5′ or 3′ end of the oligonucleotide, and a fluorescence quencher at the other end. The aforementioned uses are merely non-limiting examples.
In one embodiment, a method for preventing, treating or intervening in the recurrence of a cancer in a subject is provided. The method comprises administering to the subject an agent that blocks eIF4a helicase activity, thereby preventing, treating or intervening in the recurrence of the cancer. In one embodiment, the agent that blocks eIF4A helicase inhibits the translation of an oncogenic mRNA. In one embodiment, the oncogenic mRNA comprises an eIF4A-dependent translation-controlling motif. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the oncogenic mRNA comprises a G-quadruplex motif. In one embodiment, the oncogenic mRNA is from an oncogene, which by way of non-limiting example is selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In the foregoing embodiments, the cancer is, by way of non-limiting examples, T-cell acute lymphoblastic leukemia, small cell lung cancer, renal cell carcinoma, squamous cell carcinoma of the head and neck, neuroblastoma and pancreatic cancer. In one embodiment the subject has cancer. In one embodiment, the subject is at risk for developing cancer. In one embodiment, the subject is in remission from cancer. In other embodiments, the cancer is transformed follicular lymphoma, mantel cell lymphoma, breast cancer, ovarian cancer, hepatocellular carcinoma, and non-small cell lung cancer, as well as gastric cancer, Ewing sarcoma and lung adenocarcinoma.
In one embodiment, a method is provided for preventing, treating or intervening in the recurrence of a cancer in a subject having an eIF4A dependent cancer. The method comprises administering to the subject an agent that blocks eIF4a helicase activity, thereby preventing, treating or intervening in the recurrence of the cancer. In one embodiment, the agent that blocks eIF4A helicase inhibits the translation of an oncogenic mRNA. In one embodiment, the oncogenic mRNA comprises an eIF4A-dependent translation-controlling motif. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the oncogenic mRNA comprises a G-quadruplex motif. In one embodiment, the oncogenic mRNA is from an oncogene. In one embodiment, the oncogene is selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In the foregoing embodiments, the cancer is, by way of non-limiting examples, T-cell acute lymphoblastic leukemia, small cell lung cancer, renal cell carcinoma, squamous cell carcinoma of the head and neck, neuroblastoma and pancreatic cancer. In one embodiment the subject has cancer. In one embodiment, the subject is at risk for developing cancer. In one embodiment, the subject is in remission from cancer. In other embodiments, the cancer is transformed follicular lymphoma, mantel cell lymphoma, breast cancer, ovarian cancer, hepatocellular carcinoma, and non-small cell lung cancer, as well as gastric cancer, Ewing sarcoma and lung adenocarcinoma.
In another embodiment, a method is provided for inhibiting in a subject the translation of an oncogene that comprises an eIF4A-dependent translation-controlling motif. The method comprises administering to the subject an agent that blocks eIF4a helicase, thereby inhibiting translation of the oncogene. In one embodiment, translation of the oncogene causes cancer in the subject. In another embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In this embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the mRNA of the oncogene comprises a G-quadruplex motif. In one embodiment, the oncogene is selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In the foregoing embodiments, the cancer is, by way of non-limiting examples, T-cell acute lymphoblastic leukemia, small cell lung cancer, renal cell carcinoma, squamous cell carcinoma of the head and neck, neuroblastoma and pancreatic cancer. In one embodiment the subject has cancer. In one embodiment, the subject is at risk for developing cancer. In one embodiment, the subject is in remission from cancer. In other embodiments, the cancer is transformed follicular lymphoma, mantel cell lymphoma, breast cancer, ovarian cancer, hepatocellular carcinoma, and non-small cell lung cancer, as well as gastric cancer, Ewing sarcoma and lung adenocarcinoma.
In one embodiment, a method for inhibiting in a subject eIF4A dependent mRNA translation is provided. The method comprises administering to the subject an agent that blocks eIF4a helicase, thereby inhibiting mRNA translation. In one embodiment, the mRNA translation causes cancer in the subject. In one embodiment, the mRNA comprises an eIF4A-dependent translation-controlling motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the mRNA encodes an oncogenic protein. In one embodiment, the oncogenic protein is encoded by an oncogene selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In the foregoing embodiments, the cancer is, by way of non-limiting examples, T-cell acute lymphoblastic leukemia, small cell lung cancer, renal cell carcinoma, squamous cell carcinoma of the head and neck, neuroblastoma and pancreatic cancer. In one embodiment the subject has cancer. In one embodiment, the subject is at risk for developing cancer. In one embodiment, the subject is in remission from cancer. In other embodiments, the cancer is transformed follicular lymphoma, mantel cell lymphoma, breast cancer, ovarian cancer, hepatocellular carcinoma, and non-small cell lung cancer, as well as gastric cancer, Ewing sarcoma and lung adenocarcinoma.
In one embodiment, a method for preventing in a subject the translation of an mRNA comprising an eIF4A-dependent translation-controlling motif. The method comprises administering to the subject an agent that blocks eIF4a helicase activity, thereby inhibiting translation of the mRNA. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the mRNA is from an oncogene selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2. In one embodiment, the translation of the mRNA causes cancer.
In the foregoing embodiments, the cancer is, by way of non-limiting examples, T-cell acute lymphoblastic leukemia, small cell lung cancer, renal cell carcinoma, squamous cell carcinoma of the head and neck, neuroblastoma and pancreatic cancer. In one embodiment the subject has cancer. In one embodiment, the subject is at risk for developing cancer. In one embodiment, the subject is in remission from cancer. In other embodiments, the cancer is transformed follicular lymphoma, mantel cell lymphoma, breast cancer, ovarian cancer, hepatocellular carcinoma, and non-small cell lung cancer, as well as gastric cancer, Ewing sarcoma and lung adenocarcinoma.
In any of the foregoing embodiments, the agent blocks the activity of eIF4A helicase. In any of the foregoing embodiments, the agent blocks the translation of an mRNA comprising an eIF4A-dependent translation-controlling motif. In any of the foregoing embodiments, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In any of the foregoing embodiments, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62.
Non-limiting examples of aforementioned agents include a rocaglamide, such as silvestrol, CR-31-B, or an analogue or derivative thereof. In other embodiments, the agent is hippuristanol, pateamine A, or an analogue or derivative thereof.
U.S. Patent Application Ser. No. 61/912,420, filed Dec. 5, 2013, is incorporated herein by reference in its entirety.
A mechanism of translational control has been identified that is characterized by a requirement for eIF4A/DDX2 RNA helicase activity and underlies the anticancer effects of silvestrol and related compounds. eIF4A refers to eIF4A1 or eIF4A2, and RNA helicases include, but are not limited to, eIF4A1, eIF4A2, DHX9 or DHX36. In one embodiment, activation of cap-dependent translation contributes to T-cell leukemia (T-ALL) development and maintenance. Accordingly, inhibition of the translation initiation factor eIF4A with silvestrol produces powerful therapeutic effects. By using transcriptome-scale ribosome footprinting on silvestrol-treated T-ALL cells to identify silvestrol-sensitive transcripts, the features of eIF4A-dependent translation embodied herein were identified. These features include, in one embodiment, a long 5′UTR and a 12-mer sequence motif that encodes a guanine quartet (GGC)4. RNA folding algorithms pinpoint the (GGC)4 motif as a common site of RNA G-quadruplex structures within the 5′UTR. In T-ALL these structures mark highly silvestrol-sensitive transcripts that include key oncogenes and transcription factors and contribute to the drug's anti-leukemic action. Hence, the eIF4A-dependent translation of G-quadruplex containing transcripts is shown as a gene-selective and therapeutically targetable mechanism of translational control.
