Synthesis and activity of a novel inhibitor of nonsense-mediated mRNA decay

Article abstract of DOI:10.1039/c5ob02482j

During efforts to prepare the known compound NMDI1, a new tetracyclic compound, called VG1, was prepared in six steps. This compound was found to have good activity as an inhibitor of nonsense-mediated mRNA decay.

Full text of DOI:10.1039/c5ob02482j

Organic &
Biomolecular Chemistry
View Article Online
View Journal
COMMUNICATION
Synthesis and activity of a novel inhibitor of
nonsense-mediated mRNA decay†
Cite this: DOI: 10.1039/c5ob02482j
Victoria J. B. Gotham,a,b Melanie C. Hobbs,a Ryan Burgin,a David Turton,b
Received 4th December 2015,
Accepted 24th December 2015
Carl Smythe*b and Iain Coldham*a
DOI: 10.1039/c5ob02482j
www.rsc.org/obc
During efforts to prepare the known compound NMDI1, a new (Fig. 1) was able to inhibit both nucleus-associated and cyto-
tetracyclic compound, called VG1, was prepared in six steps. This plasmic NMD. The compound appears to be a specific inhi-
compound was found to have good activity as an inhibitor of non- bitor of NMD: it does not affect the splicing of pre-mRNA
sense-mediated mRNA decay.
reporter transcripts, the amount of pre-mRNA, translation
efficiency or the miRNA decay pathway.5 It is non-cytotoxic,
even at high concentrations, and does not induce the for-
mation of stress granules.5
The synthesis of NMDI1 was reported by Bisagni and co-
workers.6,7 The overall yield was poor (0.5% over 11 steps) and
we had difficulties on attempted repetition of the published
procedure (see ESI†). Keeling and co-workers have recently
reported a synthesis of NMDI1 by a modified route (total 13
steps) in which the final 7 steps were identical to those used
by Bisagni and co-workers; however, they report no experi-
mental details.8
There is clearly a need for a NMD inhibitor whose synthesis
is more efficient and reproducible. We have synthesised in
only 6 steps a close structural analogue of NMDI1, called VG1,
by using a different route from that described above.6–8 We
found that VG1 inhibits NMD equally well as NMDI1 and is
therefore a suitable candidate for use as a tool to investigate
NMD and related mechanisms, and possibly for development
as a drug for diseases caused by PTC mutations.
VG1 was synthesised from commercially available 2,3-di-
bromopyridine and 2,5-dimethylaniline (Scheme 1). These were
heated at 140 °C for 16 h to give bromide 1 in 76% yield. An
aldehyde was then installed para to the amino group. The use
of POCl3 and DMF (Vilsmeier reaction)9,10 failed to give the
desired aldehyde, but the use of AlCl3 and MeOCHCl2 (Rieche
formylation)11,12 was successful and gave aldehyde 2. This
A significant proportion of genetic diseases, as well as many
cancers, are associated with a particular type of mutation
which gives rise to a premature termination codon (PTC)
before the genuine stop codon.1 PTC-associated diseases
include cases of Duchenne muscular dystrophy, cystic fibrosis,
β-thalassaemia, retinitis pigmentosa, von Willebrand disease,
Robinow syndrome, haemophilia, spinal muscular atrophy
and Rett syndrome. mRNAs containing PTCs are detected and
degraded by a process known as nonsense-mediated mRNA
decay (NMD),2–4 which reduces the expression of the corres-
ponding gene to a low level. It is the subsequent lack of
protein that causes the symptoms of PTC-associated diseases.
