1
FTO suppresses STAT3 activation and modulates proinflammatory interferon-stimulated
1
gene expression
2
3
Michael J. McFadden
1, a*
, Matthew T. Sacco
1*
, Kristen A. Murphy
1
, Moonhee Park
1
, Nandan S.
4
Gokhale
2
, Kim Y. Somfleth
2
, Stacy M. Horner
†1,3
5
6
Affiliations:
7
1
Department of Molecular Genetics and Microbiology, Duke University Medical Center,
8
Durham, NC 27710, USA.
9
Michael J. McFadden; mcfaddmj@umich.edu, Matthew T. Sacco;
10
matthew.sacco@duke.edu, Kristen A. Murphy; kristen.a.murphy@duke.edu, Moonhee
11
Park; moonhee.park@duke.edu, Stacy M. Horner; stacy.horner@duke.edu
12
2
Department of Immunology, University of Washington, Seattle, WA 98109, USA.
13
Nandan S. Gokhale; ngokhale@uw.edu, Kim Y. Somfleth; ksomf@uw.edu
14
3
Department of Medicine, Duke University Medical Center, Durham, NC 27710, USA.
15
16
17
Current address:
18
a
Department of Microbiology and Immunology, University of Michigan, Ann Arbor, MI
19
48109, USA.
20
21
22
23
24
* These authors contributed equally
25
26
†
Corresponding author
27
Correspondence to Stacy M. Horner (stacy.horner@duke.edu)
28
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 24, 2021. ; https://doi.org/10.1101/2021.07.23.453596doi: bioRxiv preprint
2
Abstract
29
Signaling initiated by type I interferon (IFN) results in the induction of hundreds of IFN-
30
stimulated genes (ISGs). The type I IFN response is important for antiviral restriction, but aberrant
31
activation of this response can lead to inflammation and autoimmunity. Regulation of this
32
response is incompletely understood. We previously reported that the mRNA modification m
6
A
33
and its deposition enzymes, METTL3 and METTL14 (METTL3/14), promote the type I IFN
34
response by directly modifying the mRNA of a subset of ISGs to enhance their translation. Here,
35
we determined the role of the RNA demethylase FTO in the type I IFN response. FTO, which can
36
remove either m
6
A or the cap-adjacent m
6
Am RNA modifications, has previously been associated
37
with obesity and body mass index, type 2 diabetes, cardiovascular disease, and inflammation.
38
We found that FTO suppresses the transcription of a distinct set of ISGs, including many known
39
pro-inflammatory genes, and that this regulation is not through the actions of FTO on m
6
Am.
40
Further, we found that depletion of FTO led to activation of STAT3, a transcription factor that
41
mediates responses to various cytokines, but whose role in the type I IFN response is not well
42
understood. This activation of STAT3 increased the expression of a subset of ISGs. Importantly,
43
this increased ISG induction resulting from FTO depletion was partially ablated by depletion of
44
STAT3. Together, these results reveal that FTO negatively regulates STAT3-mediated signaling
45
that induces proinflammatory ISGs during the IFN response, highlighting an important role for
46
FTO in suppression of inflammatory genes.
