Article
Cytosolic Iron-Sulfur Asse
mbly Is EvolutionarilyTuned by a Cancer-Amplified Ubiquitin LigaseGraphical Abstract
Highlights
d MAGE-F1 forms a complex with the NSE1 E3 ubiquitin ligase
d MAGE-F1 inhibits the cytosolic Fe-S assembly pathway
through degrading MMS19
d Downregulation of MMS19 by MAGE-F1-NSE1 reduces DNA
repair capacity
d MAGE-F1 is copy-number amplified in lung cancers and
contributes to tumorigenesis
Weon et al., 2018, Molecular Cell 69, 1–13January 4, 2018 ª 2017 Elsevier Inc.https://doi.org/10.1016/j.molcel.2017.11.010
Authors
Jenny L. Weon, Seung Wook Yang,
Patrick Ryan Potts
In Brief
Weon et al. show that MAGE-F1 specifies
the cytosolic iron-sulfur assembly (CIA)
pathway protein MMS19 for
ubiquitination and degradation. MAGE-
F1 is amplified in lung cancers, where it
promotes tumor growth and increased
mutational burden. These studies provide
evidence for alteration of the CIA pathway
in cancer through post-translational
regulation.
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Molecular Cell
Article
Cytosolic Iron-Sulfur AssemblyIs Evolutionarily Tunedby a Cancer-Amplified Ubiquitin LigaseJenny L. Weon,1 Seung Wook Yang,1 and Patrick Ryan Potts1,2,*1Department of Cell andMolecular Biology, St. JudeChildren’s ResearchHospital, 262Danny ThomasPlace,Memphis, TN 38105-3678, USA2Lead Contact
*Correspondence: [email protected]://doi.org/10.1016/j.molcel.2017.11.010
SUMMARY
The cytosolic iron-sulfur (Fe-S) cluster assembly (CIA)pathway functions to incorporate inorganic Fe-S co-factors into a variety of proteins, including severalDNA repair enzymes.However, themechanisms regu-lating theCIApathwayareunknown.Wedescribeherethat the MAGE-F1-NSE1 E3 ubiquitin ligase regulatesthe CIA pathway through ubiquitination and degrada-tion of the CIA-targeting protein MMS19. Over-expression or knockout of MAGE-F1 altered Fe-Sincorporation into MMS19-dependent DNA repairenzymes, DNA repair capacity, sensitivity to DNA-damaging agents, and iron homeostasis. Intriguingly,MAGE-F1 has undergone adaptive pseudogenizationin select mammalian lineages. In contrast, MAGE-F1is highly amplified in multiple human cancer typesand amplified tumors have increased mutationalburden. Thus, flux through the CIA pathway can beregulated by degradation of the substrate-specifyingMMS19 protein and its downregulation is a commonfeature in cancer and is evolutionarily controlled.
INTRODUCTION
Iron-sulfur (Fe-S) clusters are one of the most ancient inorganic
cofactors utilized by proteins from bacteria to humans (Lill,
2009). Generation of Fe-S, typically in the form of [2Fe-2S] or
[4Fe-4S] (Rouault, 2015), requires the coordinated activity of
members of the mitochondrial iron-sulfur cluster assembly
(ISC) machinery for mitochondrial Fe-S proteins or both the
ISC and the cytosolic iron-sulfur cluster assembly (CIA) machin-
ery for cytoplasmic and nuclear Fe-S proteins (Netz et al., 2014).
The CIA pathway, consisting of at least 9 components in humans
(Paul and Lill, 2015), serves as a conduit by which iron- and sul-
fur-containing cofactors are generated from a precursor product
of the mitochondria ISC and passed through a series of proteins
to ultimately be incorporated into cytosolic or nuclear proteins
that require the Fe-S cluster as a structural (Stehling et al.,
2008, 2012), enzymatic (Beinert, 2000), or electron transfer
component (Rouault, 2015). One of the critical end-target-bind-
ing adaptors for this process isMMS19, which binds a number of
proteins, including those important for DNA repair processes,
such as FANCJ, POLD1, XPD, and RTEL1, among others (Gari
et al., 2012; Stehling et al., 2012, 2013). Characterization of the
loss of MMS19 has been well documented in yeast and more
recently in human cells as displaying greater susceptibility to
DNA-damaging agents (Gari et al., 2012; Lauder et al., 1996; Pra-
kash and Prakash, 1979; Stehling et al., 2012). Although new dis-
coveries in the CIA pathway have focused on identifying the core
assembly proteins, binding modalities of members of the
pathway, and the specific Fe-S proteins targeted, there remains
a large gap in understanding how the CIA pathway is regulated
and altered in disease.
Here, we report that an E3 ubiquitin ligase in the MAGE-RING
ligase (MRL) family controls flux through the CIA pathway through
ubiquitination anddegradation ofMMS19.MRLs are a family of E3
ubiquitin ligases that consist of a complex between an E3 RING
ubiquitin ligase and a modulatory MAGE protein, which can func-
tion to specify substrates for the ligase (Doyle et al., 2010; Hao
et al., 2013, 2015; Lee and Potts, 2017; Pineda et al., 2015). There
are >40 MAGE proteins in humans, each containing a MAGE ho-
mology domain (MHD) thatmediates binding to distinct E3 ubiqui-
tin ligases (Doyle et al., 2010; Weon and Potts, 2015). The specific
cellular function of the majority of MAGE proteins, including
MAGE-F1, has not been elucidated. In this study, we identify a
function for the orphanMAGE-F1 in specifyingMMS19 for ubiqui-
tination and degradation by the NSE1 E3 ubiquitin ligase. This re-
sults in decreased Fe-S cluster assembly intoMMS19 targets that
functionally renders cells less competent to repair a spectrum of
DNA damage. Interestingly, MAGE-F1 is copy-number amplified
in several types of cancers, resulting in increasedmutation burden
in tumors, suggesting downregulation of the CIA pathway may be
an important event in tumorigenesis.
RESULTS
MAGE-F1 Binds the NSE1 E3 Ubiquitin LigaseNo known functions for MAGE-F1 have been reported. Thus, we
searched for MAGE-F1-interacting partners by performing a
large-scale pull-down of FLAG-MAGE-F1 from HEK293 cells
and identified a single E3 RING ubiquitin ligase, NSE1. We
confirmed that MAGE-F1 interacts with NSE1 by performing co-
immunoprecipitation (co-IP) studies fromcells (Figure1A). Further-
more, MYC-MAGE-F1 immunoprecipitated with endogenous
Molecular Cell 69, 1–13, January 4, 2018 ª 2017 Elsevier Inc. 1
C D
E
A B Figure 1. MAGE-F1 Binds the E3 Ligase
NSE1 but Does Not Incorporate into the
SMC5/6 Complex
(A) Immunoprecipitation (IP) of HA-NSE1 shows it
binds MYC-MAGE-F1, similar to MYC-MAGE-G1,
in HeLa cells.
(B) Endogenous NSE1 binds to MYC-MAGE-F1.
HeLa cells were transfected with MYC-vector or
MYC-MAGE-F1 constructs for 48 hr before
immunoprecipitation with anti-MYC followed by
SDS-PAGE and immunoblotting for endogenous
NSE1.
(C) Pull-down of recombinant GST-NSE1 demon-
strates binding to MYC-MAGE-F1, similar to MYC-
MAGE-G1, but not the MYC-MAGE-F1 L87-88A
dileucine mutant in vitro.
(D) Pull-down of MYC-SMC5 or MYC-SMC6 from
HeLa cells demonstrates that FLAG-MAGE-F1
does not have the robust affinity for SMC5 or
SMC6 as does FLAG-MAGE-G1.
(E) Pull-down of FLAG-MAGE-F1 or FLAG-
MAGE-G1 in HeLa cells shows that endogenous
SMC5 and SMC6 bind to MAGE-G1, but not
MAGE-F1.
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
NSE1 (Figure 1B). In addition, in vitro binding assays revealed that
bacterially purified glutathione S-transferase (GST)-NSE1 directly
bindsMYC-MAGE-F1 (Figure1C).Aconserveddileucinemotif has
been identified in theMHD ofmostMAGEs and is critical for inter-
actionwith their cognateE3 ligases (Doyle et al., 2010).Mutationof
thismotif inMAGE-F1 (L87-88A) abolishedbinding ofMAGE-F1 to
NSE1 (Figure 1C). NSE1 has previously been shown to interact
with a closely related MAGE protein, MAGE-G1 (Taylor et al.,
2008). The MAGE-G1-NSE1 MRL incorporates into the SMC5/6
complex to facilitate homologous recombination (HR) between
sister chromatids and telomeres (Potts, 2009; Potts et al., 2006;
Potts and Yu, 2007). Thus, we examinedwhether MAGE-F1 could
simply replace MAGE-G1 in the SMC5/6 complex. However,
MAGE-F1, unlike MAGE-G1, failed to interact with co-expressed
or endogenous SMC5 or SMC6 (Figures 1D and 1E). Together,
these results suggest that MAGE-F1 forms a complex with the
NSE1ubiquitin ligasebut doesnot integrate into theSMC5/6com-
plex like MAGE-G1-NSE1. Consistent with this, it has been
observed that a fraction of cellular NSE1 exists outside the
SMC5/6 complex (Taylor et al., 2008). Thus, the function of an
E3 ligase (NSE1 in this case) can be diversified by association
with different MAGE proteins.
MAGE-F1-NSE1 MRL Targets MMS19 for Ubiquitinationand DegradationNext, we searched for targets of the MAGE-F1-NSE1 ubiquitin
ligase to provide insights into its cellular function. Analysis of
2 Molecular Cell 69, 1–13, January 4, 2018
the MAGE-F1 interactome (Table S1) re-
vealed three high-confidence substrates
(MMS19, MAT2a, and PPM1G) that we
confirmed bind to and are ubiquitinated
by MAGE-F1-NSE1 (Figures 2A and
S1A–S1D). Of these three substrates,
MAGE-F1-NSE1 bound to and ubiquiti-
nated MMS19 most robustly (Figures 2A
and S1B–S1D). MMS19 ubiquitination was unaffected by
MAGE-G1 expression (Figure 2A). MAGE-F1most robustly ubiq-
uitinated the C-terminal domain (amino acids 857–1,030) of
MMS19 (Figure S1E), which contains a cluster of five lysine res-
idues (K993, K1002, K1007, K1008, and K1013). Mutation of
these lysines to arginine (referred to herein as 5KRmutant) signif-
icantly blocked MMS19 ubiquitination by MAGE-F1 (Figure 2B).
Furthermore, MAGE-F1 L87-88A mutant incapable of binding
NSE1 failed to induce MMS19 ubiquitination (Figure 2C). Knock-
down of NSE1 severely diminished MAGE-F1’s ability to pro-
mote MMS19 ubiquitination (Figure 2D). These results suggest
that MAGE-F1 in conjunction with NSE1 promotes MMS19 ubiq-
uitination. Consistent with this, MMS19 bound recombinant
His-MAGE-F1-NSE1 complex in vitro, but not His-NSE1 alone
(Figure 2E).