The PTCs that trigger NMD can occur due to single base pair
mutations, reading frame shifts due to deletions or insertions,
or mutations in splice sites or splicing regulatory sequences
that give rise to aberrant splicing.2–4 At the mRNA level, PTCs
can arise from errors in transcription, transcription initiation
upstream of the correct site coupled with the location of a
termination codon in the region before the proper start
site, alternative splicing, or programmed frameshifts during
translation.2–4
One potential approach to the treatment of diseases caused
by PTC mutations is the inhibition of NMD. Therefore there is
a need to discover and develop drugs that inhibit NMD selec-
tively. In addition, NMD inhibitors can be useful tools for
studying the NMD mechanism and related pathways that
utilise some of the same factors. In 2007, Lejeune and co-
workers5 found that a tetracyclic compound called NMDI1
aDepartment of Chemistry, University of Sheffield, Sheffield, S3 7HF, UK.
E-mail: i.coldham@sheffield.ac.uk
bDepartment of Biomedical Science, University of Sheffield, Sheffield, S10 2TN, UK
†Electronic supplementary information (ESI) available: Experimental procedures
and spectroscopic data. See DOI: 10.1039/c5ob02482j
Fig. 1 Structure of NMD inhibitor NMDI1.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
Scheme 1 Synthesis of VG1.
compound was not purified but was treated directly with The cloning of NORM and TER sequences into the mRNA pro-
NaClO2 and NH2SO3H as a chlorine scavenger,13–15 to give the ducing plasmid pCMV, and the characterisation of resultant
carboxylic acid 3 in 66% yield from bromide 1. The acid was mRNA has been previously described.22 TER and NORM tran-
converted to amide 4 in 94% yield by using the commercially scripts are subject to splicing, and TER, but not NORM, spliced
available aminoacetaldehyde diethylacetal and the coupling mRNA is subject to NMD.22 Fig. 2 shows a plasmid map and
reagent COMU.16 Deprotection of the acetal with polyphospho- the structure of the hybrid genes. In order to control for trans-
ric acid was followed by in situ cyclisation and dehydration17 to fection efficiency, cells were also transfected with the reference
give compound 5 in 70% yield. An intramolecular palladium- plasmid phCMV-MUP (referred to as MUP), which encodes
catalysed cross-coupling reaction18 was used to convert mouse major urinary protein.25 After 24 h, the transfection
bromide 5 into tetracycle VG1 in 54% yield.
medium was removed from cells and replaced with culture
Attempted conversion of the lactam 5 to the chloropyridine medium containing either DMSO or NMDI1 or VG1 at a final
was unsuccessful by using POCl3, PhP(O)Cl2, PCl3, or PCl5 in concentration of 5 μM made up from a 5 mM stock solution in
POCl3. Chlorination of VG1 was also attempted but the only DMSO. Each compound was added both to cells transfected
isolable products obtained, after treatment of either com- with NORM and cells transfected with TER. After 20 h, total
pound 5 or VG1 with PCl5 in POCl3, were N-chloro derivatives RNA was extracted and reverse transcribed. The amount of
(see ESI†). Chlorination on the nitrogen atom of an amide has NORM, TER and MUP cDNA was determined by qPCR, where
literature precedent.19–21 Unfortunately, this meant that we the amount of cDNA serves as a proxy measurement for mRNA
were not able to prepare NMDI1 using this synthesis. However, as has been described previously.27,28 To ensure that the qPCR
we had prepared a close analogue in VG1 and decided to test reaction products reflected the appropriately spliced mature
the activity of this compound.
mRNA, the cDNA generated from RNA extracted from both
To test the ability of VG1 to inhibit NMD we utilised a well- NORM and TER transfected cells was amplified by standard
established assay for the measurement of NMD within cells.22 PCR using identical primers and DNA products which corre-
HeLa cells were transfected with either one of two reporter spond to their cognate mRNAs were analysed for size by
plasmids, pmCMV-Gl (NORM) or pmCMV-Gl (TER),22–24 agarose gel electrophoresis. These showed that PCR products
referred to as NORM and TER. The NORM plasmid contains a correspond to correctly spliced mRNA (Fig. S2 in ESI†).