47
48
Keywords: Fat mass and obesity-associated (FTO); N6-methyladenosine (m
6
A); Interferon (IFN);
49
Interferon-stimulated gene (ISG); Inflammation
50
51
Introduction
52
The type I interferon (IFN) response induces the expression of hundreds of IFN-stimulated
53
genes (ISGs), many of which have antiviral and proinflammatory functions, and thus is crucial for
54
the early response to viral infection [1]. Type I IFNs signal through a dimeric receptor composed
55
of IFNAR1 and IFNAR2 to activate the transcription factors STAT1 and STAT2, which
56
heterodimerize with IRF9 to form the transcription factor complex ISGF3 [2]. ISGF3 then binds to
57
the promoters of ISGs to induce their transcription, resulting in the establishment an antiviral
58
cellular state [3]. While transcriptional induction of ISGs by ISGF3 is the primary driver of the type
59
I IFN response, regulation of this response by other transcription factors or by post-transcriptional
60
controls is incompletely understood. Specifically, the functions of the transcription factor STAT3
61
in the type I IFN response are not clear, although it has been shown to have functions either in
62
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 24, 2021. ; https://doi.org/10.1101/2021.07.23.453596doi: bioRxiv preprint
3
suppressing this response or promoting the expression of certain ISGs, likely in a cell type-specific
63
fashion [4]. We previously established that the addition of the mRNA modification m
6
A to a subset
64
of ISGs by the m
6
A-methyltransferase complex proteins METTL3 and METTL14 (METTL3/14)
65
increases their translation by recruiting the m
6
A-binding protein YTHDF1 and that this promotes
66
an antiviral cellular state [5]. However, the role of the m
6
A demethylase FTO in the type I IFN
67
response has not yet been described. The functions of FTO are important for several aspects of
68
human health, and genetic variants of this gene are associated with obesity and body mass index,
69
type 2 diabetes, cardiovascular disease, and inflammation [6-9]. Despite its essential functions
70
for human development and health [10], the mechanisms by which FTO regulates these biological
71
processes are incompletely understood.
72
A major molecular function of FTO, which has sequence homology to alpha-ketoglutarate
73
oxygenase enzymes such as the AlkB homolog family (ALKBH) proteins, is RNA demethylation
74
[11]. Importantly, FTO has been shown to be involved in the demethylation of both m
6
A [12], a
75
function shared by ALKBH5 [13], and the cap-adjacent m
6
Am modification [14]. m
6
A is deposited
76
by a complex of proteins, including METTL3/14 and others [15], and can regulate many aspects
77
of RNA metabolism, including degradation and translation, among other processes [16-19]. m
6
Am
78
addition is catalyzed by the enzyme PCIF1 at the first transcribed nucleotide in an mRNA and
79
may be involved in regulating mRNA stability and translation [20-22]. Through its RNA
80
demethylase functions, FTO can regulate many cellular and biological processes, including
81
neurogenesis, dopamine signaling and appetite regulation, adipogenesis, oncogenesis, and viral
82
infection [23-29]. Interestingly, FTO depletion was recently shown to sensitize melanoma cells to
83
IFN-γ treatment, suggesting a role for FTO in the response to IFNs [28]. Further, FTO has been
84
shown to regulate infection by a number of viruses, presumably by demethylating viral RNA [29],
85
but its role in regulation of host responses to viral infection has not been elucidated. Therefore,
86
we sought to define the role of FTO in the type I IFN response.
87
Here, we found that depletion of FTO results in increased production of a subset of ISGs.
88
However, whereas METTL3/14 promotes the translation of a subset of ISGs without affecting their
89
mRNA levels [5], FTO regulates a distinct subset of ISGs at the mRNA level. By labeling nascent
90
RNA, we found that FTO inhibits the transcription of these ISGs and that FTO-depleted cells are
91
primed to respond to type I IFN. FTO regulation of ISGs is not through demethylation of m
6
Am,
92
as deletion of the m
6
Am writer enzyme PCIF1 did not impact the phenotypic effect of FTO
93
depletion on ISG expression. Interestingly, we found that FTO depletion results in increased
94
phosphorylation and thus activation of transcription factor STAT3. Additionally, STAT3 activation
95
through treatment with the cytokine IL6 or expression of the Salmonella effector protein SarA,
96
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 24, 2021. ; https://doi.org/10.1101/2021.07.23.453596doi: bioRxiv preprint
4
known to induce STAT3 phosphorylation, recapitulated the phenotypic effects of STAT3 activation
97
by FTO depletion [30, 31], suggesting that suppression of STAT3 activation by FTO represses
98
transcription of a specific subset of ISGs. In support of this hypothesis, depletion of STAT3 led to
99
partial ablation of FTO-mediated suppression of ISG induction. Taken together, these results
100
reveal a novel role for FTO in the transcriptional suppression of STAT3-regulated ISGs, which
101
has important implications for our understanding of the role of FTO in disease.