Ubiquitination of MMS19 by MAGE-F1-NSE1 resulted in its
degradation as knockdown of MAGE-F1 (Figure S1F) or NSE1
increased MMS19 protein levels (Figure 2F) without affecting
MMS19 mRNA levels (Figure 2G). In contrast, expression of
wild-type, but not MAGE-F1 L87-88A mutant, reduced
endogenous MMS19 protein levels (Figure 2H). Importantly,
downregulation of MMS19 by MAGE-F1 could be rescued by
the proteasome inhibitor MG132 (Figure 2H). Additionally,
knockout of MAGE-F1 in HeLa cells significantly increased
MMS19 protein half-life (Figure 2I). Consistent with these find-
ings, using ubiquitin Lys to Arg mutants that disrupt specific
chain linkage types, we found that mutation of K48, but not K6,
HF
A
I
J
DC
B
E
G
Figure 2. MAGE-F1-NSE1 Targets MMS19 for Ubiquitination and Degradation
(A) HA-MAGE-F1, but not HA-MAGE-G1, robustly increases ubiquitination of MMS19. HeLa cells were transfected with the indicated constructs for 48 hr and
treated with 10 mMMG132 for 4 hr followed by anti-MYC IP to pull-down ubiquitinated proteins followed by SDS-PAGE and immunoblotting for FLAG-MMS19.
(B) Mutation of five lysines in the C terminus of MMS19 (K993, K1002, K1007, K1008, and K1013; 5KR) disrupts MAGE-F1-induced ubiquitination. Cells were
treated and lysates were prepared as noted in (A).
(C) Wild-type MAGE-F1, but not NSE1 binding defective mutant MAGE-F1 L87-88A (LA), induces ubiquitination of MMS19. Cells were treated and lysates were
prepared as noted in (A).
(legend continued on next page)
Molecular Cell 69, 1–13, January 4, 2018 3
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
K11, K27, K29, K33, or K63, of ubiquitin impaired MAGE-F1-
induced MMS19 ubiquitination in cells (Figure 2J). Finally, the
MMS19 5KR mutant that fails to be robustly ubiquitinated by
MAGE-F1 was resistant to MAGE-F1-induced degradation (Fig-
ures 3D and 4D, insets). These results suggest that MAGE-F1
specifies MMS19 for ubiquitination by NSE1, leading to its
degradation.
MAGE-F1-NSE1 MRL Decreases Iron Incorporation intoMMS19 TargetsMMS19, in conjunction with CIA1 and CIA2B, acts late in the CIA
pathway to recruit apoproteins to the CIA machinery (namely
IOP1/Nar1) for incorporation of the Fe-S cluster to form func-
tional holoproteins (Gari et al., 2012; Lill et al., 2015; Seki et al.,
2013; Stehling et al., 2012, 2013; van Wietmarschen et al.,
2012). In order to directly determine whether MAGE-F1-NSE1-
mediated ubiquitination and degradation of MMS19 has an ef-
fect on CIA pathway flux and incorporation of Fe-S cluster into
downstream targets, we labeled cells with 55Fe, immunoprecip-
itated specific Fe-S proteins, and quantitated the amount of
radioactivity present via scintillation counting, as has been previ-
ously described (Stehling et al., 2004; Teichmann and Stremmel,
1990). We found that expression of MAGE-F1 decreased Fe-S
cluster incorporation into downstream MMS19-dependent tar-
gets POLD1, FANCJ, XPD, DPYD, and RTEL1 to a similar extent
as MMS19 knockdown by small interfering RNAs (siRNAs) (Fig-
ure 3A). Importantly, MAGE-F1 did not affect 55Fe incorporation
into a MMS19-independent Fe-S protein, PPAT (Figure 3A), sug-
gesting MAGE-F1 specifically modulates the MMS19 arm of the
CIA pathway. Changes in 55Fe incorporation upon MAGE-F1
expression were not due to altered 55Fe cellular uptake or un-
equal immunoprecipitation of proteins (Figures S2A and S2B).
Consistent with MAGE-F1 affecting the MMS19 CIA pathway
in conjunction with NSE1, knockdown of NSE1 abrogated the ef-
fects of MAGE-F1 expression on 55Fe incorporation into FANCJ
(Figure 3B). Furthermore, MAGE-F1 L87-88A that fails to bind
NSE1 was incompetent to alter 55Fe incorporation into FANCJ
and XPD (Figure 3C). In contrast to MAGE-F1, no significant
changes were found with 55Fe incorporation into targets upon
expression of MAGE-G1, again suggesting distinct functions of
MAGE-G1-NSE1 and MAGE-F1-NSE1 (Figure S2C). Impor-
tantly, rescuing MMS19 protein levels in MAGE-F1-expressing
cells by expressing the MMS19 5KR mutant returned CIA
pathway activity to normal (Figure 3D). These effects of
(D) MAGE-F1-induced ubiquitination ofMMS19 depends on the NSE1 ligase. Cells
for 72 hr and treated with 10 mM MG132 for 4 hr followed by anti-MYC IP to p
for MMS19.
(E) MYC-MMS19 binds bacterially purified recombinant His-MAGE-F1-NSE1 com
(F) Knockdown of endogenous MAGE-F1 or NSE1 in HeLa cells with two differen
(G) qRT-PCR analysis of MMS19 mRNA levels normalized to 18S rRNA in HeLa c
are mean ± SD.
(H) Expression of MYC-MAGE-F1 wild-type (WT), but not L87-88A (LA) mutant, i
manner. Cells were treated with 10 mM MG132 (or DMSO control) for 4 hr, 72 hr
(I) Knockout ofMAGE-F1 increasesMMS19 protein stability in HeLa cells. MAGE-F
for the indicated times. Cell lysates were prepared, separated by SDS-PAGE, an
(J) Denaturing His-ubiquitin pull-down from HeLa with ubiquitin mutants shows ro
K48R ubiquitin mutant. Cells were treated with 10 mM MG132 for 5 hr, 48 hr afte
See also Figure S1.
4 Molecular Cell 69, 1–13, January 4, 2018
MAGE-F1 are not restricted to overexpression as knockout of
MAGE-F1 increased levels of 55Fe incorporation into FANCJ,
POLD1, DPYD, and RTEL1 (Figure 3E). It has previously been
shown that knockdown of MMS19 destabilizes a number of
downstream target proteins due to their reliance on the Fe-S
cluster for stability (Gari et al., 2012; Stehling et al., 2012). Simi-
larly to these findings, MAGE-F1 expression decreased FANCJ
levels and to a smaller extent XPD, POLD1, and DPYD (Figures
3F and 3G). Rescuing MMS19 proteins levels in MAGE-F1-ex-
pressing cells returned FANCJ levels to baseline (Figure 3H).
Consistent with MAGE-F1-NSE1 abrogating MMS19 function,
we also observed defects in iron homeostasis previously re-
ported upon knockdown of MMS19 (Stehling et al., 2013).
Specifically, knockdown of MMS19 by siRNAs or expression of
MAGE-F1 in HeLa cells showed robust increases in uptake
of fluorescein isothiocyanate (FITC)-labeled transferrin (Fig-
ure S2D). In contrast, knockout of MAGE-F1 decreased trans-
ferrin uptake, which could be rescued by re-expression of
MAGE-F1 (Figure S2E). These results suggest that MAGE-
F1-NSE1 ubiquitination and degradation of MMS19 impairs
MMS19 functions, including flux through the CIA pathway and
regulation of iron homeostasis.
MAGE-F1 Inhibits Homologous Recombination andSensitizes Cells to DNA-Damaging AgentsBecause MAGE-F1-mediated degradation of MMS19 leads to
decreased incorporation of 55Fe into several DNA repair en-
zymes, we examined whether this would alter cellular DNA repair
mechanisms. To test this hypothesis, we utilized a previously
described homologous recombination assay (Porteus and Balti-
more, 2003; Potts et al., 2006), in which a non-functional GFP
containing in-frame stop codons can be repaired upon induction
of a double-strand break by the I-SceI endonuclease and recom-
bination with an episomal repair template (Figure S3A). We found
that dose-dependent expression of wild-type MAGE-F1, but not
MAGE-F1 L87-88A, reduced homologous recombination rates
(Figures 4A, 4B, and S3B). Similarly, knockdown of MMS19 in
a dose-dependent manner phenocopied MAGE-F1 expression,
resulting in decreased rates of homologous recombination (Fig-
ures 4C and S3C). Importantly, MAGE-F1-induced reduction
in homologous recombination was reversed upon rescuing
MMS19 protein levels by expression of the non-ubiquitinatable
MMS19 5KR mutant (Figure 4D). In order to determine what
particular downstream MMS19 target(s) may be mediating its
were transfectedwith the indicated siRNAs (siControl or siNSE1) and plasmids
ull-down ubiquitinated proteins followed by SDS-PAGE and immunoblotting
plex, but not His-NSE1 alone, in vitro.
t siRNAs increases MMS19 protein levels.
ells transfected with siMAGE-F1 or siNSE1 siRNAs. ns indicates p > 0.05. Data
n HEK293 cells decreases MMS19 protein levels in a proteasome-dependent
after transfection.
1 wild-type or knockout HeLa cells were treatedwith 100 mg/mL cycloheximide
d immunoblotted for the indicated proteins.
bust reduction of MMS19 ubiquitination in the presence of MAGE-F1 with the
r transfection.
A
GF
EDC
B
H
Figure 3. MAGE-F1 Decreases Iron Incorporation into MMS19 Targets
(A) Immunoprecipitation of endogenous MMS19 target proteins after 55Fe treatment in HeLa cells demonstrates decreased incorporation of 55Fe either with
MAGE-F1 overexpression orMMS19 knockdown, but not a non-MMS19 target protein (PPAT). Immunoblots below showMAGE-F1 transgene levels andMMS19
knockdown efficiency.
(B) Knockdown of NSE1 in HeLa cells abrogatesMAGE-F1-mediated decrease of 55Fe incorporation into a target of MMS19. Immunoblots below showMAGE-F1
transgene levels and NSE1 knockdown efficiency.
(C) MAGE-F1 dileucine mutant is incapable of altering 55Fe incorporation into MMS19 target proteins in HeLa cells. Immunoblots below show MAGE-F1
transgene levels.
(D) MAGE-F1 effects on 55Fe incorporation into MMS19 targets are dependent on degradation of MMS19. HeLa cells were transfected with MAGE-F1 alone or in
combination with wild-type MMS19 or non-ubiquitinatable MMS19 5KR. Incorporation of 55Fe into FANCJ and XPDwas determined by immunoprecipitation and
scintillation counting. Immunoblots below show expression of HA-MAGE-F1 and MMS19 transgenes.
(E) HeLa-Cas9 MAGE-F1 knockout cells exhibit greater 55Fe incorporation into MMS19-dependent proteins than their wild-type counterparts.