Thus, in one embodiment, a method for identifying an agent capable of modulating cap-dependent mRNA translation is provided, the method comprising comparing translation efficiency in the presence and absence of the agent in an in-vitro or in-vivo translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs, wherein the modulation of translation in the presence of the agent indicates the agent as capable of modulating cap-dependent mRNA translation. In some embodiments, modulating is decreasing, suppressing or inhibiting cap-dependent mRNA translation.
eIF4A-dependent translation-controlling motifs are typically present in the 5′ UTR of the mRNA. In certain embodiments, the eIF4A-dependent translation-controlling motif comprises a G-quadruplex structure. In some embodiments, the G-quadruplex structure is a (GGC/A)4 motif (i.e., four occurrences of (G, G, C or A), each occurrence independently selected from either GGC or GGA). In some embodiments, the (GGC/A)4 motif is GGCGGCGGCGGC (SEQ ID NO:1). In some embodiments, the eIF4A-dependent translation-controlling motif comprises GGGAC (SEQ ID NO:2) motif or GGGCC (SEQ ID NO:3). In other embodiments the eIF4A-dependent translation-controlling motif comprises SEQ ID NO:4, SEQ ID NO:5, SEQ ID NO:6, SEQ ID NO:7, SEQ ID NO:8, SEQ ID NO:9 or SEQ ID NO:10. In other embodiments, the eIF4A-dependent translation-controlling motif comprises a sequence selected from among SEQ ID NO:10 to SEQ ID NO:62. In other embodiments, the eIF4A-dependent translation-controlling motif is at least one sequence selected from SEQ ID NO:1 or from SEQ ID NO:4 to SEQ ID NO:62.
The mRNA may have one or more eIF4A-dependent translation-controlling motifs. In one embodiment, the eIF4A-dependent translation-controlling motif is at least one (GGC/A)4 motif. In another embodiment, the eIF4A-dependent translation-controlling motif is at least one GGGAC (SEQ ID NO:2) motif. In another embodiment, the eIF4A-dependent translation-controlling motif is at least one GGGCC (SEQ ID NO:3) motif. In another embodiment, the eIF4A-dependent translation-controlling motif is at least one 12-mer motif. In other embodiments, the mRNA may comprise 2, 3, 4, 5, 6, 7, 8, 9, 10, or more eIF4A-dependent translation-controlling motifs. In another embodiment, each eIF4A-dependent translation-controlling motif is independently selected from among SEQ ID NO:1 through and including SEQ ID NO:62.
In one embodiment, an agent identified by the methods of the invention may interfere with eIF4A activity. In one embodiment, the agent may increase eIF4A activity. In one embodiment, the agent may inhibit eIF4A helicase activity. In another embodiment, the agent may increase eIF4A helicase activity. In another embodiment, the agent can promote the stabilizing the binding of eIF4A with an eIF4A-dependent translation-controlling motif.
In another embodiment, the agent does not trigger feedback activation of Akt.
In another embodiment, the mRNA encodes a transcription factor. In another embodiment, the mRNA encodes an oncogene. In another embodiment, the mRNA encodes NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2. In another embodiment the mRNA is from a gene selected from Table 3A. In another embodiment, the mRNA is from a gene selected from Table 3B. In another embodiment, the mRNA is from a gene selected from Table 3C.
The agent identified by the methods herein may be used to treat cancer. In one embodiment, the cancer is a result of the overexpression an oncogene or transcription factor. The oncogene or transcription factor may be selected from those described herein, such as but not limited to NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2, or any described in Table 3A, 3B or 3C.
Cancer includes cancerous and precancerous conditions, including, for example, premalignant and malignant hyperproliferative diseases such as cancers of the breast, ovary, germ cell, skin, prostate, colon, bladder, cervix, uterus, stomach, lung, esophagus, blood and lymphatic system, larynx, oral cavity, as well as metaplasias, dysplasias, neoplasias, leukoplakias and papillomas of the mucous membranes, and in the treatment of Kaposi's sarcoma. These are also referred to herein as dysproliferative diseases or dysproliferation. Non-limiting examples of other cancers, tumors, malignancies, neoplasms, and other dysproliferative diseases that can be treated according to the invention include leukemias, such as myeloid and lymphocytic leukemias, lymphomas, myeloproliferative diseases, and solid tumors, such as but not limited to sarcomas and carcinomas such as fibrosarcoma, myxosarcoma, liposarcoma, chondrosarcoma, osteogenic sarcoma, chordoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, mesothelioma, Ewing's tumor, leiomyosarcoma, rhabdomyosarcoma, colon carcinoma, pancreatic cancer, breast cancer, ovarian cancer, prostate cancer, squamous cell carcinoma, basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinomas, cystadenocarcinoma, medullary carcinoma, bronchogenic carcinoma, renal cell carcinoma, hepatoma, bile duct carcinoma, choriocarcinoma, seminoma, embryonal carcinoma, Wilms' tumor, cervical cancer, testicular tumor, lung carcinoma, small cell lung carcinoma, bladder carcinoma, epithelial carcinoma, glioma, astrocytoma, medulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, meningioma, melanoma, neuroblastoma, and retinoblastoma.
In one embodiment, the compounds and uses embodied herein are directed to small cell lung cancer. In one embodiment, the compounds and uses embodied herein are directed to renal cancers. In one embodiment, the compounds and uses embodied herein are directed to neuroblastoma. In one embodiment, the compounds and uses embodied herein are directed to pancreatic cancers.
In one embodiment the agent suppresses the growth of cancer cells in vitro or in vivo.
The method of carrying out the translation assay using an in-vitro or in-vivo assay described herein may be accomplished by any of a number of methods know in the art. In one embodiment, the modulation of translation is measured by a fluorescence reporter assay. In one embodiment, the fluorescence reporter assay comprises renilla luciferase expression.
As mentioned above, certain mRNAs have longer 5′ UTRs and the eIF4A-dependent translation-controlling motif is present in the 5′ UTR. In one embodiment, the eIF4A-dependent translation-controlling motif comprises a 12-mer and the mRNA is from a gene selected from Table 3A. In another embodiment, the eIF4A-dependent translation-controlling motif comprises a 9-mer and the mRNA is from a gene selected from Table 3B. In another embodiment, eIF4A-dependent translation-controlling motif comprises a (GGC)4 motif and the mRNA is from a gene selected from Table 3C.
In another embodiment, a method for identifying an agent that modulates eIF4A activity is provided. The method comprises comparing translation efficiency in the presence and absence of the agent in an in-vitro or in-vivo translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs. An increase or decrease in translation efficiency in the presence of the agent indicates the agent as capable of increasing or decreasing eIF4A activity, respectively. The in-vitro or in-vivo translation system may be one from among those described here. The mRNA may be among those described herein. The eIF4A-dependent translation-controlling motifs may be among those described herein.
In another embodiment, a method is provided for identifying an agent that inhibits eIF4A activity, the method comprising comparing translation efficiency in the presence and absence of the agent in an in-vitro translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs, wherein a decrease in translation efficiency in the presence of the agent indicates the agent as capable of inhibiting eIF4A activity. The in-vitro or in-vivo translation system may be one from among those described here. The mRNA may be among those described herein. The eIF4A-dependent translation-controlling motifs may be among those described herein.
