MSH6- or PMS2-deficiency causes re-replication in DT40 B cells, but it has little effect on immunoglobulin gene conversion or on repair of AID-generated uracils.

1.7 MB·English

by  Campo, Vanina A.

3032–3046 Nucleic Acids Research, 2013, Vol. 41, No. 5 Published online 11 January 2013
doi:10.1093/nar/gks1470
MSH6- or PMS2-deficiency causes re-replication
in DT40 B cells, but it has little effect on
immunoglobulin gene conversion or on
repair of AID-generated uracils
Vanina A. Campo1,2,3, Anne-Marie Patenaude1, Svenja Kaden4, Lori Horb1, Daniel Firka5,
Josef Jiricny4 and Javier M. Di Noia1,2,3,6,*
1Institut de Recherches Cliniques de Montre´al, Division of Immunity and Viral Infections, Montre´al, H2W 1R7
Que´bec, Canada, 2Department of Microbiology and Immunology, Universite´ de Montre´al, Montre´al, H3T 1J4
Que´bec, Canada, 3Department of Medicine, Universite´ de Montre´al, Montre´al, H3T 1J4 Que´bec, Canada,
4Institute of Molecular Cancer Research of the University of Zurich and the ETH Zurich, 8057 Zurich,
Switzerland, 5Druida Software and Consulting, M5R 1X5 Toronto, Ontario, Canada and 6Division of
Experimental Medicine, Department of Medicine, McGill University, Montre´al, H3A 1A3 Que´bec, Canada
Received September 25, 2012; Revised December 17, 2012; Accepted December 18, 2012
ABSTRACT initiating SHM and CSR in mammalian cells.
Moreover, as MMR does not counteract or govern
The mammalian antibody repertoire is shaped by
Ig GC, we report a rare example of ‘homeologous’
somatic hypermutation (SHM) and class switch
recombination insensitive to MMR.
recombination (CSR) of the immunoglobulin (Ig)
loci of B lymphocytes. SHM and CSR are triggered
by non-canonical, error-prone processing of G/U INTRODUCTION
mismatches generated by activation-induced
The antibody repertoire of higher organisms is firstly
deaminase (AID). In birds, AID does not trigger
generated by VDJ recombination, followed by additional
SHM, but it triggers Ig gene conversion (GC), a
genetic modification through somatic hypermutation
‘homeologous’ recombination process involving
(SHM), immunoglobulin gene conversion (Ig GC) and
the Ig variable region and proximal pseudogenes. class switch recombination (CSR). During SHM, nucleo-
Because recombination fidelity is controlled by the tide changes are introduced into the exons encoding the
mismatch repair (MMR) system, we investigated variable (IgV), N-terminal domain of the antibody heavy
whether MMR affects GC in the chicken B cell line and light chains by a mechanism involving error-prone
DT40. We show here that Msh6(cid:2)/(cid:2) and Pms2(cid:2)/(cid:2) DNA polymerases [reviewed in (1,2)]. Although Ig GC
DT40 cells display cell cycle defects, including serves the same purpose, mutations are not introduced
directly but are copied from numerous pseudogene
genomic re-replication. However, although IgVk GC
sequences located upstream on the same chromosome
tracts in MMR-deficient cells were slightly longer
(3,4). In contrast, CSR involves the fusion of the IgV
than in normal cells, Ig GC frequency, donor
to a different constant (effector) region by double-strand
choice or the number of mutations per sequence
break-induced region-specific recombination [reviewed
remained unaltered. The finding that the avian
in (5,6)].
MMR system, unlike that of mammals, does not
SHM, Ig GC and CSR are all initiated by activation-
seem to contribute towards the processing of G/U induced deaminase (AID) (7–9), an enzyme expressed
mismatches in vitro could explain why MMR is in antigen-stimulated B cells, which typically converts
unable to initiate Ig GC in this species, despite multiple cytosines in the Ig loci into uracils (2).
*To whom correspondence should be addressed. Tel:[phone]; Fax:[phone]; Email: [email protected]
Present addresses:
Vanina A. Campo, IIB-INTECH, San Martin, Buenos Aires, Argentina.
Lori Horb, Marine Biological Laboratory, Bell Center for Regenerative Biology and Tissue Engineering, Woods Hole, MA, USA.
(cid:2)TheAuthor(s)2013.PublishedbyOxfordUniversityPress.
ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionNon-CommercialLicense(http://creativecommons.org/licenses/
by-nc/3.0/),whichpermitsunrestrictednon-commercialuse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
NucleicAcidsResearch,2013,Vol.41,No.5 3033
Although uracil is generally highly efficiently repaired by Tmutations(38,39).Thisimpliesthat,unlikethemamma-
base excision repair (BER), this process seems to be inef- lian enzyme (11,24), chicken MutSa does not recognize
ficient in antigen-stimulated B cells. Thus, some uracils AID-generated G/U mismatches, that MMR-dependent
persist until the next round of replication to give rise to processing of G/U mispairs does not take place in this
C:GtoT:Atransitionmutations(1,2).Othersareremoved systemintheabsenceofUNGorthatMMR-mediatedpro-
byuracil–DNAglycosylase(UNG)(10–12),buttheresult- cessingofG/UmispairsinDT40cellsismostlyerror-free,
ing abasic sites persist and are bypassed by translesion asseeninaproportionofAID-generateduracilsinmouseB
polymerases to yield all types of mutations at C:G base cells(40).However,evenifMMRdoesnottriggerIgGC,it
pairs (2,13). A third group of uracils is addressed by a would be still predicted to affect its outcome; it should
poorly defined pathway, which involves MutSa (11,14), restrict itby preventingrecombination betweensequences
a heterodimer of mutS homologue 2 (MSH2) and MSH6 that are too diverged (41–45), while, on the other hand,
that normally initiates DNA mismatch repair (MMR) helping to repair mismatches arising through annealing
(15,16). It was proposed that MutSa detects G/U ofnon-identical(homeologous)donorandrecipientDNA
mismatches generated by AID and triggers an error- sequences (31,33,46,47). Although MSH6-, MSH4- or
prone, long-patch repair process that introduces muta- MSH3-deficientcellswerereportedtoundergoIgGC[pre-
tions at sites distal to those deaminated by AID (1,2). A viouslymentioned(7)asdatanotshown],theroleofMMR
related mechanism that involves MutSa and other factors inthis processhasnot beenstudied indetail or compared
was postulated to act at Ig switch regions to give withHR.