(F and G) MYC-MAGE-F1 expression in HEK293 cells causes decreased protein levels of MMS19 and downstream target proteins. Representative immunoblots
shown in (F) and quantitation is shown in (G).
(H) Overexpression of MMS19 in the context of MAGE-F1 rescues FANCJ protein levels in HEK293 cells.
Data are mean ± SD. Asterisks indicate p < 0.05. ns indicates p > 0.05. See also Figure S2.
Molecular Cell 69, 1–13, January 4, 2018 5
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
A
G
E
DCB
F H
Figure 4. MAGE-F1 Inhibits Homologous Recombination and Sensitizes Cells to DNA-Damaging Agents
(A) Increasing MAGE-F1 expression levels in HEK293 cells decreases rates of homologous recombination in a dose-dependent manner. Immunoblots below
indicate MAGE-F1 transgene expression levels.
(B) The dileucine mutant of MAGE-F1 is incapable of inhibiting homologous recombination in HEK293 cells. Immunoblots below indicate MAGE-F1 transgene
expression levels.
(C) Knockdown of MMS19 decreases rates of homologous recombination in a dose-dependent manner in HEK293 cells. Immunoblots below indicate MMS19
knockdown efficiency.
(D) MAGE-F1 effects on homologous recombination are dependent on ubiquitination of MMS19. HEK293 cells were transfected with MAGE-F1 alone or in
combination with wild-type MMS19 or non-ubiquitinatable MMS19 5KR for 72 hr before homologous recombination rates were measured. Immunoblots below
show levels of MMS19 and transgene-expressed MAGE-F1.
(E) Knockdown of downstreamMMS19 targets shows that FANCJ and POLD1 phenocopy and likely mediate the defect in homologous recombination seen with
MMS19 knockdown in HEK293 cells. Immunoblots shown to the right denote knockdown efficiency of siRNA targets using pooled siRNAs.
(F) MYC-MAGE-F1-expressing HeLa-Cas9 cells exhibit increased sensitivity to DNA damage byMMS compared toMYC-vector cells. Viability was normalized to
untreated samples for both cell lines. Immunoblots below indicated levels of MAGE-F1 transgene expression.
(G) HeLa-Cas9 MAGE-F1 knockout cells are less sensitive than HeLa-Cas9 WT cells to MMS. Viability was normalized relative to WT cells.
(H) HeLa-Cas9 MAGE-F1 knockout cells are less sensitive than HeLa-Cas9 WT cells to UV. Viability was normalized relative to WT cells.
Data are mean ± SD. Asterisks indicate p < 0.05. ns indicates p > 0.05. See also Figure S3.
6 Molecular Cell 69, 1–13, January 4, 2018
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
effect on homologous recombination, we measured homolo-
gous recombination rates in cells depleted of several MMS19
targets affecting DNA metabolism, including FANCJ, POLD1,
RTEL1, and XPD (Figure 4E). Consistent with previous reports
demonstrating that FANCJ and POLD1 play important roles in
homologous recombination (Litman et al., 2005; Maloisel et al.,
2008), knockdown of either, but not XPD or RTEL1, reduced ho-
mologous recombination rates similarly to MMS19 knockdown
or MAGE-F1 expression (Figure 4E). Previous studies have
also shown that depletion of MMS19 results in increased sensi-
tivity to DNA-damaging agents (Lauder et al., 1996; Seki et al.,
2013; Stehling et al., 2012). Consistent with MAGE-F1 inhibition
of the MMS19 pathway, stable expression of MAGE-F1 reduced
cellular viability upon exposure to methyl methanesulfonate
(MMS) (Figure 4F) and knockout of MAGE-F1 reduced sensitivity
to both MMS and UV (Figures 4G and 4H) without affecting the
cell cycle (Figures S3D and S3E). These results suggest that
downregulation of MMS19 CIA pathway by MAGE-F1-NSE1 re-
duces the DNA repair capacity of cells.
Adaptive Pseudogenization of MAGE-F1 in SpecificMammalian LineagesIn the process of evaluating the function of MAGE-F1-NSE1MRL
in the mouse, we unanticipatedly found that the MAGE-F1 gene
in mouse has a single-nucleotide frameshifting deletion leading
to an in-frame stop codon that we confirmed by Sanger
sequencing (Figure S4). Like the majority of other MAGE genes,
MAGE-F1 is only conserved in placental mammals (De Donato
et al., 2017; Katsura and Satta, 2011; Zhao et al., 2012). Surpris-
ingly, we found that a large subset of mammals has a non-coding
MAGE-F1 gene due to insertions, deletions, or mutations that
lead to in-frame stop codons (Figure 5A). Pseudogenization of
the MAGE-F1 gene has apparently happened multiple times,
as several different mammalian lineages have mutations (Fig-
ure 5A), with some specific hotspots for mutations in MAGE-F1
gene (Figure 5B). Genes that acquire mutations leading to non-
coding RNAs often acquire additional mutations, given the relax-
ation on evolutionary pressure to keep the protein-coding gene
intact. As a result, these pseudogenized genes build up muta-
tions and the ratio of non-synonymous (dN) to synonymous
(dS) mutations increases toward neutral evolution. Consistent
with pseudogenization of MAGE-F1, we find multiple insertions,
deletions, or mutations that lead to in-frame stops across most
non-coding MAGE-F1 genes (Figure 5A). Furthermore, evalu-
ating rates of dN and dS inMAGE-F1 from 60mammals revealed
a significant (p = 2.43 10�14) increase in the dN/dS ratio inmam-
mals with MAGE-F1 pseudogene (Figure 5C). Thus, MAGE-F1
has undergone adaptive pseudogenization in specific mamma-
lian lineages to possibly alter MSM19 and CIA pathway function.
MAGE-F1 Is Amplified in Multiple Cancers and IsAssociatedwith IncreasedMutational Burden in TumorsGiven the importance of DNA repair mechanisms to maintain
genomic stability and that genomic instability is a hallmark of
cancers, we examined whether the MAGE-F1-NSE1 pathway
may be altered in tumors. Based on genomic analysis of TCGA
(The Cancer Genome Atlas) tumors, several cancer types,
including >40% of lung squamous cell carcinomas, have high
amplification of MAGE-F1, but not other MAGEs (Figures 6A
and S5A). Copy number amplification of MAGE-F1 correlated
with increased mRNA levels (Pearson coefficient 0.773;
Spearman coefficient 0.822) in multiple cancer types (Figures
6B and S5B). Of note, amplification of MAGE-F1 does not
happen in isolation and likely cooperates with other oncogenic
drivers, as it is co-amplified with other known oncogenes on
chr 3q, including PIK3CA, SOX2, and TP63 (Figure S5C).
Given our findings that MAGE-F1 specifies downregulation of
the MMS19 CIA pathway and reduces DNA repair capacity of
cells, we examinedwhether tumors that harbor high amplification
of MAGE-F1 exhibit greater genomic instability. Indeed, lung
squamous cell carcinoma patient tumors with high levels of
MAGE-F1 amplification and mRNA harbored a greater number
of mutations per tumor (Figures 6C, 6D, and S5D), with a wide
spectrum of mutations as would be consistent with modulation
of a large number of DNA enzymes that function in multiple path-
ways (Figure 6E). Importantly, high expression of other MAGEs,
such as MAGE-A3, did not alter tumor mutation burden (Fig-
ure S5E) and the effects of MAGE-F1 on mutational burden was
not driven by difference in exposure to smoke-related carcino-
gens (Figure S5F). We similarly found increased total mutational
burden inMAGE-F1-amplified head and neck squamous cell car-
cinomas as compared to non-MAGE-F1-amplified cases (Figures
6F, 6G, and S5G). These results are likely clinically significant, as
MAGE-F1 amplification, NSE1 expression levels, and the combi-
nation all showed robust decreases in survival of head and neck
squamous carcinoma patients (Figures 6H, S5H, and S5I).
To experimentally determine whether MAGE-F1 overexpres-
sion is both necessary and sufficient to drive tumor growth, we
manipulated MAGE-F1 levels in lung squamous carcinoma cell
lines and measured their xenograft tumor growth properties in
mice. We identified two lung squamous carcinoma cell lines
with naturally occurring copy number amplification and overex-
pression of MAGE-F1, HCC95, and H520. Strikingly, knockdown
of MAGE-F1 in both HCC95 and H520 slowed xenograft tumor
growth in mice (Figures 6I, 6J, and S5J). To further test the suf-
ficiency of MAGE-F1 to drive tumor growth, we expressed
MAGE-F1 in a H2170 lung squamous carcinoma cell line that
normally does not have copy number amplification of and over-
expression of MAGE-F1 (Figure S5K). Using this model for
MAGE-F1 amplification in lung squamous tumors, we measured
xenograft growth rates in mice. In contrast to MAGE-F1 knock-
down, overexpression ofMAGE-F1 drove increased tumor xeno-
graft tumor growth in mice (Figure 6K). These results suggest
that MAGE-F1 copy number amplification and overexpression
significantly contribute to tumorigenesis.
DISCUSSION
In summary, our results define a function for the orphan
MAGE-F1 protein, provide evidence for diversification of E3
ligase function by MAGE proteins, and more importantly illus-
trate how post-translational regulation of the CIA Fe-S pathway
can be achieved through ubiquitination and degradation of
MMS19. These findings have a number of implications, including
establishing a role for alteration of the CIA pathway in cancer and
genomic stability.
Molecular Cell 69, 1–13, January 4, 2018 7
A
C
B
Figure 5. Adaptive Pseudogenization of MAGE-F1 in Specific Mammalian Lineages
(A) Mammalian phylogenic tree indicating species with protein-coding or pseudogene (red) MAGE-F1. Note multiple gains and/or losses of MAGE-F1. Right:
schematic of MAGE-F1 genes indicating sites of insertions, deletions, or mutations leading to in-frame stops is shown (asterisks).
(B) Sequences of those genetic alterations denoted with black asterisks in (A). Human, cat, and cow protein-coding MAGE-F1 sequences are shown for
comparison.
(C) dN/dS ratio indicates those MAGE-F1 pseudogenes (24 species) are drifting toward neutral selection compared to protein-coding MAGE-F1 genes
(36 species). See also Figure S4.
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
MAGE-Dependent Switching of E3 Ligase FunctionMAGEs bind to E3 ubiquitin ligases through their MHD
(Doyle et al., 2010). The specific sequence and conformational
8 Molecular Cell 69, 1–13, January 4, 2018
dynamics likely dictate which specific E3 RING ligase a given
MAGE will bind. We and others have reported previously that
similar MAGEs can, in some cases, bind to the same RING
A B
C D E F G
H J KI
Figure 6. MAGE-F1 Is Amplified in Multiple Cancers and Is Associated with Increased Tumor Mutational Burden
(A) MAGE-F1 is highly amplified in multiple cancer types.
(B) Amplification of MAGE-F1 is associated with increased MAGE-F1 mRNA levels in multiple tumor types.
(C) Significantly greater copy numbers of MAGE-F1 are observed in MAGE-F1-amplified lung squamous cell carcinoma cases.