In another embodiment, a method is described for determining whether an mRNA sequence comprises at least one eIF4A-dependent translation-controlling motif. In this method, translation efficiency is compared in the presence and absence of an agent that inhibits eIF4A activity in an in-vitro translation system comprising eIF4A and an mRNA having one or more eIF4A-dependent translation-controlling motifs, wherein a decrease in translation efficiency in the presence of the agent indicates the mRNA sequence possesses at least one eIF4A-dependent translation-controlling motif. By way of non-limiting example, the agent is selected from among silvestrol(methyl(1R,2R,3S,3aR,8bS)-6-[[(2S,3R,6R)-6-R1R)-1,2-dihydroxyethyl]-3-methoxy-1,4-dioxan-2-yl]oxy]-1,8b-dihydroxy-8-methoxy-3a-(4-methoxyphenyl)-3-phenyl-2,3-dihydro-1H-cyclopenta[b][1]benzofuran-2-carboxylate), pateamine A ((3S,6Z,8E,11S,15R,17S)-15-amino-3-[(1E,3E,5E)-7-(dimethylamino)-2,5-dimethylhepta-1,3,5-trienyl]-9,11,17-trimethyl-4,12-dioxa-20-thia-21-azabicyclo[16.2.1]henicosa-1(21),6,8,18-tetraene-5,13-dione), hippuristanol, (±)-CR-31-B, among other rocaglamide((1R,2R,3S,3aR,8bS)-1,8b-dihydroxy-6,8-dimethoxy-3a-(4-methoxyphenyl)-N,N-dimethyl-3-phenyl-2,3-dihydro-1H-cyclopenta[b][1]benzofuran-2-carboxamide) derivatives.
Methods are also provided for determining whether a cancer or tumor is susceptible to an agent that inhibits eIF4A activity. In one embodiment, the method comprising identifying the presence of at least one eIF4A-dependent translation-controlling motif in mRNA from the cancer or tumor, wherein the presence of the at least one eIF4A-dependent translation-controlling motif indicates susceptibility of the cancer or tumor to the agent. In other embodiments, the eIF4A-dependent translation-controlling motifs are among those described herein above. In one embodiment, the presence of MYC is not predictive of the susceptibility of a cancer or tumor to an agent that inhibits eIF4A activity.
In another embodiment, a method for determining whether a patient having cancer or a tumor will respond to treatment with an eIF4A inhibitor is provided comprising the steps of 1) obtaining a sample of the cancer or tumor from the patient; and 2) identifying the presence of at least one eIF4A-dependent translation-controlling motif in mRNA from the cancer or tumor, wherein the presence of the at least one eIF4A-dependent translation-controlling motif indicates that the patient will respond to the treatment. In the foregoing embodiments, identifying the presence of at least one eIF4A-dependent translation-controlling motif in mRNA from the cancer or tumor can be performed by comparing translation efficiency in the presence and absence of an eIF4A inhibitor agent in an in-vitro or in-vivo translation system comprising eIF4A and mRNA from the cancer or tumor, wherein a decrease in translation efficiency in the presence of the agent indicates the presence of an eIF4A-dependent translation-controlling motif in mRNA from the cancer or tumor. In another embodiment, identifying the presence of at least one eIF4A-dependent translation-controlling motif in mRNA from the cancer or tumor can be performed by identifying a G-quadruplex motif in at least one oncogene in the cancer or tumor. In certain embodiments, the motif is selected from among those described in SEQ ID NO:1 and in any one of SEQ ID NO:4-62. In certain embodiments, the expression of MYC is not correlated with responsiveness or sensitivity of a patient's cancer or tumor to an agent that inhibits eIF4A activity.
In another embodiment, a method is provided for determining whether a patient having cancer or a tumor will respond to treatment with an eIF4A inhibitor comprising the steps of 1) obtaining a sample of the cancer or tumor from the patient; and 2) identifying the presence of at least one oncogene in the cancer or tumor described in Table 3A, 3B or 3C herein, wherein the presence of said at least one oncogene indicates that the patient will respond to the treatment. In one embodiment, the presence or expression of MYC is not correlated with responsiveness or sensitivity to the treatment.
Furthermore, in other embodiments, methods to determine the level of expression of eIF4E, eIF4A, eIF4G, or eIF4B, and presence of the eIF4F complex indicate sensitivity to silvestrol and other eIF4A inhibitors, and such methods carried out in any format will be useful or determining if a tumor or patient's cancer will be sensitive to silvestrol. In another embodiment, measuring the expression of Mdr1/p-glycoprotein, a resistance marker for silvestrol, indicates the eIF4A inhibitors may be less effective and require a different dosing regimen, such as but not limited to dose level and dosing frequency. In another embodiment, expression of other helicases, e.g. DHX9 and DHX36, may causes resistance to silvestrol and thus useful in identifying cancers or tumors that may not be sensitive to silvestrol, to guide the chemotherapeutic regimen to the optimal benefit of the patient.
In one embodiment, methods are provided for 1) measuring the effect of known RNA helicases such as eIF4A, DHX9 or DHX36 on G-quadruplex unwinding; 2) investigating the effect of other cofactors/inhibitors required for eIF4A activity; 3) a screening method to identify other proteins that can unwind G-quadruplexes; and 4) identifying and establishing the effect of small molecules that stabilize the G-quadruplex structure. These methods among others may be achieved by use of a fluorescence resonance energy transfer (FRET)-based assay utilizing an oligonucleotide comprising a G-quadruplex labeled with a fluorophore at the 5′ or 3′ end of the oligonucleotide, and a fluorescence quencher at the other. In one non-limiting example, a FRET-labeled GC-quadruplex is 5′-UAGAA ACUAC GGCGG CGGCG GAAUC GUAGA (SEQ ID NO:65) and a mutant oligonucleotide without the G-quadruplex is UAGACCCUGCAACGUCAGCGUAGUCGUAGC (SEQ ID NO:66). The 5′-end is labeled with fluorophore FAM and quencher BHQ1 on the 3′end. When folded, the labeled G-quadruplex RNA oligonucleotide will exhibit minimum baseline fluorescence. Addition of specific RNA helicase such as EIF4A with ATP and/or small molecules results in unwinding and increase in fluorescence signal measured in real time. The aforementioned FRET-labeled G-quadruplex containing oligonucleotide is merely one example and those comprising other G-quadruplexes such as but not limited to SEQ ID NOS:1-64, and in particular SEQ ID NOS:1-62 may be employed for this purpose, with other fluorophores and quencher pairs well known in the art.
This assay can therefore be used for the aforementioned purpose as well as various other purposes such as but not limited to 1) measuring the effect of known RNA helicases such as eIF4A, DHX9 or DHX36 on G-quadruplex unwinding; 2) investigating the effect of other cofactors/inhibitors required for eIF4A activity; 3) a screening method to identify other proteins that can unwind G-quadruplexes; and 4) identifying and establishing the effect of small molecules that stabilize the G-quadruplex structure.
In addition to the various embodiments described above, methods are also provided for treating a subject having cancer, and for preventing cancer in a subject at risk or recurrence in a patient in remission. Based on the findings herein that translation of oncogenes comprising an eIF4A-dependent translation-controlling motifs is dependent on eIF4A helicase activity, blocking eIF4A helicase activity is a means to prevent oncogenic protein production and prevent oncogenesis. As described herein, numerous cancer-related genes including oncogenes and transcription factors are dependent on eIF4A for translation. Heretofore, the role of eIF4A was unclear but the present studies show, inter alia, that specific motifs on oncogenic mRNAs depend on eIF4A for translation, thus blocking eIF4A helicase is a heretofore unappreciated anti-cancer mechanism. Use of agents that target eIF4A dependent translation can thus stop translation of oncogenic mRNA sequences.