rise to double-strand breaks that trigger CSR in the In this study, we describe the effect of MutSa and
absence of UNG (11,17). The molecular mechanism MutLa deficiency on the frequency, fidelity and overall
of MMR-mediated diversification of Ig genes remains quality of Ig GC in DT40 cells. Our data reveal that
to be elucidated, but genetic experiments implicated MMR is largely dispensable in this process, possibly
exonuclease I (18), DNA polymerase Z (19,20) and because of its limited contribution towards processing of
monoubiquitylated proliferating cell nuclear antigen AID-generated uracils at the Ig locus. Moreover, we find
(21,22) in this process. Interestingly, MutLa, a thatIgGCrepresentsararecaseof‘homeologous’recom-
heterodimer of mutL homologue 1/postmeiotic segre- bination that is insensitive to the anti-recombination
gation increased S cerevisiae 2 that acts immediately effects of MMR.
downstream of MutSa during MMR (15), plays no role
in SHM [reviewed in (1)], although it can influence
the chromosome rejoining pathway during CSR (23). MATERIALS AND METHODS
The roles of UNG and MutSa in mammalian antibody Cell lines and transfections
diversification seem to be partially redundant, given that
DT40IgM(cid:2)CL18(28)oritsIgM+derivativeCL18c4(34)
only their combined deficiency abrogates both CSR and
SHM. Thus, in Ung(cid:2)/(cid:2) Msh2(cid:2)/(cid:2) or Ung(cid:2)/(cid:2) Msh6(cid:2)/(cid:2) was grown in 5% CO2 at 41(cid:3)C in RPMI 1640 with glu-
tamine (Wisent) supplemented with 10% FCS (Wisent),
mice, Ig lesions are limited to C:G to T:A transitions
1% chicken serum (Invitrogen), 50mM b-MSH and anti-
(11,24).
biotics. DT40 Aid(cid:2)/(cid:2) was a kind gift of Dr H. Arakawa
That antibody diversification can also involve Ig GC
and Dr J.M. Buerstedde (7). Transfections for gene tar-
was first shown in chickens (3,4), and probably contrib-
geting and MSH6 complementation were as previously
utes to antibody diversification in most bird species (25)
described(10,34). Retroviral particlesfor DT40transduc-
and rabbits (26), and perhaps in other species (25). The
tion wereproduced inHEK293T cellsco-transfected with
IgV exons, VJ in the light chain and VDJ in the heavy
aplasmidforVSV-G,aplasmidforMLVgag/polandthe
chain, are located downstream from an array of V, VJ
retroviral vector pMIG or pMIG-Ugi. HEK293T super-
and VDJ pseudogenes (referred to as cV) that serve as
natant was harvested 48h post-transfection, filtered and
donors in the gene conversion reactions. Ig GC replaces
added to 106 DT40 cells in the presence of 8mg/ml of
acontiguousstretchof8to>200nucleotidesandcanthus
polybrene in 24-well plates.
introduce multiple base changes into the recipient V(D)J
sequence (27), which may translate into amino acid re-
Gene targeting
placements affecting the specificity and/or affinity of the
antibody. The structure of the chicken Msh6 and Pms2 genes was
ThechickenDT40Bcelllymphomalineundergoescon- determined by assembling the sequence of multiple poly-
stitutive AID-dependent Ig GC (28,29), and it is widely merase chain reaction (PCR) products amplified from
usedtostudyantibodydiversification(25,30,31)aswellas DT40 genomic DNA using oligonucleotides designed to
DNA repair (31–33). Ig GC in DT40 cells is commonly span one or more introns based on the gene structure of
used as a model for homologous recombination (HR) the murine genes. The oligonucleotide sequences were
repair because the initiating event is well defined and based on the corresponding chicken cDNAs assembled
requires HR factors, including the RAD51 paralogues from multiple expression sequence tags obtained from
XRCC2, XRCC3 and RAD51B (34), BRCA1 (35), GeneBankTM. The structures of the genes, targeting and
BRCA2 (36) and RAD54 (37). screeningstrategieswereconfirmedbySouthernblot.The
Unexpectedly, MMR does not seem to initiate Ig GC targeting constructs were assembled in pBluescript II
in DT40 cells, given that UNG inhibition or knockout (SupplementaryFigureS1).Antibioticresistancecassettes
largely eliminated Ig GC, accumulating instead C to were described earlier (48). The Msh6 targeting construct
3034 NucleicAcidsResearch,2013,Vol.41,No.5
eliminates the last 20 amino acids encoded by exon 4 and was supplemented with 1.25nM Trichostatin A (TSA)
most of the 30-adjacent intron, including the splice donor (Sigma). Cells expressing Ugi or the control retrovirus
site, and creates a frameshift. The Pms2 targeting con- were selected by GFP expression and stained using
struct eliminates exons 7 and 8 of the gene. Genomic anti-chicken IgM-PE mAb (Southern biotech). Reduced
DNA was purified using Puregene DNA isolation kit UNG activity was confirmed by enzymatic assays as
(Gentra) and digested with BamHI–EcoRI for screening described previously (10).
Msh6 targeting or BamHI–XbaI for Pms2. Probes were
labelled by PCR using 2mM dATP/dAGT/dTTP and Sequence analysis
50mCi of [a-32P]dCTP and oligonucleotides OV1 (TGG IgM+or IgM(cid:2) cells (1–5(cid:4)105) from the expanded DT40
GAGCAGGTAGTTTTGTG) and OV2 (CTCTGGGAT
populations were sorted by FACS, resuspended in
GTCAGCAAGTC) for Msh6 and OJ43 (TGACAGGTT
DirectPCR (Viagen) and extracted with proteinase K.
CTGCGTTCATA) and OJ44 (TAGGGCAGCATTCCA
The variable region of the rearranged DT40 (cid:2) light
TTC) for Pms2. Probes were purified using Illustra
chain locus was PCR amplified using KOD DNA poly-
MicroSpin G-50 columns (GE Healthcare), and filters
merase (Novagen) [95(cid:3)C 2min+8(cid:4)(95(cid:3)C 20s, touch
were hybridized in ExpressHyb (Clontech) and developed
down 68 >60(cid:3)C 10s, 70(cid:3)C 1min)+23(cid:4)(95(cid:3)C 15s, 60(cid:3)C
by PhosphorImager (GE Heathcare).