hybrid human/mouse β-globin gene. This was constructed by
The inferred amounts of NORM and TER mRNA, each
Maquat and Kinniburgh23 by substituting sequences in the expressed as a ratio of the respective inferred amount of MUP
human β-globin gene extending from the BamH1 site in exon mRNA in each sample, were compared in order to determine
II with analogous sequences from the mouse β-globin gene. the NMD efficiency. As expected, for cells treated with DMSO
The resulting hybrid gene contains exon I, intron I and most only, the amount of TER mRNA was significantly lower than
of exon II of the human β-globin gene, and 18 base pairs of the amount of NORM mRNA (0.22 times the amount of NORM
exon II, intron II and exon III of the mouse β-globin gene. The mRNA), indicating that TER mRNA was efficiently degraded by
TER plasmid contains the same chimaeric gene but with a NMD (Fig. 3A, first and second bars from left). When cells
PTC mutation (CAG to TAG) in codon 39, located in exon II. were treated with VG1, the amount of TER mRNA increased to
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Communication
Fig. 2 pmCMV-Gl (NORM) and pmCMV-Gl (TER) plasmids. (A) Plasmid map. The plasmids contain a mouse cytomegalovirus (mCMV) promoter, a
hybrid human/mouse β-globin gene and a gene for ampicillin resistance. The map was created using the PlasMapper program.26 (B) Structure of the
hybrid human/mouse β-globin gene. Grey boxes are introns. Start (ATG) and stop (TAA) codons are shown. The BamH1 site marks the boundary
between sequences from the human and mouse genes. The TER plasmid is identical to NORM, except that a PTC mutation (CAG to TAG), has been
introduced in codon 39 located in exon II.
Fig. 3 NMD inhibition by VG1. (A) Cells transfected with the NMD reporter constructs were treated with either DMSO or 5 μM NMDI1 or VG1 for
20 h. The amount of reporter β-globin mRNA (NORM or TER) is controlled to the amount of MUP mRNA. The data are normalised to the amount of
NORM mRNA (controlled to the amount of MUP mRNA) for DMSO-treated cells. Error bars show the standard error in the mean obtained from three
independent experiments. (B) Cells transfected with the NMD reporter constructs were treated with VG1 at the concentrations indicated for 20 h.
The amount of TER β-globin mRNA is expressed relative to the amount of NORM β-globin mRNA (both controlled to the amount of MUP mRNA) for
each concentration. Error bars show the standard error in the mean obtained from three independent experiments for 0.2, 1 and 5 μM VG1 and from
two independent experiments for 25 and 125 μM VG1.
0.70 times the amount of NORM mRNA, showing that NMD order to determine the NMD efficiency. To ensure that the
was inhibited (Fig. 3A, fifth and sixth bars from left). VG1 PCR primers used generated a reaction product corresponding
inhibited NMD to the same extent as NMDI1 (Fig. 3A, third to spliced RNA and not unspliced pre-mRNA, the amplified
and fourth bars from left).
PCR products from RNA extracted from cells transfected with
The effect of the dose of VG1 on NMD inhibition was then either NORM, or TER, were examined by DNA agarose gel elec-
investigated. Cells were transfected with the NMD reporter trophoresis (Fig. S2†). In each case, these showed the presence
plasmids, as described above. After 24 h, the transfection of a DNA fragment of 150 bp corresponding to correctly
medium was removed from cells and replaced with culture spliced mRNA. In qPCR analyses, the amount of TER mRNA
medium containing either DMSO or VG1 at a final concen- relative to NORM mRNA, and therefore NMD inhibition,
tration of 0.2, 1, 5, 25 or 125 μM made up from a stock solu- increased as VG1 concentration increased, reaching
a
tion in DMSO. After 20 h, total RNA was extracted and reverse maximum at 25 μM, at which concentration the amount of
transcribed. The amount of NORM, TER and MUP cDNA was TER mRNA was 0.94 times the amount of NORM mRNA
determined by qPCR and the inferred amounts of NORM and (Fig. 3B). There was no further NMD inhibition at 125 μM.