(D) Total mutation burden is higher in MAGE-F1-amplified lung squamous cell carcinomas than in cases without amplification.
(E) A variety of mutational spectra are observed at higher frequencies in MAGE-F1-amplified lung squamous cell carcinomas.
(F) Significantly greater copy numbers of MAGE-F1 are observed in MAGE-F1-amplified head and neck squamous cell carcinoma cases.
(G) Total mutation burden is higher in MAGE-F1-amplified head and neck squamous cell carcinomas than in cases without amplification.
(H) Head and neck squamous cell carcinomas with high expression of both MAGE-F1 and NSE1 exhibit significantly worse clinical prognosis than those with low
expression of both genes.
(I and J) Knockdown of MAGE-F1 in H520 (I) and HCC95 (J) lung squamous cell carcinoma cell lines with copy-number-amplified MAGE-F1 decreases xenograft
tumor growth in mice (n = 6 per group).
(K) Stable expression of MAGE-F1 in H2170 lung squamous cell carcinoma cells without natural MAGE-F1 copy number amplification increases xenograft tumor
growth rates in mice (n = 6 per group).
Data are mean ± SD. Asterisks indicate p < 0.05. See also Figure S5.
Molecular Cell 69, 1–13, January 4, 2018 9
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
protein, such as MAGE-A3 and MAGE-A6 binding TRIM28
(Doyle et al., 2010; Pineda and Potts, 2015; Pineda et al.,
2015; Yang et al., 2007). However, in those cases, the MAGE
proteins are very similar and have redundant activity. We
describe here a case where two MAGEs, MAGE-F1 and
MAGE-G1, can bind the same NSE1 RING protein but have
non-overlapping functions. Mechanistically, how may the bind-
ing of two different MAGEs switch the function of an E3 RING
ligase? One possibility is that the regions outside of the MHD
in the N or C terminus may differ enough to bind different
substrates and expand the substrate repertoire for NSE1. Alter-
natively, the binding of MAGE-F1 and MAGE-G1 may exert
differential conformations on NSE1 that may promote different
activities. Additionally, other binding partners may hinder inter-
action with substrates, such as binding of MAGE-G1-NSE1
to the SMC5/6 complex, which may hinder interaction with
MMS19. Another possibility is that the different pools of
MAGE-G1-NSE1 and MAGE-F1-NSE1 localize to distinct sub-
cellular compartments. Regardless of the mechanism, diversifi-
cation of E3 ligase function by MAGEs expands the regulatory
potential of these proteins and allows contextual changes in
multiple pathways through altering formation of distinct
MRLs. In the case of NSE1, differential binding of MAGE-G1
and MAGE-F1 not only diversifies function but also promotes
opposing activities, with MAGE-G1-NSE1 promoting DNA
repair through SMC5/6 complex and MAGE-F1-NSE1 inhibiting
DNA repair through degradation of MMS19.
Context-Dependent Regulation of Flux through the CIAPathwayUnder what specific contexts may cells (or organisms) need to
regulate MMS19 and the CIA pathway? Insufficient cellular iron
content may be one setting in which it would be beneficial to
decrease the production of MMS19. In the case of E. coli,
when iron becomes limiting, the bacterium reduces the tran-
script levels of highly expressed Fe-S proteins (Imlay, 2006).
Experiments from yeast demonstrated that limiting available
iron using the cell-membrane-impermeable ferrous (Fe2+) iron
chelator bathophenanthroline disulfonate caused a robust
decrease in MMS19 protein levels (Lev et al., 2013). This
decrease could be rescued by addition of the MG132 protea-
some inhibitor, suggesting that MMS19 is post-translationally
regulated in the context of low iron availability in yeast (Lev
et al., 2013). Alternatively, acute insults to DNA by carcinogens
or the environment may necessitate the further stabilization of
MMS19 levels to support the repair of widespread DNA damage.
Consistent with this, MMS19 protein levels have been shown to
be increased after UV damage (Ito et al., 2010).
In addition, it is important to consider that regulation of
MMS19 alters not only production of DNA repair enzymes but
also iron uptake by the transferrin receptor. Thus, regulation of
MMS19 allows cells (and organisms) to balance iron uptake
and DNA repair capacity. One setting in which generation of
DNA repair machinery components may be inhibited in favor of
iron uptake is when host defense against pathogens is neces-
sary. Many bacteria and viruses require iron to proliferate, and
thus, a key host-defense mechanism is to sequester iron away
from pathogen access (Ganz and Nemeth, 2015). Normally,
10 Molecular Cell 69, 1–13, January 4, 2018
upon infection, iron is sequestered by preventing cellular iron
export, sequestration of plasma iron into proteins, and import
of iron into cells (Ganz and Nemeth, 2015). Import of iron into
cells away from extracellular pathogens can occur in part
through uptake of iron from transferrin through the transferrin re-
ceptor, which is impacted by MMS19 (Soares and Weiss, 2015).
Indeed, rats treated with lipopolysaccharide (LPS) to promote an
immune response were found to have increased transferrin re-
ceptor mRNA and protein levels in the lung (Upton et al., 2003).
The importance of this defense mechanism is highlighted by
the high pathogen susceptibility of patients with hereditary he-
mochromatosis, characterized by systemic iron overload due
to mutations in various genes required to regulate iron homeo-
stasis, including TFR2 (Johnson and Wessling-Resnick, 2012).
Therefore, in the setting of extracellular infection, it may be bene-
ficial for cells to upregulate the transferrin receptor via downre-
gulation of MMS19.
The CIA Pathway, MAGE-F1, and Smoking-InducedCancersMany of the cancers in which MAGE-F1 is amplified are
frequently associated with smoking: lung squamous cell carci-
noma, esophageal carcinoma (Cook et al., 2010), and head
and neck squamous cell carcinoma (Mashberg et al., 1993).
Lesions caused by smoking include bulky adduct formation
on purines, such as 8-oxo-guanine, which may be repaired
through nucleotide excision repair pathways (Pleasance
et al., 2010a). It has been observed that, indeed, there are
more unrepaired lesions in cancers caused by smoking (Alex-
androv et al., 2013; Martincorena and Campbell, 2015; Pleas-
ance et al., 2010b). MMS19 was initially identified as a
gene important for both global nucleotide excision repair
and transcription-coupled repair. Thus, MAGE-F1 amplification
and overexpression may contribute to increased tumor muta-
tional burden through downregulation of MMS19. Given that
MMS19 interacts with and possibly regulates several core
DNA replication polymerases, it at first seems counter-intuitive
to the notion that the MMS19 ubiquitin ligase MAGE-F1-NSE1
is amplified in cancer. However, MAGE-F1-NSE1 downregula-
tion of MMS19 and the CIA pathway may in part foster tumor-
igenesis through promoting replication stress, leading to
genomic instability, but not to a degree that impairs DNA repli-
cation. Thus, dampening of the CIA pathway by MAGE-F1
amplification may be an important factor in promoting genomic
instability through replicative stress and downregulation of
DNA repair pathways.
Extensive efforts have been made to develop immune check-
point inhibitors, as cancer therapeutics and their clinical benefit
have been unprecedented. However, precisely why some pa-
tients respond remarkably well and others more modestly has
been of intensive study. Several studies have attributed these
differences to the abundance of neoantigens presented to the
competent immune cells after delivery of checkpoint inhibitors
(Rizvi et al., 2015; Van Allen et al., 2015). Given our findings
that MAGE-F1 amplification is associated with increased muta-
tional burden (increased neoantigens) in tumors, MAGE-F1
amplification could be a predictive biomarker for immune check-
point inhibitors.
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Evolution of MAGE-F1 Gene to Alter MMS19, CIAPathway, and Iron HomeostasisThroughout the course of evolution, the MAGE-F1 gene appears
to have undergone multiple insertion and deletion events to
convert it from a protein-coding gene to a pseudogene. What
is the cause for this rapid gain and loss across species? One
plausible theory is that, because many immune-response-
related genes evolve quickly due to the nature of host-pathogen
competition (Barreiro andQuintana-Murci, 2010), MAGE-F1may
harbor roles in pathogen defense through altering cellular iron
homeostasis. With these considerations, MAGE-F1 may be
required in species that are exposed to certain extracellular bac-
terial pathogens. Alternatively, MAGE-F1may have been actively
selected against to protect against heavy exposure to intracel-
lular viruses or to preserve a more efficient DNA repair process
in species with high or long-term exposure to DNA damage
over time. An interesting observation is that species with
MAGE-F1 pseudogenes include those that have been noted to
have relatively long lifespans with low observed cancer-associ-
ated mortality, such as elephants (lifespan: 60–80+ years; mor-
tality from cancer: 5%; Abegglen et al., 2015; Lahdenper€a
et al., 2014), whales (bowhead lifespan: �200 years; mortality
from cancer: unknown but is presumed to be low due to their
incredible lifespan; Keane et al., 2015), and naked mole rats (life-
span:�32 years; mortality from cancer: only 2 natural cases ever
recorded; Rodriguez et al., 2012), whereas animals, such as cats
and dogs, that possess the full-length coding MAGE-F1 appear
to have incidences of cancers similar to humans (maximum life-
span recorded: 122 years; mortality from cancer: 11%–25%;
Abegglen et al., 2015; Dong et al., 2016). Although part of the un-
derlying reason for cancer resistance in some species has been
noted, such as abundance of TP53 copies in elephants (Abeg-
glen et al., 2015), it will be interesting to see whether the adaptive
pseudogenization of MAGE-F1 and its effect on DNA repair ca-
pacity and iron homeostasis also contribute to lower incidence
of cancer in these animals.
STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
d KEY RESOURCES TABLE
d CONTACT FOR REAGENT AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
B Cell Lines and Culture Conditions
B Animals
d METHOD DETAILS
B Antibodies Production and CRISPR/Cas9 Knockouts
B RNA Preparation and Quantitative Reverse Transcrip-
tion PCR Analysis (qRT-PCR)
B 3xFLAG-Tagged Pull-Down
B Co-immunoprecipitation and Immunoblotting
B Purification of Recombinant Proteins
B In Vitro Binding Assays
B Ubiquitination Assays
B Homologous Recombination Assay
B55Fe Incorporation Assay
B Transferrin Uptake Assay
B Cell Viability Assays
B MAGE-F1 Sequence Analysis
B Assessment of mRNA/Copy-Number Analysis in Hu-
man Tumors and Statistical Analysis
B Xenograft Tumor Growth Assays
d QUANTIFICATION AND STATISTICAL ANALYSIS
d DATA AND SOFTWARE AVAILABILITY
SUPPLEMENTAL INFORMATION
Supplemental Information includes five figures and one table and can be found
with this article online at https://doi.org/10.1016/j.molcel.2017.11.010.
AUTHOR CONTRIBUTIONS
J.L.W., S.W.Y., and P.R.P. conducted the experiments; J.L.W., S.W.Y., and
P.R.P. designed the experiments; and J.L.W. and P.R.P. wrote the paper.