10s, 70(cid:3)C 1min)+72(cid:3)C 5 min]. PCR products were
A-tailed, cloned into pGEM-T-Easy (Promega) and
Growth and cell cycle analysis
sequenced by Macrogen Inc (Seoul, Korea) using the
Growth curves were generated by diluting exponentially SP6 oligonucleotide. Sequences were aligned using
growing cultures to 5–10(cid:4)105 cells/ml that were kept in Sequencher (Genecodes). Mutations were scored and
the log phase by appropriate dilutions. Trypan assigned to all possible donor pseudogenes using
blue-excluded cells were periodically counted in a haemo- Mutdet, a tailor-made software (D.F. and J.M.D.N., un-
cytometer and corrected for dilution factor. Population published data) that assigned all possible gene conversion
doubling times were calculated from exponential curve events following described criteria (10,34,38). Mutdet
fits of cell numbers versus time plots using Prism 5 results were then manually verified. The pseudogene that
(GraphPad software Inc). Cloning efficiency was couldexplainthemostconsecutivemutationsandhadthe
estimated by fluorescence-activated cell sorting (FACS)- longest homology encompassing all mutations was scored
depositing single cells into 150ml of normal culture as the donor. The maximum gene conversion tract was
medium in 96-well plates and counting the number of defined as the longest sequence fragment sharing perfect
viable clones per plate after 2 weeks. DNA content homology between each IgV(cid:2) sequence analysed and the
profiles were determined by flow cytometry of 2(cid:4)106 pseudogene explaining the most consecutive mutations.
cells fixed in cold 70% ethanol, and stained in 0.5%
Triton X-100, 100mg/ml of RNase A and 60mg/ml MMR and gel-shift assays
propidium iodide. S-phase profiles were determined by
DT40 cell extracts were prepared by lysing the cells in
flow cytometry in cells pulsed for 30min at 37(cid:3)C in
isotonic lysis buffer [10 mM Tris–HCl, pH 7.5, 2mM
medium containing 10mM BrdU (Sigma). Cells were
MgCl ,3mMCaCl 0.32Msucrose,1mMdithiothreitol,
fixed as described earlier in the text, and DNA was 2 2,
0.1mM spermine, 0.5mM spermidine, 0.3% IGEPAL
denatured in 2N HCl 0.5% Triton X100, neutralized
CA-630 and complete inhibitor cocktail (Roche)]. For
with 1M Na B O pH 8.5 and labelled with anti-BrdU-
2 4 7 gel-shift assays, 10mg of the respective DT40 cell
fluorescein isothiocyanate (FITC) antibody (BD extracts was incubated with 40fmol of [32P]-labelled G/
Biosciences) and propidium iodide. Apoptosis was
T,G/Cor+1duplexes(generated,respectively,byanneal-
determined by Annexin V-FITC staining (Sigma). Drug
ingtheoligonucleotidesCCAGACGTCTGTTGACGT
sensitivity was measured by culturing 50000 cells for
TGG GAA GCT TGA G, CCA GAC GTC TGT CGA
24h in 96-well plates in increasing concentrations of
CGTTGGGAAGCTTGAGorCCAGACGTCTGT
6-thioguanine (Sigma). Viability was estimated by MTS
CTG ACG TTG GGA AGC TTG A to CTC AAG CTT
reduction using Celltiter 96(cid:3) Aqueous non-radioactive
CCCAACGTCGACAGACGTCTGGorCCAGAC
cell proliferation assay (Promega).
GTCTGTCAAUGTTGG GAAGCTTGAGtoCTC
AAGCTTCCCAACGTTGACAGACGTCTGG,all
Antibody diversification assays
written 50–30) in 25mM HEPES–KOH, pH 8.0, 0.5mM
For IgM phenotype fluctuation analysis, homogeneous ethylenediaminetetraacetic acid, 10% (v/v) glycerol,
IgM+ or IgM(cid:2) populations were obtained by FACS 0.5mM dithiothreitol and 1mg of Poly[d(I:C)(cid:2)d(I:C)]
after blocking for 10min in phosphate-buffered saline, for 20min at room temperature in a final volume
1% bovine serum albumin, 3% normal goat serum and of 20ml. Samples were separated on 5 or 6% TAE–
staining for 30min with anti-chicken IgM-FITC (Bethyl, polyacrylamide gels. The dried gels were exposed to a
1:200). Dead cells were excluded by propidium iodide PhosphorImager screens.
staining (10mg/ml). Multiple 500000 cell populations A detailed procedure for the MMR assays has been
were cultured in 24-well plates for a given number of describedpreviously(49).Briefly,heteroduplexDNAsub-
population doubling times, splitting 1:2 every 1 or strates containing a G/T mismatch or a one nucleotide
2 days, and their surface IgM phenotype determined by insertion (+1) within a SalI restriction site, or a G/U
flow cytometry. Where indicated, the culture medium mispair within an AclI restriction site in the 46-bp
NucleicAcidsResearch,2013,Vol.41,No.5 3035
polylinker of a pGEM13Zf(+) derivative, were con- toverifythattheywereindeedaconsequenceoftheMMR
structed by primer extension, using the mismatch- deficiency. We, therefore, complemented the DT40
containing 31mer oligonucleotides 50-CCA GAC GTC Msh6(cid:2)/(cid:2) line with human MSH6. This largely rescued
TGT TGA CGT TGG GAA GCT TGA G-30 for the growth and cloning efficiency (Figure 1A–C), which
G/T mismatch substrate, 50-CCA GAC GTC TGT CTG suggests that the phenotype is indeed caused by the
ACG TTG GGA AGC TTG A-30 for the+1 substrate or MMR gene defect. The difference between wild type
50CCA GAC GTC TGT CAA UGT TGG GAA GCT (WT) and hMSH6-complemented Msh6(cid:2)/(cid:2) cells could
TGA G-30 for the G/U substrate (the mispaired or be caused by sequence divergence between chicken and
inserted residue is highlighted in bold) and the single- human MSH6 (Supplementary Figure S2) and/or insuffi-
strandedphagemidDNAastemplate.Thestranddiscrim- cient expression levels of the complemented cells. The
ination signal was introduced by incubation with heterozyogous Msh6+/(cid:2) and Pms2+/(cid:2) cells displayed
N.BstNBI, which introduced a specific nick into the com- only small increases in population doubling times, but
plementary strand (30 from the mismatch) at position 350 both showed reduced cloning efficiency, albeit to a much
of the duplex. Isolation of the desired supercoiled lesser extent than the homozygous knockouts (Figure 1B
heteroduplex substrates and the MMR assays were and C). It seemed conceivable that the observed growth
carried out as described, using 100ng (47.5fmol) defects were related to DNA damage caused by AID.
