International Journal of
Molecular Sciences
Article
Effects of Defective Unloading and Recycling of PCNA
Revealed by the Analysis of ELG1 Mutants
Ziv Itzkovich † , Karan Choudhary † , Matan Arbel * and Martin Kupiec *
Citation: Itzkovich, Z.; Choudhary,
K.; Arbel, M.; Kupiec, M. Effects of
Defective Unloading and Recycling
of PCNA Revealed by the Analysis of
ELG1 Mutants. Int. J. Mol. Sci. 2023,
24, 1568. https://doi.org/10.3390/
ijms24021568
Academic Editors: Anne Donaldson
and Takashi Kubota
Received: 5 December 2022
Revised: 10 January 2023
Accepted: 11 January 2023
Published: 13 January 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Int. J. Mol. Sci. 2023, 24, 1568. https://doi.org/10.3390/ijms24021568
The Shmunis School of Biomedicine and Cancer Research, The George S. Wise Faculty of Life Sciences,
Tel Aviv University, Tel Aviv 69978, Israel
* Correspondence: matanarble@gmail.com (M.A.); martin@tauex.tau.ac.il (M.K.)
† These authors contributed equally to this work.
Abstract: Timely and complete replication of the genome is essential for life. The PCNA ring plays an
essential role in DNA replication and repair by contributing to the processivity of DNA polymerases
and by recruiting proteins that act in DNA replication-associated processes. The ELG1 gene encodes
a protein that works, together with the Rfc2-5 subunits (shared by the replication factor C complex),
to unload PCNA from chromatin. While ELG1 is not essential for life, deletion of the gene has strong
consequences for the stability of the genome, and elg1 mutants exhibit sensitivity to DNA damaging
agents, defects in genomic silencing, high mutation rates, and other striking phenotypes. Here, we
sought to understand whether all the roles attributed to Elg1 in genome stability maintenance are
due to its effects on PCNA unloading, or whether they are due to additional functions of the protein.
By using a battery of mutants that affect PCNA accumulation at various degrees, we show that all
the phenotypes measured correlate with the amount of PCNA left at the chromatin. Our results
thus demonstrate the importance of Elg1 and of PCNA unloading in promoting proper chromatin
structure and in maintaining a stable genome.
Keywords: genome stability; Saccharomyces cerevisiae; DNA replication; DNA repair; mutagenesis;
SRS2; gene silencing
1. Introduction
DNA replication is one of the central processes of life. The robust and fast replication
of the entirety of the genome is essential for life and must be well regulated. If the
replication takes too long, the cell cycle may become disrupted; a fast but inaccurate DNA
replication, on the other hand, would lead to accumulation of mutations, most of which
will be deleterious to the next generation [1,2]. Thus, a well-kept balance exists between
the accuracy and the speed of replication in all organisms investigated. This balance is
maintained by a group of proteins involved in DNA replication, DNA damage avoidance,
and DNA damage repair [3–5]. The full and accurate replication of the genome poses
another challenge, which is the unpacking and repacking of DNA during replication [6].
The genomic DNA is neatly organized within the nucleus [7,8]. DNA is wrapped around
nucleosomes, and post-translational modifications of the histones that compose these basic
structures determines whether chromatin is in a (mostly) active (euchromatin) or inactive
(heterochromatin) configuration [9,10]. Since DNA replication requires the opening of
double-stranded DNA, nucleosomes must be evicted from DNA. In order to maintain the
same epigenetic state, the two newly replicated DNA molecules have to be repacked around
nucleosomes containing the same histone modifications present before DNA replication.
This is achieved by the activity of a large number of proteins involved in the transfer
of modified histones to both leading and lagging strands on the wake of the moving
fork [11]. The activity of all these important factors is coordinated and regulated by the
DNA replication clamp, PCNA [12–14].
https://www.mdpi.com/journal/ijms
Int. J. Mol. Sci. 2023, 24, 1568 2 of 13

PCNA is a homotrimeric complex that acts as a sliding clamp, enclosing DNA [15]; it
is a master regulator for different DNA repair and chromatin assembly processes [16,17].
PCNA coordinates the work of many proteins involved in many cellular processes. PCNA
functions as a processivity factor for DNA polymerases [18]; it helps recruiting Okazaki
fragment maturation factors such as Rad27 and Cdc9 (FEN1 and LIG1 in humans) [19,20];
it is involved in the recruitment of DNA damage tolerance proteins such as Rev3 and
Rad5 [21,22]. In addition, PCNA has physical interactions with proteins involved in histone
deposition and chromatin assembly at the fork, such as the CAF-1 subunit Cac2 [17]. Thus,
understanding the regulation of PCNA and the dynamics of its loading and unloading
from chromatin is of utmost importance.
PCNA is recruited and loaded onto chromatin by the RFC complex [23], composed of a
large subunit (Rfc1) and four small subunits (Rfc2-5). Three additional RFC-like complexes
(RLCs) share the small subunits, but each one carries a different, unique protein instead
of Rfc1. Whereas one of them (carrying the Rad24 protein in S. cerevisiae and RAD17
in humans) oversees the loading of the 9–1–1 checkpoint complex, the two other RLCs,
carrying Elg1/ATAD5 or Ctf18/CHTF18, interact with PCNA. The focus of the current
study is the Elg1 RLC, which has been shown to work as the main unloader of PCNA
(reviewed in [13]). Elg1 is a nonessential protein, and its deletion causes a wide range of
phenotypes, including increased chromosome loss and homologous recombination, [24],
increased sensitivity to DNA damage [25], and defects in chromatin maintenance [26]. In
addition to this, ELG1 also genetically interacts with the genes encoding the sister chromatid
cohesion machinery [27,28]. Over the years, there has been much debate on the possible
roles of Elg1, and whether it carries any role outside of PCNA unloading, as hinted by
its importance in many cellular processes. Here, we take advantage of the availability
of a collection of elg1 mutants to probe their effect on various phenotypes. Our results
suggest that all the phenotypes of elg1 mutants are due to the accumulation of PCNA on the
chromatin, suggesting that Elg1’s sole role is the timely unloading and recycling of PCNA.