ACKNOWLEDGMENTS
We thank members of the Potts lab for helpful discussions and critical reading
of the manuscript. We would like to thank Dr. Zhijian ‘‘James’’ Chen for ubiq-
uitin constructs, Dr. Matthew Porteus for homologous recombination-gene-
targeting cells and constructs, Dr. John Minna for lung cancer cell lines, and
Dr. Ezra Burstein for HeLa-Cas9 cells.We also thank Dr. Yi Liu, Dr. JohnMinna,
and Dr. Benjamin Tu for guidance. This work was supported by Worldwide
Cancer Research grant 15-0177 (P.R.P.), American Cancer Society Research
Scholar Award 181691010 (P.R.P.), and UT Southwestern Medical Scientist
Training Program (J.L.W.).
Received: July 7, 2017
Revised: October 4, 2017
Accepted: November 8, 2017
Published: December 7, 2017
REFERENCES
Abegglen, L.M., Caulin, A.F., Chan, A., Lee, K., Robinson, R., Campbell, M.S.,
Kiso, W.K., Schmitt, D.L., Waddell, P.J., Bhaskara, S., et al. (2015). Potential
mechanisms for cancer resistance in elephants and comparative cellular
response to DNA damage in humans. JAMA 314, 1850–1860.
Alexandrov, L.B., Nik-Zainal, S., Wedge, D.C., Aparicio, S.A., Behjati, S.,
Biankin, A.V., Bignell, G.R., Bolli, N., Borg, A., Børresen-Dale, A.L., et al.;
Australian Pancreatic Cancer Genome Initiative; ICGC Breast Cancer
Consortium; ICGC MMML-Seq Consortium; ICGC PedBrain (2013).
Signatures of mutational processes in human cancer. Nature 500, 415–421.
Barreiro, L.B., and Quintana-Murci, L. (2010). From evolutionary genetics to
human immunology: how selection shapes host defence genes. Nat. Rev.
Genet. 11, 17–30.
Beinert, H. (2000). Iron-sulfur proteins: ancient structures, still full of surprises.
J. Biol. Inorg. Chem. 5, 2–15.
Cook, J.D., Kondapalli, K.C., Rawat, S., Childs, W.C., Murugesan, Y., Dancis,
A., and Stemmler, T.L. (2010). Molecular details of the yeast frataxin-Isu1 inter-
action during mitochondrial Fe-S cluster assembly. Biochemistry 49,
8756–8765.
De Donato, M., Peters, S.O., Hussain, T., Rodulfo, H., Thomas, B.N., Babar,
M.E., and Imumorin, I.G. (2017). Molecular evolution of type II MAGE genes
from ancestral MAGED2 gene and their phylogenetic resolution of basal
mammalian clades. Mamm. Genome. Published online May 17, 2017.
https://doi.org/10.1007/s00335-017-9695-6.
Dong, X., Milholland, B., and Vijg, J. (2016). Evidence for a limit to human life-
span. Nature 538, 257–259.
Molecular Cell 69, 1–13, January 4, 2018 11
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Doyle, J.M., Gao, J., Wang, J., Yang, M., and Potts, P.R. (2010). MAGE-RING
protein complexes comprise a family of E3 ubiquitin ligases. Mol. Cell 39,
963–974.
Ganz, T., and Nemeth, E. (2015). Iron homeostasis in host defence and inflam-
mation. Nat. Rev. Immunol. 15, 500–510.
Gari, K., Leon Ortiz, A.M., Borel, V., Flynn, H., Skehel, J.M., and Boulton, S.J.
(2012). MMS19 links cytoplasmic iron-sulfur cluster assembly to DNA meta-
bolism. Science 337, 243–245.
Hao, Y.H., Doyle, J.M., Ramanathan, S., Gomez, T.S., Jia, D., Xu, M., Chen,
Z.J., Billadeau, D.D., Rosen, M.K., and Potts, P.R. (2013). Regulation of
WASH-dependent actin polymerization and protein trafficking by ubiquitina-
tion. Cell 152, 1051–1064.
Hao, Y.H., Fountain, M.D., Jr., Fon Tacer, K., Xia, F., Bi, W., Kang, S.H., Patel,
A., Rosenfeld, J.A., Le Caignec, C., Isidor, B., et al. (2015). USP7 acts as a mo-
lecular rheostat to promote WASH-dependent endosomal protein recycling
and is mutated in a human neurodevelopmental disorder. Mol. Cell 59,
956–969.
Hocke, S., Guo, Y., Job, A., Orth, M., Ziesch, A., Lauber, K., De Toni, E.N.,
Gress, T.M., Herbst, A., Goke, B., and Gallmeier, E. (2016). A synthetic lethal
screen identifies ATR-inhibition as a novel therapeutic approach for POLD1-
deficient cancers. Oncotarget 7, 7080–7095.
Imlay, J.A. (2006). Iron-sulphur clusters and the problem with oxygen. Mol.
Microbiol. 59, 1073–1082.
Ito, S., Tan, L.J., Andoh, D., Narita, T., Seki, M., Hirano, Y., Narita, K., Kuraoka,
I., Hiraoka, Y., and Tanaka, K. (2010). MMXD, a TFIIH-independent XPD-
MMS19 protein complex involved in chromosome segregation. Mol. Cell 39,
632–640.
Johnson, E.E., and Wessling-Resnick, M. (2012). Iron metabolism and the
innate immune response to infection. Microbes Infect. 14, 207–216.
Katsura, Y., and Satta, Y. (2011). Evolutionary history of the cancer immunity
antigen MAGE gene family. PLoS ONE 6, e20365.
Keane, M., Semeiks, J., Webb, A.E., Li, Y.I., Quesada, V., Craig, T., Madsen,
L.B., van Dam, S., Brawand, D., Marques, P.I., et al. (2015). Insights into the
evolution of longevity from the bowhead whale genome. Cell Rep. 10,
112–122.
Korber, B. (2000). HIV Signature and Sequence Variation Analysis, Chapter 4
(Kluwer Academic Publishers).
Lahdenper€a, M., Mar, K.U., and Lummaa, V. (2014). Reproductive cessation
and post-reproductive lifespan in Asian elephants and pre-industrial humans.
Front. Zool. 11, 54.
Lauder, S., Bankmann, M., Guzder, S.N., Sung, P., Prakash, L., and Prakash,
S. (1996). Dual requirement for the yeast MMS19 gene in DNA repair and RNA
polymerase II transcription. Mol. Cell. Biol. 16, 6783–6793.
Lee, A.K., and Potts, P.R. (2017). A comprehensive guide to the MAGE family
of ubiquitin ligases. J. Mol. Biol. 429, 1114–1142.
Lev, I., Volpe, M., Goor, L., Levinton, N., Emuna, L., and Ben-Aroya, S. (2013).
Reverse PCA, a systematic approach for identifying genes important for the
physical interaction between protein pairs. PLoS Genet. 9, e1003838.
Lill, R. (2009). Function and biogenesis of iron-sulphur proteins. Nature 460,
831–838.
Lill, R., Dutkiewicz, R., Freibert, S.A., Heidenreich, T., Mascarenhas, J., Netz,
D.J., Paul, V.D., Pierik, A.J., Richter, N., St€umpfig, M., et al. (2015). The role of
mitochondria and the CIAmachinery in the maturation of cytosolic and nuclear
iron-sulfur proteins. Eur. J. Cell Biol. 94, 280–291.
Litman, R., Peng, M., Jin, Z., Zhang, F., Zhang, J., Powell, S., Andreassen,
P.R., and Cantor, S.B. (2005). BACH1 is critical for homologous recombination
and appears to be the Fanconi anemia gene product FANCJ. Cancer Cell 8,
255–265.
Maloisel, L., Fabre, F., and Gangloff, S. (2008). DNA polymerase delta is pref-
erentially recruited during homologous recombination to promote heterodu-
plex DNA extension. Mol. Cell. Biol. 28, 1373–1382.
12 Molecular Cell 69, 1–13, January 4, 2018
Martincorena, I., and Campbell, P.J. (2015). Somatic mutation in cancer and
normal cells. Science 349, 1483–1489.
Mashberg, A., Boffetta, P., Winkelman, R., and Garfinkel, L. (1993). Tobacco
smoking, alcohol drinking, and cancer of the oral cavity and oropharynx
among U.S. veterans. Cancer 72, 1369–1375.
Nei, M., and Gojobori, T. (1986). Simple methods for estimating the numbers of
synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3,
418–426.
Netz, D.J., Mascarenhas, J., Stehling, O., Pierik, A.J., and Lill, R. (2014).
Maturation of cytosolic and nuclear iron-sulfur proteins. Trends Cell Biol. 24,
303–312.
Paul, V.D., and Lill, R. (2015). Biogenesis of cytosolic and nuclear iron-sulfur
proteins and their role in genome stability. Biochim. Biophys. Acta 1853,
1528–1539.
Pineda, C.T., and Potts, P.R. (2015). Oncogenic MAGEA-TRIM28 ubiquitin
ligase downregulates autophagy by ubiquitinating and degrading AMPK in
cancer. Autophagy 11, 844–846.
Pineda, C.T., Ramanathan, S., Fon Tacer, K.,Weon, J.L., Potts,M.B., Ou, Y.H.,
White, M.A., and Potts, P.R. (2015). Degradation of AMPK by a cancer-specific
ubiquitin ligase. Cell 160, 715–728.
Pleasance, E.D., Stephens, P.J., O’Meara, S., McBride, D.J., Meynert, A.,
Jones, D., Lin, M.L., Beare, D., Lau, K.W., Greenman, C., et al. (2010a).
A small-cell lung cancer genome with complex signatures of tobacco expo-
sure. Nature 463, 184–190.
Pleasance, E.D., Cheetham, R.K., Stephens, P.J., McBride, D.J., Humphray,
S.J., Greenman, C.D., Varela, I., Lin, M.L., Ordonez, G.R., Bignell, G.R.,
et al. (2010b). A comprehensive catalogue of somatic mutations from a human
cancer genome. Nature 463, 191–196.
Porteus, M.H., and Baltimore, D. (2003). Chimeric nucleases stimulate gene
targeting in human cells. Science 300, 763.
Potts, P.R. (2009). The Yin and Yang of the MMS21-SMC5/6 SUMO ligase
complex in homologous recombination. DNA Repair (Amst.) 8, 499–506.
Potts, P.R., and Yu, H. (2005). Human MMS21/NSE2 is a SUMO ligase
required for DNA repair. Mol. Cell Biol. 25, 7021–7032.
Potts, P.R., and Yu, H. (2007). The SMC5/6 complexmaintains telomere length
in ALT cancer cells through SUMOylation of telomere-binding proteins. Nat.
Struct. Mol. Biol. 14, 581–590.
Potts, P.R., Porteus, M.H., and Yu, H. (2006). Human SMC5/6 complex pro-
motes sister chromatid homologous recombination by recruiting the
SMC1/3 cohesin complex to double-strand breaks. EMBO J. 25, 3377–3388.