heteroduplex DNA substrate and 80mg of DT40 cell However, as doubly deficient Msh6(cid:2)/(cid:2) Aid(cid:2)/(cid:2) DT40
extracts in a volume of 25ml. After 45min of incubation cells had similar growth defects to Msh6(cid:2)/(cid:2) cells
at37(cid:3)C,the reactions were terminated byadding 10mg of (Figure 1D), this was not the case. Ablation of MSH6
proteinase K and heat inactivation at 55(cid:3)C for 2h. After or PMS2 in a second, IgM+, DT40 line CL18c4 led to
purification of the DNA on a Qiagen MinElute Reaction similarphenotypesandshowedanIgMdullsubpopulation,
CleanupKit,samplesweredigestedwithSalIandDraIor which was constant over time and likely, another mani-
AclI,respectively,treatedwith10mgofRNaseAovernight festationofthegrowthdefects(Figure1E).We,therefore,
and then run on a 1% TAE–agarose gel containing conclude that normal DT40 growth is dependent on the
0.5(cid:4)Gel-Red (Biotium). To inhibit BER in assays using MMR pathway.
theG/Usubstrate,ThymineDNAglycosylase(TDG)was
immunodepleted as previously described (49), and the Cell cycle alterations in MSH6- and
extracts were pretreated for 10min with 4U of Ugi PMS2-deficient DT40
(NEB) at 37(cid:3)C to inhibit UNG.
Unlike parental DT40, logarithmically growing Msh6(cid:2)/(cid:2)
and Pms2(cid:2)/(cid:2) cells showed many Trypan blue-positive
Statistics
cells indicative of necrotic cell death (data not shown).
Statistical analysis was performed using Prism 5 Although we observed a trend towards higher spontan-
(GraphPad software Inc). eous apoptosis in the Msh6(cid:2)/(cid:2) and Pms2(cid:2)/(cid:2) cells
(Figure 2A), this was too small to account for the
growth defects. Microscopic examination showed mor-
RESULTS phologically abnormalities in the Msh6(cid:2)/(cid:2) and Pms2(cid:2)/(cid:2)
Growth defects in MSH6- and PMS2-deficient DT40 cells cultures, with numerous cells larger than normal. These
cells had larger, often bi- or multilobular nuclei
WetargetedtheMMRpathwayinDT40cellsbydisrupt-
(Figure 2B), which suggested cell cycle checkpoint
ingeithertheMsh6orPms2genes(SupplementaryFigure
defects. Indeed, analysis of DNA content confirmed cell
S1). Loss of MSH6 selectively depletes MutSa, which
cycle alterations, with reduced proportion of cells in
plays an important role in both SHM and CSR, but not
S-phase and a significant number of cells with
MutSb, which is not involved in SHM or CSR (50–52),
>4n DNA content (Figure 2C). Complementation of
but which might affect other recombinogenic events.
Msh6(cid:2)/(cid:2) DT40 cells with hMSH6 corrected these defects
Similarly, loss of PMS2 depletes MutLa, which plays a
to a large extent (Figure 2C). 2D cell cycle analysis of
role in CSR [reviewed in (1,2,5)], but it should not affect
Msh6(cid:2)/(cid:2) cells using BrdU incorporation to distinguish
the other MLH1-containing heterodimer MutLg (MLH1/
cellsinS-phaserevealedadefectinDNAdamage-induced
MLH3), which participates in recombination (53). The
G /S checkpoint (Figure 2D), as previously described
chicken MSH6 and PMS2 proteins have not been 2
for MMR-deficient cells (54). More importantly, a
characterized but share high similarity in primary
second cycle of replication and a distinct cell population
sequence as well as in their structural and functional
with 8n DNA content were visible in Msh6(cid:2)/(cid:2) cells,
features with their mouse and human orthologues
indicative of endore-duplication (Figure 2D). Thus,
(Supplementary Figure S2).
To our surprise, Msh6(cid:2)/(cid:2) and Pms2(cid:2)/(cid:2) DT40 cells dis- MMR-deficiency in DT40 causes cell cycle defects and
re-replication that are most likely responsible for the
playedmajorgrowthdefects.Bothcelllineshadconsider-
impaired growth.
ablylongerlagphases(notshown)anda(cid:5)2-foldincrease
in the population doubling times during exponential
MMR deficiency does not increase Ig gene conversion
growth (Figure 1A and B). In addition, their cloning
frequency
efficiencies were extremely low (Figure 1C). Because
such drastic growth phenotypes have not been described We analysed the frequency of Ig GC to test whether
in MMR-deficient cells from other organisms, we wanted MMRcounteractedthisprocess,eitherthrougherror-free
3036 NucleicAcidsResearch,2013,Vol.41,No.5
A B
20 Mean DT(h) 9.7 11.4 24.5 14.0 10.0 19.1
∗
CL18 h) 40 ∗
6)15 Msh6-/- me (
ells (x1010 Msh6-/- hMsh6 ubling ti 2300
C 5 o
d
n
0 atio 10
0 20 40 60 80 100 ul
Time (h) op 0
P
CL18 +/- -/- -/- +/- -/-
C
hMsh6
60
Msh6 Pms2
%) ∗
y ( D
nc 40 25
e DT40 cre1 (11h)
ng effici 20 ∗ ∗ 610) 1250 AAiidd--//-- (M1s2hh6)+/- (16h)
Cloni ∗ ∗ ells (x 10 Aid-/- Msh6-/- (19h)
0 C
5
CL18 +/- -/- -/- +/- -/-
+hMsh6 Pms2 0
Msh6 0 20 40 60 80
Time (h)
E DT40 Cl18c4 DT40 Cl18c4
Msh6+/+ Msh6+/- Msh6-/- Pms2+/+ Pms2+/- Pms2-/-
0.15 99 0.63 96.8 1.6 89.9 0.58 96.7 0.12 95.4 0.87 66.6
PI
IgM
Figure 1. ImpairedproliferationinMMR-deficientDT40cells.(A)Growthcurvesoflog-phaseculturesofDT40CL18cells,itsMsh6(cid:2)/(cid:2)derivative
1015andthelattercloneexpressinghumanMsh6(hMsh6).Averagecellnumbers,correctedbydilutionfactor,ofduplicatecultures±standarderror
ofthemeanareplottedovertime.(B)Populationdoublingtimeswerecalculatedbyexponentialcurvefit(R2(cid:6)0.95ineverycase)ofgrowthcurves
as in (A). Multiple measurements for both DT40 parental lines used in this work (CL18 and CL18c4) are plotted, as well as one or more meas-
urementsofatleasttwoindependentcelllinesforeachoftheindicatedgenotypes.Meanpopulationdoublingtimes (DT)areindicatedatthetop.
(C)Cloningefficiencydeterminedastheproportionofclonesarisingper96-wellplateaftersinglecelldeposition.Mean+standarddeviation(SD)of
—six to nine plates from two to three experiments are plotted. (D) Growth curves as in (A) with DT indicated in brackets next to each cell line.