2. Results
2.1. A Collection of elg1 Mutants
Deletion of the ELG1 gene is not lethal, but results in a vast array of phenotypes,
including sensitivity to DNA damaging agents, hyper-recombination, chromosome loss,
and decreased chromatin silencing (reviewed in [13]). However, most experiments were
carried out with strains bearing a total deletion of the ELG1 gene. To investigate whether
all the phenotypes of elg1 mutants were due to a defect in PCNA unloading activity, we
created a collection of point mutations in the ELG1 gene, which affect Elg1’s unloading
activity to different degrees, and studied their effect on a number of phenotypes. We created
mutations at different motifs in the protein, as follows (Figure 1):
(1) A sequence at the N-terminus of Elg1 bears resemblance to the PCNA-interacting
peptide (PIP) of Rfc1 [29]. We mutated amino acids N55 and S57 to alanine (elg1-
NS55,57AA; this allele is hereafter referred to as elg1-PIP).
(2) Three SUMO-interacting motifs were found in the same region, and point mutations
in them (I28A I93K II121/2AA) reduced the interaction between Elg1 and SUMO [29].
We refer to this allele as elg1-SIM.
(3) The third region mutated in the N terminus was an unstructured loop region, spanning
from aa 290 to aa 319 and containing a hydrophobic patch. This site is a possible
protein–protein interaction site and is conserved throughout all Elg1 orthologs but is
absent in other components of the RFC complex and in the other Rfc1-like proteins [30].
We deleted this loop and replaced it by a short peptide. This allele is referred to as
elg1-loop.
(4) By homology modeling of the Elg1 protein on the solved crystal structure of the
RFC complex together with PCNA, it was determined that threonines 386/7 in Elg1
correspond to asparagines 694 and 695 of the human Rfc1, which were shown to be at
 complex together with PCNA, it was determined that threonines 386/7 in Elg1 corre-
spond to asparagines 694 and 695 of the human Rfc1, which were shown to be at the
interface of the two proteins [30]. These residues were mutated to either aspartic acid
(hereafter referred to as elg1-TT386/7DD) or to alanine (elg1-TT386/7AA).
Int. J. Mol. Sci. 2023, 24, 1568 (5) These mutants were also combined with the SIM mutation, creating SIM+AA3and of 13
SIM+DD alleles.
(6) Elg1 and all members of the AAA+ family contain Walker A and Walker B motifs
[31]. The current models suggest that ATP binding is crucial for the association of
the interface of the two proteins [30]. These residues were mutated to either aspartic
Elg1-RLC with PCNA for its unloading, and that the hydrolysis of ATP is important
acid (hereafter referred to as elg1-TT386/7DD) or to alanine (elg1-TT386/7AA).
for the subsequent detachment of Elg1-RFC from PCNA, allowing its recycling
(5) These mutants were also combined with the SIM mutation, creating SIM+AA and
[23,32]. We mutated the two lysines (at positions 343, 344) either to aspartic acid (elg1-
SIM+DD alleles.
KK343/4DD) or to alanine (elg1-KK343/4AA). These are referred to hereafter as Walk-
(6) Elg1 and all members of the AAA+ family contain Walker A and Walker B motifs [31].
erA-DD (WA-DD)
The current modelsand Walkerthat
suggest A-AA
ATP(WA-AA).
binding is crucial for the association of Elg1-RLC
(7) We also mutated two aspartic acids
with PCNA for its unloading, and that at positions 407,409 that
the hydrolysis lie within
of ATP the Walker
is important for theB
motif, which detachment
subsequent are conserved of between
Elg1-RFC Elg1
fromand its human
PCNA, ortholog
allowing ATAD5 [23].
its recycling These
[23,32]. We
residues are thought to be crucial for ATP binding and/or hydrolysis. They
mutated the two lysines (at positions 343, 344) either to aspartic acid (elg1-KK343/4DD) were mu-
tated
or to either
alanine to(elg1-KK343/4AA).
lysine (elg1-DD407/409KK)
These areor to alanine
referred (elg1-DD407/9AA),
to hereafter as WalkerA-DD hereafter
(WA-
referred to as Walker B-KK
DD) and Walker A-AA (WA-AA). (WB-KK) or Walker B-AA (WB-AA).
(8)
(7) We alsomutated
We also created a double
two aspartic acidsWalker
at positionsA 407,409
+ Walker B mutant:
that lie within elg1-B
the Walker
KK343/4DD+DD407/409AA,
motif, which are conserved hereafter referred
between Elg1 and to
itsas elg1-WAB.
human ortholog ATAD5 [23]. These
(9) Finally,
residues are thought to be crucial for ATP binding and/or (elg1-V390D)
we mutated position 390 from valine to aspartic acid hydrolysis. They and to al-
were
anine (elg1-V390A). These mutations disrupt the similarity in Elg1 at
mutated either to lysine (elg1-DD407/409KK) or to alanine (elg1-DD407/9AA), hereafter positions 381-
390 (LLDFTTTHYV)
referred to as Walkerto a small
B-KK patch or
(WB-KK) of Walker
the DNA B-AArepair and telomeric protein Yku80
(WB-AA).
(aa 450-456 LLDrTTTsgV). Moreover, the valine is conserved
(8) We also created a double Walker A + Walker B mutant: elg1-KK343/4DD+DD407/409AA, throughout Elg1
orthologs
hereafter [23].
referred to as elg1-WAB.
(9) All
Finally, we
the strains mutated
carryingposition
the ELG1390alleles
from valine to asparticplasmids,
on centromeric acid (elg1-V390D)
or replacingandthe to
alanine (elg1-V390A). These mutations disrupt the similarity in
ELG1 gene in the genome, showed the same cell cycle distribution (Figure S1). We tested Elg1 at positions
381-390
our battery of (LLDFTTTHYV) to a small
mutants for a number patch of the
of phenotypes DNA
that repair and
characterize telomeric
the full ELG1 protein
dele-
tion. Yku80 (aa 450-456 LLDrTTTsgV). Moreover, the valine is conserved throughout Elg1
orthologs [23].

Figure 1. Schematic representation of the Elg1 protein showing domains and mutation sites. The
Figure 1. Schematic
ELG1 gene representation
was divided of the Elg1
into three domains. proteindescribed
Mutations showingindomains and mutation
the text affect the N-tersites. The
and AAA
ELG1 gene was divided into three domains. Mutations described in the text affect the N-ter and
domains. All proteins are expressed at similar levels.
AAA domains. All proteins are expressed at similar levels.
All the strains carrying the ELG1 alleles on centromeric plasmids, or replacing the
2.2.
ELG1 Sensitivity to DNA
gene in the Damage
genome, showed the same cell cycle distribution (Figure S1). We tested
Methylof
our battery methanesulfonate (MMS)
mutants for a number ofisphenotypes
a DNA alkylating agent thatthe
that characterize impairs DNA
full ELG1 repli-
deletion.
cation (as alkylated DNA is poorly replicated by DNA polymerases and must be effi-
2.2. Sensitivity
ciently repaired)to [33].
DNAAllDamage
elg1 mutants, as well as the wild-type control, were cloned into
a centromeric plasmid (pRS415)(MMS)
Methyl methanesulfonate and introduced into a elg1Δagent
is a DNA alkylating yeastthat
strain. An empty
impairs vector
DNA replica-
tion (as alkylated DNA is poorly replicated by DNA polymerases and must be efficiently
repaired) [33]. All elg1 mutants, as well as the wild-type control, were cloned into a cen-
tromeric plasmid (pRS415) and introduced into a elg1∆ yeast strain. An empty vector
served as an additional control. Serial dilutions of the different strains were spotted on
SD-complete plates or on similar plates with varying amounts of MMS (Figure 2). Wild-
type cells were able to grow at all MMS concentrations used, while elg1∆ cells exhibited
increased sensitivity to the DNA damaging agent.

Int. J. Mol. Sci. 2023, 24, 1568
served as an additional control. Serial dilutions of the different strains were spotted on
SD-complete plates or on similar plates with varying amounts of MMS (Figure 2). Wild-
type cells were able to grow at all MMS concentrations used, while elg1Δ cells exhibited
4 of 13
increased sensitivity to the DNA damaging agent.

Figure 2.
Figure 2. DNA
DNA damage
damage sensitivity
sensitivity assay
assay for different ELG1
for different ELG1 mutations. Serial fivefold
mutations. Serial fivefold dilutions
dilutions of
of
yeast cultures on SD-complete without and with methyl methanesulphonate (MMS) at the
yeast cultures on SD-complete without and with methyl methanesulphonate (MMS) at the indicated indicated
concentrations show different sensitivities of elg1 mutants. elg1Δ
concentrations elg1∆ strains were transformed with a
centromeric pRS415
pRS415 plasmid
plasmidcollection
collectioncontaining
containingdifferent
differentELG1
ELG1 mutations.
mutations. The
The strain
strain described
described as
as “Wild
“Wild type”
type” carried
carried thethe ELG1
ELG1 gene,
gene, andand
the the
oneone described
described as “elg1Δ”
as “elg1∆” carried
carried an empty
an empty vector.
vector.