Prakash, L., and Prakash, S. (1979). Three additional genes involved in pyrim-
idine dimer removal in Saccharomyces cerevisiae: RAD7, RAD14 andMMS19.
Mol. Gen. Genet. 176, 351–359.
Rizvi, N.A., Hellmann, M.D., Snyder, A., Kvistborg, P., Makarov, V., Havel, J.J.,
Lee, W., Yuan, J., Wong, P., Ho, T.S., et al. (2015). Cancer immunology.
Mutational landscape determines sensitivity to PD-1 blockade in non-small
cell lung cancer. Science 348, 124–128.
Rodriguez, K.A., Edrey, Y.H., Osmulski, P., Gaczynska, M., and Buffenstein, R.
(2012). Altered composition of liver proteasome assemblies contributes to
enhanced proteasome activity in the exceptionally long-lived naked mole-rat.
PLoS ONE 7, e35890.
Rouault, T.A. (2015). Mammalian iron-sulphur proteins: novel insights into
biogenesis and function. Nat. Rev. Mol. Cell Biol. 16, 45–55.
Schertzer, M., Jouravleva, K., Perderiset, M., Dingli, F., Loew, D., Le Guen,
T., Bardoni, B., de Villartay, J.P., Revy, P., and Londono-Vallejo, A. (2015).
Human regulator of telomere elongation helicase 1 (RTEL1) is required for
the nuclear and cytoplasmic trafficking of pre-U2 RNA. Nucleic Acids Res.
43, 1834–1847.
Seki, M., Takeda, Y., Iwai, K., and Tanaka, K. (2013). IOP1 protein is an external
component of the human cytosolic iron-sulfur cluster assembly (CIA) machin-
ery and functions in the MMS19 protein-dependent CIA pathway. J. Biol.
Chem. 288, 16680–16689.
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Soares, M.P., and Weiss, G. (2015). The iron age of host-microbe interactions.
EMBO Rep. 16, 1482–1500.
Stehling, O., Els€asser, H.P., Br€uckel, B., M€uhlenhoff, U., and Lill, R. (2004).
Iron-sulfur protein maturation in human cells: evidence for a function of fra-
taxin. Hum. Mol. Genet. 13, 3007–3015.
Stehling, O., Netz, D.J., Niggemeyer, B., Rosser, R., Eisenstein, R.S., Puccio,
H., Pierik, A.J., and Lill, R. (2008). Human Nbp35 is essential for both cytosolic
iron-sulfur protein assembly and iron homeostasis. Mol. Cell. Biol. 28,
5517–5528.
Stehling, O., Vashisht, A.A., Mascarenhas, J., Jonsson, Z.O., Sharma, T., Netz,
D.J., Pierik, A.J., Wohlschlegel, J.A., and Lill, R. (2012). MMS19 assembles
iron-sulfur proteins required for DNA metabolism and genomic integrity.
Science 337, 195–199.
Stehling, O., Mascarenhas, J., Vashisht, A.A., Sheftel, A.D., Niggemeyer, B.,
Rosser, R., Pierik, A.J., Wohlschlegel, J.A., and Lill, R. (2013). Human
CIA2A-FAM96A and CIA2B-FAM96B integrate iron homeostasis and matura-
tion of different subsets of cytosolic-nuclear iron-sulfur proteins. Cell Metab.
18, 187–198.
Taylor, E.M., Copsey, A.C., Hudson, J.J., Vidot, S., and Lehmann, A.R. (2008).
Identification of the proteins, including MAGEG1, that make up the human
SMC5-6 protein complex. Mol. Cell. Biol. 28, 1197–1206.
Teichmann, R., and Stremmel, W. (1990). Iron uptake by human upper small
intestine microvillous membrane vesicles. Indication for a facilitated transport
mechanism mediated by a membrane iron-binding protein. J. Clin. Invest. 86,
2145–2153.
Upton, R.L., Chen, Y., Mumby, S., Gutteridge, J.M., Anning, P.B., Nicholson,
A.G., Evans, T.W., and Quinlan, G.J. (2003). Variable tissue expression of
transferrin receptors: relevance to acute respiratory distress syndrome. Eur.
Respir. J. 22, 335–341.
Van Allen, E.M., Miao, D., Schilling, B., Shukla, S.A., Blank, C., Zimmer, L.,
Sucker, A., Hillen, U., Foppen, M.H.G., Goldinger, S.M., et al. (2015).
Genomic correlates of response to CTLA-4 blockade in metastatic melanoma.
Science 350, 207–211.
vanWietmarschen, N., Moradian, A., Morin, G.B., Lansdorp, P.M., and Uringa,
E.J. (2012). Themammalian proteinsMMS19,MIP18, and ANT2 are involved in
cytoplasmic iron-sulfur cluster protein assembly. J. Biol. Chem. 287,
43351–43358.
Weon, J.L., and Potts, P.R. (2015). The MAGE protein family and cancer. Curr.
Opin. Cell Biol. 37, 1–8.
Yalcin, B., Wong, K., Agam, A., Goodson, M., Keane, T.M., Gan, X., Nellaker,
C., Goodstadt, L., Nicod, J., Bhomra, A., et al. (2011). Sequence-based char-
acterization of structural variation in the mouse genome. Nature 477, 326–329.
Yang, B., O’Herrin, S.M., Wu, J., Reagan-Shaw, S., Ma, Y., Bhat, K.M.,
Gravekamp, C., Setaluri, V., Peters, N., Hoffmann, F.M., et al. (2007).
MAGE-A, mMage-b, and MAGE-C proteins form complexes with KAP1 and
suppress p53-dependent apoptosis in MAGE-positive cell lines. Cancer
Res. 67, 9954–9962.
Zhao, Q., Caballero, O.L., Simpson, A.J., and Strausberg, R.L. (2012).
Differential evolution of MAGE genes based on expression pattern and selec-
tion pressure. PLoS ONE 7, e48240.
Zou, J., Tian, F., Li, J., Pickner, W., Long, M., Rezvani, K., Wang, H., and
Zhang, D. (2013). FancJ regulates interstrand crosslinker induced centrosome
amplification through the activation of polo-like kinase 1. Biol. Open 2,
1022–1031.
Molecular Cell 69, 1–13, January 4, 2018 13
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
STAR+METHODS
KEY RESOURCES TABLE
REAGENT or RESOURCE SOURCE IDENTIFIER
Antibodies
Anti-Actin Abcam AB6276; RRID: AB_2223210
Anti-DPYD Santa Cruz Biotechnology SC-50521; RRID: AB_2094323
Anti-FANCJ Bethyl Laboratories A300-561A; RRID: AB_2066311
Anti-FLAG Sigma F3165; RRID: AB_259529
Anti-GAPDH Cell Signaling Technology 2118; RRID: AB_561053
Anti-HA Roche 11666606001; RRID: AB_514506
Anti-IOP1 Sigma SAB4502760; RRID: AB_10746098
Anti-MMS19 Proteintech 16015-1-AP; RRID: AB_2145043
Anti-Myc Roche 9E10
Anti-POLD1 Santa Cruz Biotechnology SC-17776; RRID: AB_675487
Anti-RTEL1 Santa Cruz Biotechnology SC-85900; RRID: AB_2301126
Anti-SMC5 Bethyl Laboratories A300-236A; RRID: AB_2192785
Anti-SMC6 Bethyl Laboratories A300-237A; RRID: AB_263353
Anti-XPD Abcam AB47186; RRID: AB_1143931
Anti-MAGE-F1 This paper N/A
Donkey Anti-Rabbit IgG GE NA934V
Sheep Anti-Mouse IgG GE NA931V
Bacterial and Virus Strains
DH5a Competent Cells Thermo Fisher Scientific 18265017
XL1-blue Competent Cells Agilent 200130
Chemicals, Peptides, and Recombinant Proteins
ECL detection reagent GE RPN2209
ECL prime detection reagent GE RPN2236
Protein A beads Bio-Rad 1560005
Effectene transfection reagent QIAGEN 301425
RNAStat60 TelTest Cs-112
Anti-FLAG M2 Beads Sigma A2220
PR-619 Sigma SML0430
TALON Metal Affinity Resin Clontech 635501
Lipofectamine RNAiMAX Thermo Fisher Scientific 13778030
Critical Commercial Assays
High Capacity cDNA Reverse Transcription kit Thermo Fisher Scientific 4368813
Wizard Genomic DNA Purification Kit Promega A1120
TNT SP6 Quick In Vitro Transcription/Translation Kit Promega L2080
BCA protein assay kit Thermo Fisher Scientific 23227
Experimental Models: Cell Lines
HEK293 ATCC CRL-1573
HEK293/A658 Potts and Yu, 2005 N/A
HeLa Tet-ON Clontech 631183
HeLa-Cas9 Gift from Ezra Burstein UT
Southwestern
N/A
HCC95 Gift John Minna UT Southwestern N/A
H520 Gift John Minna UT Southwestern N/A
(Continued on next page)
e1 Molecular Cell 69, 1–13.e1–e6, January 4, 2018
Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
H2170 Gift John Minna UT
Southwestern
N/A
MAGE-F1 KO HeLa This paper N/A
Experimental Models: Organisms/Strains
NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ: NOD scid gamma mice The Jackson Laboratory 005557; RRID: IMSR_JAX:005557
Oligonucleotides
siRNA targeting sequence: siControl: ACUACAUCGUGA
UUCAAACUU
This paper N/A
siRNA targeting sequence: siFANCJ #1: AGUCAAGAGU
CAUCGAAUA
Zou et al., 2013 N/A
siRNA targeting sequence: siFANCJ #2: UAACCCAAGU
CGCUAUAUA
Zou et al., 2013 N/A
siRNA targeting sequence: siMAGE-F1 #1: GGUGCAAC
CCUCAAAGUAU
This paper N/A
siRNA targeting sequence: siMAGE-F1 #2: CGAAGAGG
CUUAUUAUGGA
This paper N/A
siRNA targeting sequence: siMMS19 #1: AGGCCCUAG
UGCUCAGAUA
This paper N/A
siRNA targeting sequence: siMMS19 #2: GACUCUGAA
UGCUUGCUGU
This paper N/A
siRNA targeting sequence: siNSE1 #1: GGAACUGAUUA
UUGACUCA
This paper N/A
siRNA targeting sequence: siPOLD1 #1: CGGGACCAGG
GAGAAUUAAUA
Hocke et al., 2016 N/A
siRNA targeting sequence: siPOLD1 #2: CAGUUGGAGA
UUGACCAUUAU
Hocke et al., 2016 N/A
siRNA targeting sequence: siRTEL1 #1: GCCUGUGUGU
GGAGUAUGA
Schertzer et al., 2015 N/A
siRNA targeting sequence: siRTEL1 #2: GACCAUCAGU
GCUUACUAU
Schertzer et al., 2015 N/A
siRNA targeting sequence: siXPD1 #1: GAAGUUGCUC
AACUUCUA
This paper N/A
siRNA targeting sequence: siXPD1 #2: CAUACUUCCU
UGCUCGAUA
This paper N/A
siRNA targeting sequence: siXPD1 #3: CAGAGAUUGA
GAAGGUGAU
This paper N/A
sgRNA targeting sequence: MAGE-F1 #1: CUCCCGGUC
CCGCAGGCCGAGUUUUAGAGCUAUGCUGUUUUG
This paper N/A
sgRNA targeting sequence: MAGE-F1 #1: CUAGGGCCG
GCAUCCACCUCGUUUUAGAGCUAUGCUGUUUUG
This paper N/A
Amplification primers: human MAGE-F1 Forward: AGCTC
CCGCTGCCATTGCTCCTTGTAC
This paper N/A
Amplification primers: human MAGE-F1 Reverse: TCGCC
CCACCCATATTACTTATGACTCAGG
This paper N/A
Amplification primers: mouse MAGE-F1 Forward: GATCG
GCCGGCCAATGATGCTCCCGCCGCCGTC
This paper N/A
Amplification primers: mouse MAGE-F1 Reverse: GATCG
GCGCGCCTCACCATTTTCTCAGAAATG
This paper N/A
qPCR primers: MAGE-F1 Forward: AGTACCGTGAGGCC
CTAGC
This paper N/A
qPCR primers: MAGE-F1 Reverse: TCATACTGGCTTCAG
CTCTGG
This paper N/A
(Continued on next page)
Molecular Cell 69, 1–13.e1–e6, January 4, 2018 e2
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
qPCR primers: MMS19 Forward: AATCCAGCTTTTGTCAC
AGGTG
This paper N/A
qPCR primers: MMS19 Reverse: AGTATCAGGTGTACCAC
TTCCTT
This paper N/A
qPCR primers: 18S rRNA Forward: ACCGCAGCTAG
GAATAATGGA
This paper N/A
qPCR primers: 18S rRNA Reverse: GCCTCAGTTCC
GAAAACCA
This paper N/A
Software and Algorithms
GraphPad Prism 5 GraphPad https://www.graphpad.com
Deposited Data
Original imaging data Mendeley https://doi.org/10.17632/f8842jrv8z.1
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
CONTACT FOR REAGENT AND RESOURCE SHARING
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, P. Ryan
Potts ([email protected]).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Cell Lines and Culture ConditionsHEK293, HEK293/A658, HeLa Tet-ON (Clontech), and HeLa-Cas9 stable cells were grown in DMEM supplemented with 10% FBS,
2mML-glutamine, 100 units/mL penicillin, 100mg/mL streptomycin, and 0.25mg/mL amphotericin B. HeLa-Cas9 stable cells were a
gift from Dr. Ezra Burstein (UT Southwestern). HCC95, H520, and H2170 cells were grown in RPMI supplemented with 5% heat
inactivated serum.