In further embodiments, methods are provided for reducing or preventing recurrence of cancer in a patient in remission or otherwise considered cured. In these embodiments, the cancer is any among those described herein among others, and by way of non-limiting examples, T-cell acute lymphoblastic leukemia, small cell lung cancer, renal cell carcinoma, squamous cell carcinoma of the head and neck, neuroblastoma and pancreatic cancer. In other embodiments, the cancer is transformed follicular lymphoma, mantel cell lymphoma, breast cancer, ovarian cancer, hepatocellular carcinoma, and non-small cell lung cancer, as well as gastric cancer, Ewing sarcoma and lung adenocarcinoma. In one embodiment the subject has cancer. Other cancers are described in
Among these methods, administering to the subject an agent that blocks eIF4a helicase activity prevents, treats or intervenes in the recurrence of the cancer. In one embodiment, a method for preventing, treating or intervening in the recurrence of a cancer in a subject is provided. The method comprises administering to the subject an agent that blocks eIF4a helicase activity, thereby preventing, treating or intervening in the recurrence of the cancer. In one embodiment, the agent that blocks eIF4A helicase inhibits the translation of an oncogenic mRNA. In one embodiment, the oncogenic mRNA comprises an eIF4A-dependent translation-controlling motif. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the oncogenic mRNA comprises a G-quadruplex motif. In one embodiment, the oncogenic mRNA is from an oncogene, which by way of non-limiting example is selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In one embodiment, a method is provided for preventing, treating or intervening in the recurrence of a cancer in a subject having an eIF4A dependent cancer. The method comprises administering to the subject an agent that blocks eIF4a helicase activity, thereby preventing, treating or intervening in the recurrence of the cancer. In one embodiment, the agent that blocks eIF4A helicase inhibits the translation of an oncogenic mRNA. In one embodiment, the oncogenic mRNA comprises an eIF4A-dependent translation-controlling motif. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the oncogenic mRNA comprises a G-quadruplex motif. In one embodiment, the oncogenic mRNA is from an oncogene. In one embodiment, the oncogene is selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In another embodiment, a method is provided for inhibiting in a subject the translation of an oncogene that comprises an eIF4A-dependent translation-controlling motif. The method comprises administering to the subject an agent that blocks eIF4a helicase, thereby inhibiting translation of the oncogene. In one embodiment, translation of the oncogene causes cancer in the subject. In another embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In this embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the mRNA of the oncogene comprises a G-quadruplex motif. In one embodiment, the oncogene is selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In one embodiment, a method for inhibiting in a subject eIF4A dependent mRNA translation is provided. The method comprises administering to the subject an agent that blocks eIF4a helicase, thereby inhibiting mRNA translation. In one embodiment, the mRNA translation causes cancer in the subject. In one embodiment, the mRNA comprises an eIF4A-dependent translation-controlling motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the mRNA encodes an oncogenic protein. In one embodiment, the oncogenic protein is encoded by an oncogene selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2.
In one embodiment, a method for preventing in a subject the translation of an mRNA comprising an eIF4A-dependent translation-controlling motif. The method comprises administering to the subject an agent that blocks eIF4a helicase activity, thereby inhibiting translation of the mRNA. In one embodiment, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In one embodiment, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62. In one embodiment, the mRNA is from an oncogene selected from among Tables 3A, 3B and 3C. In one embodiment, the oncogene is NOTCH1, BCL11B, MYC, CDK6, RUNX1, BCL2 or MDM2. In one embodiment, the translation of the mRNA causes cancer.
In any of these embodiments, the agent blocks the activity of eIF4A helicase. In any of the foregoing embodiments, the agent blocks the translation of an mRNA comprising an eIF4A-dependent translation-controlling motif. In any of the foregoing embodiments, the eIF4A-dependent translation-controlling motif is a G-quadruplex motif. In any of the foregoing embodiments, the eIF4A-dependent translation-controlling motif is selected from among SEQ ID NOs:1-62.
Non-limiting examples of aforementioned agents include a rocaglamide, such as silvestrol, CR-31-B, or any active analogue or derivative thereof. In other embodiments, the agent is hippuristanol, pateamine A, or any active analogue or derivative thereof. Other examples of suitable agents include those described in WO2011/140334 (based on PCT/US2011/035351).
Hallmark features are described here of eIF4A-dependent translation and defines specific 5′UTR elements that confer a requirement for that RNA helicase. The key features are longer 5′UTRs, a 12-mer (GGC)4 motif, and related 9-mer variant motifs. Importantly, the 12-mer and 9-mer motifs precisely localize to between 53% and 65% of all predicted RNA G-quadruplex structures (depending on the analysis tool). The 9-mer sequences require neighboring nucleotides to complete the structure as the minimal number is 12 nucleotides, and it was frequently observed that more than 12 nucleotides contribute to the G-quadruplex. Moreover, most of the remaining G-quadruplexes are based on highly similar sequence elements. On the other hand IRES mRNAs are somewhat protected, while TOP, TOP-like, or PRTE elements do not appear to influence the eIF4A requirement. This is distinct from mTORC1 inhibition, which affects a different set of transcripts marked by TOP and TOP-like elements (Thoreen et al., 2012). These findings identify sequence motifs that represent translational control elements encoded in the 5′UTR of several hundred transcripts and that confer a requirement for eIF4A RNA helicase action.
RNA G-quadruplex structures are typically made from at least two stacks of four guanosines exhibiting non-Watson-Crick interactions (e.g. hydrogen bonds) and connected by one or more linker nucleotides (reviewed in (Bugaut and Balasubramanian, 2012)). In the examples herein, the linker is most often a cytosine and less frequently an adenosine. There is variation in the exact structural composition and sequence requirement as our examples illustrate. The minimum requirement for the structure is a (GGC/A)4 sequence and neighboring nucleotides can complete the structure.
The cap-binding protein eIF4E is limiting for cap-dependent translation and its signaling control by mTORC1 and 4E-BP has been studied in great detail (Jackson et al., 2010). The results described here indicate that for a set of mRNAs the eIF4A helicase activity is required and represents the point of attack for three natural compounds, silvestrol, hippuristanol, and pateamine (Cencic et al., 2007). Moving forward, an intriguing question concerns the physiological control of eIF4A activity (Parsyan et al., 2011). In this regard, recent studies have defined the mechanics of eIF4A action (Marintchev, 2013; Marintchev et al., 2009), identified mutually exclusive potentially regulatory interactions between eIF4A and the eIF4B, eIF4G, and eIF4H factors (Rozovsky et al., 2008), and further implicated S6 kinase in the phosphorylation and signaling control of eIF4B (Kroczynska, 2009; Shahbazian et al., 2010; Shahbazian et al., 2006). The data herein indicate that these interactions define a broadly relevant layer of translational control that is distinct from the control of eIF4E by 4E-BP and mTORC1, and that is specifically aimed at a subset of transcripts.
In one embodiment, the novel sequence motifs and/or G-quadruplex structures are present in a large number of transcription factors, several known oncogenes, but also some tumor suppressor genes. A number of examples are listed and suggest that an eIF4A dependent program of translational control may have broad ramification on a cell's biology. Several genetic lesions implicated in translational activation can promote T-ALL development (e.g. PTEN, IL7R) (Palomero et al., 2007; Zenatti et al., 2011; Zhang et al., 2012).
Ribosome Footprinting. KOPTK1 cells were treated with silvestrol or DMSO for 45 minutes, followed by cycloheximide treatment for 10 minutes and then harvested for total RNA and ribosome footprint fragment isolation. Total RNA was isolated using RNA isolation kit from Qiagen (74104) and subjected to RNA sequencing. Ribosome protected fragments were isolated following published protocol (Ingolia et al., 2009). Briefly cell lysates were subjected to ribosome footprinting by nuclease treatment. Footprint fragments were purified by one step sucrose cushion and gel extraction. Deep sequencing libraries were generated from these fragments. Both total RNA and footprint fragment libraries were analyzed by sequencing on the HiSeq 2000 platform.