TER mRNA, each expressed as a ratio of the respective inferred Thus high levels of inhibition of NMD were achieved by treat-
amount of MUP mRNA in each sample, were compared in ing cells with 5–25 μM VG1, assuming that if NMD were fully
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
inhibited, the amount of TER mRNA would be equal to the through. This approach has been used by Lejeune and co-
amount of NORM mRNA.
workers,33 who found that the effect of 5 μM NMD inhibitor
NMDI1 has been reported to have little effect on cell viabi- amlexanox combined with 25 μM readthrough promoter
lity.5 When cells were seeded as normal (104 cells per cm2) and PTC124 was greater than that of either drug on its own in
treated with 5 μM VG1 for 24 h, the rate of proliferation increasing iodide efflux in 6CFSMEo cystic fibrosis cells.
appeared similar to cells treated with DMSO alone (Fig. S1A Keeling and co-workers have shown that NMDI1 was able to
in ESI†) and there was little evidence of increased pyknosis or enhance the therapeutic effect of at least a subset of read-
karyorrhexis (indicators of necrosis or late-stage apoptosis) through promoters called aminoglycosides in IduaW392X mice
suggesting that, as with NMDI1, effective doses on this (which have a premature stop codon in the gene encoding α-L-
timescale are not toxic for cell survival. The effect of VG1 dose iduronidase).8
on cells after 20 h exposure was also investigated (Fig. S1B†).
In this paper, we have developed a synthesis of an analogue
As before, at concentrations up to 5 μM, there was no of NMDI1 and have found that our compound, VG1, inhibited
obvious effect on cell number or morphology. When VG1 NMD to the same extent as NMDI1. Our synthesis requires
was added to cell media at 25 μM and cells were incubated only 6 steps and proceeds in 18% overall yield. Lejeune and
with compound for 20 h, there was some indication of altered co-workers presented evidence that NMDI1 inhibits the inter-
cell morphology in a minority of cells, suggestive of some tox- action between two key NMD factors, UPF1 and SMG5.5 UPF1
icity at higher concentrations and possible induction of apop- forms part of a complex that recognises the PTC-containing
tosis (Fig. S1B†). Further analyses of the effects of longer mRNA after the first round of translation terminates at the
exposure at 25 μM or the use of higher concentrations were PTC. The protein undergoes cycles of phosphorylation and de-
not possible due to insolubility of VG1 in cell media at high phosphorylation, which allow it to take part in multiple
concentrations. Exploration of the induction of apoptosis at an rounds of NMD. SMG5 is one of the proteins that mediates the
intermediate concentration of VG1 (10 μM, Fig. S1C†) was eval- de-phosphorylation of UPF1, and inhibition of the UPF1-SMG5
uated by immunoblotting total cell lysates for the appearance interaction may cause UPF1 to become trapped in a hyper-
of the cell marker of apoptotic induction, activated (cleaved) phosphorylated state. Given the similarity between the struc-
Caspase 3.29 These data (Fig. S1C†) indicated that, at this con- tures of VG1 and NMDI1, it is likely that the compounds
centration and time scale, there was no evidence of elevated inhibit NMD by the same mechanism, and therefore, that VG1
levels of apoptotic cell death under these conditions. Taken also inhibits the interaction between UPF1 and SMG5,
together our data suggest that, as with NMDI1, effective doses although this still needs to be confirmed. It is conceivable that
of VG1 over a 24 h time scale are not toxic for cell survival.