Animals6-8 week old male NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NOD scid gamma) mice from Jackson Labs were used for xenograft growth
assays. All studies were approved by the St. Jude Children’s Research Hospital institutional review committee on animal safety.
METHOD DETAILS
Antibodies Production and CRISPR/Cas9 KnockoutsRabbit polyclonal antibodies were generated against MAGE-F1 (first 59 amino acids fromN terminus) (YenZym) andNSE1 (full-length
protein) (Cocalico Biologicals, Inc). For generation of HeLa-Cas9 MAGE-F1 knockout cells, synthetic tracrRNA
(Dharmacon #U-002000-20) and custom crRNAs targeting near the 30 and 50 ends of MAGE-F1 were purchased from Dharmacon.
Both tracrRNA and custom crRNAs were resuspended in sterile Tris buffer pH 7.4 to 10 uM. HeLa-Cas9 MAGE-F1 knockout cells
were generated by plating 1.2x106 HeLa-Cas9 stable cells in 10 cm plates. 24 hours later, the tracrRNA and 2 crRNAs were trans-
fected using Lipofectamine RNAiMAX. A 1:2 ratio of RNA:RNAiMAX was utilized, and tracrRNA and total crRNA were utilized in a
1:1 ratio (total 20 nM final concentration in 10 cm plate) and transfected as per manufacturer’s instructions. Media was changed
at 24 hours after transfection and cell were expanded after 48 hours. Cells were then diluted to single cell density per 2 wells in
96-well plates. Clones were expanded and then tested for knockout by harvesting genomic DNA via Wizard Genomic DNA Purifica-
tion Kit (Promega) as per manufacturer’s instructions and PCR amplifying for MAGE-F1. Loss of MAGE-F1 expression was also vali-
dated by qPCR.
RNA Preparation and Quantitative Reverse Transcription PCR Analysis (qRT-PCR)RNAwas extracted from cultured cells using RNAStat60 (TelTest) according tomanufacturer’s instructions and subsequently treated
with DNase I (Roche) and converted to cDNA utilizing reagents from High Capacity cDNA Reverse Transcription kit (Life Technolo-
gies). cDNA from cells were plated in triplicate in a 384-well plate and expression of genes of interest was measured using SYBR
Green. Primer validation and analysis of qRT-PCR data were performed as previously published (Pineda et al., 2015).
3xFLAG-Tagged Pull-DownHEK293 cells stably expressing 3xFLAG-MAGE-F1 or 3xFLAG-vector were grown to confluency in ten 15 cm plates each. Cells
were washed and scraped in cold PBS, spun down, and resuspended in FLAG buffer (50 mM Tris pH 7.4, 150 mM NaCl, 1 mM
e3 Molecular Cell 69, 1–13.e1–e6, January 4, 2018
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
EDTA, 1% Triton X-100, 1X protease inhibitor cocktail) for 30minutes on ice. Lysates were dounced, spun down, and filtered through
a 0.45 mmfilter and incubated with anti-FLAGM2 beads (Sigma-Aldrich A2220) for 1 hour rotating at 4�C. Beads were then washed in
1X TBS 4 times and eluted for 5 fractions with elution buffer (150 pg/ml FLAG peptide in 1X TBS) at 4�C rotating, 30 minutes each.
Eluted fractions with protein were then prepared with SDS sample buffer and run on SDS-PAGE gel until minimally resolved prior
to submitting entire sample for mass spec analysis.
Co-immunoprecipitation and Immunoblotting4x105 HeLa cells were plated in 6 cm dishes and transfected 24 hours later with Effectene (QIAGEN) according to manufacturers’
protocol. 16 hr post-transfection, media was changed. 48 hr post-transfection, cells were washed and scraped in cold PBS, spun
down, and resuspended in NP-40 lysis buffer (50 mM Tris-HCl pH 7.7, 150 mM NaCl, 0.5% NP-40 (v/v), 1 mM dithiothreitol (DTT),
1X protease inhibitor cocktail) for 45 minutes on ice prior to spinning down insoluble material. Soluble lysate was incubated with
respective antibodies conjugated to Protein A beads (Bio-Rad) for 2 hours at 4�C while rotating. Beads were then washed in
NP-40 lysis buffer 3-5 times and eluted with 2X sodium dodecyl sulfate (SDS) sample buffer.
For immunoblotting, samples prepared in SDS sample buffer were resolved on SDS-PAGE gels and then transferred to nitrocel-
lulose membranes prior to blocking in TBST with 5%milk (w/v) or 5% bovine serum albumin (w/v) and probing with primary (as indi-
cated above) and secondary antibodies (donkey anti-rabbit IgG, GE Healthcare, NA934V; sheep anti-mouse IgG, GE Healthcare,
NA931V). Protein signal was visualized after addition of ECL detection reagent (GE Healthcare, RPN2209; GE Healthcare,
RPN2236) as per manufacturer’s instructions.
Purification of Recombinant ProteinsHis-NSE1 and His-MAGE-F1-NSE1 complex were purified from 6 L LB cultures of BL21(DE3) or Rosetta 2(DE3) competent cells
(EMD Millipore), respectively. Bacterial pellets were lysed in high salt lysis buffer (50 mM Tris-HCl pH 7.7, 500 mM NaCl, 100 mM
ZnCl2, 10 mM imidazole), sonicated, and spun down at 2x104 RPM for 1 hour. Soluble lysate was filtered through a 0.45 mm filter
and incubated with His Select Nickel Affinity Gel (Sigma, P6611) for 1 hour at 4�C rotating. Beads were then washed with high
salt wash buffer (50 mM Tris-HCl pH 7.7, 500 mM NaCl, 100 mM ZnCl2, 20 mM imidazole) and eluted with 2 different elution buffers
(50 mM Tris-HCl pH 7.7, 500 mM NaCl, 100 mM ZnCl2, 140 or 200 mM imidazole).
In Vitro Binding Assays30 mg of either His-NSE1 or His-MAGE-F1-NSE1 complex were incubated with 8.75 ml TALONMetal Affinity Resin (Clontech) in TBST
buffer (25 mM Tris pH 8.0, 2.7 mM KCl, 137 mM NaCl, 0.05% Tween-20 (v/v), and 10 mM 2-mercaptoethanol) for 1 hour vibrating at
room temperature. Beads were then blocked with 5% non-fat milk (w/v) powder in TBST for 1 hour vibrating at room temperature.
MYC-MMS19 was in vitro translated using TNT SP6 Quick In Vitro Transcription/Translation Kit (Promega) and added to the recom-
binant proteins in 5%milk (w/v) in TBST buffer vibrating for 1 hour. Beadswerewashed four timeswith TBST buffer prior to addition of
2X SDS sample buffer to elute.
Ubiquitination Assays4x105 HeLa cells were plated in 6 cmdishes and transfected 24 hours later with Effectene according tomanufacturers’ protocol. 48 hr
post-transfection, cells were treated for 4 hr with 10 mMMG132 before being collected in ice cold PBS. For anti-MYC-ubiquitin pull-
downs, cells were resuspended in RIPA buffer (150 mM NaCl, 5 mM EDTA pH 8.0, 50 mM Tris-HCl pH 8.0, 1% NP-40 (v/v), 0.5%
sodiumdeoxycholate (w/v), 0.1%SDS (w/v), 1mMDTT, and 1X protease inhibitor cocktail (Roche)) containing 10 mMPR-619 (Sigma),
insoluble fraction removed by centrifugation, and soluble protein incubated with anti-MYC conjugated Protein A beads (Bio-Rad) for
2 hr at 4�C while rotating, washed, and eluted with 2X SDS sample buffer. For denaturing His-ubiquitin pull-down cell pellets were
resuspended in denaturing lysis buffer (6M guanidinium-HCl, 100 mMNa2HPO4-NaH2PO4, 10 mM Tris-HCl pH 8.0, 5 mM imidazole,
10 mM 2-mercaptoethanol) and rotated for four hours at room temperature. Beads were spun down and washed successively with
the following buffers for 5 minutes each at room temperature: 1) denaturing lysis buffer without imidazole 2) buffer A pH 8.0 (8M urea,
100 mM Na2HPO4-NaH2PO4, 10 mM Tris-HCl pH 8.0, 10 mM 2-mercaptoethanol) 3) buffer A pH 6.3 with 0.2% Triton X-100 (v/v)
(8M urea, 100 mM Na2HPO4-NaH2PO4, 10 mM Tris-HCl pH 6.3, 10 mM beta-mercaptoethanol, 0.2% Triton X-100 (v/v)) 4) buffer
A pH 6.3 with 0.1% Triton X-100 (v/v) (8M urea, 100 mMNa2HPO4-NaH2PO4, 10 mM Tris-HCl pH 6.3, 10 mM beta-mercaptoethanol,
0.1% Triton X-100 (v/v)). Beads were then incubated with elution buffer (200 mM imidazole, 150 mM Tris-HCl pH 6.7, 30% glycerol
(v/v), 5% SDS (w/v), 720 mM 2-mercaptoethanol) for 20 minutes at room temperature. Eluted fraction was then prepared with SDS
sample buffer prior to running samples on SDS-PAGE gel.
