Sudemycins, novel small molecule analogues of FR901464, induce alternative gene splicing

Abstract

Two unrelated bacterial natural products, FR901464 and pladienolide B, have previously been shown to have significant antitumor activity in vivo. These compounds target the SF3b subunit of the spliceosome, with a derivative of pladienolide (E7107) entering clinical trials for cancer. However, due to the structural complexity of these molecules, their research and development has been significantly constrained. We have generated a set of novel analogues (Sudemycins) that possess the pharmacophore that is common to FR901464 and pladienolide, via a flexible enantioselective route, which allows for the production of gram quantities of drug. These compounds demonstrate cytotoxicity toward human tumor cell lines in culture and exhibit antitumor activity in a xenograft model. Here, we present evidence that Sudemycins are potent modulators of alternative splicing in human cells, both of endogenous genes and from minigene constructs. Furthermore, levels of alternative splicing are increased in tumor cells relative to normal cells, and these modifications can be observed in human tumor xenografts in vivo following exposure of animals to the drug. In addition, the change in the splicing pattern observed with the Sudemycins are similar to that observed with Spliceostatin A, a molecule known to interact with the SF3b subunit of the spliceosome. Hence, we conclude that Sudemycins can regulate the production of alternatively spliced RNA transcripts and these alterations are more prevalent in tumors, as compared to normal cells, following drug exposure. These studies suggest that modulation of alternative splicing may play a role in the antitumor activity of this class of agents.

Figure 1

Structures of pladienolide B, E7107, FR901464, Spliceostatin A and the Sudemycins. The lower structure indicates a summary of the design of concise analogs of FR901464: Showing Sudemycin E with color-coded highlighting of the atoms that have been modified in order to remove chirality and to give a less complex, and more chemically stable structure when compared to FR901464. Red – gain of symmetry and loss of a chemically destabilizing OH group; green – loss of a methyl group and/or atom type change to enhance solubility; cyan – gain of symmetry.

Figure 2

Sudemycin C1 modulates alternate splicing of MDM2 in Rh18 cells. Panel A – PCR analysis of MDM2 transcripts present in Rh18 cells following exposure to drug. Cells were exposed to 0.1, 1 or 10μM Sudemycin C1 for 2, 8 or 24 hours prior to analysis. The PCR products are indicated by the arrows and their sizes (in bp) are included in parentheses. M – markers (1kb ladder); ND – no drug. Panel B – Analysis of ubiquitin transcripts present in the same cells as analyzed in panel A. M – markers (100bp ladder). Panel C – Schematics of the truncated Mdm2 proteins that would be encoded by the alternatively spliced transcripts identified in panel A. The different regions with the protein are identified as: p53 – p53 binding domain (red); NLS – nuclear localization signal (green); NES – nuclear export signal (green); Ac – acidic domain (dark blue); Zn – zinc finger domains (pale blue); and RING - ring finger domain (yellow). Panel D – Western analysis of Mdm2 and β-Tubulin expression in Rh18 cells following exposure to Sudemycin C1. It is presumed that the signal at ~40kDa in the upper panel represents Mdm2B protein, however no definitive evidence has been obtained to confirm that this antibody cross reacts with this isoform.

Figure 3

Sudemycin modulates alternative splicing from MDM2 minigenes. Panel A – The structure of the introns and exons from the MDM2 gene present within the two minigene plasmids. Panel B – PCR analysis of MDM2 transcripts observed in Sudemycin C1 treated Rh18 cells following transfection with either the 3-4 or 3-12 minigene plasmid. The identity of the PCR products corresponding to the alternatively spliced transcripts is indicated at the side of the figure. These bands are identified by the arrows and their sizes (in bp) are included in parentheses. M – markers (100bp ladder); 3-4 – cells transfected with the 3-4 minigene; 3-4 + C - cells transfected with the 3-4 minigene and subsequently treated with 10μM Sudemycin C1; 3-10 – cells transfected with the 3-10 minigene; 3-10 + C - cells transfected with the 3-4 minigene and subsequently treated with 10μM Sudemycin C1. Panel C – A time course of MDM2 minigene splicing in Rh18 cells. Cells were transfected with either the 3-4 or the 3-10 minigene and at various times after exposure to 10μM Sudemycin C1, RNA was prepared and the presence of the different minigene transcripts was assessed (upper panel). The identity of the transcripts, indicated at the side of the figure, was validated by DNA sequencing. The lower panel indicates the analysis of the ubiquitin transcripts in the same samples. As above, products are identified by arrows and sizes are indicated in parentheses. M – markers (upper panel, 100bp ladder; lower panel, 1kb ladder); ND – no drug.

Figure 4

Modulation of MDM2 minigene splicing in LHCN-M2 and Rh18 cells following exposure to the different Sudemycin analogues. Following transfection, cells were exposed to 10μM drug for 24 hours and the presence of the different transcripts was determined by PCR and DNA sequencing. The identity of the products is indicated at the sides of the figure. The lower panel indicates the analysis of the ubiquitin transcripts in the same samples. The PCR products are indicated by the arrows and their sizes (in bp) are included in parentheses. M – markers (100bp ladder); ND – no drug; C1 – Sudemycin C1; E – Sudemycin E; F – Sudemycin F; L – LHCN-M2 cells; R – Rh18 cells.

Figure 5

Spliceostatin modulates MDM2 splicing in an identical fashion to that observed with Sudemycin. Panel A – Rh18 cells were treated with 100nM Spliceostatin A for up to 24 hours, and samples were harvested and MDM2 RNA splicing was evaluated by PCR. The lower panel indicates the analysis of the ubiquitin transcripts in the same samples. The PCR products are indicated by the arrows and their sizes (in bp) are included in parentheses. M – markers (1kb ladder); ND – no drug; S-A – Spliceostatin A. Panel B – LHCN-M2 and Rh18 cells were transfected with the 3-4 MDM2 minigene and the effects of Spliceostatin A on RNA splicing were assessed by PCR. The lower panel indicates the analysis of the ubiquitin transcripts in the same samples. As above, products are identified by arrows and sizes are indicated in parentheses. M – markers (100bp ladder); ‘−’ – no drug; ‘+’ – cells treated with 100nM Spliceostatin A; L – LHCN-M2 cells; R – Rh18 cells.

Figure 6

Effects of Sudemycin on the splicing of other genes in Rh18 cells. Cells were treated with 0.1, 1 or 10μM Sudemycin C1 for 2, 8 or 24 hours and splicing of transcripts for caspase 2, 9 and BCL-X were assessed by PCR. The validity of all Caspase 2S and 2L transcripts was determined by DNA sequencing. Ubiquitin was used as a control for these studies. In all panels, the PCR products are indicated by the arrows and their sizes (in bp) are included in parentheses. M – Markers (100bp ladder); C – no drug.

Figure 7

Sudemycin modulates MDM2 splicing in Rh18 xenografts. Mice bearing subcutaneous Rh18 xenografts were dosed with 50mg/kg Sudemycin E and 24 hours later tumors were harvested and MDM2 RNA splicing was determined by PCR. The lower panel indicates the analysis of the ubiquitin transcripts in the same samples. The PCR products are indicated by the arrows and their sizes (in bp) are included in parentheses. M – markers (upper panel, 1kb ladder; lower panel 100bp ladder); C – no drug; T – Rh18 tumor treated with Sudemycin.