Sequence Alignment. Sequences were aligned to the transcripts available from the human genome sequence hg19 from UCSC public database. Ribosome footprint (RF) reads were aligned to reference genome hg19 using PALMapper (Jean et al., 2010). Only the uniquely aligned reads were used for analysis. Read length of 25- to 35-bp was selected and used to analyze the translation effect of silvestrol. Total mRNA sequencing reads were aligned to the hg19 reference using STAR (Dobin et al., 2013). The splice alignment was used, and only used the uniquely aligned reads with maximum 3 mismatches.
Footprint Profile Analysis. The genome annotation was from GENCODE project (http://www.gencodegenes.org/releases/14.html). Ribosome footprint intensity (reads per million, RPM) and the expression value (reads per kilobase per million, RPKM) were measured from total mRNA-seq data and translation values were measured from ribosome footprint data. To evaluate the translation efficiency (TE) change between silvestrol- and vehicle-treated samples, TE was calculated as RPKMfootprint/RPKMmRNA (as Thoreen et al. did recently (Thoreen et al., 2012)). Changes in ribosome footprint profiles were determined by using DEXSeq algorithm (Anders et al., 2012). DEXSeq accounts for the discrete nature of the read counts and models biological variability to avoid false positives. Ratio of TEsilvestrol/TEcontrol of all the genes was plotted and color-highlighted according to the statistical significance of DEXSeq test.
Ribosome distribution analysis. The ribosomal distribution change was evaluated between silvestrol treated samples and controls. A BED file containing all non-overlapped exonic regions was generated based on genome annotation. Then the BED file and footprint BAM files were given as an input to SAMTOOLS (Li et al., 2009a) to generate new BAM files that only included exonic alignment. The exonic BAM files were input for two conditions to rDiff (Drewe et al., 2013) to identify genes that presented significant change in ribosomal distribution.
(Non-radioactive) Metabolic labeling of nascent protein. KOPTK1 cells were labeled for nascent protein synthesis using Click-iTR AHA (L-azidohomoalanine) metabolic labeling reagent obtained from Invitrogen (cat no. C10102) as per manufacturer's instructions. Briefly, following silvestrol, Cycloheximide or DMSO treated cells were incubated in methionine free medium for 30 min prior to AHA labeling for 1 hr. Cells were fixed with 4% paraformaldehyde in PBS for 15 min, permeablized with 0.25% Triton X-100 in PBS for 15 min followed by one wash with 3% BSA. Cells were then stained using Alexa Fluor 488 Alkyne (Invitrogen cat no. A10267) with Click-iT Cell reaction Buffer Kit (Invitrogen cat no. C10269). Changes in mean fluorescence intensity as a measure of newly synthesized protein was detected by Flow cytometry analysis.
Polysome profiling. KOPTK1 cells were treated with silvestrol or DMSO for 45 minutes, followed by cycloheximide treatment for 10 minutes. Cell pellet was lysed in polysome lysis buffer (300 mM NaCl, 15 mM Tris-HCl (pH 7.5), 15 mM MgCl2, 1% TritonX-100, 0.1 mg/ml Cycloheximide, 1 mg/ml Heparin). Polysome fractions were isolated using 4 ml 10-50% sucrose density gradients (300 mM NaCl, 100 mM MgCl2, 15 mM Tris-HCl (pH 7.5), 1 mg/ml Cycloheximide, 10 mg/ml Heparin). Gradients were centrifuged in an SW40Ti rotor at 35,000 rpm for 2 hrs. Fractions of 100 ul were collected manually from the top, and optical density (OD) at 254 nM was measured.
Sequence Alignment. The human genome sequence hg19 was downloaded from UCSC public database: http://hgdownload.cse.ucsc.edu/goldenPath/hg19/chromosomes. Ribosome footprint (RF) reads were aligned to reference genome hg19 using PALMapper (Jean et al., 2010). PALMapper clips the linker sequence (5′-CTGTAGGCACCATCAAT-3′), which is technically introduced during RF library construction, and trims the remaining sequence from the 3′ end while aligning the reads to reference sequence. Briefly, the parameters for PALMapper were set as follows: maximum number of mismatches: 2; maximum number of gaps: 0; minimum aligning length: 15; maximum intron length (splice alignment): 10000; minimum length of a splicing read aligned to either side of the intron boundary: 10. Only the uniquely aligned reads were used for further analysis.
To remove ribosome RNA contamination, the footprint reads were also aligned to a ribosome sequence database using PALMapper with the same parameters except allowing splice alignment. The human ribosome sequences were retrieved from BioMart Ensembl (Flicek et al., 2013) and SILVA (Quast et al., 2013) databases and merged the results into a single FASTA file, which was used as reference sequence to align against. The rRNA-aligned reads were filtered out from hg19-aligned reads.
After removing the rRNA contamination, a portion of reads were observed that were dominated by linker sequence and Illumina P7 adapter. These reads can also be trimmed during mapping and cause false alignment. Therefore, a search was undertaken for a string of 1˜8 nt from linker sequence around the trimming site (±2 bp) allowing 1 nt mismatch. The read was removed if there was no such linker sequence. Finally, reads ≦24-bp and ≧36-bp were filtered out, and the remaining reads with aligned length from 25- to 35-bp were used to analyze the translational effects of silvestrol.
Total mRNA sequencing reads were aligned to the hg19 reference using STAR (Dobin et al., 2013). The splice alignment was performed and only use the uniquely aligned reads with maximum 3 mismatches. rRNA contaminating reads were also filtered out using the same strategy described before.
Footprint Profile Analysis. For each gene, only the number of aligned reads were counted that were mapped within exonic regions. The genome annotation was downloaded from GENCODE project (http://www.gencodegenes.org/releases/14.html). Ribosome footprint intensity (reads per million, RPM) was calculated as RPM=Ci/(N/106), where Ci is the read count for gene i, and N is the library size of silvestrol- or vehicle-treated samples. In order to eliminate the effluence of rRNA contamination, the library size was calculated after read filtering described previously. Similarly, the expression value measured from total mRNA-seq data and translation value measured from ribosome footprint data (both were referred as reads per kilobase per million, RPKM) were calculated as RPKM=Ci/(Ki·N/106), where Ki is the non-overlapped exonic region of each gene. To evaluate the translation efficiency (TE) change between silvestrol- and vehicle-treated samples, TE=RPKMfootprint/RPKMmRNA was calculated as Thoreen et al did recently (Thoreen et al., 2012).
To detect the genes that ribosome footprint profiles were significantly changed between silvestrol treated sample and control, DEXSeq (Anders et al., 2012) was used to perform the statistical test. DEXSeq accounts for the discrete nature of the read counts and it also models the biological variability which has been demonstrated in other applications to be crucial to avoid a great number of false positives. Here, DEXSeq was used in a specific way: the footprint and mRNA-seq read counts were fit into DEXseq framework, in which silvestrol treatment and control are two biological conditions, and then tested whether footprint (consisting 2 replicates for each condition) and mRNA-seq (The 3 replicates were split and recombined into two combinations such that each of them consists of two replicates) read counts were significantly different in the two conditions. The log-ratio of normalized read counts of silvestrol treated sample to control indicated whether ribosome footprint profile was increased or decreased. In the end, the ratio of TEsilvestrol/TEcontrol of all the genes was plotted, and color-highlighted them according to the statistical significance of the DEXSeq test.
