PROJECT SUMMARY
Pediatric high grade gliomas (pHGGs) represent 10-15% of all brain tumors diagnosed in children. One subtype
of pHGGs, diffuse intrinsic pontine glioma (DIPG), is especially deadly, with a five-year survival rate of < 1%.
Current treatment options are non-curative; surgical resection, localized radiation, and chemotherapy are com-
plicated by the brainstem location of the tumor and accompanying negative side-effects. Thus, more effective
strategies are urgently needed. Approximately 70-80% of all DIPG tumors are marked by a dominant heterozy-
gous point mutation in H3F3A, which codes for the non-canonical histone H3.3. This toxic gain-of-function mu-
tation replaces lysine 27 with methionine (K27M), preventing trimethylation of lysine 27 (H3K27me3). H3K27M
has also been shown to interfere with Polycomb repressive complex 2 (PRC2), leading to global reduction of di-
and tri-methylation on histone proteins. This mutant H3.3 histone is predicted to be a major driver of tumorigen-
esis in H3K27M-mutated DIPG by disrupting normal neural differentiation. Antisense oligonucleotides (ASOs)
offer a unique method to target mRNA through Watson-Crick base pairing with high specificity and low toxicity.
In a recently submitted manuscript, my lab developed a “gapmer” ASO which targets H3F3A for RNaseH deg-
radation, inducing neural differentiation and prolonging survival. To further these results by pursuing a parallel
ASO modality, I will develop a splice-switching ASO that reduces translation of mutant H3K27M RNA by inducing
skipping of H3F3A exon 2, which contains the only in-frame start codon for H3F3A and the K27M mutation. This
splice-switching modality is significant because uniformly-modified ASOs, unlike “gapmer” ASOs, exhibit longer
half-lives and reduced off-target liability in the central nervous system, and ASOs of this chemistry are already
FDA-approved treatments. Furthermore, H3F3A and its paralogous gene H3F3B both encode identical H3.3
histone proteins, H3F3B is sufficient to compensate for loss of H3F3A, and H3F3B should remain unaffected by
my ASO strategy due to sequence dissimilarities. I hypothesize that the reduction of H3K27M will limit DIPG
tumor growth and prolong survival by promoting differentiation of the tumor cells. The aims of my pro-
posal are to identify a lead splice-switching ASO that reduces H3K27M and restores H3K27me3 and to charac-
terize the role of the RNA-binding protein RBFOX3 in reinforcing neural differentiation and H3F3A exon 2 skip-
ping following ASO injection. Methods to achieve will include establishing a preclinical in vivo mouse model with
H3.3K27M DIPG xenografts to study splicing changes of H3F3A and H3F3B through RT-PCR, protein expres-
sion of H3K27M and H3K27me3 through immunoblot, and neural cell differentiation through immunofluores-
cence following ASO injection. Additionally, splicing and analytical RNA-protein binding assays will be used to
characterize the role and binding kinetics of RBFOX3 with H3F3A RNA. The proposed research is significant
because it will: (i) help identify and provide mechanistic explanation for a new clinical candidate for drug devel-
opment; and (ii) provide further justification for use of ASOs in the treatment of neuro-oncological disorders.