Homologous Recombination AssayHR assays were performed essentially as described previously (Porteus and Baltimore, 2003; Potts et al., 2006). Briefly, a 293 cell line
(293/A658) expressing aGFP gene containing in-frame stop codons and a I-SceI recognition site (50-TAGGATAACAGGGTAAT-30) atbp 327. The I-SceI/repair plasmid contained: an I-SceI expression cassette and a 2100 bp repair substrate that contains a truncated
GFP gene (truncGFP) followed by an additional 1300 bp of 30-homology to the mutated GFP genomic target. HR was measured by
transfecting 293/A658 cells with the I-SceI/repair plasmid and the indicated plasmids or siRNA oligonucleotides. Cells were grown for
Molecular Cell 69, 1–13.e1–e6, January 4, 2018 e4
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
three days and the percentage of GFP-positive cells was measured by flow cytometry and normalized to transfection efficiency
controls.
55Fe Incorporation Assay55Fe incorporation assays were performed as described previously (Stehling et al., 2004; Teichmann and Stremmel, 1990). Briefly,
3.5x105 HeLa cells were plated in 6 cm dishes and transfected 24 hours later with Effectene according to manufacturer’s protocol.
72 hours after transfection cells were labeled with 2 mCi/mL 55Fe-NTA for 18 hours in DMEM supplemented with 3.75% FBS (v/v) and
150 mM ascorbate. 55Fe-NTA was prepared by incubation of 16 mM 55FeCl3 (Perkin Elmer) in 100 mM HCl, 63 mM nitrilotriacetic acid
(NTA), and 20 mM HEPES pH to 6.0 with Tris followed by titration to pH 7.0 with 100 mM NaOH. After 55Fe-NTA labeling, cells were
washed in complete media, ice-cold PBS, and cell lysates were prepared in RIPA buffer (150 mM NaCl, 5 mM EDTA pH 8.0, 50 mM
Tris-HCl pH 8.0, 1% NP-40 (v/v), 0.5% sodium deoxycholate (w/v), 0.1% SDS (w/v), 1mM DTT, and 1X protease inhibitor cocktail
(Roche)). Fe-S proteins were immunoprecipitated by incubation of cell lysates with 5 mg antibody for 1 hour on ice followed by im-
munocomplex capture by addition of protein A agarose beads (Bio-Rad) for 1 hour rotating at 4�C. Beads were washed three times
with RIPA buffer followed by scintillation counting. Data were normalized to protein lysate concentrations as determined by BCA
analysis (Pierce).
Transferrin Uptake AssayTransferrin uptake was performed as described previously (Stehling et al., 2008). Briefly, 3.5x105 HeLa cells were plated in 6 cm
dishes or 2x105 HeLa-Cas9 cells were plated in 6-well dishes and transfected 24 hours later with Effectene according to manufac-
turers’ protocol. 72 hours after transfection cells were washed once in PBS followed by incubation with 0.1 mg/mL FITC-Transferrin
(Invitrogen) in PBS for 1 hour at 37�C. Cells were then washed in PBS, cell lysates prepared in RIPA buffer, and FITC-Transferrin pre-
sent in cell lysatewas determinedwith a 96-well Enspire plate reader. Data were normalized to protein lysate concentrations as deter-
mined by BCA analysis (Pierce) and background signal from cells not treated with FITC-Transferrin.
Cell Viability AssaysTo assess cell viability after siRNA knockdown, cells were reverse transfected with 20-35 nM siRNA and Lipofectamine RNAiMAX
according to manufacturers’ protocol and left to incubate for 72-120 hours prior to changing the media and adding MTT (Life Tech-
nologies) and incubating for 4 hours at 37�C in the dark. Media was then removed and DMSO was added to each well to solubilize
crystals for 20 minutes. Plates were read at 540 nm on an Enspire plate reader. For MMS treatment, HeLa-Cas9 cells were plated at
low density in 6-well plates and 16 hours later MMS was added at specified concentrations diluted in media. After 11 days fresh me-
dia was added and MTT (Life Technologies) was added according to manufactures’ instructions. Cells were incubated for 4 hours at
37�C in the dark, media was then removed and DMSOwas added to each well to solubilize crystals for 20 minutes. Plates were read
at 540 nm on an Enspire plate reader. For UV treatment, 5x105 HeLa-Cas9 cells were plated in 6 cm plates. 16 hours later the cells
were washed with PBS then directly irradiated using a Stratagene Stratalinker 2400 and immediately trypsinized and plated at low
density in 6-well plates. Cells were then left to grow for 11 days and collected with MTT as described above.
MAGE-F1 Sequence AnalysisFor evolutionary analysis, nucleotide sequences of MAGE-F1from the selected mammals were aligned with CLUSTALW, codon
aligned with Codon Alignment v2.1.0 (http://www.hiv.lanl.gov/content/index), and confirmed by deduced amino acid sequence
alignment. We analyzed MAGE-F1 sequences available in NCBI database, including human (NM_022149), cow (NM_001102049),
mouse (NR_131983), rhesus monkey (XM_001097423), pig (XM_003483283), sheep (XM_004003083), white-tufted-ear marmoset
(XM_002758151), olive baboon (XM_003894729), gorilla (XM_004038139), goat (XM_005675179), tiger (XM_007083609), damara
mole-rat (XM_010624239), panda (XM_011220317), cat (XM_011286042), mouse lemur (XM_012736854), horse (XM_014732786),
donkey (XM_014846151), cheetah (XM_015081991), alpaca (XM_015236516), alpine marmot (XM_015499947), black flying fox
(XM_015599859), Egyptian fruit bat (XM_016159705), Natal long-fingered bat (XM_016200140), European rabbit (XM_017346949),
white-headed capuchin (XM017538378), black snub-nosed monkey (XM_017856339), leopard (XM_019463388), great roundleaf
bat (XM_019667946), Chinese rufous horseshoe bat (XM_019735753), common bottlenose dolphin (XM_019946054), yak
(XM_005909144, white-tailed deer (XM_020153210), naked mole-rat (XM_021252936), crab-eating macaque (XM_005546582),
ferret (XM_004745168), Coquerel’s sifaka (XM_012640473), Ma’s night monkey (XM_ 012457656), sooty mangabey
(XM_ 012034417), mandrill (XM_ 011981865), Angola colobus (XM_ 011945854), Arabian camel (XM_ 010976182), Bactrian camel
(XM_ 010959175), bison (XM_ 010853859), golden snub-nosedmonkey (XM_ 010366158), polar bear (XM_ 008705235), Przewalski’s
horse (XM_ 008531962), big brown bat (XM_ 008157575), green monkey (XM_ 007971872), killer whale (XM_012536258), Chinese
tree shrew (XM_ 014592510), little brown bat (XM_ 014452355), Brandt’s bat (XM_ 014530702), long-tailed chinchilla
(XM_ 013513325), thirteen-lined ground squirrel (XM_ 005331029), lesser Egyptian jerboa (XM_ 012947874), degu
(XM_ 012516798), nine-banded armadillo (XM_ 012531083), Pacific walrus (XM_ 004406749), small-eared galago
(XM_012813408), domestic guinea pig (XM_ 013160097), African savanna elephant (XM_ 010592823), Sumatran orangutan
(XM_ 002814368), and chimpanzee (XM_516919).
e5 Molecular Cell 69, 1–13.e1–e6, January 4, 2018
Please cite this article in press as: Weon et al., Cytosolic Iron-Sulfur Assembly Is Evolutionarily Tuned by a Cancer-Amplified Ubiquitin Ligase, MolecularCell (2017), https://doi.org/10.1016/j.molcel.2017.11.010
Sequences of MAGE-F1 from eight mouse stains were obtained from the Mouse Genome Project (Yalcin et al., 2011). MAGE-F1
sequence was validated by Sanger sequencing by PCR amplification of MAGE-F1 from C57BL/6 mouse testis cDNA.
Assessment of mRNA/Copy-Number Analysis in Human Tumors and Statistical AnalysismRNA levels (RNA-seq) and copy-number variation from tumors were determined by examining the cancer genome atlas (TCGA).
Tumors were stratified into diploid, gain (low level amplification), and amplified (high level amplification) from normalized segmenta-
tion values by the GISTIC method. Correlation of patient survival to copy-number variation status was similarly analyzed from TCGA
data and plotted as a Kaplan-Meier survival curve for head and neck squamous cell carcinoma. Mutational burden was determined
by stratification of tumors into amplified and non-amplified and total number of de novo intragenic mutations and specific types of
mutations (insertion, deletion, A:T > C:G, A:T > G:C, A:T > T:A, G:C > A:T, G:C > C:G, G:C > T:A) were determined from the TCGA
datasets for lung and head and neck squamous cell carcinoma tumors.
Xenograft Tumor Growth AssaysHCC95 and H520 lung squamous carcinoma cells with natural copy-number amplification of MAGE-F1 were transfected with
MAGE-F1 siRNA 6 hr before injection of 53 106 cells mixed with matrigel into NOD scid gammamice (Jackson Lab) (n = 6 per group).
H2170 lung squamous carcinoma cells without natural copy-number amplification of MAGE-F1 were made to stably express MYC-
MAGE-F1 or MYC-Vector before injection of 5 3 106 cells in PBS into NOD scid gamma mice (n = 6 per group). Tumor size was
measured three times a week during the duration of the study.
QUANTIFICATION AND STATISTICAL ANALYSIS
Data are represented as mean ± SEM unless otherwise indicated, and Student’s t test was used for all statistical analysis witih
GraphPad Prism 5 Software unless noted. TCGA data analysis was analyzed with Chi-square (c2), Pearson correlation (r), and
ANOVA using GraphPad Prism 5 Software. Nonsynonymous and synonymous nucleotide substitution rates were calculated using
the maximum likelihood method (Nei and Gojobori, 1986) implemented in the SNAP program (Korber, 2000) with filtering out of
insertions, deletions, and stop codons.
DATA AND SOFTWARE AVAILABILITY
Original imaging data have been deposited to Mendeley Data at https://doi.org/10.17632/f8842jrv8z.1.
Molecular Cell 69, 1–13.e1–e6, January 4, 2018 e6