Letter
https://doi.org/10.1038/s41586-019-1374-1
m
6
A enhances the phase separation potential of mRNAr yan J. ries^1 , Sara Zaccara^1 , Pierre Klein^1 , Anthony Olarerin-George^1 , Sim Namkoong^2 , Brian F. Pickering^1 , Deepak P. Patil^1 ,
Hojoong Kwak^3 , Jun Hee Lee^2 & Samie r. Jaffrey^1 *
N^6 -methyladenosine (m^6 A) is the most prevalent modified
nucleotide in mRNA^1 ,^2 , with around 25% of mRNAs containing at
least one m^6 A. Methylation of mRNA to form m^6 A is required for
diverse cellular and physiological processes^3. Although the presence
of m^6 A in an mRNA can affect its fate in different ways, it is unclear
how m^6 A directs this process and why the effects of m^6 A can vary
in different cellular contexts. Here we show that the cytosolic
m^6 A-binding proteins—YTHDF1, YTHDF2 and YTHDF3—
undergo liquid–liquid phase separation in vitro and in cells. This
phase separation is markedly enhanced by mRNAs that contain
multiple, but not single, m^6 A residues. Polymethylated mRNAs
act as a multivalent scaffold for the binding of YTHDF proteins,
juxtaposing their low-complexity domains and thereby leading to
phase separation. The resulting mRNA–YTHDF complexes then
partition into different endogenous phase-separated compartments,
such as P-bodies, stress granules or neuronal RNA granules. m^6 A-
mRNA is subject to compartment-specific regulation, including a
reduction in the stability and translation of mRNA. These studies
reveal that the number and distribution of m^6 A sites in cellular
mRNAs can regulate and influence the composition of the phase-
separated transcriptome, and suggest that the cellular properties
of m^6 A-modified mRNAs are governed by liquid–liquid phase
separation principles.
To understand how m^6 A affects the fate of mRNA, we considered
the biochemical properties of the major cytosolic m^6 A-binding pro-
teins YTHDF1, YTHDF2 and YTHDF3 (hereafter denoted DF1, DF2
and DF3, respectively). These paralogous proteins have high sequence
identity and comprise a YTH domain of around 15 kDa that binds m^6 A,
and a low-complexity domain of around 40 kDa that includes prion-like
domains^4 (Extended Data Fig. 1a).
Some low-complexity amino acid sequences form fibrils, hydrogels
or liquid droplets as a result of phase separation^5 ,^6. To test whether DF(^1) Department of Pharmacology, Weill-Cornell Medical College, Cornell University, New York, NY, USA. (^2) Department of Molecular and Integrative Physiology, University of Michigan, Ann Arbor, MI,
USA.^3 Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY, USA. *e-mail: [email protected]
22 °C 37 °C 22 °C
4 °C 37 °C 4 °C
DF2
a b DF2 (75 μM)
2 4 6 8
0
100
200
300
DF2 (μM)
NaCl (mM)
No droplets Droplets
c
30 s 45 s
0 s 15 s
60 s
d
Bleach
0 s
Postbleach
10 s
e
Prebleach
–3 s
g
Alexa488–DF2
10 m^6 A RNA
No RNA
(^0102030)
0.5
1.0
Time (s)
Normalized intensity
0
Alexa488–DF2
Alexa488–DF2
DF2
0 m^6 A RNA
1 m^6 A RNA
10 m^6 A RNA
f
DFs,
no RNA10 m (^6) A RNA
DF1 DF2DF3
0.0
0.5
1.0
1.5
2.0
Partition coef
cient (AU)
+++ ****
+–
+– +– +–
- –+–
––+
Fig. 1 | Polymethylated m^6 A RNAs trigger liquid–liquid phase separation
of DF proteins. a, Tubes containing either buffer only or recombinant
DF2 were heated from 4 °C to 37 °C. DF2 undergoes phase separation
when heated; this is reversible upon cooling. b, Time-lapse of bright-field
microscopy images of DF2 droplets (75 μM) subjected to a temperature
gradient. The temperature was increased at a rate of 1 °C per minute from
22 °C to 37 °C, enabling the formation of protein droplets. Lowering the
temperature back to 22 °C causes disassembly of the droplets. c, Phase
diagram of DF2 in the presence of different concentrations of NaCl, showing
that salt dampens the phase-separation potential of the protein. Green
circles indicate that protein droplets were present; pink squares indicate
that no protein droplets were observed in the buffer. d, Alexa488–DF2
(75 μM) was imaged by fluorescence microscopy over 1 min. The fusion
of Alexa488–DF2 droplets can be seen in Supplementary Video 1. e, Top,
changes in the fluorescence intensity of Alexa488–DF2 droplets after
photobleaching were plotted over time. The background was subtracted
from the fluorescence measurement. The black curve represents the mean of
the fluorescence intensity in the photobleached region of interest in distinct
droplets (n = 8); grey bars indicate s.e.m. Bottom, representative images
of fluorescence recovery. f, A 65-nucleotide RNA containing
10 m^6 A nucleotides (570 nM) induces DF2 (25 μM) to rapidly form small
liquid droplets, whereas RNA containing one m^6 A nucleotide or no m^6 A
nucleotides does not cause substantial DF2 phase separation. g, Left, the
addition of RNA containing 10 m^6 A sites (425 nM) enhances the phase
separation of DFs (15 μM) in solution. Right, partition coefficients (PCs)
for DFs in the presence and absence of RNA containing 10 m^6 A sites. For
the no-RNA condition, partition coefficients were calculated immediately
before the addition of m^6 A-containing RNA (right panel: DFs with no
RNA, mean PC = 1.0; DF1, n = 8; DF2, n = 10; DF3, n = 9; total, n = 27).
Partition coefficients for the DF proteins were measured shortly after the
addition of RNA containing 10 m^6 A nucleotides and the mean DF partition
coefficients increased notably (right panel: DF1 mean PC = 1.40, n = 14;
DF2 mean PC = 1.67, n = 14; DF3 mean PC = 1.41, n = 14 droplets) within
minutes of RNA addition. Error bars represent s.e.m. n represents the
number of droplets from technical replicates. ****P < 0.0001, two-sided
Mann–Whitney test. Scale bars, 10 μm (d–g).424 | NAtUre | VOL 571 | 18 JULY 2019