Science - USA (2018-12-21)

MORN domains from JP2NT (JP2NTDMORNs)did
not influence its distribution in the nucleus or its
association with chromatin (Fig. 2B, ii, b). In
contrast, deletion of the bNLS-like sequence
from this construct (JP2NTDMORNs/DbNLS)signifi-
cantly reduced the association of JP2NT with
chromatin (Fig. 2B, ii, c). Deletion of the alanine-
rich domain in combination with the bNLS
(JP2NTDMORNs/DbNLS/DARRor JP2NTDbNLS/DARR)
completely prevented presence of JP2NT in
the MNASE-releasable chromatin fraction (Fig.
2B, ii, d to f). On the basis of these data, we
conclude that JP2NT associates with chromatin
via a domain located at residues ~345 to 402,
which is highly evolutionarily conserved in mam-

malian species as well as in vertebrates such as

fish and birds (fig. S3C).

JP2NT is enriched at transcription

start sites

To systematically study the genomic targets of

JP2NT, we generated transgenic mice with cardiac-

specific overexpression of hemagglutinin (HA)–

tagged JP2NT. In these mice, JP2NT is predomi-

nantly localized in the nuclei of cardiomyocytes

(fig. S4A). The level of JP2NT transgene in nu-

clearfractionisofthesameorderofmagnitude,

relative to the peak level of endogenous JP2NT

induced by TAB (Fig. 1F). Hearts of the JP2NT-

overexpressing mice (JP2NT-OE) were subjected

to chromatin immunoprecipitation sequenc-

ing (ChIP-seq) analysis using antibody to HA.

Two replicates of ChIP-seq analyses were per-

formed using different batches of JP2NT heart

samples. Only the DNA peaks (P<10−^10 ) that

were detected in both replicates of ChIP-seq

analyses were considered as JP2NT-binding DNA

regions. We identified 9414 JP2NT-binding ge-

nomic DNA regions encompassing 7398 genes.

The DNA binding profile revealed that JP2NT

is concentrated in gene-enriched regions, espe-

cially the promoter and 5′untranslated regions

(Fig. 3A). Moreover, JP2NT is preferentially en-

riched at transcription start sites (Fig. 3B), a

characteristic of transcription regulators. We

Guoet al.,Science 362 , eaan3303 (2018) 21 December 2018 5of9

Fig. 4. JP2NT represses MEF2-mediated transcription by competing
for the MEF2 response element (MRE).(A) Enrichment of MEF2 binding
motifs in ChIP-seq dataset. (B) Gel shift assay of JP2NT binding to a
desmin enhancer–derived DNA sequence containing a WT or mutant MRE
in vitro. (C) Coimmunoprecipitation of HA-tagged JP2NT binding to MEF2C
or Histone H3 in HEK293T cells transfected with JP2NT and MEF2C as well
as in JP2NT-OE hearts (postnatal 2 weeks). (DandE) MEF2 activity assays in

HEK293T cells cotransfected with PMEF2-firefly, PSV40-renilla, MEF2C,

and WT JP2NT (D) or a mutant lacking the ARR (E, JP2NTDARR)(n≥ 3

independent batches of experiments, three transfection replicates in-

cluded in each batch; **P< 0.01 versus MEF2C+empty pcDNA). (F)Overlap

chart of MEF2C (left) and TBP (right) DNA binding peaks compared

between control and JP2NT-OE hearts, as well as compared with JP2NT

binding peaks from JP2NT-OE hearts.

RESEARCH | RESEARCH ARTICLE

on December 25, 2018^

http://science.sciencemag.org/

Downloaded from