Copyright ? 2010 by the Genetics Society of America
Trivalent Arsenic Inhibits the Functions of Chaperonin Complex
Xuewen Pan,*,†,1Stefanie Reissman,‡Nick R. Douglas,‡Zhiwei Huang,†Daniel S. Yuan,*
Xiaoling Wang,* J. Michael McCaffery,§Judith Frydman‡and Jef D. Boeke*,1
*Department of Molecular Biology and Genetics and The High Throughput Biology Center, The Johns Hopkins University School of Medicine,
Baltimore, Maryland 21205,†Verna and Marrs McLean Department of Biochemistry and Molecular Biology and Department of Molecular
and Human Genetics, Baylor College of Medicine, Houston, Texas 77030,‡Department of Biological Sciences, Bio-X Program and
Cancer Biology Program, Stanford University, Stanford, California 94305 and§Department of Biology and
Integrated Imaging Center, The Johns Hopkins University, Baltimore, Maryland 21218
Manuscript received April 10, 2010
Accepted for publication July 14, 2010
The exact molecular mechanisms by which the environmental pollutant arsenic works in biological
systems are not completely understood. Using an unbiased chemogenomics approach in Saccharomyces
cerevisiae, we found that mutants of the chaperonin complex TRiC and the functionally related prefoldin
complex are all hypersensitive to arsenic compared to a wild-type strain. In contrast, mutants with
impaired ribosome functions were highly arsenic resistant. These observations led us to hypothesize that
arsenic might inhibit TRiC function, required for folding of actin, tubulin, and other proteins
postsynthesis. Consistent with this hypothesis, we found that arsenic treatment distorted morphology of
both actin and microtubule filaments. Moreover, arsenic impaired substrate folding by both bovine and
archaeal TRiC complexes in vitro. These results together indicate that TRiC is a conserved target of arsenic
inhibition in various biological systems.
is also used as an effective therapeutic agent against
various diseases and infections. Using advanced geno-
mic tools in the model organism yeast and biochemical
experiments, we demonstrated that arsenic disturbs
functions of the chaperonin complex required for
proper folding and maturation of a large number of
proteins. This mechanism of action by arsenic is
conserved in various biological systems ranging from
archaeal bacteria to mammals. Such an understanding
should help in exploring possible ways to overcome
toxic effects caused by exposure to arsenic.
Trivalent inorganic arsenic is among the most signif-
icant environmental hazards affecting the health of
millions of people worldwide (Nordstrom 2002). Par-
ticularly, inorganic trivalent arsenic [As(III)] in under-
ground drinking water and some mining environments
is recognized as the cause of various cancers affecting
the skin, lung, urinary tract, bladder, liver, and kidney
(Tapio and Grosche 2006), as well as being implicated
in several other disorders such as diabetes, hyperten-
sion, neuropathy, and vascular diseases (Tseng 2004).
RSENIC is a ubiquitous environmental pollutant
that causes severe health problems in humans. It
Interestingly, As(III) is also an effective therapeutic
agent against cancer and human pathogens. A number
of models have been proposed to explain the biological
effects of As(III), including stimulation of reactive
oxygen species (ROS) production (Miller et al. 2002;
Tapio and Grosche 2006) and inhibition of tubulin
polymerization (Ramirez et al. 1997; Li and Broome
1999). However, exactly how As(III) disturbs biological
systems is still not clear.
The eukaryotic chaperonin TRiC (TCP1-ring com-
of two apposed heterooligomeric protein rings. Each
ring, constituted by eight homologous subunits (en-
coded by the essential CCT1–CCT8 genes in budding
yeast), contains a central cavity in which unfolded poly-
peptide substrates attain a properly folded state in an
ATP-requiring reaction (Bukau and Horwich 1998;
Gutsche et al. 1999). TRiC is required for the proper
folding of an important subset of cytosolic proteins,
including cytoskeleton components, cell cycle regula-
tors, and tumor suppressor proteins (Spiess et al. 2004).
Some of these protein substrates are themselves en-
coded by essential genes; thus TRiC is indispensable
for eukaryotic cell survival. Many TRiC substrates are
subunits of oligomeric complexes and their assembly
into functional multisubunit complexes also requires
TRiC (Spiess et al. 2004). Assembly of such macromo-
lecular complexes in some cases eliminates the accu-
mulation of toxic subunits such as free b-tubulin
molecules, which can bind to g-tubulin and thereby
1Corresponding authors: 733 NorthBroadway, Room339, Baltimore, MD
21205.Email: firstname.lastname@example.org; and Department of Biochemistry,
BCM 125, Houston, TX 77030. Email: email@example.com
Genetics 186: 725–734 (October 2010)
disrupt the formation of mitotic spindles in the yeast
S. cerevisiae (Archer et al. 1995). Folding of yeast
actin, a-tubulin, and b-tubulin and their oligomeriza-
tion require TRiC and GimC (also known as prefoldin),
a nonessential protein complex of six distinct but
structurally related subunits of 13–23 kDa (Geiser
et al. 1997; Vainberg et al. 1998). Mutational loss of
GimC function substantially reduces actin and tubulin
folding efficiency although it does not cause obvious
growth defects in yeast. However, deletion of various
GimC subunits strongly reduces the viability of condi-
tional-lethal alleles of TRiC subunits under permissive
conditions (Siegers et al. 1999).
To elucidate the mechanisms of inorganic As(III)’s
action(s) in a eukaryotic system, we first took an un-
biased functional chemogenomics approach in yeast to
systematically probe for the genetic determinants of
arsenic sensitivity. These genetic and subsequent bio-
chemical results point to the conclusion that As(III)
inhibits the yeast TRiC complex. This mechanism of
action is apparently conserved because the activities of
both a mammalian TRiC complex and an archaeal TRiC-
like chaperonin are significantly inhibited by arsenic
in vitro. Given that mammalian TRiC and some of its
substrates are implicated in tumor suppression, angio-
Bouhouche et al. 2006), TRiC is likely an important
protein mediator of As(III)’s effects on human health.
MATERIALS AND METHODS
Profilingof thesensitivity of genome-wide deletionmutants
to arsenic and individual validation were carried out as
described in supporting information, File S1. As2O3(Sigma-
stock solution of sodium arsenite.
Yeast strains and plasmids: Individual yeast deletion mutant
strains used in this study were all obtained or derived from the
yeast knockout mutants constructed by the Saccharomyces
Genome Deletion Project (Giaever et al. 1999). The DAmP
alleles of TUB2, CCT1, and CCT2 were obtained from Open
Biosystems (Breslow et al. 2008). The Cdc55p-3HA expression
strain was a gift from Katja Siegers. Plasmids expressing ACT1
and TUB2 from a centromere-based plasmid were constructed
by in vivo homologous recombination into YCplac33 (CEN and
URA3) (Gietz and Sugino 1988). The CCT1–CCT7, TUB1, and
TUB3 overexpression plasmids were similarly constructed
into YEplac195 (2 m, URA3) (Gietz and Sugino 1988). Either
YCplac33 or YEplac195 was used as the vector control.