In addition to studying the translation efficiency, the ribosomal distribution change was also evaluated between silvestrol treated sample and control. First, a BED file contained all non-overlapped exonic regions was generated based on genome annotation. Then the BED file and footprint BAM files were given as an input to SAMTOOLS (Li et al., 2009) to generate new BAM files only included exonic alignment. The exonic BAM files of two conditions to rDiff (Drewe et al., 2013) were input to identify genes that presented significant change in ribosomal distribution. In detail, a nonparametric test was performed implemented in rDiff to detect differential read densities. rDiff takes relevant read information, such as the mapping location and the read structure, to measure the significance of changes in the read density within a given gene between two conditions. The minimal read length was set to 25-bp, and number of permutation was set to 10000.
To plot the ribosomal distribution curves for multiple genes, read coverage of each transcript was normalized by the mean coverage value of that particular transcript. Then the UTR and coding exon length were normalized in proportion to the overall average length of corresponding regions of a group of genes. Finally all the normalized transcripts were averaged together in a vectorized way to plot the coverage distribution. The ribosomal distribution curves for a single gene were plotted in a similar way but without normalizing the read coverage, and the coverage was smoothed using ‘moving average’ smoothing algorithm.
Motif analysis. The transcripts of each gene were quantified based on the total mRNA-seq data using MISO (Katz et al., 2010). The 5′UTR of most abundant transcript was collected for predicting motifs. Both the significant genes with increased or decreased TE and altered ribosomal distribution and the corresponding background gene sets were predicted by DREME (Bailey, 2011). Over- and under-represented motifs were determined with three different settings: searching for motifs of length greater than or equal to six, nine and twelve base pairs. The predicted consensus sequences with P<1×10-4 were considered as significant motifs. The secondary structure of different gene sets was predicted using RNAfold (Hofacker, 2003) based on the same 5′UTR prepared before.
5′UTR sequences for respective group of targets were subjected to motif prediction using online available program RegRNA (A Regulatory RNA motifs and Elements Finder) (http://regrna.mbc.nctu.edu.tw/html/prediction.html) and looked specifically for motifs that occur in 5′UTR. Statistical significance for the results obtained was calculated using Fisher's exact test for count data.
T-ALL samples. Thirty-six bone marrow biopsies were collected from patients with T-ALL at multiple organizations (Universitair Ziekenhuis (UZ) Ghent, Ghent, Belgium; UZ Leuven, Leuven, Belgium; Hôpital Purpan, Toulouse, France; Centre Hospitalier Universitaire (CHU) de Nancy-Brabois, Vandoeuvre-Les-Nancy, France). The QIAamp DNA Mini kit was used to obtain genomic DNA (Qiagen 51304). The Medical Ethical Commission of Ghent University Hospital (Ghent, Belgium, B67020084745) approved this study.
Mutation analysis. NOTCH1 (exons 26, 27, 28 and 34), FBXW7 (exons 7, 8, 9, 10 and 11), PTEN (exons 1 till 9) and IL7R (exon 6) were amplified and sequenced using primers as reported in (Mavrakis et al., 2011; Shochat et al., 2011; Zuurbier et al., 2012). FBXW7, PTEN and IL7R amplification were performed using 20 ng of genomic DNA, 1×KapaTaq reaction buffer (KapaBiosystems), 1U KapaTaq DNA polymerase, 0.2 mM dNTP, 2.5 μM MgCl2, 0.2 mM forward and reverse primer in a 25 μl PCR reaction. For NOTCH1 amplification, the PCRx enhancer system (Invitrogen) was used for the PCR reaction. Reactions contained 20 ng of DNA, 2.5 U KapaTaq DNA Polymerase, 1×PCRx Amplification Buffer, 2×PCRx Enhancer Solution, 0.2 mM dNTP, 1.5 mM MgSO4 and 0.2 mM of each primer. The PCR steps were: 95° C. for 10 minutes, (96° C. for 15 sec, 57° C. for 1 minute, then 72° C. for 1 min) for 40 cycles, then 72° C. for 10 minutes. Purified PCR products were analyzed using the Applied Biosystems 3730XL DNA Analyze.
Array Complete Genomic Hybridization. PTEN deletions and MYC amplifications were detected by array CGH analysis using SurePrint G3 Human 4×180K CGH Microarrays (Agilent Technologies). First, random prime labeling of the T-ALL DNA sample and a control human reference DNA was performed with Cy3 and Cy5 dyes (Perkin Elmer), respectively. The subsequent hybridization protocol was performed according to the manufacturer's instructions (Agilent Technologies). The data was analyzed using arrayCGHbase (Menten et al., 2005).
Immunohistochemistry and Tissue Microarrays. T-cell acute lymphoblastic leukemia tissue microarrays were made as previously published (Schatz et al., 2011) using an automated tissue arrayer (Beecher Instruments, ATA-27). T-ALL samples were ascertained at Memorial Sloan-Kettering Cancer Center and were approved with an Institutional Review Board Waiver and approval of the Human Biospecimen Utilization Committee. All cancer biopsies were evaluated at MSKCC, and the histological diagnoses were based on haematoxylin and eosin (H&E) staining. TMAs were stained with the c-MYC polyclonal antibody (Epitomics 51242) using Discovery XT (Ventana) for 1 hour and a secondary anti-rabbit antibody (Vector Laboratories) for 1 hour. Histological images were captured using a Zeiss Axiocam MRc through a Zeiss Achropla lens on an Axioskop 40 microscope. Images were processed for brightness and contrast using Axiovision Rel. 4.6. Cores were scored as 0, 1, or 2 reflecting the fraction of positive cells.
Generation of mice. The ICN-driven mouse T-ALL model has been reported (Pear et al., 1996; Wendel et al., 2004). Data were analyzed in Kaplan-Meier format using the log-rank (Mantel-Cox) test for statistical significance. The surface marker analysis was as described (Wendel et al., 2004). ShRNAs against Pten and Fbxw7 have been reported in (Mavrakis et al., 2011).
Tumor transplantation. Leukemic bone marrow from mice expressing the ICN and IK6 was infected with OMOMYC and selected using puromycin. 2,000,000 cells were injected into syngeneic recipients via tail vein. Mice were monitored by blood analysis. Upon leukemia detection, tamoxifen (50 mg/kg) or vehicle treatment was performed on alternating days until mice were moribund. Severe leukemia reflects >100,000 blasts/μl and led to rapid demise of animals if untreated, whereas complete remission was defined as the absence of GFP positive leukemic blasts in the blood and bone marrow.
Real-Time Quantitative PCR. Total RNA was extracted using AllPrep DNA/RNA/Protein Mini Kit (Qiagen 80004). Normal CD3+ T-cell RNA mixed from healthy donors was purchased from Miltenyi Biotec (130-093-164). cDNA was made using SuperScript III First-Strand (Invitrogen 18080-400). Analysis was performed by ΔΔCt. Applied Biosystems Taqman GeneExpression Assays: human Myc Hs00153408_m1, hsa-miR-19b RT and TM 396, Rnu6b RT and TM 001093, and mouse Myc Mm00487804_m1.
T-ALL cell lines. T-ALL cell lines were cultured in RPMI-1640 (Invitrogen, CA), 20% fetal calf serum, 1% penicillin/streptomycin, and 1% L-glutamine. The MOHITO line was supplemented with 5 ng/mL IL2 (Fitzgerald 30R-A1022 and 10 ng/mL of IL7 (Fitzgerald 30R-AI084X).