In accordance
accordance with
withprevious
previousresults
results[34],
[34],nono sensitivity
sensitivity to to
MMS MMSwaswas observed
observed for
for the
the PIP
PIP or SIM
or SIM mutants.
mutants. AmongAmongthe the different
different Elg1Elg1 mutants,
mutants, MMS MMS sensitivity
sensitivity waswas
seenseen
for
for TT386/7
TT386/7 whenwhen mutated,
mutated, especially
especially to aspartic
to aspartic acid,
acid, but theybut
alsothey also exhibited
exhibited a mild
a mild sensitiv-
sensitivity
ity in the highest
in the highest of MMSofdoses
MMStested
doseswhen
testedchanged
when changed to alanine.
to alanine. This sensitivity
This sensitivity was in-
was increased
creased when combined
when combined with thewith
SIMthe SIM mutations.
mutations. MutationMutation
at V390atalso
V390 also to
seems seems
conferto
confer increased sensitivity when mutated to aspartic acid but not to alanine.
increased sensitivity when mutated to aspartic acid but not to alanine. Growth sensitivity Growth
sensitivity
was wasfor
also seen also seenWalker
both for both Walker mutations,
mutations, with the
with the lysines at lysines at 343/4
343/4 (Walker A (Walker
region)
A region) showing sensitivity when mutated to aspartic acid but not to
showing sensitivity when mutated to aspartic acid but not to alanine, and the aspartic alanine, and the
aspartic
acids acids at positions
at positions 407/9B (Walker
407/9 (Walker B region)
region) showed showedsensitivity
increased increased when
sensitivity when
mutated to
mutated
both to both
lysine lysine and alanine.
and alanine.
Apart from
Apart from the
the mutations
mutations mentioned,
mentioned, no no other elg1 mutation
other elg1 mutation within
within the
the group
group in-in-
creased MMS sensitivity.
creased MMS sensitivity.
2.3. Mutation Rate
2.3. Mutation Rate
The rate of genomic mutation is usually assessed by either measuring the rate of
The rate of genomic mutation is usually assessed by either measuring the rate of in-
inactivation of a gene that confers sensitivity to a drug (e.g., resistance to canavanine,
activation of a gene that confers sensitivity to a drug (e.g., resistance to canavanine, ac-
acquired by inactivating the arginine pump encoded by the CAN1 gene), or by reversion of
quired by inactivating the arginine pump encoded by the CAN1 gene), or by reversion of
a particular auxotrophic mutation. In the absence of the ELG1 gene, the rate of mutation is
a particular auxotrophic mutation. In the absence of the ELG1 gene, the rate of mutation
not greatly altered at the CAN1 locus, but slippage mutations are strikingly elevated [35].
is not greatly altered at the CAN1 locus, but slippage mutations are strikingly elevated
These can be easily scored using strains in which a row of 14 adenine residues is inserted
[35]. These can be easily scored using strains in which a row of 14 adenine residues is
into the LYS2 gene [36]. Addition of a single residue or deletion of two nucleotides restores
the reading frame and confers a Lys+ phenotype. elg1∆ strains exhibit roughly a ~5 fold
increase in the slippage rate using this assay. A possible explanation is that over-retention
of PCNA on the chromatin in the deletion mutant results in untimely or reduced activity of
the mismatch repair complex [37].
 inserted into the LYS2 gene [36]. Addition of a single residue or deletion of two nucleo-
tides restores the reading frame and confers a Lys+ phenotype. elg1Δ strains exhibit
Int. J. Mol. Sci. 2023, 24, 1568 roughly a ~5 fold increase in the slippage rate using this assay. A possible explanation 5 ofis13
that over-retention of PCNA on the chromatin in the deletion mutant results in untimely
or reduced activity of the mismatch repair complex [37].
We introduced all the elg1 mutations into a strain carrying the lys2-14A allele and
We introduced all the elg1 mutations into a strain carrying the lys2-14A allele and
measured,
measured,bybyfluctuation
fluctuationrate,
rate,their
theirrate
rateofofmutagenesis
mutagenesis(Figure
(Figure3).
3).The
Theelg1
elg1mutant
mutantpanel
panel
could be divided roughly in two groups, one that exhibited a wildtype level
could be divided roughly in two groups, one that exhibited a wildtype level of mutagenesisof mutagen-
esis
andand another
another thatthat had
had levelssimilar
levels similartotothe elg1Δ allele.
theelg1∆ allele. The
The first
firstgroup
groupincludes
includesthe
thePIP,
PIP,
SIM,
SIM, TT386/7AA, loop, V390A, and WA-AA alleles. Members of the second group arethe
TT386/7AA, loop, V390A, and WA-AA alleles. Members of the second group are the
TT386/7DD,
TT386/7DD,SIM+AA,
SIM+AA,SIM+DD,
SIM+DD,V390D,V390D, WA-DD,
WA-DD,WAB, WAB,WB-KK,
WB-KK,and andWB-AA.
WB-AA.

Figure 3. Slippage mutation rates of elg1 alleles. Reversion of the lys2-14A allele was measured by
Figure 3. Slippage mutation rates of elg1 alleles. Reversion of the lys2-14A allele was measured by
fluctuationtest
fluctuation testininall
allisogenic elg1mutants.
isogenicelg1 mutants.The Therates
rates were
were compared
compared to to that
that of
of the
the wildtype,
wildtype,which
whichis
isconsidered
consideredasas 1.
1. At least three
At least three experiments
experiments were wereused
usedfor
forquantitation
quantitationandandthe
theerror
errorbars
barsrepresent
represent
thestandard
the standarderror
errorofofthethemean. 0.05> >p p> >
mean.* *0.05 0.01;
0.01; ** ** > p> >p 0.001.
0.008
0.008 > 0.001.

2.4. Rescue of the Synthetic Lethality of Elg1∆ Srs2∆ Double Mutants

2.4. Rescue of the Synthetic Lethality of Elg1Δ Srs2Δ Double Mutants
The Srs2 helicase plays a central role during the repair of double-stranded breaks by
The Srs2 helicase plays a central role during the repair of double-stranded breaks by
homologous recombination [38] and has been shown to prevent the bypass of genomic
homologous recombination [38] and has been shown to prevent the bypass of genomic
lesions using this mechanism [39]. Double mutants elg1∆ srs2∆ grow very slowly as
lesions using this mechanism [39]. Double mutants elg1Δ srs2Δ grow very slowly as hap-
haploids and are essentially dead as diploids [40]. The reason for this lethality is still
loids and are essentially dead as diploids [40]. The reason for this lethality is still un-
unknown. We used our battery of mutants to ask which of them was able to suppress the
known. We used our battery of mutants to ask which of them was able to suppress the
lethality. We created a diploid stain carrying genomic deletions of the ELG1 and SRS2 genes,
lethality. We created a diploid stain carrying genomic deletions of the ELG1 and SRS2
which is kept alive by the presence of a URA3-marked plasmid bearing the ELG1 gene.
genes, which is kept alive by the presence of a URA3-marked plasmid bearing the ELG1
These cells cannot lose the plasmid, as the diploid double mutant without it is inviable.
gene. Theseagainst
Selection cells cannot lose marker
the URA3 the plasmid,
carriedasbythethe
diploid double
covering mutant
plasmid (onwithout it is invi-
5-FOA-containing
able. Selection against the URA3 marker carried by the covering plasmid (on
medium) allows us to test whether the elg1 allele introduced can complement the activity 5-FOA-con-
taining
requiredmedium) allowsofusthe
for survival to test whether
strain. Figure the4 elg1
showsallele introduced
that a control can complement
plasmid carryingthethe
activity required for survival of the strain. Figure 4 shows that a control
wildtype ELG1 gene complements, allowing growth on 5FOA, whereas an empty vector plasmid carrying
the wildtype
is unable to ELG1
allowgene complements,
plasmid loss. Again, allowing
the elg1growth
mutantsoncan5FOA, whereas
be divided inantwoempty vec-
categories,
tor is unable to allow plasmid loss. Again, the elg1 mutants
which coincide with the results observed in the slippage assay above. can be divided in two catego-
ries, which coincide with the results observed in the slippage assay above.
Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 6 of 13