(E)FlowcytometryprofilesoftheIgM+DT40cell lineCL18c4(Msh6+/+andPms2+/+panels)alongwithrepresentativeMsh6+/(cid:2)andMsh6(cid:2)/(cid:2)or
Pms2+/(cid:2) and Pms2(cid:2)/(cid:2) derivatives. In (B) an (C), *P<0.05 ANOVA with Bonferroni’s post-test.
repair of G/U mispairs or through its anti-recombination growth density ((cid:5)0.5(cid:4)106 sIgM(cid:2) cells/ml). These large
activity. The frequency of Ig GC can be estimated from starting populations bypassed the defect in cloning effi-
fluctuationanalysisofthesurfaceIgMphenotypeinDT40 ciency and minimized the distorting effects that the
(Figure3A)(34).Inbrief,multiplesubclonesoftheIgM(cid:2) much longer lag phase in more diluted Msh6(cid:2)/(cid:2) and
DT40 CL18 cell line, which has a frameshift in IgV(cid:2) Pms2(cid:2)/(cid:2) cultures would have on Ig GC frequency.
preventing IgM expression (28), are expanded over time. To account for the different growth rates, we expanded
Reversion of the frameshift by constitutive Ig GC allows the cell populations for the same number of generations,
re-expressionofIgMwiththemedianproportionofIgM+ which were calculated from the respective population
cells in the clonal populations providing an estimation doublingtimes.Weconfirmedthatthemedianproportion
of Ig GC frequency. As the proportion of IgM+ cells ofIgM+cellsincreased linearlywiththenumberofpopu-
increases with clonal expansion time, any difference lation doublings for DT40 WT as well as Msh6(cid:2)/(cid:2) cells
in growth kinetics would affect these assays, which (Figure3BandC).Whenweusedthismethodtocompare
was a major concern in this case. Moreover, we could the DT40 CL18 to multiple independent Msh6(cid:2)/(cid:2) or
not obtain Msh6(cid:2)/(cid:2) or Pms2(cid:2)/(cid:2) single-cell clones Pms2(cid:2)/(cid:2) derivatives, we observed interclonal variability.
(Figure 1C). We, therefore, modified the fluctuation Nevertheless, although the median proportion of IgM+
assay by setting up multiple populations at exponential cells arising from DT40 Msh6(cid:2)/(cid:2) was reduced for one
NucleicAcidsResearch,2013,Vol.41,No.5 3037
A B
5 P=0.037
%) 4
+ (
V 3
n
xi 2
e
n 1
n
A
0
CL18 Msh6-/-Pms2-/- CL18 Msh6-/- Msh6-/- Pms2-/-
1015 1C1 c44
C D
DT40 CL18 Time post-Cisplatin
60
>G2 0 5 24
40
CL18
20 65 68 15
0
Msh6-/-
60
>G2
10 21 4 23 11 66
40
20 Msh6+/-
Cells (n) >G2 46000 Msh6-/- hMsh6 Cells (%) BrdU 62 77 48
14 21 7 14 1136
20
0 Msh6-/-
Pms2-/- 31 27 32 33 31 13
60
>G2
40
8 6 22
20 9 22 1820 1022
0 2n 4n 8n
2n4n G1 S G2 >G2
DNA content /M DNA content
Figure 2. Cell cycle defects in MMR-deficient DT40 cells. (A) Proportion of AnnexinV-positive cells in exponentially growing cultures of the
indicated DT40 cell lines. Mean+SD of —five to seven measurements done on CL18 and two independently derived knockout cell lines are
plotted. Statistical significance by Student’s unpaired two-tailed t-test. (B) Representative H&E staining images illustrating the presence of large
cells with abnormal nuclear morphology in MSH6- and PMS2-deficient DT40. Bar, 20mm. (C) Cell cycle profile of exponentially growing popula-
tions of DT40 cells. DNA was stained with propidium iodide and analysed by flow cytometry. 2n and 4n indicate the G and G/M peaks,
1 2
respectively(leftpanels).ThepopulationofcellswithDNAcontent>4nwasdefinedas>G.Thebargraphssummarizethemean+SDproportion
2
of cells in each cell cycle stage obtained from seven replicates for CL18 and Msh6(cid:2)/(cid:2), two replicates for Msh6(cid:2)/(cid:2) hMsh6 and three replicates for
Pms2(cid:2)/(cid:2)(rightpanels).(D)2DcellcycleanalysisofDT40cellspulsedwithBrdUat0,5or24hpost-cisplatintreatment.BrdUincorporationinto
DNAwasprobedwithFITC-labelledanti-BrdUantibodyandDNAcontentbypropidiumiodidestainingbyflowcytometry.Theproportionofcells
in each gate is indicated. One representative of two experiments performed is shown.
line, it was not significantly different in three others of DT40 Msh6(cid:2)/(cid:2) Aid(cid:2)/(cid:2) lines showed that ablation of
(Figure 3D). Similarly, one DT40 Pms2(cid:2)/(cid:2) line generated Msh6(cid:2)/(cid:2) did not result in spontaneous Ig GC in the
onlyfew IgM+cells, but two otherswere significantly dif- absence of the initiating lesion even on treatment with
ferentneitherfromCL18norfromtheirPms2+/(cid:2)parental Trichostatin A (Figure 3G), which exacerbates Ig GC in
lines. The variability in IgM+ proportion between lines DT40 cells (55).
could not be accounted for by variations in AID levels The frequency of Ig GC could be overestimated
(Figure 3E; data not shown) nor did it correlate with when normalized to population doubling times, which is
their relative population doubling times. We suspect influenced by cell death. Nevertheless, the BrdU pulse-
these variations to be because of different compensatory chase experiments suggest that the cell cycle is longer in
adaptations arising in the MMR-deficient lines during MMR-deficient than in WT DT40 cells given the persist-
clonal expansion and growth. Importantly, Ig GC in enceofahigherproportionofBrdU+cellsinMsh6(cid:2)/(cid:2)and
Msh6(cid:2)/(cid:2) cells proceeded through its canonical mechan- Msh6+/(cid:2)cells24hpost-pulse(Figure2D;datanotshown).
ism. First, UNG deficiency caused by expressing the In any case, when logarithmically growing cells were
protein inhibitor Ugi in two independent DT40 Msh6(cid:2)/(cid:2) expanded for the same length of time, Ig GC was still
cell lines inhibited Ig GC as shown by a decrease in the apparent,althoughitsfrequencyinMMR-deficientDT40
frequencyofIgM+cells(Figure3F)(38).Second,analysis cellswas(cid:5)2-to4-foldlowerthaninthecontrolpopulation
3038 NucleicAcidsResearch,2013,Vol.41,No.5
A B
40
Gene DT40
conversion 30 R2 = 0.97
IgM- IgM+ %)
+ ( 20
M
ΨV Vλ C g
I 10
Frameshift 0
20 40 60
Generation (n)
C
30 30 -/-
CL18 Msh6 1015
R2 = 0.92 R2 = 0.95
%) 20 %) 20
+ ( + (
M M
g 10 g 10
I 7.5 I
6.1
0 0
20 40 60 20 40 60
Generation (n) Generation (n)
D
10.5 1.4 5.1 5.7 14.6 0.9 0.7 0.9 12.8 10.3 10.7 13.6
∗ ∗ ∗ ∗
55
30
%)
+ ( 20
M
g
I
10
0
CL18 102 1015 1C1 1A5 A5 B1 c20 c5 c11 c21 c44
wt Msh6-/- Msh6-/- Aid-/- Pms2-/- Pms2+/-
E F 15 G +TSA
DT40 Msh6-/-Pms2-/- 43.3 18 1.0 1.0
60
Aid-/-10151A51C1c20 + (%) 5 %) 40
3276 AID IgM 3.6 gM+ ( 20
I
1
48
0 0
Actin Msh6-/- Msh6-/- CL18 1015 A5 B1
Vector Ugi wt Msh6-/- Msh6-/- Aid-/-
Figure 3. Noincrease inIgGCfrequencyinMsh6(cid:2)/(cid:2)orPms2(cid:2)/(cid:2)DT40 cells.(A)Scheme oftheframeshiftreversion assayused toestimateIgVl
gene conversion frequency by measuring the proportion of DT40 cells gaining surface IgM expression over time. (B) Linear correlation between
medianproportionofIgM-gaincellsandgenerationnumber(definedbypopulationdoublingtime)formultipleDT40CL18subpopulationskeptin
exponentialgrowth.TheproportionofIgM+cellsarisingfrom24subpopulationsofIgM(cid:2)cellswasdeterminedbyflowcytometryateachindicated
number of generations. Median values and data range for two independent experiments (circles and triangles) are plotted with linear fit shown.
(C)MultipleIgM(cid:2)subpopulationsofDT40CL18(n=24)anditsMsh6(cid:2)/(cid:2)derivativeline1015(n=12)weresorted,independentlyexpandedand
analysedforIgM-gainasin(B)after15,30,45and60generations.TheIgM+proportionvaluesareindicatedforeachsubpopulationwithmedians
plottedashorizontalbars.Linearfitofthemediansisshown.(D)IgM-gaininmultiplesubpopulationsofCL18(n=19),itsMsh6(cid:2)/(cid:2)derivativelines
102(n=12),1015(n=9),1C1(n=19)and1A5(n=9),Msh6(cid:2)/(cid:2)Aid(cid:2)/(cid:2)A5(n=12)andB1(n=12),Pms2(cid:2)/(cid:2)linesc20(n=20),c5(n=12)and
c11 (n=12)and the Pms2+/(cid:2)c21 (n=12) andc44 (n=12)(parentals to Pms2(cid:2)/(cid:2)c5 andc11) after70 generationswas determined asin (C). The
proportionofIgM+cellsineachsubpopulationisplottedwithmediansindicatedbyhorizontallinesandthevalueabove.*P<0.05Kruskal–Walis
non-parametric testwith post-test. (E)Total lysates of DT40 cell lines were separated by sodiumdodecyl sulphate–polyacrylamide gel electrophor-
esis,andAIDandactinlevelswereanalysedbywesternblotting.(F)Populationsof50IgM(cid:2)CL18Msh6(cid:2)/(cid:2)cellsexpressingornotUgiweresorted
into 96-well plates and expanded for 45 generations. The proportion of IgM+cells is plotted for each population (two shades of grey distinguish
independent Msh6(cid:2)/(cid:2) lines) with median values indicated. (G) Three IgM(cid:2) subpopulations of DT40 CL18, Msh6(cid:2)/(cid:2) line 1015 or Aid(cid:2)/(cid:2) Msh6(cid:2)/(cid:2)
linesA5andB1were expandedfor45generationsinthepresenceof1.25nMTSA.TheproportionofIgM+cells ineachsubpopulation isplotted
with median values indicated.
NucleicAcidsResearch,2013,Vol.41,No.5 3039
(e.g.compareIgGCforCL18at60generationstoMsh6(cid:2)/(cid:2) frequencies (Table 1), but not beyond what others and
1015at30generationsinFigure2C;datanotshown). us have observed in UNG-deficient DT40 cells (38,39).
WeconcludethatMSH6andPMS2aredispensablefor Thus, MSH6 or PMS2 deficiency seems to lead to a
Ig GC, and that their deficiency does not result in any small increase in the length of the GC tracts, but does
obvious change in Ig GC mechanism (i.e dependence on not affect the donor preference or the number of muta-
AIDandUNG).WecannotformallyruleourthatMSH6 genic events per sequence at the IgVl locus of DT40 cells
andPMS2 mightbe requiredforoptimalIg GC(see later during Ig GC.
in the text), but our results show that their absence does
not lead to an increase in Ig GC frequency. MMR in DT40 cells is functional, and homologous
recombination is heterology sensitive
MMR deficiency does not significantly affect the outcome The lack of effect of MMR-deficiency on Ig GC or
Ig gene conversion mutation frequency was striking. We, therefore, asked
As the growth defects in MMR-deficient DT40 cells whether MMR was active in DT40 cells. We measured
hinder accurate calculation of Ig GC frequency, we their sensitivity to 6-thioguanine (6-TG), which, once
analysed the GC events by sequencing the IgVl region incorporated into DNA and methylated, generates
amplified from IgM+ cells. This analysis provides infor- mismatches with C that cause futile MMR cycles, DNA
mation about multiple parameters that could be affected fragmentation and cytotoxicity (56). As anticipated, WT
by MMR (Supplementary Figure S3), can be unequivo- DT40 cells were sensitive to 6-TG, whereas Msh6(cid:2)/(cid:2) cells
cally scored and are less influenced by cell growth. Thus, were resistant (Figure 5A). Pms2(cid:2)/(cid:2) cells were also more
we analysed the GC donor pseudogene usage, GC tract resistant to 6-TG than WT cells, albeit to a lesser extent
length, number of GC events per sequence and frequency than Msh6(cid:2)/(cid:2) cells (Figure 5A). Together with the bio-
of untemplated point mutations. We compared independ- chemical evidence described later in the text, these
ent (not clonally related) Ig GC events obtained from results show that MMR is active in DT40 cells. Despite
this, DT40 cells fail to reject Ig GC intermediates, which
multiple subpopulations of various DT40 lines grown
must have heterologies. We, therefore, set out to test
for the same number of generations.
whether HR in DT40 cells might generally tolerate heter-
IgGCinchickensprefersusingthemostsimilarpseudo-
ology, especially given that one of the key hallmarks of
geneasdonor(27).The pseudogene cVL8ismost closely
related in sequence to DT40 V(cid:2) ((cid:5)90% overall, (cid:5)97% DT40 cells is their high efficiency of targeted integration
(57) mediated by HR (32). To this end, we used gene tar-
around the frameshift insertion) (Supplementary Figure
geting as an assay. The Msh6 disruption construct
S4), and, as anticipated, it was most frequently used to
inactivated the first Msh6 allele with 35% efficiency. We
correct the frameshift in DT40 CL18 (Figure 4A). This
preference was retained in the Msh6(cid:2)/(cid:2) or Pms2(cid:2)/(cid:2) cells generated a modified version of this construct containing
15 mutations within the 1764-bp-long 50-arm (Figure 5B).
(Figure 4A). We observed slightly longer GC tracts in
DT40 Msh6(cid:2)/(cid:2) cells than in CL18, but this difference This represents only 0.8% of heterology, well below the
(cid:5)10% divergence between Ig VL and the donor pseudo-
was not significant, irrespective of whether the data from
eachindependentMsh6(cid:2)/(cid:2)linewereanalysedseparatelyor genes (3); yet, it was sufficient to completely abolish tar-
geting (Figure 5C and D). Because of their negligible
pooled(Figure4BandC).Aclearertrendtowardslonger
Ig GC tract lengths in the Pms2(cid:2)/(cid:2) cells became statistic- cloning efficiency, we were unable to formally demon-
strate that gene targeting sensitivity to heterology in
ally significant when the analysis was restricted to Ig GC
DT40 Msh6(cid:2)/(cid:2) and Pms2(cid:2)/(cid:2) cells is MMR-dependent.
events templated on cVL8 (Figure 4C). This eliminates
However, this has been well demonstrated in bacteria
confounding effects, such as differences in the length,
(58), yeast (59) and mammalian cells (44,60,61). Thus,
relative orientation or degree of homology of different
our findings show that MMR is functional in DT40
pseudogeneswiththeIgV(cid:2).
cells, and that these cells tightly control the fidelity of
Wethencomparedthenumberofmutageniceventsper
HR in vivo, most likely through MMR.
sequence.ShouldMMR-deficientcellsbemorepermissive
forrecombination, onewould expect more GCevents per
Extracts of DT40 cells process G/U mispairs efficiently
sequence. However, this was not the case (Figure 4D).
but predominantly by BER
Alternatively, if MutSa was recognizing AID-generated
G/U mispairs to initiate faithful MMR, Msh6(cid:2)/(cid:2) cells To unequivocally demonstrate that MMR is biochem-
would be expected to display more single base pair muta- ically active in DT40 cells and in an attempt to identify
tionsattheIgV.Thiswasalsonotthecase.Thedatasetof the possible cause(s) of the lack of effect of MutSa and
sequences obtained from IgM+ cells arising from IgM(cid:2) MutLa loss on GC, we examined the substrate specificity
populations had a similar ratio of point mutations per of avian MMR. We first carried out gel-shift experi-
mutated sequence in CL18 and Msh6(cid:2)/(cid:2) or Pms2(cid:2)/(cid:2) ments with DT40 extracts. As shown in Figure 6A,
lines (Figure 4E). A complementary experiment analysing mismatch-specific complexes were formed on incubation
mutationsintheIgVofIgM-losscellsoriginatingfromthe of the WT extracts with oligonucleotide heteroduplexes
IgM+ Msh6(cid:2)/(cid:2) did not show any increase in mutations containing a G/T mispair or one extrahelical nucleotide
either (Table 1). Combining MSH6 deficiency with UNG (+1). The homoduplex G/C formed only unspecific
inhibition by expressing Ugi in two independent CL18c4 complexes in these extracts (lane 1) (62). The fact that
Msh6(cid:2)/(cid:2) lines increased untemplated point mutation only the unspecific complexes were observed in extracts
3040 NucleicAcidsResearch,2013,Vol.41,No.5
A 100
CL18 n=54
80
Msh6-/- n=57
%) 60 Pms2-/- n=38
(
e
g 40
a
s
U
20
0
1 2 3 4 5 6 7 8 9 10 1112 13 14 15 16 17 18 19 20 21 22 23 24 25
Donor pseudogene
B
88 81 88 89 118 96 83 71 105 121 Mean
300 86 84 86 79 88 84 84 84 114 121 Median
)
p
b
(
ct
a
r 200
C t
G
m
u
m 100
xi
a
M
0
E1 E2 E3 102 1015 1c1 1a5 c20 c5 c11
CL18 Msh6-/- Pms2-/-
C D E
All ΨVλ ΨVλ8 q
300 e
s 1.0
p) 71 d
b * e
ct ( utat 0.8
a200 CL18 m
C tr s / 0.6
G n
um 100 115 atio 0.4
m ut 0.2
xi Msh6-/- m
Ma nt 0.0
0 CL18sh6-/-ms2-/- CL18sh6-/-ms2-/- 76 Poi CL18 Msh6-/- P ms2-/-
M P M P
Pms2-/-
Figure 4. UnaffectedIggeneconversionfidelityinMsh6(cid:2)/(cid:2)andPms2(cid:2)/(cid:2)DT40.(A)Donorpseudogeneusage.ProportionofIgVlconversionevents
templated on each of the 25 IgV pseudogenes plotted for DT40 CL18, Msh6(cid:2)/(cid:2) and Pms2(cid:2)/(cid:2). The data compile sequences from subpopulations
comingfromseveralfluctuationexperiments (CL18 n= 3,Msh6(cid:2)/(cid:2)n=6andfourindependentlines, Pms(cid:2)/(cid:2)n=3andthreeindependent lines).
Thedistributionwasnotsignificantlydifferentbetweenthedifferentgenotypes(two-wayANOVA).(B)Maximumgeneconversiontractlengthwas
determined for each event obtained from CL18 (three independent experiments, E1–E3) as well as from independent lines of Msh6(cid:2)/(cid:2) (two fluc-
tuations compiled for lines 1015 and 1C1) or Pms2(cid:2)/(cid:2). Tract length for each event is plotted with the mean and median (horizontal bars) values
above.DifferencesarenotsignificantbyKruskal–Waliesnon-parametrictest.(C)Alldatafrom(B)wereaggregatedbygenotypeinboxplotswith
25–75%percentilesdividedbythemedianandwhiskersrepresentingtheminimumandmaximumvalues.Ontheright,asimilarplotconsideredonly
conversiontractstemplatedonpseudogeneVL8.*P<0.05Kruskal–Walisnon-parametrictestwithpost-test.In(A–C),onlyindependenteventswere
considered (i.e. identical events found in sequences coming from the same subpopulation were excluded to prevent counting dynastically related
events more than once). (D) The number of sequences containing one (white), two (grey) or three (black) conversion events were plotted as
proportion of all sequences containing gene conversion events (n, indicated in the central circle). (E) The ratio of point mutations per mutated
sequencewascalculatedformultiple datasets.ThehighestvalueinthePms2(cid:2)/(cid:2)datasetisbiasedbythepresenceoffoursequenceswithmultiple
clonallyrelatedmutations.Theemptytriangleisthevalueforthesamedatasetafterremovingthesesequences.Clonallyrelatedsequenceswerenot
excluded in (D) and (E).
NucleicAcidsResearch,2013,Vol.41,No.5 3041
Table 1. Mutations in CL18c4-derived Msh6(cid:2)/(cid:2) DT40 cells sorted for IgM-loss
Msh6+/+ Msh6(cid:2)/(cid:2)
Clone CL18c4 c414 c429
Population analysed IgM(cid:2) IgMlow IgMlow IgMlow IgMlow
Ugia Ugi
Sequences n 83 69 24 72 24
Mutated n (%) 39 (47) 10 (14) 12 (50) 3 (4) 14 (58)
Gene conversions Independentb 15 6 2 0 1
Total 15 6 3 0 1
Per mut seqc 0.38 0.6 0.25 - 0.1
Point mutations Independentc 15 3 10 1 14
Total 16 4 12 1 15
Per mut seqc 0.4 0.4 1 0.3 1.1
Ambiguousd 8 4 2 2 2
Insertion/deletions 8 0 0 0 0
aThe endogenous UNG activity was inhibited in vivo during expansion of the populations by expressing the Ugi protein.
bIdentical events within a population were excluded to avoid recounting dynastically related events.
cAll, rather than independent, events were considered to calculate mutation load per mutated sequence.
dSingle base pair substitutions that could have been templated or untemplated.
of the Msh6(cid:2)/(cid:2) cells (lanes 2, 3, 5, 6, 8, 9) confirmed that experiments and implies that MutSb in DT40 extracts is
thecellslackedMutSa.Moreover,thisresultshowedthat either absent, or that it does not address an insertion/
MutSb in DT40 cell extracts either does not bind the+1 deletionloopofoneextrahelicalnucleotide,anintermedi-
substrateorispresentinundetectableamounts.Wefailed ate arising during Ig GC, the processing of which is
to detect specific complexes with the G/U substrate even required to convert IgM(cid:2) cells to IgM+phenotype.
onsupplementationoftheextractswithUgi.Thisimplied We also wanted to learn how the extracts processed a
either that, unlike the human protein (63), avian MutSa G/U mispair. As shown in Figure 6D (lanes 2–6), this
does not recognize this substrate, or that the uracil was substrate was quantitatively converted to G/C in all
removed from the oligonucleotide substrate by a three extracts. The repair efficiency was diminished on
glycosylase other than UNG. Pms2(cid:2)/(cid:2) extracts contained TDGdepletionandUgiaddition(lanes7–11),butsupple-
similar levels of MutSa to WT extracts (Supplementary mentation with recombinant MutSa or MutLa failed to
Figure S5). To confirm that the binding affinity was re- increase it, which implied that MMR does not contribute
flected in repair efficiency, we carried out in vitro MMR to the processing of G/U mispairs in our assay, and that
assays using DT40 cell extracts and a phagemid these mispairs are addressed almost exclusively—and effi-
heteroduplex containing a G/T mispair in a SalI restric- ciently—by BER. Intriguingly, the limited effect of Ugi
tion site, which makes it refractory to cleavage with this addition and TDG depletion suggested that the extracts
enzyme. The phagemid also contained a nick as a strand containanotherefficienturacilDNAglycosylase.Itcould
discrimination signal 30 from the mispaired T. In this be TDG because we cannot estimate efficiency of its
scenario, both MutSa and MutLa are absolutely immunodepletion by the anti-human TDG antibody
requiredforsuccessfulMMR.Digestionoftheunrepaired other than from the additive effect of UNG inhibition
phagemid with DraI and SalI gives rise to fragments of by Ugi and TDG immunodepletion on the in vitro
2484 and 694bp. On correction of the G/T mismatch to MMR assay (Supplementary Figure S6), but we cannot
G/C,theSalIsiteisrestored,suchthatthe2484fragment exclude the presence of another enzyme, such as MBD4
iscutintofragmentsof1324and1160bp.Repairefficiency (64), given that DT40 cells lack SMUG1 (65).
canthusbeestimatedfromtheratioofthe2484and1324
(or1160)fragments.AsshowninFigure6B,(cid:5)30%ofthe
DISCUSSION
G/TsubstratewasrepairedinDT40WTextracts(lane2).
In contrast, extracts of Pms2(cid:2)/(cid:2) (lane 3) and Msh6(cid:2)/(cid:2) The near absence of Ig GC in UNG-deficient DT40 cells
(lane 5) cells were repair deficient. The lack of activity implied that MMR does not activate an alternative
was not caused by poor quality of the extracts, given pathway of uracil processing for initiating Ig GC
that MMR could be restored by the addition of recom- (10,39). This was unexpected, given that MMR compen-
binanthumanMutLaorMutSa,respectively(lanes4and sates to a substantial extent for UNG-deficiency for
6). Importantly, although the WT extracts addressed the SHM and CSR (11,24). Moreover, Ig GC is a process of
+1 substrate with similar efficiency to G/T (Figure 6C, homeologous recombination, which is believed to involve
compare lanes 2 and 4), we failed to detect repair of MMR in at least two stages (Supplementary Figure S2).
either substrate in extracts of Msh6(cid:2)/(cid:2) cells (lane 5), To gain insights into the criteria that govern Ig GC, we
unless they were supplemented with recombinant MutSa disrupted the Msh6 or Pms2 genes in DT40 cells. In this
(lane 6). This confirms the results of the gel-shift way, we selectively inactivated two key MMR complexes,