Immunoblots. Lysates were made using Laemli lysis buffer. 30 ug of protein was loaded onto SDS-PAGE gels then transferred onto Immobilon-FL Transfer Membranes (Millipore IPFL00010). The antibodies used were α-Tubulin (Sigma T5168), β-actin (Sigma A5316), Myc (Santa Cruz Biotechnology sc-40), p-Akt 308 (Cell Signaling 9275), Akt (Cell Signaling 9272), S6 (Cell Signaling 2317), and p-S6 (Cell Signaling 2215), Notch1 (Cell signaling 3608), Myb (Santa Cruz Biotechnology, sc-517), CDK6 (Cell Signaling 3136), EZH2 (Cell Signaling 5246), Mdm2 (Santa Cruz Biotechnology, sc-965), Bcl2 (Santa Cruz Biotechnology, sc-509), Run×1 (Cell Signaling 4336), and GAPDH (Cell Signaling 5174).
Luciferase assays. Four tandem repeats of the (CGG)4 12-mer motif (GQs) or random sequence matched for length and GC content (random) were cloned into the 5′UTR of Renilla luciferase plasmid pGL4.73. Empty firefly luciferase plasmid pGL4.13 or HCV-IRES firefly were used as internal controls. Luciferase assays were performed using Dual-Luciferase Reporter Assay System (Promega E1960) following the manufacturer's instructions. GQs sequence:
Random Sequence:
Statistical analysis. All Kaplan-Meier curves were analyzed using the Mantel-Cox test. The significance of xenografted tumor size differences was determined using two-way repeated measures ANOVA tests. RT-PCRs were analyzed with two tailed t-tests.
Xenografts. 5,000,000 KOPT-K1 cells in 30% matrigel (BD 354234) were injected subcutaneously into C.B-17 scid mice. When tumors were readily visible, the mice were injected on 7 consecutive days with either 0.5 mg/kg silvestrol, 0.2 mg/kg (±)-CR-31-B, or every other day with 1 mg tamoxifen. Tumor size was measured daily by caliper. P-values were calculated using 2-way repeated measures ANOVA.
Silvestrol and (±)-CR-31-B. Each was suspended in DMSO for in vitro experiments and 5.2% Tween 80 5.2% PEG 400 for in vivo experiments. Cycloheximide (C7698) and Rapamycin (R8781) were purchased from Sigma.
Toxicity studies. Eight week-old C57Bl/6NTac female mice were randomly assigned to either control or treatment groups. Each treatment group received one daily dose of test article through i.p. injection over 5 consecutive days. Toxicity was monitored by weight loss and daily clinical observation for the 14 days following test article administration. 24 hours after the last test article administration 4 mice in each group were sacrificed and clinical chemistry, hematology and tissue specific histopathology were done at autopsy. The remaining mice (n=2 per group) were kept under observation for an additional 13 days; at that point all mice were sacrificed and clinical chemistry, hematology and tissue specific histopathology were done at time of autopsy.
References for Materials and Methods:
NOTCH-driven T-ALL exemplifies the frequent activation of AKT/mTORC1 and cap-dependent translation seen in cancer. For example, in a small series of pediatric T-ALLs the common NOTCH1 HD and PEST domain mutations were confirmed (56%; 20/36 samples) (O'Neil et al., 2007; Weng et al., 2006), PTEN mutations (14%; 5/36), and PTEN deletions (11%; 4/36), resulting in mono- (16%) or bi-allelic (6%) PTEN loss (Gutierrez et al., 2009; Palomero et al., 2007; Zhang et al., 2012), and occasional IL7R mutation (3%) (Zenatti et al., 2011) (
These mutations contribute to T-cell leukemogenesis. Briefly, murine hematopoietic precursor cells (HPCs) expressing Notchl intracellular fragment (ICN) alone or in combination with additional alleles were transplanted and disease latency measured in recipient animals (
A genetic approach was then used to test the requirement for eIF4E in maintaining the leukemic cells. Briefly, the 4E-binding protein (4E-BP) sequesters eIF4E and blocks cap-dependent translation (Rousseau et al., 1996). 4E-BP is negatively regulated by sequential phosphorylation at several serine residues by mTORC1, and mutation of these sites results in a constitutively active 4E-BP1 (4E-BP1(4A)) allele (Rong et al., 2008). In mixed populations of murine T-ALL cells where a fraction of cells express either 4E-BP1(4A) and GFP or an empty vector and compete with un-transduced parental cells, rapid elimination was seen of 4E-BP1(4A)/GFP expressing cells from the culture (FIG. 1D/E). Hence, eIF4E activity is required to maintain T-ALL, which indicates that targeting translation might be a therapeutic strategy.
Based on this genetic evidence a pharmacological inhibitor was then tested. Silvestrol is perhaps the best-characterized inhibitor of the eIF4F complex, it does not target eIF4E and instead blocks the eIF4A RNA helicase by stabilizing its mRNA bound form (Bordeleau et al., 2008; Cencic, 2009). Silvestrol, and a synthetic rocaglamide analogue (±)-CR-31-B (CR) bind the same site on eIF4A (Rodrigo et al., 2012; Sadlish et al., 2013). In a dual-luciferase reporter assay, where renilla and firefly luciferase are either capped or under control of an internal ribosomal entry site (IRES) element, both drugs were confirmed to preferentially block cap-dependent over IRES-dependent translation (Bordeleau et al., 2006) (
Silvestrol has excellent single-agent activity against T-ALL in vitro and in vivo. Silvestrol was tested against primary human T-ALL samples in vitro and observed efficient apoptosis induction with IC50 values ranging from 3 to 13 nM; and confirmed activity in established cell lines (
Silvestrol acts in a manner that is distinct from mTORC1 inhibitors. For example, the predominant inhibition of S6 kinase instead of 4E-BP and feedback activation of AKT (S308 phosphorylation) are thought to hinder the therapeutic effect of rapamycin (
Next, use of the recently developed ribosome footprinting technology (Ingolia et al., 2009) was employed to measure precisely how silvestrol affected protein translation. Briefly, KOPT-K1 cells were treated with 25 nM of silvestrol or vehicle, cells collected after 45 minutes, then isolated and deep-sequenced total RNA and ribosome footprints (RFs) prepared (
Silvestrol produced an immediate and broad inhibitory effect on cap-dependent translation. RF reads were fewer in number and showed a wider variation between control and silvestrol than total RNA sequences indicating minimal transcriptional variation (
Silvestrol affected the translational efficiency of specific sets of mRNAs. To calculate the translational efficiency (TE) for each mRNA the RF frequency was normalized to the length of the corresponding mRNA yielding an RF density (expressed as RPKM: reads per kilobase per million reads), and was corrected for total mRNA expression. Overall RPKM values for RF from vehicle and silvestrol treated samples were significantly correlated (R2=0.94) indicating a broad inhibitory effect on translation (
5′UTR length has been implicated in translational control (Hay and Sonenberg, 2004), although a recent study found no effects of UTR length on mTORC1-dependent translation (Thoreen et al., 2012). Comparing the 5′UTR length across TE up, TE down, and background groups (as described in U.S. application Ser. No. 61/912,420, filed Dec. 5, 2013; and Wolfe et al., Nature. 2014 Sep. 4; 513(7516):65-70), it was observed that mRNAs with longer 5′UTRs were significantly enriched among the most silvestrol-sensitive mRNAs (TE down: mean UTR length 368 nucleotides; background: mean 250 nucleotides; p(Silvestrol vs. Control)=7.6×10−12 using two-sample Kolmogorov-Smirnov) (
Known translation regulatory elements were sought. For example, TOP sequences (cytidine in pos. 2 followed by 4-14 pyrimidines) (Meyuhas, 2000), TOP-like sequences (cytidine in pos. 1-4 and >5 pyrimidines) (Thoreen et al., 2012), internal ribosome entry sites (IRES) (Pelletier and Sonenberg, 1988), and pyrimidine rich translational elements (PRTEs) (Meyuhas, 2000). Comparing TE down and the background lists no predilection was found for TOP, TOP-like, PRTE, or IRES elements (