Int. J. Mol. Sci. 2023, 24, 1568 6 of 13

Figure 4.
Figure Rescue of
4. Rescue of the
the synthetic
synthetic lethality
lethality of elg1∆ srs2∆.
of elg1∆ srs2∆. Diploid
Diploid stains
stains homozygous
homozygous for ELG1 and
for ELG1 and
SRS2 genomic deletions are kept alive by the presence of a plasmid carrying the ELG1
SRS2 genomic deletions are kept alive by the presence of a plasmid carrying the ELG1 and URA3 and URA3
genes. We
genes. We introduced
introduced into
into these
these strains
strains the elg1 allele
the elg1 allele collection
collection on LEU2-marked pRS415
on LEU2-marked pRS415 vectors.
vectors.
5-FOA-containing medium selects for cells able to lose the URA3-ELG1 covering plasmid. The strain
described as “Wild type” carried the ELG1 gene, and the one described as “elg1Δ” “elg1∆” carried an empty
empty
pRS415
pRS415 vector.
vector.

2.5. Gene
2.5. Gene Silencing
Silencing in in elg1
elg1 Mutants
Mutants
Regions of
Regions of the
thegenome
genomethatthatare
are silenced,
silenced, such
such as the
as the mating
mating typetype HMLHML
cassettes
cassettes and
HMR
HMR in yeast, are not expressed, and are maintained in a heterochromatin-like statestate
and in yeast, are not expressed, and are maintained in a heterochromatin-like de-
despite
spite thethe
factfact that
that cells
cells proliferate.
proliferate. Thus,
Thus, thethe cells
cells areare able
able to duplicate
to duplicate thethe genome
genome andand
to
to maintain
maintain an an epigenetic
epigenetic memory
memory of the
of the locallocal chromatin
chromatin state.
state. We have
We have shown shown
in theinpast
the
that thatΔelg1∆
past elg1 cells
cells are are defective
defective in thisinprocess,
this process,
using using the CRASH
the CRASH (Cre-reported
(Cre-reported alteredaltered
states
of heterochromatin) assay [17]. This system, which monitors silencing at HML, at
states of heterochromatin) assay [17]. This system, which monitors silencing hasHML,
two
has two genomic
genomic components: components: thea first
the first is Cre is a Cre recombinase
recombinase gene inserted
gene inserted at HML.atThe HML. The
second
second component is an RFP gene, placed between two LoxP sites,
component is an RFP gene, placed between two LoxP sites, downstream to a strong con- downstream to a strong
constitutive
stitutive GPD GPD promoter
promoter andand upstream
upstream to atopromoterless
a promoterlessGFPGFP gene.
gene. If silencing
If silencing at at
HMLHML is
is lost, even transiently, and the Cre gene is expressed, leading to site-specific
lost, even transiently, and the Cre gene is expressed, leading to site-specific recombination recombination
between the LoxP sites, it results in a permanent switch in expression from RFP to GFP
between the LoxP sites, it results in a permanent switch in expression from RFP to GFP
(Figure 5A). Such silencing-loss events are manifested as green sectors in colonies and
(Figure 5A). Such silencing-loss events are manifested as green sectors in colonies and
green cells in flow cytometry and appear at low levels in wildtype cells (Figure 5B).
green cells in flow cytometry and appear at low levels in wildtype cells (Figure 5B).
Flow cytometry analysis was used to quantify the rate of RFP to GFP switching. Cells
that have recently switched express both RFP (remaining from the previous generations)
and GFP (newly switched). Thus, the number of both RFP- and GFP-positive cells compared
to the total number of cells shows the apparent rate of silencing loss. We call this rate
“apparent” because RFP molecules may be still present in cells for more than one generation
after switching from RFP to GFP expression. However, this calculation allows a quantitative
comparison between the different mutants. We tested all the mutants in the CRASH assay,
and the results of three experiments for each strain are shown in Figure 5C.
 Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 7 of 13
Int. J. Mol. Sci. 2023, 24, 1568 7 of 13

Figure 5. The apparent silencing-loss rates of the different elg1 mutations. (A) Schematic descrip-
Figure 5. The apparent silencing-loss rates of the different elg1 mutations. (A) Schematic descrip-
tion of the CRASH assay. (B) Colony sectoring of wildtype and elg1∆ alleles. (C) Genomic versions of
tion of the CRASH assay. (B) Colony sectoring of wildtype and elg1Δ alleles. (C) Genomic versions
the elg1 allele collection were created in CRASH-containing strains. Cre activity was quantified using
of the elg1 allele collection were created in CRASH-containing strains. Cre activity was quantified
flow
usingcytometry, as described
flow cytometry, in Materials
as described and Methods.
in Materials **** p < 0.0001
and Methods. **** p was calculated
< 0.0001 using ANOVA
was calculated using
and Dunnett’s test post-hoc analysis.
ANOVA and Dunnett’s test post-hoc analysis.