NMDI1 is converted into VG1 in the cell and that the active
Inhibiting NMD is a promising approach towards treating form of the compounds is the same. The compound VG1
diseases caused by PTC mutations as, in many cases, the trun- passed the filter on an online PAINS (pan assay interference
cated protein that would result if the PTC-containing mRNA compounds) false positive remover tool,39 which helps to vali-
were to be translated could be at least partially functional and date VG1 as a genuine NMD inhibitor. Work is required to
could relieve some of the symptoms of the disease. Obviously, establish the mechanism of action of VG1. Whatever the
the function of the truncated protein would depend on the outcome, VG1 will be a useful tool for developing an under-
location of the PTC within the mRNA. As an example of the standing of both the mechanism of NMD and of other related
potential of this approach, when cultured fibroblasts from a mRNA decay pathways.
patient with Ullrich’s disease (which is caused by a PTC in the
gene for the extracellular matrix protein collagen VI α2) were
treated with inhibitors of the NMD factor SMG1, there was an
increase in the expression of truncated collagen VI α2 which
Acknowledgements
partially restored the function of the extracellular matrix.30 We thank the EPSRC and the University of Sheffield for
However, SMG1 inhibitors are unlikely to be able to be used funding.
clinically as they are toxic and are not specific for SMG1. Other
small molecules that have been used to inhibit NMD have not
yet reached clinical trials of NMD inhibition.31–36
Notes and references
One approach for treating PTC-associated diseases involves
the use of a class of compounds called ribosome readthrough
promoters.37,38 These compounds induce the ribosome to
“read through” a PTC, allowing a full-length protein to be pro-
duced from the transcript. However, the efficacy of read-
through promoters is limited by a paucity of PTC-containing
mRNAs available for readthrough, as most are degraded
rapidly by NMD. Inhibiting NMD, in combination with read-
through therapy, should mean that fewer PTC-containing
mRNAs are degraded and therefore more are available for read-
1 P. A. Frischmeyer and H. C. Dietz, Hum. Mol. Genet., 1999,
8, 1893.
2 O. Isken and L. E. Maquat, Genes Dev., 2007, 21, 1833.
3 P. Nicholson, H. Yepiskoposyan, S. Metze, R. Z. Orozco,
N. Kleinschmidt and O. Mühlemann, Cell. Mol. Life Sci.,
2010, 67, 677.
4 C. Schweingruber, S. C. Rufener, D. Zünd, A. Yamashita
and O. Mühlemann, Biochim. Biophys. Acta, 2013, 1829,
612.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Communication
5 S. Durand, N. Cougot, F. Mahuteau-Betzer, C. H. Nguyen, 24 J. Cheng, M. Fogel-Petrovic and L. E. Maquat, Mol. Cell.
D. S. Grierson, E. Bertrand, J. Tazi and F. Lejeune, J. Cell
Biol., 2007, 178, 1145.
Biol., 1990, 10, 5215.
25 K. R. Prowse and H. Baumann, Mol. Cell. Biol., 1988, 8, 42.
6 C. Rivalle, C. Ducrocq, J.-M. Lhoste and E. Bisagni, J. Org. 26 X. Dong, P. Stothard, I. J. Forsythe and D. S. Wishart,
Chem., 1980, 45, 2176. Nucleic Acids Res., 2004, 32, W660.
7 C. Rivalle, C. Ducrocq, J.-M. Lhoste, F. Wendling, E. Bisagni 27 T. F. Brazão, J. Demmers, W. van IJcken, J. Strouboulis,
and J.-C. Cherman, Tetrahedron, 1981, 37, 2097.
8 K. M. Keeling, D. Wang, Y. Dai, S. Murugesan, B. Chenna,
M. Fornerod, L. Romão and F. G. Grosveld, FEBS Lett.,
2012, 586, 1101.
J. Clark, V. Belakhov, J. Kandasamy, S. E. Velu, T. Baasov 28 L. Linde, S. Boelz, G. Neu-Yilik, A. E. Kulozik and B. Kerem,
and D. M. Bedwell, PLoS One, 2013, 8, 60478.