Immunofluorescence: A 100-ml culture of the wild-type
phase and split. Sodium arsenite was added into one part of
the culture at a final concentration of 1 mm. The other part
served as a nontreatment control. Both were incubated at 30?
for 3 additional hr. Immunofluorescence analyses of actin and
microtubule morphology were performed as previously de-
phalloidin and microtubule with an anti-a–tubulin antibody.
Immunoprecipitation and Western blotting: The yeast
strain expressing Cdc55p–3HA was grown in 100 ml YPD at
30? until mid-log phase. The culture was split and sodium
arsenite was added into one aliquot at a final concentration
of 1 mm. The other aliquot served as a nontreatment control.
Both were subsequently incubated at 30? for 1 hr before cells
were harvested. Cell homogenization and immunoprecipita-
tion were carried out essentially as described (Pan and
Heitman 2002). An anti HA and anti-actin antibody were
used to immunoprecipitate Cdc55p–3HA and actin, respec-
tively. Cdc55p–3HA, actin andTRiC on the Western blots were
detected with anti-HA, anti-actin, and anti-Cct5p antibodies.
In vitro actin folding assay: The actin-folding assay was
carried out as described by Frydman and Hartl (1996).
KOH (pH 7.5), 100 mm KOAc, 5 mm MgCl2, 1 mm DTT, 10%
glycerol, and 1% PEG 8000]. Subsequently [35S]-actin, which
was denatured in 6 m guanidine/HCl, was rapidly diluted
1:100 to a final concentration of 30 mm into the reaction mix.
for 10 min to remove aggregated actin, the reaction was
supplemented with 1 mm ATP andincubated for 40 min at 30?
to allow time for ATP-dependent actin refolding by the
chaperonin. A total of 1 mm arsenite was added as indicated
in the figure. Generation of native [35S]-actin was determined
by native gel electrophoresis using folded [35S]-actin as a
control as described previously (Frydman and Hartl 1996).
The gel was exposed on a phosphor storage screen (Kodak)
and scanned in a Typhoon 9410 imager (GE Healthcare).
Archaeal chaperonin folding assays: Purification of the
archaeal chaperonin from Mm-Cpn was carried out by conven-
tional chromatography essentially as described (Kusmierczyk
and Martin 2003). Rhodanese folding by the archaeal chaper-
onin Mm-Cpn was assayed as described (Weber and Hayer-
Hartl 2000). In brief, 0.25 mm protein was incubated in
Cpn-buffer supplemented with 20 mm sodium thiosulfate.
Purified rhodanese was denatured in 6 m guanidinium/HCl
containing 5 mm DTT and rapidly diluted 1:100 to a final
for 5 min at 37?, the reaction was started by addition of 2 mm
ATP and allowed to proceed for 50 min at 37?. To detect the
presence of refolded rhodanese, 10 ml of the reaction was
as described (Weber and Hayer-Hartl 2000).
Mutants of the chaperonin pathway are As(III) hyper-
to increased sensitivity to the cognate antiproliferation
cytotoxin or drug (Giaever et al. 1999; Lum et al. 2004).
To probe molecular mechanisms of As(III) actions in
yeast, we investigated As(III) sensitivity of genome-wide
heterozygous diploid yeast knockout (YKO) mutants
using a TAG array-based analysis (Giaever et al. 1999;
Lum et al. 2004) followed by individual validation. We
found that 33 heterozygous diploid YKOs were signifi-
cantly more sensitivethan an isogenic wild-type strain to
450 mm sodium arsenite (Figure 1A and Table S1).
Because As(III) completely inhibits yeast cell growth at
higher concentrations, we reasoned that it might in-
activate at least one essential protein or protein com-
plex. A total of 15 of these 33 As(III)-hypersensitive
mutants were heterozygous for essential genes, includ-
ing five (CCT1, CCT4, CCT5, CCT7, and CCT8) encod-
ing subunits of the TRiC complex and two more (SPC97
726 X. Pan et al.
and TUB4) that are directly involved in microtubule
biogenesis and function (Figure 1A and Table S1). We
directly tested three other heterozygous YKOs (CCT2,
CCT3, and CCT6) of TRiC that were missed in the initial
array-based screen due to low microarray hybridization
signal intensities and found that they were also more
sensitive to As(III) than a control strain (Figure 1B).
Yeast cells overexpressing one of the TRiC complex
genes, CCT1, were also partially arsenic resistant on
solid medium (Figure 1C). Such an effect was also ob-
served in liquid culture. Under unperturbed condi-
tions, the growth rates of hoD mutants carrying an
empty vector and overexpressing TCC1 in a liquid
synthetic medium were at 113 min/division and 115
min/division, respectively. In the presence of 450 mm
of As(III), their growth rates were 181 min/division
and 137 min/division, respectively. These results to-
gether suggest that As(III) might inhibit TRiC com-
Synthetic lethality interactions were previously ob-
served between CCT1 partial loss-of-function alleles and
deletion mutations of the GimC complex, which acts
as a cochaperone in TRiC-mediated actin and tubulin
folding (Siegers et al. 1999). Thus, if As(III) inhibits
TRiC, GimC deletion mutants should be arsenic hyper-
sensitive. To test this prediction and to further extend
the study of cellular response to As(III), we systemati-
sensitivity using dSLAM, a barcode microarray-based
method for detecting gene-compound and gene–gene
interactions (Pan et al. 2004). Upon individual valida-
tion, we found that 191 haploid YKOs were sensitive to
nine most sensitive ones directly affected either As(III)
efflux (arr1D and arr3D) or the GimC complex (gim3D,
gim4D, gim5D, pac10D, pfd1D, yke2D, and yml094c-aD,
which deletes part of GIM5) (data not shown) and all
GimC mutants were more sensitive than the ARR
mutants when tested at 40 mm As(III) (Figure 2B and
data not shown), further supporting the model that
arsenic inhibits TRiC.
Most other As(III)-sensitive mutants were susceptible
only to relatively high concentrations (Table S2). These
included additional mutants affecting actin and micro-
tubule biogenesis and those affecting peroxisome bio-
genesis, mitotic chromosome segregation, cell cycle
Figure 1.—Genetic alterations in the
TRiC complex affect arsenic susceptibility.
(A) As(III)-sensitive heterozygous diploid
YKOs identified by genome-wide mutant
fitness profiling. Essential genes are col-
ored in red and nonessential genes in
black. The network diagram was created
with Cytoscape 2.0 (Shannon et al. 2003).
(B) Sensitivity of the heterozygous diploid
YKOs of all eight TRiC subunits to As(III)
at 450 mm. An ho/hoD mutant phenotypi-
cally indistinguishable from a wild-type
yeast served as the control strain. (C) An
hoD mutant harboring a vector or a plasmid
overexpressing a TRiC subunit as indicated
was tested for growth on solid SC ?Ura that
either contained or lacked 1 mm As(III).
Arsenic Inhibits TRiC 727
progression, histone modification, ergosterol biosynthe-
sis, oxidative stress response, DNA repair, and others
(Figure 2A and Table S2), consistent with results of
recently performed screens using the haploid or homo-
zygous diploid knockout mutants (Dilda et al. 2008; Jin
et al. 2008; Thorsen et al. 2009). That mutants defective
in oxidative stress response and DNA repair were sen-
sitive to As(III) also agrees with a previous finding that
As(III) stimulates the production of ROS (Tapio and
Grosche 2006), which damage DNA molecules. The
arsenic-sensitive phenotypes of some of these other
mutants are likely related to TRiC inhibition because
CCT1 overexpression partially suppressed arsenic sensi-
tivity in various mutants, including those of oxidative
affected pathways are likely required for buffering the
effects of arsenic inhibition of TRiC. Others might well
Slowed protein synthesis confers As(III) resistance:
We also identified 109 arsenic-resistant haploid YKOs
and about 62.4% of them affected either ribosomal
protein genes or those involved in ribosomal biogenesis
mutations in ribosomal biogenesis and arsenic resis-
tance was also recently observed by others (Dilda et al.
2008). Although ribosomal proteins are essential, some
are encoded by duplicated genes in yeast and deleting
the ribosomal protein mutants largely correlated with
their fitness defects under normal conditions (data not
shown). We suspected that these mutants might have
compromised capacity in protein synthesis, which leads
to As(III) resistance. To test this hypothesis, we in-
vestigated the effects of inhibiting protein synthesis
with a sublethal concentration of cycloheximide. Simi-
lar to the rpl19bD and rps23bD mutations that affect
ribosomal protein genes, treatment with cycloheximide
conferred yeast cells resistance to As(III) (Figure 3B).
This relationship between protein synthesis and arsenic
cytotoxicity is consistent with the model that As(III)
inhibits TRiC, which facilitates the folding of newly
synthesized polypeptides and their subsequent assem-
bly into oligomeric complexes (Spiess et al. 2004).
Figure 2.—Arsenic-sensitive haploid YKOs
and their genetic relationships with CCT1. (A)
As(III)-sensitive haploid deletion mutants identi-
fied by genome-wide mutant fitness profiling. A
total of 191 were individually verified to be sensi-
tive to As(III) at 400 mm. The number of mutants
in each biological process affected and the corre-
sponding percentage among all mutations iden-
tified were listed. This plot was derived from
Table S2. (B) Growth of representative arsenic-
sensitive haploid YKOs of indicated genotypes
on a solid synthetic complete (SC) medium that
either lacked or contained As(III) at 40 mm,
150 mm, and 800 mm. (C) Partial suppression
of the arsenic sensitivity by CCT1 overexpression
in various mutants.
728X. Pan et al.
Inhibition of TRiC by As(III) likely has two distinct
typesof effects: (1)production ofinsufficient quantities
of correctly folded substrates and assembled complexes
that are essential for viability and (2) accumulation of
incorrectly folded substrates or unassembled subunits
that are toxic.
b-Tubulin contributes to As(III) toxicity: One such
TRiC substrate isb-tubulin, which causes growth defects
a-tubulin (Archer et al. 1995). Presumably, b-tubulin
polypeptides produced at lower rates (resulting from
slowed protein translation) in ribosomal protein mu-
tants or cycloheximide-treated cells were more compat-
ible with the capacity of arsenic-crippled TRiC complex.
As a result, unfolded polypeptides and/or unassembled
toxic free b-tubulin molecules might have been accu-
mulated at lower levels, mitigating As(III) toxicity. To
test this hypothesis, we investigated whether b-tubulin
contributes to As(III) toxicity by genetically perturbing
the relative ratio between a- and b-tubulins. We first
a centromeric plasmid under control of its endogenous
promoter and tested its effects on the growth and
arsenic sensitivity of HO/hoD (a surrogate wild type)
and CCT6/ cct6D heterozygous diploid mutants. Mod-
est overexpression of TUB2 made the HO/hoD mu-
tant slightly yet reproducibly more sensitive to As(III)
(Figure 3D). It also noticeably hampered growth of the
CCT6/cct6D mutant in the absence of As(III) and even
more so in its presence (Figure 3D). In comparison,
modest overexpression of ACT1, which encodes another
important cytoskeletal substrate of TRiC, actin, had no
effect (Figure 3C). In agreement with these results,
caused As(III) hypersensitivity (Figures 2B and 3C).
Overexpression of TUB3 from a high copy plasmid also
suppressed As(III) toxicity in cin2D, gim3D, and pfd1D
mutants (Figure 3C), which are defective in tubulin
folding and dimerization. Similar results were observed
when TUB1, an essential a-tubulin gene highly homol-
ogous to TUB3, was overexpressed (data not shown). In
moreefficient folding and/ orincorporation ofb-tubulin
molecules into nontoxic heterodimers due to the in-
creased abundance of available a-tubulin molecules.
These results together indicated that yeast cells with a
relatively high b/ a-tubulin ratio are sensitized to As(III)
and those with a low b/a-tubulin ratio are more resis-
tant to the drug. Thus unfolded and/ or folded yet free
b-tubulin molecules apparently contribute to As(III)
Arsenic unlikely directly inhibits microtubule func-
tion in yeast: It has been proposed that As(III) directly
binds to tubulins and disrupts their functions in
mammalian cells (Li and Broome 1999). In particular,
As(III) was shown to directly bind to b-tubulin isolated
from a human breast cancer cell line and structural
modeling suggested the amino acid residue Cys12 is key
to such a physical interaction (Zhang et al. 2007). This
Figure 3.—The rate of
protein synthesis and free
b-tubulin modulate yeast ar-
haploid YKOs according
to biological processes af-
fected. This plot was de-
rived from Table S3. (B)
were grown in an SC me-
dium that either contained
or lacked 800 mm of As(III).
Cycloheximide was used at
10 ng/ml. (C) One of the
a-tubulin genes TUB3 was
overexpressed from a 2m
plasmid in mutants defec-
tive in microtubule biogen-
esis. Cells were grown on
solid SC ?Ura either in
the presence or absence of
As(III). As(III) concentration
used was 600 mm for the
hoD, tub3D, and cin2D mu-
tants and 50 mm for the
gim3D and pfd1D mutants. (D) The b-tubulin gene TUB2 and the actin gene ACT1 were expressed in mutants of indicated
genotypes from a centromere-based plasmid. Cells were grown on solid SC that lacked uracil (SC ?Ura) either in the presence
or absence of 200 mm of As(III). (E) Strains of indicated genotypes were grown on solid YPD with or without As(III) (1 mm) or
benomyl (20 mg/ml).
Arsenic Inhibits TRiC 729
the genetic evidence described above. However, it is not
consistent with other genetic evidence. We found that a
tub2–DAmP mutant (Breslow et al. 2008), which pre-
sumably expresses TUB2 at a lower level as compare to a
wild type due to mRNA instability, responds to As(III)
and the well-known microtubule poison benomyl very
differently. Consistent with the idea that benomyl
directly binds to and inihibits microtubules, the tub2–
DAmP was hypersensitive to this drug (Figure 3E). The
latter result also implies that TUB2 expression is indeed
reduced in the DAmP allele. In contrast to the hypoth-
esis that As(III) directly affects microtubules, the tub2–
DAmP strain was slightly more resistant to As(III) than
the wild type (Figure 3E), a highly reproducible phe-
similarly hypersensitive to benomyl but slightly more
resistant to As(III) as compared to a HO/ hoD control
strain (data not shown). In contrast to the tub2–DAmP
mutant, both the cct1–DAmP and cct2–DAmP mutants
were hypersensitive to both As(III) and benomyl. These
results suggest that As(III) unlikely inhibit yeast growth
by binding to microbutules as benomyl does. Of course,
but has a different effect than benomyl. However, the
Cys12 residue critical for arsenic binding to human
b-tubulin is not conserved in its yeast ortholog. The
genetic results are also more consistent with the model
that As(III) inhibits the TRiC complex, which is re-
quired to compensate for the loss of microtubule func-
tions in the presence of benomyl due to its essential
functions in microtubule biogenesis (Ursic et al. 1994;
Siegers et al. 2003).
Arsenic affects TRiC function in vivo: Defects in
TRiC were previously shown to distort morphological
organization of both actin and microtubule filaments
(Ursic etal.1994;Siegers et al.2003).IfAs(III)inhibits
TRiC, we expected that arsenic treatment of wild-type
to emerging buds under normal conditions (Figure 4A)
(Adams and Pringle 1984; Kilmartin and Adams
1984). In the presence of As(III), such polarization
disappeared (Figure 4A). Arsenic treatment also dis-
torted the mitotic microtubule structures (Figure 4A).
Similar observations were recently made in another
study ( Jin et al. 2008), although the exact microtubule
morphology change caused by arsenic treatment shown
there was different from what we saw. Such difference
this other study, the intrinsic fluorescence of GFP
tagged a-tubulin was observed, whereas we used immu-
nostaining. Despite this, the fact that arsenic treatment
affects the morphology of both actin and microtubule
filaments strongly support the model that As(III)
inhibits TRiC functions in vivo.
We next investigated whether As(III) might interfere
with the physical interaction between TRiC and its sub-
strates. In addition to the major substrates actin and
tubulins, TRiC is required for folding of cytoplasmic
proteins such as Cdc55p in yeast (Siegers et al. 2003).
We found that As(III) reproducibly and significantly
reducedphysicalinteractionbetweenTRiC and Cdc55p
in coimmunoprecipitation assays (Figure 4B), further
indicating that As(III) directly affects TRiC functions.
However, arsenic had no effect on the physical in-
and its effect on the interaction betweenTRiC and actin
was inconclusive. Out of three independent coimmu-
noprecipitation experiments, we found that As(III)
partially reduced TRiC binding to actin in one experi-
ment and did not see any effect in the other two (data
not shown). Thus, As(III) inhibits TRiC binding to
some substrates (i.e., Cdc55p) but not to others in vivo
Arsenic inhibits TRiC activity in vitro: Eukaryotic
pyruvate dehydrogenase (PDH), which consumes pyru-
vate to form acetyl-CoA while reducing NAD1to NADH,
has long been considered a primary target for arsenic
lowered intracellular ATP/ ADP ratio. TRiC folding of
substrates absolutely requires ATP hydrolysis (Spiess
et al. 2004) and As(III) might inhibit TRiC function
indirectly by inactivating PDH and thereby damping
intracellular ATP concentration. However, we consider
this unlikely because PDH is nonessential in yeast. In
addition, we did not observe a synthetic lethality or slow
growth interaction between a PDH (pda1D) mutation
and a GimC mutation (pfd1D) (data not shown).
We next tested the possibility that As(III) directly
inactivates TRiC using a well-characterized in vitro assay
that monitors the ability of purified chaperonin to fold
chemically denatured35S-actin (Meyer et al. 2003). We
using purified bovine TRiC significantly inhibited actin
folding. Such inhibition was observed both when the
inhibitor was added before and shortly after substrate
binding to TRiC. Here binding of TRiC to actin was not
inhibited by As(III) (Figure 4C). Yet actin folding
activity was still inhibited, suggesting that As(III) in-
activates TRiC after substrate binding. Similarly, As(III)
inhibited substrate folding by a TRiC-like archaeal
complex, the Mm-Cpn chaperonin from Methanococcus
maripaludis, which is ?40% identical to its human
counterpart (Figure 4D). Together, these results sup-
port the idea that As(III) directly inhibits the folding
activity of TRiC and TRiC-like chaperones.
We also found that inhibition of TRiC by arsenic
in vitro was reversible because its activity was recovered
after gel filtration (data not shown), ruling out direct
covalent modification of TRiC by As(III) as a mecha-
nism of action. Such reversibility was also observed with
arsenic inhibition of yeast growth; nearly 100% of yeast
cells regained colony formation capacity after being
730X. Pan et al.
incubated in the presence of inhibitory concentration
of the drug (.2 mm) for over 24 hr at 30? (data not
Detoxification mutations sensitize TRiC defective
cells toward As(III): Arsenic concentrations required
for inhibiting growth of wild-type yeast cells are much
higher than needed for arsenic chemotherapy and
toxicity in human cells. To mitigate such difference,
we next investigated whether arsenic inhibition of TRiC
at clinically achievable doses could inhibit growth of
relatively healthy yeast cells. To do this, we first per-
screen to identify secondary mutations that further sen-
mediated actin and tubulin folding (Geiser et al. 1997;
such mutations (Table S4). Among them, both arr1D
5 mm of As(III) without affecting its growth rate under
normal conditions (Figure 5B and Table S4). Growth of
the arr1D pfd1D and arr3D pfd1D double mutants was
also significantly inhibited by a lower concentration of
arsenic (2 mm) (Figure 5B). Thus, at clinically effective
doses, arsenic inactivation of TRiC indeed significantly
inhibited the growth of otherwise healthy yeast cells.
These results indicate that, at the mechanistic level,
arsenic modes of actions may well be the same in both
yeast and human, even though the wild-type yeast cells
seem to be more resistant to As(III) than human cells.
Interestingly,Arr1p isatranscription factorrequired for
Figure 4.—As(III) inhibits TRiC functions. (A) Wild-type yeast (BY4743) cells were grown in liquid YPD either in the absence or
presence of 1 mm of As(III) for 3 hr. Actin was stained with rhodamine-phalloidin and microtubules were visualized with an anti-
a–tubulin antibody. (B) ATP-depleted cell lysates prepared from yeast cells expressing Cdc55–3HA grown in liquid YPD that either
contained or lacked 1 mm As(III) were subjected to immunoprecipitation with anti-HA antibody. Western blots of the total cell
extracts (5mg) and the immunoprecipitates (IP) from lysates containing 200 mg total protein were analyzed with an anti-HA an-
tibody and an antibody against one of the TRiC subunits Cct5. (C) In vitro binding and folding of denatured [35S]-actin by bovine
TRiC both in the presence and absence of 1 mm As(III) were assessed by native gel analysis followed by autoradiography. Three (2,
3, and 4) different schemes of arsenic treatment were tested. Native [35S]-actin samples incubated with or without As(III) were
included as controls (right two lanes). The extent of As(III) inhibition in each condition was quantified for three independent
experiments and expressed as percentage of actin folding relative to the untreated control. (D) In vitro rhodanese folding by the
M. maripaludis TRiC-like chaperonin (Mm-Cpn) was assessed in the presence and absence of 1 mm sodium arsenite as described
(Kusmierczyk and Martin 2003).
Arsenic Inhibits TRiC731
pumps the drug out of yeast cells (Ghosh et al. 1999).
Mammalian genomes do not seem to encode an Arr3p-
like arsenic transporter, the absence of which might
allow for more effective accumulation of intracellular
By using a combination of functional genomic,
genetic, and biochemical studies, we have investigated
the genetic determinants of arsenic susceptibility in
yeast and elucidated a mechanism of action by arsenic
common to both eukaryotes and archaea. Our data
suggest that arsenic inhibits the function(s) of the
chaperonin complex. Similar functional genomic stud-
ies have been recently reported (Dilda et al. 2008; Jin
et al. 2008; Thorsen et al. 2009). Although there was
considerable overlap among all of these studies regard-
ing the lists of arsenic-sensitive and -resistant mutants
identified, our study is distinctly different from these
others, which studied only the haploid and/or homo-
zygous diploid knockout mutant libraries that lack
mutants of essential genes. In contrast, we first studied
the genome-wide heterozygous diploid mutants, which
directly and dramatically revealed the essential TRiC
haploinsufficiency. We subsequently corroborated this
haploid mutants. In particular, we have put an emphasis
on the most sensitive haploid knockout mutant(s) with
the premise that it likely affects functions most closely
relatedto the arsenic target(s). That all6mutants ofthe
GimC complex exhibited the highest sensitivity toward
low concentrations of As(III) (Figure 2B and Table S2)
further supported the idea that TRiC is the target. More
importantly, we provided further evidence that As(III)
inhibits TRiC functions both in vivo and in vitro. Such a
conclusion is also consistent with aprevious observation
of arsenic leads to distorted organization of both actin
subtle modification of the previously described dSLAM
methodology (Pan et al. 2004), we also demonstrated
that it is possible to systematically identify genome-wide
double mutations that confer hypersensitivity to a given
drug in a high throughput manner. This will allow for
expansion in the elucidation of genetic interaction
networks involved in drug response to beyond mono-
Despite that As(III) concentrations used in most of
our experiments were relatively high, our results are
likely relevant to arsenic effects on human health. First,
our results with the apparently healthy arr1D pfd1D and
arr3D pfd1D double mutants (Figure 5B) have demon-
strated that low concentrations of arsenic can inhibit
cell growth in yeast. That human cells are more
susceptible might also be due to stress-induced apopto-
sis that is basically lacking in yeast cells. Second, both
yeast and human cells accumulate high levels of arsenic
after exposure to low concentrations and this is more
obvious in human cells than in yeast. It was shown that
?106wild-type yeast cells exposed to 160 mm of arsenic
for 2 hr accumulate ?0.163 nmol of the drug (Dilda
Figure 5.—A genome-wide screen identifies double mu-
tants hypersensitive to low concentrations of arsenic. (A) A
high-efficiency pfd1DTURA3 gene disruption cassette was
transformed into a pool of haploid-convertible heterozygous
diploid yeast knockout mutants. After sporulation, a pool of
haploid double knockout mutants were derived from the het-
erozygous diploid double mutant pool in the presence of
5 mm of As(III). A separate pool of haploid single and double
mutants were also similarly derived either in the presence or
absence of 5 mm of As(III). Relative representation of each
knockout mutation in both pools was compared by TAG-array
analysis (Pan et al. 2004). (B) Haploid-convertible ARR1/
arr1DTkanMX PFD1/pfd1DTURA3 and ARR3/arr3DTkanMX
PFD1/pfd1DTURA3 diploid double mutants were sporulated
and spotted on a haploid selection medium that either lacked
both uracil and G418 (to select for Ura1single and double
mutants), contained both uracil and G418 (to select for
G418-resistant single and double mutants), or lacked uracil
but contained G418 (to select for the double mutants). Cell
growth under each condition was assessed both in the ab-
sence (control) and presence of 2 mm or 5 mm of As(III)
732X. Pan et al.
et al. 2008), equivalent to a final intracellular concen-
tration of ?2.3 mm, assuming an average volume of a
haploid yeast cell of ?70 fl (Sherman 2002). Similarly,
?106human APL cells (NB4) exposed to 20 mm of
arsenic for 4 hr accumulated ?0.73 nmol of the drug
(Dilda et al. 2008), equivalent to an intracellular
of of ?1040 fl for an NB4 cell (Miossec-Bartoli et al.
1999). As(III) concentrations within cells of certain
human organs and tissues or within some intracellular
compartments might be even higher. Thus our obser-
vation of TRiC inhibition by 1 mm of As(III) in vitro
exposed to relatively low levels of As(III). Third, our
in vitro assays might have underdetected the potency of
we had to include 1 mm of DTT, which is absolutely re-
quired for substrate folding by TRiC and Mm-Cpn ( J.
Frydman, unpublished observations). The presence of
DTTinthefoldingassays likely at leastpartially reversed
the inhibitory effects of arsenic on TRiC and Mm-Cpn.
Unfortunately, we could not test the effects of lower
concentrations of As(III) on TRiC activities due to this
technical limitation of the assay.
Among the heavy metals that interact with thiol
groups, As(III) inhibition of TRiC seems to be specific.
One piece of supporting evidence is that mutants of the
did not exhibit sensitivity toward Cd21when compared
to a wild-type strain (data not shown). However, cur-
rently it is not clear how As(III) inhibits the TRiC
complex at the biochemical level. It did not seem to
inhibit TRiC’s substrate binding (Figure 4C) or its
ATPase activity. In fact, arsenic stimulated the ATPase
activity of TRiC by approximately twofold (data not
shown). Given the similar inhibitory effects observed
when As(III) was added both before and shortly after
a late step of the process, for example, release of cor-
rectly folded products from the chaperonin complex.