Next it was sought to identify a sequence motif that might confer eIF4A dependence. The DREME algorithm was used to look for significantly enriched sequences in the TE down and TE up groups compared to the background list (as described in U.S. application Ser. No. 61/912,420, filed Dec. 5, 2013; and Wolfe et al., Nature. 2014 Sep. 4; 513(7516):65-70) (Bailey, 2011). No motif was found in the TE up group of mRNAs. However, the analysis revealed a 12-mer (GGC)4 motif that was significantly over represented among the TE down transcripts and present in 94 out of 220 genes (p<2.2×10-16) (
Whether silvestrol-sensitive mRNAs might have specific structural features that set them apart from less affected transcripts was considered. Using the program RNAfold (http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.cgi) the background, TE up, and TE down genes were modeled and a striking enrichment was observed for G-quadruplex structures among the TE down genes (p=2×10-11) (
Next, the distribution of ribosomes was examined along the transcript as this might provide an additional indication of eIF4A sensitive translation (
Similar to the TE down group an enrichment of longer 5′UTR in the rDiff positive set was found (rDiff pos.: n=641; mean length 271 nucleotides; Background (rDiff negative=no significant change): n=976, mean UTR length: 230 nucleotides; p=0.004) (Figure S4D). No significant enrichment for TOP, PRTE, or IRES elements was detected, however there was a small and significant drop in TOP-like sequences (Figure S4E). The DREME analysis for sequence motifs identified a significant enrichment for a 12-mer and three 9-mer motifs among rDiff positive genes (p=2.2×10-16) (FIG. 4C/D, Table 4, Table 7A/B). Among 641 genes in the rDiff group, the 12-mer motif occurred in 232, and an additional three 9-mer motifs were found in 322 genes. Notably, the motifs were nearly identical to the TE down motif (
Next, directly testing the translational effect of the 12-mer sequence motif was sought. Briefly, a luciferase reporter system was constructed to directly compare four 12-mer motifs in tandem reflecting the common occurrence of multiple motifs in sensitive mRNAs (GQ construct) to a random sequence of equal length and GC content (control construct) and using an IRES-driven firefly luciferase as an internal control (
The most silvestrol sensitive transcripts in the TE down group and the rDiff positive set include many genes with known roles in T-ALL (FIG. 5A/B). Categorization by gene ontology reveals a preponderance of transcription factors, many oncogenes, but also potential tumor suppressors (Figure S5A/5B). Sub-grouping of TE down genes by 5′UTR features (12-mer, 9-mer motif, and G-quadruplex structures) illustrates how sometimes multiple features occur in the same transcripts (Figure S5C-E). Exploring individual RF distribution graphs (normalized for mean RF count and gene length) illustrates recurrent patterns and also variations. For example, the c-MYC transcript (TE: p=1.3×10-4; rDiff: p=3×10-8) harbors six 9-mer motifs in its 5′UTR which correspond to peaks in RF density (
Given the complexity of the RF data analysis, it was important to directly confirm loss of expression for at least some of these proteins. Briefly, immunoblots on JURKAT and KOPT-K1 cells treated with silvestrol (25 nM) and loaded with equal amounts of total protein confirmed dramatic loss of MYC, NOTCH1, BCL2, and CCND3 proteins (
Genomic studies have implicated many silvestrol-sensitive genes in T-ALL and other cancers. For example oncogenic mutations of NOTCH (Weng et al., 2004), increased CDK6/CCND3 (Sawai et al., 2012), and amplifications of MYB (Lahortiga et al., 2007) have been reported in T-ALL. Similarly, a brief survey of mRNA expression using RNAseq on 9 primary T-ALL samples compared to 4 T-cell samples confirms increased expression of NOTCH, MYB, CDK6, and BCL2 in T-ALL (
Given the pleiotropic effects of eIF4A inhibition it was considered which of its target genes may account for the drug's anti-leukemia effect. The MYC oncogene is a first candidate, because of silvestrol's powerful effects on MYC levels and its known oncogenic role in this cancer (Gutierrez et al., 2011a; Palomero et al., 2006). Moreover, genetic MYC blockade using the tamoxifen-inducible OmomycER allele (Soucek et al., 2008) readily induces cell death and clears T-ALL cells from the marrow leading to an extended survival in leukemic animals (nOMO=9, ncontrol=10; p=0.002) (
A FRET-based assay was set up for measuring the effect of RNA helicases on G-quadruplex unwinding, screening proteins that can unwind G-quadruplexes and identify small molecules that stabilize the G-quadruplex structure. An RNA oligonucleotide (1XTEDownMotif 5′-UAGAA ACUAC GGCGG CGGCG GAAUC GUAGA; SEQ ID NO:65) containing the G-quadruplex motif was labeled with fluorophore FAM on the 5′ end and quencher BHQ1 on the 3′end. When folded, the labeled GQ RNA oligonucleotide will exhibit minimum baseline fluorescence. Addition of specific RNA helicase such as EIF4A with ATP and/or small molecules would result in unwinding and increase in fluorescence signal measured in real time, as shown in
Fluorescence measured as function of concentration using a mutant RNA (1XMutant; 5′-UAGACCCUGCAACGUCAGCGUAGUCGUAGC; SEQ ID NO:66) with or without KCl is shown in
In
This assay can therefore be used for the aforementioned purpose as well as various other purposes such as but not limited to 1) measuring the effect of known RNA helicases such as eIF4A, DHX9 or DHX36 on G-quadruplex unwinding; 2) investigating the effect of other cofactors/inhibitors required for eIF4A activity; 3) a screening method to identify other proteins that can unwind G-quadruplexes; and 4) identifying and establishing the effect of small molecules that stabilize the G-quadruplex structure.
The IC50 of silvestrol in several small cell lung cancer lines was evaluated. As shown in
A range of sensitivities from renal carcinoma lines ACHN, A498, CAKI-1, CAKI-2 to 786-O was demonstrated, as shown in
In addition, IC50s of 2 to 20 nM have been obtained with neuroblastoma cell lines SKNAS, CLBGA, IMR32 and N206. Pancreatic cancer line PANC-1 show sensitivity to 20 nM silvestrol and a loss of KRAS expression.
In addition to the renal cell carcinoma and small cell lung cancer lines mentioned above, about 60 cancer cell lines were evaluated for silvestrol sensitivity as shown in
Using the dual-luciferase reporter assay described above, where renilla and firefly luciferase are either capped or under control of an internal ribosomal entry site (IRES) element, both hippuristanol and pateamine A were shown to preferentially block cap-dependent over IRES-dependent translation (
The in vitro activity of silvestrol was evaluated against a panel of small cell lung cancer cell lines, shown in
The effect of silvestrol on several neuroblastoma cell lines was evaluated. The results are shown in the table below.
As shown in the four annotated transcripts of KRAS in
Similarly to that described in Example 8, silvestrol at 50 nM shows a loss of KRAS expression (
Table 7A-C. Motifs and G-quadruplexes in rDiff positive genes.
This application claims priority to International application PCT/US2014/68875, filed Dec. 5, 2014, which in turn claims priority to U.S. Patent Application Ser. No. 61/912,420, filed Dec. 5, 2013, both of which are incorporated herein by reference in their entireties.
This research was supported by funding from the National Cancer Institute Grants R01-CA142798-01 and U01CA105492-08, and National Institutes of Health Grant GM-073855. The U.S. Government has certain rights in the invention.
Number | Date | Country | |
---|---|---|---|
61912420 | Dec 2013 | US |
Number | Date | Country | |
---|---|---|---|
Parent | PCT/US14/68875 | Dec 2014 | US |
Child | 14693832 | US |