The elg1∆ mutant exhibits a 12-fold increase in the apparent rate of silencing, in com-
Flow cytometry analysis was used to quantify the rate of RFP to GFP switching. Cells
parison to the wildtype control. The PIP, SIM, loop, V390A, and Walker A-AA alleles showed
that have recently switched express both RFP (remaining from the previous generations)
wildtype levels of silencing. The elg1-TT386/7AA allele showed a slight but persistent
and GFP (newly switched). Thus, the number of both RFP- and GFP-positive cells com-
increase that was statistically significant. Addition of the SIM mutations further increased
pared to the total number of cells shows the apparent rate of silencing loss. We call this rate
the silencing twofold. Similarly, the elg1-V390D allele showed an intermediate level of si-
"apparent" because RFP molecules may be still present in cells for more than one genera-
lencing, whereas the elg1-TT386/7DD, sim+DD, WA-DD, WB-KK, WB-AA and WAB showed
tion after switching from RFP to GFP expression. However, this calculation allows a quan-
an extent of silencing loss comparable to that of the elg1∆ strain.
titative comparison between the different mutants. We tested all the mutants in the
CRASH
2.6. PCNAassay, and the results
Accumulation Levels of
in three experiments
the elg1 Mutants for each strain are shown in Figure 5C.
The elg1Δ mutant exhibits a 12-fold increase in the apparent rate of silencing, in com-
We then proceeded to measure the level of chromatin-bound PCNA levels in all
parison to the wildtype control. The PIP, SIM, loop, V390A, and Walker A-AA alleles
our different mutants. (Figure 6). A wildtype strain and a full ELG1 deletion served as
showed wildtype
appropriate levels
controls. All of silencing.
cell cultures The
wereelg1-TT386/7AA allele
grown to mid-log showed
in rich a slight
medium but and
(YPD), per-
sistent increase that was statistically significant. Addition of the SIM mutations
proteins were extracted and subjected to fractionation into chromatin and cytoplasmic further
 different mutants. (Figure 6). A wildtype strain and a full ELG1 deletion served as appro‐
priate controls. All cell cultures were grown to mid‐log in rich medium (YPD), and pro‐
teins were extracted and subjected to fractionation into chromatin and cytoplasmic frac‐
tions (see Materials and Methods). This process was repeated independently three times 
Int. J. Mol. Sci. 2023, 24, 1568 and the results were quantified after normalizing to the protein levels of the chromatin‐ 8 of 13
bound protein H3 (Figure  6). As the elg1‐sim mutants had slightly lower expression levels 
(Figure S2), we overexpressed their protein from high copy number plasmids (Figure S3). 
However, Figure S4 shows that PCNA accumulation was identical irrespective of protein 
fractions (see Materials and Methods). This process was repeated independently three times
levels, 
and theindicating 
results werethat  the  phenotypes 
quantified observed  were 
after normalizing to thenot  affected 
protein by  of
levels expression  levels. 
the chromatin-
elg1Δ cells exhibited a sevenfold increase in the level of chromatin‐bound PCNA in com‐
bound protein H3 (Figure 6). As the elg1-sim mutants had slightly lower expression levels
parison S2),
(Figure to  the  wildtype  control. 
we overexpressed theirThe  PIP  and 
protein fromWalkerA‐AA  mutations 
high copy number showed 
plasmids wildtype 
(Figure S3).
levels of PCNA. Interestingly, the Loop and V390A mutants showed lower than wildtype 
However, Figure S4 shows that PCNA accumulation was identical irrespective of protein
levels indicating
levels, (0.85  and that
0.88, 
therespectively).  The  SIM were
phenotypes observed mutant  exhibited 
not affected a  1.7‐fold  increase, 
by expression the 
levels. elg1∆
TT386/7AA mutant a 1.9‐fold elevation, which was further increased by addition of the 
cells exhibited a sevenfold increase in the level of chromatin-bound PCNA in comparison to
SIM mutations to 3‐fold. The TT386/7DD mutant also showed a 3‐fold increase in chroma‐
the wildtype control. The PIP and WalkerA-AA mutations showed wildtype levels of PCNA.
tin‐bound PCNA, and this value was further elevated to 4.63‐fold by the addition of the 
Interestingly, the Loop and V390A mutants showed lower than wildtype levels (0.85 and
SIM mutations. The V390D showed a 3.5‐fold increase, the Walker A‐DD a 6‐fold elevation, 
0.88, respectively). The SIM mutant exhibited a 1.7-fold increase, the TT386/7AA mutant a
and the Walker B and the combined Walker AB mutant a 4–5‐fold increase in chromatin‐
1.9-fold elevation, which was further increased by addition of the SIM mutations to 3-fold.
bound PCNA. 
The TT386/7DD mutant also showed a 3-fold increase in chromatin-bound PCNA, and this
value was further elevated to 4.63-fold by the addition of the SIM mutations. The V390D
showed a 3.5-fold increase, the  Walker A-DD a 6-fold elevation, and the Walker B and the
combined Walker AB mutant a 4–5-fold increase in chromatin-bound PCNA.

Figure 6. PCNA chromatin levels for the different elg1 mutations. (A) A fractionation assay fol-  
Figure 6. PCNA chromatin levels for the different elg1 mutations. (A) A fractionation assay fol‐
lowed by Western blot to detect PCNA levels on the chromatin in the various elg1 mutants. Histone
lowed by Western blot to detect PCNA levels on the chromatin in the various elg1 mutants. Histone 
H3 served as a chromatin-bound- as well as loading control. The full blots are shown as Figure S4.
H3 served as a chromatin‐bound‐ as well as loading control. The full blots are shown as Figure S4. 
(B) Relative levels of silencing loss. At least three experiments were used for quantitation and the
(B) Relative levels of silencing loss. At least three experiments were used for quantitation and the 
error bars represent the standard error of the mean. The graph represents the quantification of
Western blots of the relative chromatin-bound PCNA abundance of the different Elg1 mutants in the
cell, compared to the relative level of wildtype cells. **, 0.006 > p > 0.001; ***, p < 0.0001; values were
 
calculated using ANOVA and Dunnett’s test post-hoc analysis. (C). Correlation between PCNA
accumulation and the apparent silencing-loss rate. A correlation is observed between the levels of
PCNA and the loss of silencing for each elg1 mutant as measured by the CRASH assay. The R2 is
0.6731 (p < 0.0001).

Thus, all point mutants examined retained some level of PCNA unloading, and none
reached the accumulation seen in the elg1∆ allele. Our mutants showed a gradient of effects
on PCNA accumulation, and in general, it is clear that all the phenotypes analyzed correlate
well with the amount of PCNA accumulated, although differences could be observed
depending on the phenotype scored (see Discussion). The correlation is most clearly seen
for the results of the CRASH assay (Figures 5 and 6C), in which subtle phenotypes are
Int. J. Mol. Sci. 2023, 24, 1568 9 of 13

observed for mutants with weak effects, but it can also be seen for DNA damage sensitivity
(Figure 2), Lys+ mutation rate (Figure 3), and the suppression of synthetic lethality in
the srs2∆ elg1∆ strain (Figure 4). Our results thus further solidify the notion that Elg1’s
PCNA unloading role is the main reason for all the genomic instability effects seen in elg1
mutant cells.

3. Discussion
PCNA is a central player during DNA replication and repair. It acts as a landing
platform that directs proteins to the moving replication fork, according to need. The
proteins recruited may act on DNA (e.g., DNA polymerases, Okazaki fragment maturation
enzymes) or may affect chromatin remodeling (e.g., histone chaperones). The combined
work of loading PCNA by RFC and its unloading by the Elg1-RLC provides the cell with the
ability to recycle PCNA quickly and accurately to and from needed areas during replication.
Whereas no subunit of the RFC complex can be deleted without losing viability,
cells carrying a full deletion of the ELG1 gene are alive, but present striking phenotypes
related to genome stability: hyper-recombination, gross chromosomal rearrangements,
chromosome loss, elongated telomeres, increased loss of silencing at heterochromatic
regions, increased mutagenesis, and sensitivity to DNA damaging agents (reviewed in [13]).
Previous work has demonstrated that the phenotypes of elg1∆ related to sensitivity to DNA
damage [34,41], hyper-recombination [24], and telomere length ([42], our unpublished
results) were due to the accumulation of PCNA, and in particular its SUMOylated version,
on chromatin. However, other processes affected by Elg1 activity, such as the maintenance
of gene silencing, the synthetic lethality in the absence of Srs2, or the rate of replication
slippage, were not analyzed.
Here, we have presented a panel of elg1 mutants that display different amounts of
PCNA on the chromatin. We have systematically analyzed their performance in several
assays: sensitivity to MMS, slippage mutagenesis, synthetic sickness with srs2∆, and loss
of gene silencing. From our results, it is clear that all the phenotypes tested are correlative
with the amount of PCNA left on the chromatin. Mutants that had no effect on the level
of PCNA, such as elg1-PIP and WalkerA-AA, showed wildtype phenotypes in all assays,
whereas mutants that accumulated PCNA to 2/3 the level of the deletion allele (e.g., elg1-
SIM+DD, Walker A-DD, Walker B alleles, and the double Walker A-Walker B mutant) all
showed phenotypes indistinguishable from those of the elg1∆ allele. Thus, a ~3–4-fold
accumulation of PCNA on the chromatin is sufficient to confer all the phenotypes seen in
strains deleted for the ELG1 gene.
Especially interesting are alleles with intermediate levels of accumulation of PCNA.
These resulted in a wildtype or defective phenotype, depending on the assay. For example,
the elg1-TT386/7AA allele, which shows a 1.87-fold increase in the level of PCNA, has
wildtype levels of mutagenesis, shows no synthetic sickness when combined with srs2∆,
but shows a very mild sensitivity to MMS, and a slight elevation in the level of silencing loss.
When combined with the SIM mutations, PCNA levels rise to ~3-fold of the wildtype cells;
this confers only intermediary levels of silencing loss, but is enough to confer mutagenesis,
MMS sensitivity, and synthetic lethality to srs2∆ to the same extent as the elg1∆ allele.
Thus, a small difference in the level of PCNA can sometimes make a large difference in the
final phenotype.
Why does accumulation of PCNA lead to defective phenotypes? We propose that
PCNA that was loaded onto chromatin but was not unloaded by the Elg1 RLC does not get
recycled; this in fact results in a shortage of PCNA, required for recruiting other proteins to
the moving fork. One such example is the chromatin remodeler CAF-1, which has a known
interaction with PCNA [43,44]. We have previously shown that overexpression of CAF-1
can correct the increased silencing loss of elg1∆ mutants [17].
We created mutations at the valine residue 390 of Elg1 to test the possibility that a
small motif shared with the Yku80 protein may have functional implications. However,
the results obtained were very different if the valine was mutated to alanine, or to aspartic
Int. J. Mol. Sci. 2023, 24, 1568 10 of 13

acid. In the first case, the strain behaved as a wildtype in all assays, whereas in the second
case it was indistinguishable from an elg1 null. As the protein levels of the two mutants
are undistinguishable from those of the wildtype, the different results imply a structural
change introduced by the aspartic acid, rather than an inactivation of an important residue
in Elg1. Since the valine 390 residue is close to the threonines 386 and 387, which are
predicted to interface with PCNA, the addition of the more bulky aspartic acid may make
the interaction with PCNA sterically difficult.
A similar result was obtained for our Walker A mutations. When mutating the non-
canonical lysine to alanine, there was no detectable phenotype in any of our experiments,
but when mutating it to aspartic acid, it mimicked elg1∆. This is in contrast to Walker B
mutations, which consistently produced results similar to those of the elg1∆ allele. This
stands in line with the current model, in which Walker A is important for ATP binding, and
the conserved residue does not hold any catalytic activity, while Walker B is important for
the hydrolysis of the ATP molecule, and thus does hold a catalytic role and could not be
easily mutated without yielding a phenotype [23,31].
To summarize our results, we see a strong correlation between the amount of PCNA
left at the chromatin in elg1 point mutants, and the degree of their phenotype. This
supports the notion that all the phenotypes measured are due to improper PCNA unloading
and recycling.

4. Materials and Methods

Standard yeast media and methods were used. Standard yeast molecular biology
techniques were used to create the mutant collection.

4.1. Chromatin Fractionation Assay

Cells from 50 mL cultures (OD600 < 1.0) were collected by centrifugation, successively
washed with ddH2O, PSB (20 mM Tris–Cl pH 7.4, 2 mM EDTA, 100 mM NaCl, 10 mMb-
ME), and SB (1 M Sorbitol, 20 mM Tris–Cl pH 7.4), and transferred to a 2 mL Eppendorf
tube. Cells were suspended in 1 mL SB, 30ul Zymolase 20T (20 mg/mL in SB) was added,
and samples were incubated at 30 ◦ C with rotation until >85% spheroplasts were observed
(60–90 min). Spheroplasts were collected by centrifugation (2 K, 5 min, 4 ◦ C), washed
twice with SB, and suspended in 500 mL EBX (20mM Tris–Cl pH 7.4, 100 mM NaCl, 0.25%
Triton X-100,15 mM-ME + protease/phosphatase inhibitors). TritonX-100 was added to a
0.5% final concentration to lyse the outer cell membrane, and the samples were kept on
ice for 10 min with gentle mixing. The lysate was layered over 1 mL NIB (20 mM Tris–Cl
pH 7.4, 100 mM NaCl, 1.2 M sucrose, 15 mM-ME + protease/phosphatase inhibitors) and
centrifuged at 12 K RPM for 15 min, at 4 ◦ C. The supernatant (cytoplasm) was discarded.
The glassy white nuclear pellet was suspended in 500 uL EBX and Triton X-100 was added
to a 1% final concentration to lyse the nuclear membrane. The chromatin and nuclear debris
were collected by centrifugation (15 K, 10 min, 4 ◦ C). Chromatin was suspended in 50 uL
Tris pH 8.0 for Western blot analysis (Chromatin). To each fraction, an equal volume of
2 × SDS-PAGE loading buffer (60 mM Tris pH 6.8, 2% SDS, 10% glycerol, 0.2%bromophenol
blue, 200 mM DTT) was added; samples were incubated at 95 ◦ C for 5 min and were then
analyzed by SDS-PAGE and Western blot analyses.

4.2. Quantification of Silencing Loss Using Flow Cytometry

For each CRASH strain, 10 single colonies were inoculated separately into 2 mL of
SD-Trp-Leu medium in 96-well deep-well plates (VWR) and were grown overnight to
saturation at 30 ◦ C on a microplate orbital shaker (VWR). Overnight cultures were diluted
in 1 mL of fresh medium at a density of 105 cells/mL in 96-well deep-well plates and were
grown at 30 ◦ C on a microplate orbital shaker until mid-log phase. For each culture, a
minimum of 50,000 events were collected using an MACSQuant® flow. Gating was used to
include only unbudded and budded cells and to measure separately the number of GFP+
cells and the number of RFP+ cells. A Boolean logic gate was used to determine the number
Int. J. Mol. Sci. 2023, 24, 1568 11 of 13

of cells that were both GFP- and RFP-fluorescent. The apparent rate of silencing loss is
calculated as the number of cells that are both GFP- and RFP-fluorescent divided by the
total number of cells.

4.3. Staining for Flow Cytometry

Mid-log cells were pelleted and resuspended in 140 µL of ice-cold ethanol 100% and
60 µL of TE, and incubated overnight at 4 ◦ C. The following day, cells were arranged in a
96-well plate, centrifuged at 4000 RPM for one minute, and the supernatant was disposed
of. Cells were resuspended in 200 µL of TE with 0.5 mg/mL of RNaseA, and the plate was
incubated overnight at 37 ◦ C. Cells were again centrifuged at 4000 RPM for one minute,
and the supernatant was disposed of. Cells were resuspended in TE with 0.5 mg/mL
of proteinase K solution and incubated for 4 h at 37 ◦ C. Cells were again centrifuged at
4000 RPM for one minute, the supernatant was discarded, and the pellet was resuspended
in TE solution with 20 µg/mL of propidium iodide. Cells were then incubated overnight
at 4 ◦ C in the dark. The following day, single cells were de-clumped using three rounds
of sonication at 20% (for 5 s), then readings were taken using the MACSQuant® flow
cytometer machine.

4.4. Mutation Measurement

Mutation levels were measured in E134 by fluctuation tests as described [45]. Standard
deviations were usually <5%. Each experiment was repeated at least three times.

Supplementary Materials: The following supporting information can be downloaded at: https:
//www.mdpi.com/article/10.3390/ijms24021568/s1.
Author Contributions: Conceptualization, Z.I., K.C., M.A. and M.K.; formal analysis, Z.I., K.C., M.A.
and M.K.; investigation, Z.I., K.C., M.A. and M.K.; writing—original draft preparation, Z.I., K.C.,
M.A. and M.K.; writing—review and editing, M.A. and M.K.; supervision, M.K.; funding acquisition,
M.K. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the Israel Science Foundation, grant number 140/18, the
Israel Cancer Research Fund, grant 458, the Minerva Stiftung, grant #238 and the DFG-Middle East,
grant 188/4-1.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Fitzsimmons, W.J.; Woods, R.J.; McCrone, J.T.; Woodman, A.; Arnold, J.J.; Yennawar, M.; Evans, R.; Cameron, C.E.; Lauring, A.S.
A speed–fidelity trade-off determines the mutation rate and virulence of an RNA virus. PLoS Biol. 2018, 16, e2006459. [CrossRef]
[PubMed]
2. Eyre-Walker, A.; Keightley, P.D. The distribution of fitness effects of new mutations. Nat. Rev. Genet. 2007, 8, 610–618. [CrossRef]
[PubMed]
3. Arbel, M.; Liefshitz, B.; Kupiec, M. DNA damage bypass pathways and their effect on mutagenesis in yeast. FEMS Microbiol. Rev.
2020, 45, fuaa038. [CrossRef]
4. Bellí, G.; Colomina, N.; Castells-Roca, L.; Lorite, N.P. Post-Translational Modifications of PCNA: Guiding for the Best DNA
Damage Tolerance Choice. J. Fungi 2022, 8, 621. [CrossRef] [PubMed]
5. McClure, A.W.; Canal, B.; Diffley, J.F.X. A DNA replication fork-centric view of the budding yeast DNA damage response. DNA
Repair 2022, 119, 103393. [CrossRef]
6. Bar-Ziv, R.; Voichek, Y.; Barkai, N. Chromatin dynamics during DNA replication. Genome Res. 2016, 26, 1245–1256. [CrossRef]
7. Yuan, G.C.; Liu, Y.J.; Dion, M.F.; Slack, M.D.; Wu, L.F.; Altschuler, S.J.; Rando, O.J. Molecular biology: Genome-scale identification
of nucleosome positions in S. cerevisiae. Science 2005, 309, 1112178. [CrossRef] [PubMed]
8. Anderson, J.D.; Widom, J. Sequence and position-dependence of the equilibrium accessibility of nucleosomal DNA target sites. J.
Mol. Biol. 2000, 296, 979–987. [CrossRef]
Int. J. Mol. Sci. 2023, 24, 1568 12 of 13

9. Lu, X.; Simon, M.D.; Chodaparambil, J.V.; Hansen, J.C.; Shokat, K.M.; Luger, K. The effect of H3K79 dimethylation and H4K20
trimethylation on nucleosome and chromatin structure. Nat. Struct. Mol. Biol. 2008, 15, 1122–1124. [CrossRef]
10. Kalashnikova, A.A.; Porter-Goff, M.E.; Muthurajan, U.M.; Luger, K.; Hansen, J.C. The role of the nucleosome acidic patch in
modulating higher order chromatin structure. J. R. Soc. Interface 2013, 10, 20121022. [CrossRef]
11. Reverón-Gómez, N.; González-Aguilera, C.; Stewart-Morgan, K.R.; Petryk, N.; Flury, V.; Graziano, S.; Johansen, J.V.; Jakobsen, J.S.;
Alabert, C.; Groth, A. Accurate Recycling of Parental Histones Reproduces the Histone Modification Landscape during DNA
Replication. Mol. Cell 2018, 72, 239–249.e5. [CrossRef] [PubMed]
12. Strzalka, W.; Ziemienowicz, A. Proliferating cell nuclear antigen (PCNA): A key factor in DNA replication and cell cycle regulation.
Ann. Bot. 2010, 107, 1127–1140. [CrossRef] [PubMed]
13. Arbel, M.; Choudhary, K.; Tfilin, O.; Kupiec, M. PCNA loaders and unloaders—One ring that rules them all. Genes 2021, 12, 1812.
[CrossRef] [PubMed]
14. Paunesku, T.; Mittal, S.; Protić, M.; Oryhon, J.; Korolev, S.V.; Joachimiak, A.; Woloschak, G.E. Proliferating cell nuclear antigen
(PCNA): Ringmaster of the genome. Int. J. Radiat. Biol. 2001, 77, 1007–1021. [CrossRef]
15. Madru, C.; Henneke, G.; Raia, P.; Hugonneau-Beaufet, I.; Pehau-Arnaudet, G.; England, P.; Lindahl, E.; Delarue, M.; Carroni, M.;
Sauguet, L. Structural basis for the increased processivity of D-family DNA polymerases in complex with PCNA. Nat. Commun.
2020, 11, 1–12. [CrossRef]
16. Zhang, W.; Qin, Z.; Zhang, X.; Xiao, W. Roles of sequential ubiquitination of PCNA in DNA-damage tolerance. FEBS Lett. 2011,
585, 2786–2794. [CrossRef]
17. Janke, R.; King, G.A.; Kupiec, M.; Rine, J. Pivotal roles of PCNA loading and unloading in heterochromatin function. Proc. Natl.
Acad. Sci. USA 2018, 115, E2030–E2039. [CrossRef]
18. Lee, K.-y.; Park, S.H. Eukaryotic clamp loaders and unloaders in the maintenance of genome stability. Exp. Mol. Med. 2020, 52,
1948–1958. [CrossRef]
19. Dovrat, D.; Stodola, J.L.; Burgers, P.M.J.; Aharoni, A. Sequential switching of binding partners on PCNA during in vitro Okazaki
fragment maturation. Proc. Natl. Acad. Sci. USA 2014, 111, 14118–14123. [CrossRef]
20. Thakar, T.; Leung, W.; Nicolae, C.M.; Clements, K.E.; Shen, B.; Bielinsky, A.K.; Moldovan, G.L. Ubiquitinated-PCNA protects
replication forks from DNA2-mediated degradation by regulating Okazaki fragment maturation and chromatin assembly. Nat.
Commun. 2020, 11, 2147. [CrossRef]
21. Makarova, A.V.; Stodola, J.L.; Burgers, P.M. A four-subunit DNA polymerase ζ complex containing Pol δ accessory subunits is
essential for PCNA-mediated mutagenesis. Nucleic Acids Res. 2012, 40, 11618–11626. [CrossRef] [PubMed]
22. Motegi, A.; Liaw, H.J.; Lee, K.Y.; Roest, H.P.; Maas, A.; Wu, X.; Moinova, H.; Markowitz, S.D.; Ding, H.; Hoeijmakers, J.H.J.; et al.
Polyubiquitination of proliferating cell nuclear antigen by HLTF and SHPRH prevents genomic instability from stalled replication
forks. Proc. Natl. Acad. Sci. USA 2008, 105, 12411–12416. [CrossRef] [PubMed]
23. Kang, M.S.; Ryu, E.; Lee, S.W.; Park, J.; Ha, N.Y.; Ra, J.S.; Kim, Y.J.; Kim, J.; Abdel-Rahman, M.; Park, S.H.; et al. Regulation of
PCNA cycling on replicating DNA by RFC and RFC-like complexes. Nat. Commun. 2019, 10, 2420. [CrossRef]
24. Ben-Aroya, S.; Koren, A.; Liefshitz, B.; Steinlauf, R.; Kupiec, M. ELG1, a yeast gene required for genome stability, forms a complex
related to replication factor C. Proc. Natl. Acad. Sci. USA 2003, 100, 9906–9911. [CrossRef] [PubMed]
25. Sikdar, N.; Banerjee, S.; Lee, K.Y.; Wincovitch, S.; Pak, E.; Nakanishi, K.; Jasin, M.; Dutra, A.; Myung, K. DNA damage responses
by human ELGI in S phase are important to maintain genomic integrity. Cell Cycle 2009, 8, 3199–3207. [CrossRef]
26. Smolikov, S.; Mazor, Y.; Krauskopf, A. ELG1, a regulator of genome stability, has a role in telomere length regulation and in
silencing. Proc. Natl. Acad. Sci. USA 2004, 101, 1656–1661. [CrossRef]
27. Choudhary, K.; Itzkovich, Z.; Alonso-Perez, E.; Bishara, H.; Dunn, B.; Sherlock, G.; Kupiec, M. S. cerevisiae Cells Can Grow without
the Pds5 Cohesin Subunit. MBio 2022, 100, 9906–9911. [CrossRef]
28. Choudhary, K.; Kupiec, M. The cohesin complex of yeasts: Sister chromatid cohesion and beyond. FEMS Microbiol. Rev. 2022, 11,
1–22. [CrossRef]
29. Parnas, O.; Zipin-Roitman, A.; Pfander, B.; Liefshitz, B.; Mazor, Y.; Ben-Aroya, S.; Jentsch, S.; Kupiec, M. Elg1, an alternative
subunit of the RFC clamp loader, preferentially interacts with SUMOylated PCNA. EMBO J. 2010, 29, 2611–2622. [CrossRef]
30. Venclovas, Č.; Colvin, M.E.; Thelen, M.P. Molecular modeling-based analysis of interactions in the RFC-dependent clamp-loading
process. Protein Sci. 2008, 11, 0214302. [CrossRef]
31. Kang, S.; Warner, M.D.; Bell, S.P. Multiple Functions for Mcm2-7 ATPase Motifs during Replication Initiation. Mol. Cell 2014, 55,
655–665. [CrossRef] [PubMed]
32. Davidson, M.B.; Brown, G.W. The N- and C-termini of Elg1 contribute to the maintenance of genome stability. DNA Repair 2008,
7, 1221–1232. [CrossRef] [PubMed]
33. Lundin, C.; North, M.; Erixon, K.; Walters, K.; Jenssen, D.; Goldman, A.S.H.; Helleday, T. Methyl methanesulfonate (MMS)
produces heat-labile DNA damage but no detectable in vivo DNA double-strand breaks. Nucleic Acids Res. 2005, 33, 3799–3811.
[CrossRef] [PubMed]
34. Shemesh, K.; Sebesta, M.; Pacesa, M.; Sau, S.; Bronstein, A.; Parnas, O.; Liefshitz, B.; Venclovas, Č.; Krejci, L.; Kupiec, M. A
structure-function analysis of the yeast Elg1 protein reveals the importance of PCNA unloading in genome stability maintenance.
Nucleic Acids Res. 2017, 45, 3189–3203. [CrossRef]
Int. J. Mol. Sci. 2023, 24, 1568 13 of 13

35. Goellner, E.M.; Smith, C.E.; Campbell, C.S.; Hombauer, H.; Desai, A.; Putnam, C.D.; Kolodner, R.D. PCNA and Msh2-Msh6
Activate an Mlh1-Pms1 Endonuclease Pathway Required for Exo1-Independent Mismatch Repair. Mol. Cell 2014, 55, 291–304.
[CrossRef]
36. Tran, H.T.; Keen, J.D.; Kricker, M.; Resnick, M.A.; Gordenin, D.A. Hypermutability of homonucleotide runs in mismatch repair
and DNA polymerase proofreading yeast mutants. Mol. Cell. Biol. 1997, 17, 2859–2865. [CrossRef]
37. Paul Solomon Devakumar, L.J.; Gaubitz, C.; Lundblad, V.; Kelch, B.A.; Kubota, T. Effective mismatch repair depends on timely
control of PCNA retention on DNA by the Elg1 complex. Nucleic Acids Res. 2019, 47, 6826–6841. [CrossRef]
38. Aylon, Y.; Kupiec, M. The Checkpoint Protein Rad24 of Saccharomyces cerevisiae Is Involved in Processing Double-Strand Break
Ends and in Recombination Partner Choice. Mol. Cell. Biol. 2003, 23, 6585–6596. [CrossRef]
39. Arbel, M.; Bronstein, A.; Sau, S.; Liefshitz, B.; Kupiec, M. Access to pcna by srs2 and elg1 controls the choice between alternative
repair pathways in saccharomyces cerevisiae. MBio 2020, 11, e00705-20. [CrossRef]
40. Gazy, I.; Liefshitz, B.; Bronstein, A.; Parnas, O.; Atias, N.; Sharan, R.; Kupiec, M. A genetic screen for high copy number
suppressors of the synthetic lethality between elg1∆ and srs2∆ in yeast. G3 Genes, Genomes, Genet. 2013, 3, 917–926. [CrossRef]
41. Kubota, T.; Katou, Y.; Nakato, R.; Shirahige, K.; Donaldson, A.D. Replication-Coupled PCNA Unloading by the Elg1 Complex
Occurs Genome-wide and Requires Okazaki Fragment Ligation. Cell Rep. 2015, 12, 774–787. [CrossRef] [PubMed]
42. Johnson, C.; Gali, V.K.; Takahashi, T.S.; Kubota, T. PCNA Retention on DNA into G2/M Phase Causes Genome Instability in Cells
Lacking Elg1. Cell Rep. 2016, 16, 684–695. [CrossRef] [PubMed]
43. Ben-Shahar, T.R.; Castillo, A.G.; Osborne, M.J.; Borden, K.L.B.; Kornblatt, J.; Verreault, A. Two Fundamentally Distinct PCNA
Interaction Peptides Contribute to Chromatin Assembly Factor 1 Function. Mol. Cell. Biol. 2009, 29, 6353–6365. [CrossRef]
[PubMed]
44. Kondratick, C.M.; Litman, J.M.; Shaffer, K.V.; Washington, M.T.; Dieckman, L.M. Crystal structures of PCNA mutant proteins
defective in gene silencing suggest a novel interaction site on the front face of the PCNA ring. PLoS ONE 2018, 13, e0193333.
[CrossRef] [PubMed]
45. Liefshitz, B.; Steinlauf, R.; Friedl, A.; Eckardt-Schupp, F.; Kupiec, M. Genetic interactions between mutants of the “error-prone”
repair group of Saccharomyces cerevisiae and their effect on recombination and mutagenesis. Mutat. Res.-DNA Repair 1998, 407,
135–145. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.