This might gradually lead to accumulation of unpro-
ductive TRiC complex and reduction in its overall
productivity. This might partly explain why arsenic
reproducibly inhibits TRiC binding to Cdc55 but not
actin and tubulin. Possibly, most TRiC molecules within
arsenic-treated cells are stuck with the more abundant
substrates such as actin and tubulin.
That TRiC is required for folding and maturation of
as much as ?9–15% of all cytosolic proteins in
mammals (Thulasiraman et al. 1999) might also at
least partly explain the pleiotropic effects of arsenic on
human health. For example, exposure to arsenic has
been linked to cancers and neuropathy. The former
might be related to the fact that TRiC is required for
the assembly of the Von Hippel-Lindau (VHL) tumor
suppressor complex (Feldman et al. 1999), which plays
a positive role in stabilizing and activating p53 (Roe
et al. 2006), a major tumor suppressor commonly
mutated in various human cancers. Arsenic inhibition
of TRiC might thus indirectly downregulate p53 ac-
tivity and cause cancers, a model consistent with the
observation that arsenic inhibits p53 activation in
response to DNA damage (Tang et al. 2006). In ad-
dition, mutating TRiC subunit genes CCT5 and CCT4
are directly implicated in sensory neuropathy in both
human patients and in a rat model (Lee et al. 2003;
Bouhouche et al. 2006).
We thank Katja Siegers for providing the anti-Cct5p antibody and a
yeast strain that expresses Cdc55p–3HA (Siegers et al. 2003). We are
also grateful to David Drubin and Andy Hoyt for anti-a–tubulin
antibodies and Pamela Meluh for an anti-b–tubulin antibody. We
thank Robert Cohen and Pamela Meluh for helpful comments on the
manuscript. This work was supported in part by National Institutes of
Health (NIH) grants HG02432 and RR020839 to J.D.B, by NIH grant
HG004840 to X.P., by NIH grant RR019409 to J.M.M., and by NIH
grant GM74074 and a grant from the NIH Roadmap Initiative on
Nanomedicine to J.F.
Adams, A. E., and J. R. Pringle, 1984
mutant Saccharomyces cerevisiae. J. Cell Biol. 98: 934–945.
Archer, J. E., L. R. Vega and F. Solomon, 1995
tein that binds to beta-tubulin and participates in microtubule
function in vivo. Cell 82: 425–434.
Bouhouche, A., A. Benomar, N. Bouslam, T. Chkili and M.
Yahyaoui, 2006Mutation in the epsilon subunit of the cyto-
solic chaperonin-containing t-complex peptide-1 (Cct5) gene
causes autosomal recessive mutilating sensory neuropathy with
spastic paraplegia. J. Med. Genet. 43: 441–443.
Breslow, D. K., D. M. Cameron, S. R. Collins, M. Schuldiner,
J. Stewart-Ornstein et al., 2008
abling high-resolution functional analysis of the yeast genome.
Nat. Methods 5: 711–718.
Bukau, B., and A. L. Horwich, 1998
erone machines. Cell 92: 351–366.
Dilda, P. J., G. G. Perrone, A. Philp, R. B. Lock, I. W. Dawes et al.,
2008Insight into the selectivity of arsenic trioxide for acute
promyelocytic leukemia cells by characterizing Saccharomyces
cerevisiae deletion strains that are sensitive or resistant to the
metalloid. Int. J. Biochem. Cell Biol. 40: 1016–1029.
Feldman, D. E., V. Thulasiraman, R. G. Ferreyra and J. Frydman,
1999Formation of the VHL-elongin BC tumor suppressor com-
plex is mediated by the chaperonin TRiC. Mol. Cell 4: 1051–
Frydman, J., andF. U. Hartl, 1996
protein folding: differences between in vitro and in vivo mecha-
nisms. Science 272: 1497–1502.
Geiser, J. R., E. J. Schott, T. J. Kingsbury, N. B. Cole, L. J. Totis
et al., 1997Saccharomyces cerevisiae genes required in the ab-
sence of the CIN8-encoded spindle motor act in functionally
diverse mitotic pathways. Mol. Biol. Cell 8: 1035–1050.
Ghosh, M., J. Shen and B. P. Rosen, 1999
ification in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA
Giaever, G., D. D. Shoemaker, T. W. Jones, H. Liang, E. A. Winzeler
et al., 1999 Genomic profiling of drug sensitivities via induced
haploinsufficiency. Nat. Genet. 21: 278–283.
Gietz, R. D., and A. Sugino, 1988
vectors constructed with in vitro mutagenized yeast genes lacking
six-base pair restriction sites. Gene 74: 527–534.
Gutsche, I., L. O. Essen and W. Baumeister, 1999
eronins: new TRiC(k)s and turns of a protein folding machine.
J. Mol. Biol. 293: 295–312.
Relationship of actin and
Rbl2p, a yeast pro-
A comprehensive strategy en-
The Hsp70 and Hsp60 chap-
Pathways of As(III) detox-
New yeast-Escherichia coli shuttle
Group II chap-
Arsenic Inhibits TRiC733
Jin, Y. H., P. E. Dunlap, S. J. McBride, H. Al-Refai, P. R. Bushel et al.,
2008 Global transcriptome and deletome profiles of yeast ex-
posed to transition metals. PLoS Genet. 4: e1000053.
Kilmartin, J. V., and A. E. Adams, 1984
of tubulin and actin during the cell cycle of the yeast Saccharo-
myces. J. Cell Biol. 98: 922–933.
Kusmierczyk, A. R., and J. Martin, 2003
protein folding in the type II chaperonin from the mesophilic
archaeon Methanococcus maripaludis. Biochem. J. 371: 669–
Lee, M. J., D. A. Stephenson, M. J. Groves, M. G. Sweeney, M. B.
Davis et al., 2003 Hereditary sensory neuropathy is caused
by a mutation in the delta subunit of the cytosolic chaperonin-
containing t-complex peptide-1 (Cct4) gene. Hum. Mol. Genet.
Li, W., and I. N. Chou, 1992 Effects of sodium arsenite on the cy-
toskeleton and cellular glutathione levels in cultured cells. Tox-
icol. Appl. Pharmacol. 114: 132–139.
Li, Y. M., and J. D. Broome, 1999
apoptosis in myeloid leukemia cells. Cancer Res. 59: 776–780.
Lum, P. Y., C. D. Armour, S. B. Stepaniants, G. Cavet, M. K. Wolf
et al., 2004Discovering modes of action for therapeutic com-
pounds using a genome-wide screen of yeast heterozygotes. Cell
Meyer, A. S., J. R. Gillespie, D. Walther, I. S. Millet, S. Doniach
et al., 2003 Closing the folding chamber of the eukaryotic chap-
eronin requires the transition state of ATP hydrolysis. Cell 113:
Miller, Jr., W. H., H. M. Schipper, J. S. Lee, J. Singer and S.
Waxman, 2002 Mechanisms of action of arsenic trioxide.
Cancer Res. 62: 3893–3903.
Miossec-Bartoli, C., L. Pilatre, P. Peyron, E. N. N’Diaye, V.
Collart-Dutilleul et al., 1999
accumulates in the azurophil granules of human polymorphonu-
clear cells. Antimicrob. Agents Chemother. 43: 2457–2462.
Nordstrom, D. K., 2002 Public health. Worldwide occurrences of
arsenic in ground water. Science 296: 2143–2145.
Pan, X., and J. Heitman, 2002Protein kinase A operates a molec-
ular switch that governs yeast pseudohyphal differentiation.
Mol. Cell. Biol. 22: 3981–3993.
Pan, X., D. S. Yuan, D. Xiang, X. Wang, S. Sookhai-Mahadeo et al.,
2004A robust toolkit for functional profiling of the yeast
genome. Mol. Cell 16: 487–496.
Ramirez, P., D. A. Eastmond, J. P. Laclette and P. Ostrosky-
Wegman, 1997Disruption of microtubule assembly and spin-
dle formation as a mechanism for the induction of aneuploid
cells by sodium arsenite and vanadium pentoxide. Mutat. Res.
Rieder, S. E., L. M. Banta, K. Kohrer, J. M. McCaffery and S. D.
Emr, 1996 Multilamellar endosome-like compartment accumu-
lates in the yeast vps28 vacuolar protein sorting mutant. Mol. Biol.
Cell 7: 985–999.
Arsenic targets tubulins to induce
The new ketolide HMR3647
Roe, J. S., H. Kim, S. M. Lee, S. T. Kim, E. J. Cho et al., 2006
stabilization and transactivation by a von Hippel-Lindau protein.
Mol. Cell 22: 395–405.
Schiller, C. M., B. A. Fowler and J. S. Woods, 1977
senic on pyruvate dehydrogenase activation. Environ. Health Per-
spect. 19: 205–207.
Shannon, P., A. Markiel, O. Ozier, N. S. Baliga, J. T. Wang et al.,
2003Cytoscape: asoftwareenvironmentfor integratedmodelsof
biomolecular interaction networks. Genome Res. 13: 2498–2504.
Sherman, F., 2002 Getting started with yeast. Methods Enzymol.
Siegers, K., T. Waldmann, M. R. Leroux, K. Grein, A. Shevchenko
et al., 1999Compartmentation of protein folding in vivo: se-
questration of non-native polypeptide by the chaperonin-GimC
system. EMBO J. 18: 75–84.
Siegers, K., B. Bolter, J. P. Schwarz, U. M. Bottcher, S. Guha
et al., 2003 TRiC/CCT cooperates with different upstream
chaperones in the folding of distinct protein classes. EMBO J.
Spiess, C., A. S. Meyer, S. Reissmann and J. Frydman, 2004
anism of the eukaryotic chaperonin: protein folding in the cham-
ber of secrets. Trends Cell. Biol. 14: 598–604.
Tang, F., G. Liu, Z. He, W. Y. Ma, A. M. Bode et al., 2006
inhibits p53 phosphorylation, DNA binding activity, and p53
target gene p21 expression in mouse epidermal JB6 cells. Mol.
Carcinog. 45: 861–870.
Tapio, S., and B. Grosche, 2006
Mutat. Res. 612: 215–246.
Thorsen, M., G. G. Perrone, E. Kristiansson, M. Traini, T. Ye et al.,
2009Genetic basis of arsenite and cadmium tolerance in Sac-
charomyces cerevisiae. BMC Genomics 10: 105.
Thulasiraman, V., C. F. Yang and J. Frydman, 1999
translated polypeptides are sequestered in a protected folding
environment. EMBO J. 18: 85–95.
Tseng, C. H., 2004 The potential biological mechanisms of arsenic-
induced diabetes mellitus. Toxicol. Appl. Pharmacol. 197: 67–83.
Ursic, D., J. C. Sedbrook, K. L. Himmel and M. R. Culbertson,
1994The essential yeast Tcp1 protein affects actin and micro-
tubules. Mol. Biol. Cell 5: 1065–1080.
Vainberg, I. E., S. A. Lewis, H. Rommelaere, C. Ampe, J.
Vandekerckhove et al., 1998
ers unfolded proteins to cytosolic chaperonin. Cell 93: 863–873.
Weber, F., and M. Hayer-Hartl, 2000
chondrial rhodanese by chaperonins GroEL and GroES.
Methods Mol. Biol. 140: 117–126.
Zhang, X., F. Yang, J. Y. Shim, K. L. Kirk, D. E. Anderson et al.,
2007Identification of arsenic-binding proteins in human
breast cancer cells. Cancer Lett. 255: 95–106.
Effects of ar-
Arsenic in the aetiology of cancer.
In vivo newly
Prefoldin, a chaperone that deliv-
Refolding of bovine mito-
Communicating editor: M. D. Rose
734X. Pan et al.
Trivalent Arsenic Inhibits the Functions of Chaperonin Complex
Xuewen Pan, Stefanie Reissman, Nick R. Douglas, Zhiwei Huang, Daniel S. Yuan,
Xiaoling Wang, J. Michael McCaffery, Judith Frydman and Jef D. Boeke
Copyright ? 2010 by the Genetics Society of America
X. Pan et al.
Genome-wide haplo-insufficiency screen As2O3 (Sigma-Aldrich, St. Louis, MO) was first dissolved in NaOH as a 400
mM stock solution of sodium arsenite. A pool of 5996 haploid-convertible heterozygote diploid YKOs was constructed as
previously described (PAN et al. 2004). ~2 x 106 cells of this pool were directly inoculated into 100 ml of YPD liquid that contains
(experiment) or lacks (control) 450 mM of sodium arsenite and incubated by shaking at 30 °C for 10 generations. After this
outgrowth, cells were harvested from both the control and experimental samples for genomic DNA preparation and TAG array
analysis of relative representation of each YKO in both pools as described (PAN et al. 2004). YKOs with control/experiment
hybridization ratios ≥ 2.0 were selected as candidate drug sensitive mutant for individual validation.
Screens for arsenic-hypersensitive and -resistant haploid YKOs The screens were done essentially as described in Pan
et al. (PAN et al. 2004). Briefly, the haploid-convertible heterozygote dipoid YKO pool was sporulated and pools of isogenic
MATa haploid cells were derived by growth on a haploid selection (SC-Leu-His-Arg+canavanine+G418 with glutamic acid
instead of ammonium sulfate as the nitrogen source) that either contained (experiment) or lacked (control) sodium arsenite.
Relative representation of each YKO in drug-treated and untreated pools was compared by TAG-array analysis. The
concentrations of sodium arsenite used were 100 mM, 200 mM, 400 mM, 800 mM, and 1600 mM. Experiments with the lower
concentrations identified As(III)-sensitive mutants, whereas the highest concentration (1600 mM) revealed As(III)-resistant
mutants, which should be over-represented in the drug-treated population because growth of most other haploid YKOs is
Synthetic arsenic-hypersensitivity screen This screen was carried out essentially like a typical dSLAM screen (PAN et al.
2004) with the following modifications. A pfd1∆::URA3 construct that contains ~1.5 kb flanking targeting sequences was PCR
amplified with a pair of primers (5’ GCTGTATCGCACTCAAACAA 3’ and 5’ TACACTACTACACCTCTGCAT 3’) and en
masse transformed into a pool of haploid convertible heterozygote diploid YKOs. The resultant heterozygous diploid double
mutant pool was sporulated and a MATa haploid double mutant (pfd1∆::URA3 xxx∆::kanMX, “xxx∆” stands for any yeast gene
deletion) pool was derived by selection on a haploid selection medium that lacked uracil but contained 5 mM of sodium arsenite
(SC-Leu-His-Arg-Ura+canavanine+G418+ arsenite). As a control, a mixed population of MATa haploid single (xxx∆::kanMX)
and double (pfd1∆::URA3 xxx∆::kanMX) mutants was selected on a haploid selection medium that contained uracil but lacked
arsenic (SC-Leu-His-Arg+canavanine+G418). Relative representation of each YKO in these two populations was compared by
TAG-array analysis. This screen identified mutations that cause growth defects in a pfd1∆ mutant both in the absence and
X. Pan et al.
presence of arsenic 5 mM As(III). These were subsequently confirmed and distinguished by testing individual mutants both in the
presence and absence of the drug as described below.
Individual confirmation of arsenic-hypersensitive and -resistant YKOs To confirm the arsenic-sensitive
heterozygous diploid mutants identified by TAG-array, individual strains were spotted onto solid YPD plus G418 that either
contained or lacked 450 mM of sodium arsenite and incubated at 30°C for 3 days. The results were reported in Fig. 1a and
Supporting Table 1. To confirm arsenic-hypersensitive or -resistant haploid YKOs, individual haploid convertible heterozygous
diploid mutants were sporulated, spotted onto haploid selection media that either contained or lacked sodium arsenite at
indicated concentrations and incubated at 30°C for 3 days. The confirmed results were reported in Supporting Tables 2 and 3.
To confirm mutations that cause growth defects in a pfd1∆ mutant or those that specifically sensitizes the pfd1∆ mutant to arsenic,
individual haploid convertible heterozygote diploid mutants were transformed with a pfd1∆::URA3 construct to create diploid
double mutants. Two independent transformants for each were tested by random spore analysis (RSA) (PAN et al. 2004). Haploid
progeny were spotted as 10X serial dilutions on SC-Leu-His-Arg+canavanine+G418, which allows for growth of xxx∆::kanMX
and pfd1∆::URA3 xxx∆::kanMX cells, SC-Leu-His-Arg-Ura+canavanine, which allows for growth of pfd1∆::URA3 and pfd1∆::URA3
xxx∆::kanMX cells, and SC-Leu-His-Arg-Ura+canavanine+G418, which allows for growth of only the pfd1∆::URA3 xxx∆::kanMX
double-mutant cells) and incubated at 30°C for 2 to 3 days. This was similarly carried out in the presence of 5 µM sodium
arsenite. The confirmed results were reported in Supporting Table 4.
X. Pan et al.
Arsenic-hypersensitive heterozygous diploid YKOs
ORF Name Gene name Biological Process Essentiala
protein folding Y
protein folding Y
protein folding Y
protein folding Y
Dubious ORF overlaps with CCT8, protein folding Y
nuclear organization and biogenesis Y
sphingolipid biosynthesis Y
nuclear transportation Y
mRNA polyadenylylation Y
mRNA polyadenylylation Y
ER to Golgi vesicle-mediated transport Y
microtubule nucleation Y
chromatin remodeling & transcriptional regulation Y
microtubule nucleation Y
adenosine biosynthesis N
aromatic amino acid family biosynthesis N
arsenite efflux N
arsenite efflux N
cell ion homeostasis & cell wall biogenesis N
Intracellular copper ion transport N
chromatin remodeling & transcriptional regulation N
nuclear import & telomere maintenance N
microtubule biogenesis N
proline biosynthesis N
microtubule biogenesis N
transcriptional regulation N
chromatin remodeling N
microtubule biogenesis N
chromatin remodeling & transcriptional regulation N
Note: a “Y” means that the gene is essential to yeast cell viability and “N” means that the gene is not essential. b
“YJL009W” is a dubious ORF that overlaps with the ORF of CCT8 required for protein folding.
X. Pan et al. Download full-text
Arsenite-sensitive haploid YKOs
ORF Name Gene Name Biological Process 100 µM 400 µM
YNL153C GIM3 actin & tubulin folding SS SS
YEL003W GIM4 actin & tubulin folding SS SS
YML094W GIM5 actin & tubulin folding SS SS
YGR078C PAC10 actin & tubulin folding SS SS
YJL179W PFD1 actin & tubulin folding SS SS
YLR200W YKE2 actin & tubulin folding SS SS
YPL161C BEM4 actin cytoskeleton organization and biogenesis S SS
YLR370C ARC18 actin filament organization NO SS
YLR371W ROM2 actin filament organization SS SS
YNL271C BNI1 actin filament organization NO S
YNL084C END3 actin filament organization; endocytosis NO SS
YJR125C ENT3 actin filament organization; endocytosis NO S
YBR200W BEM1 actin organization; establishment of cell polarity NO S
YPR199C ARR1 arsenite detoxification SS SS
YPR201W ARR3 arsenite detoxification SS SS
YDR135C YCF1 arsenite detoxification NO SS
YFL023W BUD27 bud site selection SS SS
YCR063W BUD31 bud site selection NO SS
YKR061W KTR2 cell wall mannoprotein biosynthesis NO S
YCR017C CWH43 cell wall organization and biogenesis NO S
YDR293C SSD1 cell wall organization and biogenesis NO SS
YLR085C ARP6 chromatin remodeling; transcriptional regulation NO S
YOL012C HTZ1 chromatin remodeling; transcriptional regulation NO SS
YOR304W ISW2 chromatin remodeling; transcriptional regulation NO S
YNL192W CHS1 cytokinesis NO S
YOL076W MDM20 cytoskeleton organization and biogenesis NO S
YNL138W SRV2 cytoskeleton organization and biogenesis NO SS
YDL101C DUN1 DNA damage checkpoint NO S
YBL051C PIN4 DNA damage checkpoint NO S
YER177W BMH1 DNA damage checkpoint; cell polarization NO S
YLL002W RTT109 DNA damage response; DNA transpotision NO S
YDL013W HEX3 DNA repair NO S
YPR164W MMS1 DNA repair NO S
YLR320W MMS22 DNA repair NO SS
YMR224C MRE11 DNA repair NO SS
YCR066W RAD18 DNA repair NO S
YNL250W RAD50 DNA repair NO S