9 A. Vilsmeier and A. Haack, Chem. Ber., 1927, 60, 119.
10 G. Lai, X. R. Bu, J. Santos and E. A. Mintz, Synlett, 1997,
1275.
11 A. Rieche, H. Gross and E. Höft, Chem. Ber., 1960, 93, 88.
12 C. Schneider, D. Gueyrard, F. Popowycz, B. Joseph and
P. G. Goekjian, Synlett, 2007, 2237.
13 B. O. Lindgren and T. Nilsson, Acta Chem. Scand., 1973, 27,
888.
14 B. S. Bal, W. E. Childers and H. W. Pinnick, Tetrahedron,
1981, 37, 2091.
Eur. J. Hum. Genet., 2007, 15, 1156.
29 V. Ramu, M. R. Gill, P. J. Jarman, D. Turton, J. A. Thomas,
A. Das and C. Smythe, Chem. – Eur. J., 2015, 21, 9185.
30 F. Usuki, A. Yamashita, I. Higuchi, T. Ohnishi,
T. Shirigashi, M. Osame and S. Ohno, Ann. Neurol., 2004,
55, 740.
31 H. Jungwirth, H. Bergler and G. Högenauer, J. Biol. Chem.,
2001, 276, 36419.
32 Y. Dang, W.-K. Low, J. Xu, N. H. Gehring, H. C. Dietz,
D. Romo and J. O. Liu, J. Biol. Chem., 2009, 284, 23613.
33 S. Gonzalez-Hilarion, T. Beghyn, J. Jia, N. Debreuck,
G. Berte, K. Mamchaoui, V. Mouly, D. C. Gruenert,
B. Déprez and F. Lejeune, Orphanet J. Rare Dis., 2012, 7, 58.
15 L. Colombo, C. Gennari, M. Santandrea, E. Narisano and
C. Scolastico, J. Chem. Soc., Perkin Trans. 1, 1980, 136.
16 A. El-Faham, R. S. Funosas, R. Prohens and F. Albericio, 34 J. Wengrod, L. Martin, D. Wang, P. Frischmeyer-Guerrerio,
Chem. – Eur. J., 2009, 15, 9404.
17 G. Ricci, R. Ruzziconi and E. Giorgio, J. Org. Chem., 2005,
70, 1011.
18 J. K. Laha, P. Petrou and G. D. Cuny, J. Org. Chem., 2009,
74, 3152.
H. C. Dietz and L. B. Gardner, Mol. Cell. Biol., 2013, 33,
2128.
35 L. Martin, A. Grigoryan, D. Wang, J. Wang, L. Breda,
S. Rivella, T. Cardozo and L. B. Gardner, Cancer Res., 2014,
74, 3104.
19 A. Alder and D. Bellus, J. Am. Chem. Soc., 1983, 105, 6712.
20 R. G. Micetich, R. Singh and C. C. Shaw, J. Org. Chem.,
1986, 51, 1811.
21 R. G. Micetich, R. Singh, M. P. Singh and C. C. Shaw, Synth.
Commun., 1986, 16, 453.
22 J. Zhang, X. Sun, Y. Qian and L. E. Maquat, RNA, 1998, 4,
801.
23 L. E. Maquat and A. J. Kinniburgh, Nucleic Acids Res., 1985,
13, 2855.
36 M. Bhuvanagiri, J. Lewis, K. Putzker, J. P. Becker, S. Leicht,
J. Krijgsveld, R. Batra, B. Turnwald, B. Jovanovic, C. Hauer,
J. Sieber, M. W. Hentze and A. E. Kulozik, EMBO Mol. Med.,
2014, 6, 1593.
37 K. M. Keeling, D. Wang, S. E. Conard and D. M. Bedwell,
Crit. Rev. Biochem. Mol. Biol., 2012, 47, 444.
38 H. L. Lee and J. P. Dougherty, Pharmacol. Ther., 2012, 136,
227.
39 http://cbligand.org/PAINS/login.php.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem.