BLAP18/RMI2, a novel
is an essential component of the Bloom
helicase–double Holliday junction
Thiyam Ramsing Singh,1,3Abdullah Mahmood Ali,1,3Valeria Busygina,2,3Steven Raynard,2
Qiang Fan,1Chang-hu Du,1Paul R. Andreassen,1Patrick Sung,2and Amom Ruhikanta Meetei1,4
1Division of Experimental Hematology and Cancer Biology, Cincinnati Children’s Research Foundation, University
of Cincinnati College of Medicine, Cincinnati, Ohio 45229, USA;2Department of Molecular Biophysics and Biochemistry,
Yale University School of Medicine, New Haven, Connecticut 06520, USA
Bloom Syndrome is an autosomal recessive cancer-prone disorder caused by mutations in the BLM gene. BLM
encodes a DNA helicase of the RECQ family, and associates with Topo III? and BLAP75/RMI1 (BLAP for
BLM-associated polypeptide/RecQ-mediated genome instability) to form the BTB (BLM–Topo III?–BLAP75/
RMI1) complex. This complex can resolve the double Holliday junction (dHJ), a DNA intermediate generated
during homologous recombination, to yield noncrossover recombinants exclusively. This attribute of the BTB
complex likely serves to prevent chromosomal aberrations and rearrangements. Here we report the isolation
and characterization of a novel member of the BTB complex termed BLAP18/RMI2. BLAP18/RMI2 contains a
putative OB-fold domain, and several lines of evidence suggest that it is essential for BTB complex function.
First, the majority of BLAP18/RMI2 exists in complex with Topo III? and BLAP75/RMI1. Second, depletion
of BLAP18/RMI2 results in the destabilization of the BTB complex. Third, BLAP18/RMI2-depleted cells show
spontaneous chromosomal breaks and are sensitive to methyl methanesulfonate treatment. Fourth,
BLAP18/RMI2 is required to target BLM to chromatin and for the assembly of BLM foci upon hydroxyurea
treatment. Finally, BLAP18/RMI2 stimulates the dHJ resolution capability of the BTB complex. Together,
these results establish BLAP18/RMI2 as an essential member of the BTB dHJ dissolvasome that is required for
the maintenance of a stable genome.
[Keywords: RECQ; double Holliday junction; Chromatin; BLAP18/RMI2; BLM]
Supplemental material is available at http://www.genesdev.org.
Received August 8, 2008; revised version accepted August 28, 2008.
The autosomal recessive disorder Bloom syndrome (BS)
is characterized by severe growth retardation, immuno-
deficiency, anemia, reduced fertility, and cancer predis-
position (German 1969; German and Ellis 2002). BS pa-
tients show different malignancies that appear early in
life and within various tissues (German 1969; German
and Ellis 2002). Cells derived from BS patients show cy-
togenetic abnormalities, such as chromosomal breaks
and an elevated rate (10-fold) of sister chromatid ex-
changes (SCEs). These cytogenetic abnormalities are
used as a molecular diagnostic test for the disease (Ger-
man 1969; German et al. 1977c).
BS stems from mutations in the BLM gene (Ellis et al.
1995). BLM protein belongs to the RecQ helicase family,
which also includes RECQ1, WRN, RECQ4/RTS, and
RECQ5, all of which play a unique role in the mainte-
nance of genomic stability. WRN and RECQ4/RTS are
also needed for the suppression of cancer and premature
aging in humans (Ellis et al. 1995; Hanada and Hickson
2007), while the ablation of RECQ5 in mice engenders a
late-onset tumor susceptibility phenotype (Hu et al.
BLM is a structure-specific helicase that can unwind
3?-tailed duplexes, bubble structures, forked duplexes,
(D-loops), and four-way junctions that model Holliday
junction (HJ) recombination intermediates (for review,
see Hanada and Hickson 2007). The BLM–Topo III? com-
plex has been shown to resolve double Holliday junc-
3These authors contributed equally to this work.
E-MAIL firstname.lastname@example.org; FAX (513) 636-3768.
Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.1725108.
2856 GENES & DEVELOPMENT 22:2856–2868 © 2008 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/08; www.genesdev.org
tion (dHJ) in vitro in a noncrossover fashion, and the
recently discovered BLAP75/RMI1 (BLAP for BLM-asso-
ciated polypeptide/RecQ-mediated genome instability)
strongly stimulates this reaction (Raynard et al. 2006;
Wu et al. 2006). The BLM–Topo III?–BLAP75/RMI1 en-
semble has been termed the BTB (or RecQ–Topo III?–
RMI1) complex (Raynard et al. 2006; Wu et al. 2006).
BLM and Topo III? interact with the OB-fold-containing
N-terminal region of BLAP75/RMI1 (Raynard et al.
2008).The ability of the BTB complex to dissolve dHJs to
yield noncrossovers is thought to play a crucial role in
the avoidance of chromosomal rearrangements, such as
translocations, during the homology-directed repair of
chromosomal lesions and injured replication forks (Sung
and Klein 2006; Wu and Hickson 2006).
BLM has been shown to localize to promyelocytic leu-
kemia (PML) bodies in the absence of DNA damage (Bis-
chof et al. 2001). Upon the occurrence of DNA damage or
inhibition of DNA replication, however, BLM dissoci-
ates from PML bodies to form nuclear foci, where it co-
localizes with other DNA repair proteins, such as
RAD51, BRCA1, and the MRE11–RAD50–NBS1 com-
plex (Bischof et al. 2001). Consistent with these obser-
vations, BLM is recruited to laser-induced DNA double-
strand breaks (DSBs) (Dutertre et al. 2000; Karmakar et
Here, to better understand the function of BLM in
DNA damage repair and response, we sought to deter-
mine whether the BTB complex harbors other protein
components and, if so, to ascertain the function of these
novel BTB components. Earlier immunopurification ap-
proaches utilizing BLM antibody had a disadvantage in
that the IgG light chain of the antibody might have
masked BTB-associated proteins (Meetei et al. 2003). We
therefore used a recently developed two-step affinity pu-
rification approach by expressing BLM fused with a
double tag containing (His)6and Flag. This new approach
has led to the identification of BLAP18/RMI2 as a novel
component of the BTB complex. We find that BLAP18/
RMI2 forms a core complex with Topo III? and BLAP75/
RMI1. We also find that BLAP18/RMI2 is required for
the recruitment of BLM to chromatin and replication
stress-induced nuclear foci. Depletion of BLAP18/RMI2
yields a profile of chromosome instability and sensitivity
to DNA damage similar to that observed in BS cells.
These results thus help define the nature of the BLM-
associated protein complex in cells and reveal a critical
role of BLAP18/RMI2 in the promotion of BLM-depen-
dent genome maintenance pathway.
Since the Saccharomyces cerevisiae BLAP75 ortholog
is called Rmi1, we will henceforth refer to BLAP75 as
RMI1 and BLAP18 as RMI2 to be consistent with the
RMI2 is a novel component of BLM-containing
In order to gain more insight into the cellular function of
BLM-containing complexes, we used a two-step affinity
purification coupled mass spectrometry (MS) approach
to isolate and identify novel BLM-associated polypep-
tides. BLM that harbors an N-terminal Flag tag and a
C-terminal (His)6-tagged (F-BLM-H) was stably expressed
in HT1080 cells by retroviral-mediated gene transfer,
and BLM and its associated proteins were purified by a
two-step affinity chromatographic procedure as de-
scribed in the Materials and Methods. MS analysis of the
polypeptides in the purified fraction identified several
proteins that are known to interact with BLM, including
the other components of the BTB complex, Topo III?,
and RMI1. In addition, we also identified novel polypep-
tides of 18 kDa and 15 kDa molecular mass. Here we
focus on the 18-kDa polypeptide, which we named as a
BLM-associated polypeptide of 18-kDa mass, or RMI2
(Fig. 1A). The 15-kDa polypeptide, BLAP15, will not be
considered further here.
RMI2 is clearly one of the prominent polypeptide com-
ponents of the purified BLM protein complex (Fig. 1A).
RMI2 corresponds to an uncharacterized hypothetical
protein named C16orf75 (chromosome 16 ORF 75), oth-
erwise also known as MGC24665 or LOC116028. RMI2
has no known function and appears to harbor an OB-fold
domain (Supplemental Fig. 1).
We generated two different antibodies against RMI2 (A
and B), both of which proved suitable for immunoblot
analysis of RMI2 tagged with the (His)6and Flag epitopes
at its N terminus (HF-RMI2) and expressed in HeLa cells
(Fig. 1B), and of endogenous RMI2 present in the BLM
complex (Fig. 1B,C). When we immunoprecipitated
RMI2 using the anti-RMI2 antibodies (Fig. 1C), we found
by immunoblot analysis the presence of BLM, Topo III?,
and RMI1 (Fig. 1C).
Likewise, two-step affinity purification of HF-RMI2
from extracts of HeLa S3 cells under high stringency (400
mM salt) coupled with MS revealed its association with
BLM, Topo III?, and RMI1. BLAP15 was also found in the
purified fraction (Fig. 1D). We established that the poly-
peptide denoted as HF-RMI2-L stems from a leaky up-
stream start site in the retroviral construct used for
tagged RMI2 expression.
Immunoprecipitation of HF-RMI2 using anti-Flag an-
tibody significantly depleted both Topo III? and RMI1,
and to a large extent BLM, from the flow-through frac-
tion (Fig. 1E). Taken together, the results help identify
RMI2 as a novel and integral component of the BTB com-
RMI2 contains an OB-fold domain and is conserved
among higher eukaryotes
BLAST analysis of human RMI2 identified likely or-
thologs in vertebrate and invertebrate species that show
a high degree of amino acid sequence identity to the
human protein (Supplemental Fig. 1). Sequence analysis
using InterProScan revealed a putative OB-fold domain.
The OB-fold domain is a compact structural motif in-
volved in both nucleic acid-binding and protein–protein
interactions (Theobald et al. 2003; Raynard et al. 2008).
Proteins with OB-fold domains have been implicated in
BLAP18/RMI2, a novel component of the BTB complex
GENES & DEVELOPMENT 2857
DNA repair, DNA replication, DNA recombination,
transcription, translation, cold shock response, and telo-
mere maintenance. The recently discovered BLM-inter-
acting protein, RMI1, also contains OB-fold domain that
apparently mediates the interaction of RMI1 with Topo
III? and BLM (Raynard et al. 2008). Additional examples
of OB-fold-containing proteins include the single-strand-
binding protein RPA and the DNA repair protein and
tumor suppressor BRCA2. OB-fold domains from differ-
ent proteins share the characteristic feature of five ?-
sheets (Theobald et al. 2003). All five ?-sheets are pres-
ent in the putative OB-fold domain of RMI2 as predicted
using the program PSIPRED (Supplemental Fig. 1). We
did not find any RMI2 ortholog in lower organisms like
Caenorhabiditis elegans or yeast based on protein se-
quence alignment analysis.
RMI2 is required for the stability of Topo III?
The absence of a critical subunit of a multicomponent
protein complex often destabilizes the complex (Yin et
al. 2005). As shown in Figure 2A, upon depletion of RMI2
in HeLa cells by siRNA treatment, a decrease in Topo
III? and RMI1 protein level was observed (Fig. 2A, lane
2), although the level of BLM protein remained little af-
fected. Conversely, depletion of Topo III? or RMI1 by
siRNA treatment reduced the level of RMI2 (Fig. 2A,
lanes 3,4). The results thus reveal an interdependence of
RMI2, RMI1, and Topo III? for their stability in cells. On
the other hand, depletion of BLM by siRNA resulted in
no significant change in RMI2, Topo III?, or RMI1 pro-
tein level (Fig. 2A, lane 5).
We also investigated the level of RMI2, Topo III?, and
RMI1 in BLM-deficient cells (GM08505) derived from a
BS patient and the same cells reconstituted with func-
tional BLM (Fig. 2B, lanes 1,2). We observed a compa-
rable level of RMI2, Topo III?, and RMI1 in both cell
lines (Fig. 2B, lanes 1,2). SiRNA-mediated depletion of
RMI2 in BLM-reconstituted cells also led to the destabi-
lization of Topo III? and RMI1 (Fig. 2B, lane 3). These
results are consistent with earlier reports suggesting that
the stability of Topo III? is dependent on RMI1 but that
Topo III? and RMI1 are stable without BLM (Meetei et
al. 2003; Yin et al. 2005).
Since RMI2, Topo III?, and RMI1 levels were un-
changed in BLM-depleted cells, we next investigated the
role of BLM in maintaining the integrity of RMI2, Topo
III?, and the RMI1 complex. We stably expressed HF-
RMI2 in GM0067 (wild type) and GM08505 (BLM-null
cells) and purified the RMI2 complex from these cells
using anti-Flag M2 agarose. HF-RMI2 was able to inter-
act with, and form a stable complex with Topo III? and
RMI1 in these BLM-null cells (Fig. 2C). Thus, the forma-
tion of a stable RMI2, RMI1, and Topo III? complex can
occur in the absence of BLM.
plex. (A) Silver-stained gel showing the polypeptide bands
purified from the nuclear extract of HT1080 cells trans-
duced with vector alone (Mock) and cells expressing F-
BLM-H, as described in Materials and Methods. The poly-
peptides identified by MS analysis are indicated by an
arrow, and the asterisk (*) marks polypeptides also found
in the mock purification. (B) Immunoblot of complexes
purified from HeLa or HT1080 cells using anti-Flag M2
agarose. HeLa cells were transduced with vector alone or
with vector that contained HF-RMI2. HT1080 cells sta-
bly expressed F-BLM-H. Immunoblots were probed with
the antibodies indicated on the right. (C) Immunoblot of
endogenous RMI2 complex immuno-isolated using anti-
RMI2 antibodies raised against either MBP-RMI2 (RMI2-
A) or GST-RMI2 (RMI2-B) fusions and probed with anti-
bodies against known BTB members. RMI2 was detected
with RMI2-A antibody on these immunoblots. (D) Silver-
stained gel showing polypeptide bands purified from the
nuclear extract of HeLa S3 cells transduced with vector
alone (Mock) and cells stably expressing HF-RMI2. The
RMI2 complex was purified using two different NaCl
concentrations (250 and 400 mM, as indicated). Note that
the unique polypeptide with molecular mass ∼27 kDa
(denoted as HF-RMI2-L) was identified as a longer version
of RMI2 that resulted from a leaky start site in the ret-
roviral construct used. (E) Immunoblot of RMI2 complex
purified using anti-Flag M2 agarose from HeLa cells sta-
bly expressing either the vector alone or HF-RMI2. Im-
munoblots were probed with antibodies against BTB
members. The flow-through (FT) fraction is shown as a
measure of depletion of BTB proteins along with HF-
RMI2. Note that there is a substantial amount of BLM
not depleted along with HF-RMI2, Topo III?, and RMI1.
RMI2 is a novel component of the BTB com-
Singh et al.
2858GENES & DEVELOPMENT
RMI2 is phosphorylated during mitosis
BLM is phosphorylated during mitosis and becomes hy-
perphosphorylated in the presence of microtubule desta-
bilizing agents, such as demecolcine and nocodazole, or
microtubule stabilizing agents, such as taxol (Dutertre et
al. 2000; Leng et al. 2006). We found that HF-RMI2 over-
expressed in HeLa cells displayed a slower gel mobility
when cells were grown in the presence of taxol or noco-
dazole (Fig. 2D). Determination of DNA content by flow
cytometry of cells treated with demecolcine and noco-
dazole confirmed cell arrest at the G2/M phase (data not
shown). We also observed a slower moving form of en-
dogenous RMI2 upon treatment of cells with taxol and
nocodazole (Supplemental Fig. 2A). These results suggest
that RMI2, like BLM, is post-translationally modified
during mitosis (Fig. 2D). To determine whether the mo-
bility shift is due to phosphorylation, HF-RMI2 was im-
munoprecipitated with anti-Flag M2 agarose, and the im-
munoprecipitate was treated with ?-phosphatase. As
shown in Figure 2E, the slower migrating form of HF-
RMI2 disappeared upon phosphatase treatment. Block-
ing phosphatase activity with phosphatase inhibitors
prevented the disappearance of the slower migrating HF-
RMI2 form, indicating that RMI2 is indeed phosphory-
lated during mitosis. Immunoprecipitation of HF-RMI2
from taxol-treated HeLa cells expressing HF-RMI2, using
anti-Flag M2 agarose, brought down the phosphorylated
form of BLM (Supplemental Fig. 2B), suggesting that the
BLM–RMI2 complex is not affected by BLM phosphory-
RMI2-depleted cells show methyl methanesulfonate
(MMS) sensitivity and increased chromosome breaks
BS cells and BLM−/−chicken DT40 cells show hypersen-
sitivity to MMS (So et al. 2004). Since RMI2 is an integral
component of the BLM complex, we examined the sen-
sitivity of RMI2- or RMI1-depleted HEK293 cells to
MMS (Fig. 2F). Both RMI2- and RMI1-depleted cells
showed sensitivity to MMS that was comparable with
BLM-depleted cells (Fig. 2F).
We found that depletion of either RMI2 or Topo III?
engenders an increase in spontaneous chromosome gaps
and breaks in HEK293 cells (Table 1), a result that is in
congruence with the recent observation that Topo III?-
depleted DT40 cells accumulate the same chromosome
aberrations (Seki et al. 2006). Taken together, the above
results further strengthen the conclusion that RMI2 is
essential for BLM complex function.
plex. (A) Immunoblot showing levels of BTB members, RMI2,
and actin in lysates from HeLa cells that were transfected with
the indicated siRNA oligos for 72 h. HeLa cells transfected with
a scrambled siRNA oligo were used a control. Asterisk (*) de-
notes a nonspecific crossreactive band detected by anti-RMI1
antibodies. (B) Immunoblot showing levels of BTB members and
RMI2 in lysates from GM08505 (BLM−/−) cells (lane 1) and the
same cells corrected with BLM cDNA and transfected with ei-
ther the control siRNA (lane 2) or RMI2 siRNA (lane 3). (C)
Immunoblot showing input (lanes 1–3) and anti-Flag M2 agarose
immunoprecipitated complex (lanes 4–6) from HeLa S3 cells
transduced with vector alone or GM0067 (wild-type) and
GM08505 (BLM−/−) fibroblast cells transduced with HF-RMI2.
(D) Immunoblot showing the emergence of a slower mobility
species of BLM and RMI2 (lanes 2–3) after HEK293 cells stably
expressing HF-RMI2 were treated with either taxol or noco-
dazole for 16 h, as compared with untreated cells. (E) Immuno-
blot showing the effect of ?-protein phosphatase treatment on
the slower migrating form of RMI2. (F) Graph showing MMS
survival curve of RMI2, BLM. and RMI1 knockdown cells.
HEK293 cells were transfected with either scramble or siRNA
oligos targeting either BLM, RMI1, or RMI2, and were subse-
quently treated with the indicated concentration of MMS. Vis-
ible colonies from 200 cells were counted after 10 d. The data
represent the percent survival, as compared with untreated
cells. Each experiment was independently repeated three times
and representative data are shown. Each experiment was per-
formed in triplicate and mean values are shown with standard
RMI2 is essential for the integrity of the BTB com-
in RMI2 or Topo III?-depleted cells
Frequency of chromosome gaps and breaks
number Cell type and siRNA
Chromosome gaps +
293 + siControl (3 d)a
293 + siRMI2 (3 d)
293 + siTopo III? (3 d)
293 + siControl (7 d)
293 + siRMI2 (7 d)
aThe days in parentheses represent the days after transfection of
BLAP18/RMI2, a novel component of the BTB complex
GENES & DEVELOPMENT2859
RMI2 is required for chromatin targeting of BLM
in response to replication block
Having established that RMI2 forms a stable complex
with BLM, Topo III?, and RMI1 in untreated cells, we
investigated the interaction of RMI2 with BLM, Topo
III?, and RMI1 after DNA damage or blockage of DNA
replication. HeLa cells stably expressing HF-RMI2 were
treated with either hydroxyurea (HU) or MMC for 16 h,
cells were lysed, and HF-RMI2 was immunoprecipitated.
Figure 3A shows that the levels of BLM, Topo III?, RMI1,
and HF-RMI2 were not affected by DNA damage or rep-
lication stress. While a majority of the Topo III? and
RMI1 are depleted from the flow-through fraction before
and after DNA damage, there is a slight decrease of BLM
in the flow-through in extracts from treated cells (Fig.
3A). These data suggest that BLM may exist in cells apart
from the BTB–RMI2 complex. Importantly, there is a
concomitant increase of BLM in the IP fraction using
extracts derived from treated cells (Fig. 3A). These re-
sults suggest that there is an increased interaction of
BLM with HF-RMI2 after DNA damage or replication
Since RMI1 is required for BLM foci formation at the
site of DNA damage or replication block in chromatin
(Yin et al. 2005), we hypothesized that RMI2 is also re-
quired for the localization of BLM to chromatin. We first
compared the distribution of BLM, Topo III?, RMI1, and
RMI2 in untreated HeLa cells or following exposure to
HU. Cells were lysed in low-salt-containing (100 mM)
and detergent-containing buffer, yielding a soluble frac-
tion (S100) and an insoluble fraction. The S100 fraction
contained cytoplasmic and nucleoplasmic proteins, and
the insoluble fraction contained proteins stably bound to
nuclear structures (mainly chromatin and nuclear ma-
trix proteins). The insoluble fraction was further ex-
tracted with a higher salt buffer (300 mM) to give a
soluble fraction (S300: proteins loosely bound to nuclear
structures) and an insoluble fraction (P300: proteins
tightly bound to nuclear structures). Immunoblotting re-
vealed that in untreated HeLa cells, most of the BLM and
Topo III? was in the S300 and P300 fractions (Fig. 3B),
while RMI2 and RMI1 were distributed in all fractions
(S100, S300 and P300) (Fig. 3B). Interestingly, in the
HU-treated samples, BLM, Topo III?, RMI1, and RMI2
became preferentially associated with the P300 fraction.
We next investigated the role of RMI2 in targeting the
BTB complex to chromatin. We knocked down RMI2 in
HeLa cells by siRNA and treated the cells with HU for 16
h The cellular fractionation was carried out as described
above. As shown in Figure 3C, RMI2 depletion resulted
in abrogation of chromatin targeting of BLM in response
to DNA replication stress. We also established that
RMI1 is similarly critical for the targeting of BLM to
RMI2 is required for HU-induced BLM foci formation
The subcellular localization of HF-RMI2 was studied
with an antibody directed against the Flag epitope tag.
While only a subset of untreated cells displayed RMI2
foci, the percentage of cells with RMI2 foci increased
following treatment with HU for 24 h. It has been re-
ported that BLM foci localize to stalled replication forks
following treatment with HU (Davalos and Campisi
2003). As shown in Figure 4A, RMI2 foci strongly colo-
calized with BLM foci, both in untreated cells and fol-
lowing HU treatment. Since RMI2 is required for the
efficient targeting of BLM to chromatin in response to
replication stress (Fig. 3C), we sought to determine
whether RMI2 has a role in the assembly of BLM foci.
Examples of HeLa cells transfected with a siRNA di-
rected against RMI2 or a scrambled siRNA (control) are
shown in Figure 4B. In the control population, most cells
displayed BLM foci following HU exposure, whereas the
assembly of these foci was attenuated in most HeLa cells
following RMI2 depletion (Fig. 4B,C). As summarized in
in response to replication fork blockage. (A) Immunoblot show-
ing that RMI2 and BLM form a tighter complex in response to
DNA damage or replication blockage. HeLa cells stably express-
ing HF-RMI2 were left untreated (UN), or were treated with HU
(1.5 mM) or MMC (100 ng/mL) for 16 h. Cells were lysed and
HF-RMI2 was immunodepleted with anti-Flag M2 agarose. Im-
munoblot showing the levels of BTB members and RMI2 in
input and flow-through (FT) fractions and in immunoprecipi-
tates. Cell lysate from HeLa cells transduced with vector alone
was used as a control. (B,C). Immunoblot showing the levels of
BTB members and RMI2 in the following cellular fractions:
cyto-nucleoplasmic (S100), nucleoplasmic (S300), and chroma-
tin (P300) (see the Material and Methods for details). (B) Com-
parison between untreated (lanes 1–3) and HU treated (lanes
4–6) cellular fractions. (C) Comparison of the effect of a control
siRNA (scramble) (lanes 1–3), RMI2 siRNA (lanes 4–6), or RMI1
siRNA (lanes 7–9) on the distribution of BLM, Topo III?, RMI1,
and RMI2 to various fractions. Cells were treated with HU prior
to fractionation, and H2A serves as a marker for the chromatin
fraction. Asterisk (*) denotes a nonspecific crossreactive band
detected by anti-RMI1 antibodies.
RMI2 is essential for the chromatin targeting of BLM
Singh et al.
2860 GENES & DEVELOPMENT
Figure 4D, the assembly of HU-induced BLM foci in
HEK293 cells is similarly dependent on RMI2.
Purification of the RMI2/RMI1 complex
In order to elucidate the biochemical function of RMI2,
we attempted to purify N-terminally Flag-tagged RMI2
from bacterial cells. However, the resulting protein ap-
peared aggregated when analyzed by gel filtration. It also
did not interact with BLM, Topo III?, or RMI1 in our in
vitro pull-down assays (data not shown). Importantly,
upon coexpression of Flag-tagged RMI2 and MPB-tagged
RMI1 in bacteria, a stable complex of the two proteins
could be obtained. We could purify the RMI2/RMI1 com-
plex to a nearly homogeneous state using affinity chro-
matographic steps on anti-Flag and amylose matrices, as
described in the Materials and Methods. The purified
preparation consisted of a monodispersed, nonaggregated
species, as determined by sizing in a Sephracryl S200
gel filtration column (data not shown). The resulting
protein complex (Fig. 5A) appears to harbor stoichiomet-
ric amounts of RMI1 and RMI2.
We previously found that the N-terminal portion of
RMI1 (residues 1–211) was responsible for binding both
BLM and TopoIII?. Moreover, the RMI1-N fragment
stimulates the dHJ dissolution reaction as well as
the full-length protein (Raynard et al. 2008). In order
to determine which portion of RMI1 harbors the RMI2
interaction domain, we expressed the RMI1-N, RMI1-M
(residues 212–424) and RMI1-C (residues 425–625) frag-
ments (Raynard et al. 2008) in Escherichia coli, either
alone or with Flag-tagged RMI2. The cell lysates con-
taining these proteins were incubated with anti-Flag aga-
rose and the eluted proteins analyzed by SDS-PAGE
and Coomassie Blue staining. As expected, none of the
RMI1 fragments was retained on the anti-Flag beads
when expressed alone (Supplemental Fig. 3A). Upon
coexpression with Flag-RMI2, a near stoichiometric
amount of the RMI1 C-terminal fragment, but little if
any of the RMI1 N or M fragment, became associated
with RMI2 on the anti-Flag beads (Supplemental Fig. 3B).
We could purify the RMI1-C/RMI2 complexwith the
(Fig. 5B). This finding indicates that RMI1 interacts
protein and by BLM following exposure to 2 mM HU for 24 h were detected by immunofluorescence microscopy using anti-Flag and
anti-BLM antibodies, respectively. HF-RMI2 foci were detected in red, while BLM foci were detected in green. A merged image shows
colocalizing foci. The position of nuclei is indicated by a counterstain with DAPI (blue). Examples are shown for an untreated
population of cells (top panels) and for cells treated with 1 mM HU for 24 h (bottom panels). (B) Examples of BLM foci in HeLa cells
transfected with a scrambled control siRNA or with a siRNA that targeted RMI2 are shown. Cells were treated with 2 mM HU for 24
h prior to fixation and preparation for immunofluorescence microscopy. Bar, 10 µm. (C,D) Quantification of the assembly of BLM foci
in HeLa (C) and HEK293 (D) cells. Cells were transfected with a scrambled control siRNA or with a siRNA that targeted RMI2 and
were fixed for immunofluorescence microscopy at 96 h after transfection. Cells were either left untreated or were exposed to 2 mM
HU for 24 h. The mean percentage of cells with five or more BLM foci from three counts of 150 or more cells each is shown with the
standard deviation. Statistical significance (P < 0.01) using a Z-test for two sample proportions is indicated by an asterisk (*).
RMI2 regulates the assembly of BLM nuclear foci. (A) HF-RMI2 was expressed in HeLa cells. Foci assembled by this fusion
BLAP18/RMI2, a novel component of the BTB complex
GENES & DEVELOPMENT 2861
with BLM, TopoIII?, and RMI2 through different epi-
RMI2/RMI1 complex interacts with BLM and Topo
III? in vitro
We used affinity pull-down assays to confirm that our
purified RMI2/RMI1 complex interacts with BLM and
Topo III?. When amylose resin was used to pull down
the RMI2–RMI1 complex through the maltose-binding
protein (MBP) tag on RMI1, purified BLM and Topo III?
were retained on the affinity matrix (Fig. 5C, lanes 3,6).