102
103
Results
104
FTO regulates the mRNA expression of a subset of ISGs.
105
Having previously shown that the m
6
A-methyltransferase complex proteins METTL3/14
106
promote the translation of specific ISGs via the addition of m
6
A to these ISGs [5], we hypothesized
107
that the major m
6
A-demethylase FTO would suppress the translation of these ISGs by removal of
108
m
6
A. To test this hypothesis, we used siRNAs to deplete METTL3/14 or FTO in Huh7 cells and
109
induced the expression of ISGs by treatment with IFN-β, a type I IFN. Similar to our previous
110
results, depletion of METTL3/14 resulted in reduced protein levels of IFITM1, GBP1, IFIT3, and
111
MX1 [5]. Depletion of FTO showed the opposite result in that its depletion resulted in increased
112
protein expression of the METTL3/14-regulated ISGs, expect for MX1 whose expression was only
113
modestly affected by METTL3/14 depletion and not altered by FTO depletion (Figure 1A). As
114
before, the ISGs ISG15 and EIF2AK2 were unaffected by METTL3/14 depletion, and here we
115
also found that their IFN-induced expression was unaffected by FTO depletion (Figure 1A). We
116
next determined whether FTO depletion changed the mRNA induction of these ISGs.
117
Interestingly, while METTL3/14 depletion did not impact the mRNA levels of these ISGs, as
118
expected [5], FTO depletion did increase the mRNA levels of the regulated ISGs, including
119
IFITM1, GBP1, and IFIT3 (Figure 1B). These results show that FTO negatively regulates specific
120
ISGs at the mRNA level. Thus, the mechanism underlying the regulatory role of FTO on ISGs is
121
distinct from METTL3/14.
122
123
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 24, 2021. ; https://doi.org/10.1101/2021.07.23.453596doi: bioRxiv preprint
5
Figure 1: FTO regulates the mRNA expression of a subset of ISGs. (A) Immunoblot analysis
124
of extracts from Huh7 cells transfected with siRNAs to control (CTRL), METTL3/14 (M3/14), or
125
FTO prior to mock or IFN-β (12 h) treatment. Data are representative of 3 biological experiments.
126
(B) RT-qPCR analysis of ISG induction normalized to HPRT1 following IFN-β treatment (12 hours)
127
of Huh7 cells treated with non-targeting control (CTRL) or METTL3/14 siRNA plotted as fold
128
change over mock treatment for each ISG. Values are the mean ± SD of 3 technical replicates,
129
representative of 3 biological experiments. * p < 0.05, ** p < 0.01, *** p < 0.005 by one way
130
ANOVA with Dunnett’s multiple comparisons test.
131
132
FTO regulates the transcription of certain ISGs.
133
To determine how FTO regulates the mRNA expression of ISGs, we pulsed siCTRL,
134
siMETTL3/14, or siFTO-treated cells with 4-thiouridine (4-sU) and IFN-β for 1 hour to
135
metabolically label nascent transcripts produced by IFN stimulation. We then purified the nascent
136
RNA and performed RT-qPCR to quantify relative transcription. These experiments revealed that
137
depletion of FTO led to a marked increase of the transcription of the FTO-regulated ISG IFITM1
138
during the pulse. Interestingly, the transcription of the non-FTO-regulated ISGs ISG15 and
139
EIF2AK2 were also slightly increased by FTO depletion (Figure 2A). METTL3/14 depletion did
140
not affect the transcription of these ISGs nor a known m
6
A-modified gene, CREBBP, whose
141
stability is regulated by m
6
A [16] (Figure 2A). Next, to determine how FTO regulates the
142
transcription of ISGs over time, we performed the 4-sU and IFN-β pulses at both 1 and 12 hours
143
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 24, 2021. ; https://doi.org/10.1101/2021.07.23.453596doi: bioRxiv preprint