Similar results were obtained when anti-Flag agarose
was used to pull down the RMI2–RMI1 complex through
the Flag tag on RMI2 (Fig. 5D, lanes 3,6). Neither BLM
nor Topo III? bound to amylose resin or anti-Flag agarose
in the absence of the RMI2–RMI1 complex (Fig. 5C,D,
Enhancement of dHJ dissolution by RMI2
We wished to determine whether RMI2 affects the dHJ
dissolution reaction mediated by the BTB complex. For
this, a radiolabeled dHJ substrate, constructed as de-
scribed (Fu et al. 1994; Wu and Hickson 2003; Raynard et
al. 2006), was incubated with the combination of BLM,
Topo III?, and either RMI1 or RMI2–RMI1, and the dis-
solution products were resolved in a polyacrylamide gel
under denaturing conditions and then visualized by
phosphorimaging analysis (Wu and Hickson 2003; Ray-
nard et al. 2006). As expected, time course experiments
showed an enhancement of dHJ dissolution by RMI1,
with ∼40% and 65% of dissolution in 3 and 9 min, re-
spectively, compared with 18% and 24% by BLM–Topo
III? in the same time frame (Fig. 5E,F). Importantly, in-
clusion of RMI2 caused a further increase in dHJ disso-
lution catalyzed by the BTB complex, with around ∼52%
and >75% of the dHJ substrate resolved at 3 and 9 min,
respectively (Fig. 5E,F). The same level of stimulation of
dHJ dissolution by RMI2 was observed at higher KCl
concentrations (Fig. 5G).
RMI2 mutants deficient in BTB complex assembly
A highly conserved lysine residue of RMI1 (K166A) has
been shown to be required for the interaction of RMI1
with Topo III? (Raynard et al. 2008). In order to map the
residues that are important for RMI2 function and inter-
action with BTB complex, we used site-directed muta-
genesis to change conserved residues in RMI2 (Supple-
mental Fig. 1). Specifically, we substituted the following
highly conserved residues in HF-RMI2 with alanine: lys-
24 (K24A), trp-59 (W59A), lys-100 (K100A), lys-121
(K121A), and trp-135 (W135A). Since a RMI2 knockout
cell line is not available, we established a siRNA-based
complementation system using HeLa cells that specifi-
of the BTB complex. (A) Purified RMI2/RMI1 com-
plex, 3 µg, was analyzed by Coomassie Blue-stained
SDS-PAGE. (B) Purified RMI2/RMI1-(C) complex, 3
µg, was resolved by SDS-PAGE and stained with
Coomassie Blue. (C) BLM (5 µg) or Topo III? (5 µg)
was incubated with or without RMI2/RMI1 com-
plex (5 µg), and protein complexes were captured on
amylose resin, which was washed and treated with
SDS to elute bound proteins. The supernatant (S)
that contained unbound proteins, wash (W), and SDS
eluate (E) were analyzed by SDS-PAGE. (D) Experi-
ment performed as in C, but protein complexes were
captured on anti-Flag agarose. (*) IgG light chain dis-
sociated from anti-Flag M2 agarose. (E) Time course
of dHJ dissolution by combinations of BLM–Topo
III?, RMI1, and RMI2. (BTB) BLM, Topo III?, and
RMI1 combined. (F) Time course of dHJ dissolution
by combinations of BLM–Topo III?, RMI1, and
RMI2 from five independent experiments presented
in the graph. The average values ± SEM. (G) dHJ dis-
solution by the BTB complex with or without RMI2
as a function of KCl concentration. The histogram
shows the average levels of dissolution ± SEM. from
four independent experiments.
RMI2 stimulates dHJ dissolution activity
Singh et al.
2862 GENES & DEVELOPMENT
cally knock down endogenous RMI2 but permit the ex-
pression of exogenous RMI2. HeLa cells were transduced
with retroviruses expressing wild type or one of the
RMI2 variants and sorted for EGFP-positive cells by
FACS. Then, the cells were transfected with siRNA oli-
gos targeting the 3? untranslated region of RMI2
(siRMI2-3?UTR). Since the exogenous RMI2 construct
lacks the 3?UTR, this siRNA should deplete only the
endogenous RMI2. As shown in Figure 6A, transfection
of HeLa cells with siRMI2-3?UTR knocked down endog-
enous RMI2, thereby destabilizing Topo III? and RMI1.
As expected, siRMI2-3?UTR did not affect the level of
ectopically expressed HF-RMI2, and both Topo III? and
RMI1 were protected from degradation by HF-RMI2.
Immunoblot analysis confirmed the expression of the
RMI2 mutants, with the K24A, W59A, and W135A, be-
ing present at a lower level (Fig. 6B). Since the amount of
EGFP coexpressed with the RMI2 mutants from the
same bicistronic transcript appeared to be the same in
each case, we attributed the lower steady level of some of
the RMI2 mutants to reduced protein stability (Fig. 6B).
We then tested the RMI2 variants for their ability to
associate with BLM, Topo III?, and RMI1 in HeLa or
HEK293 cells by immunoprecipitation. As shown in Fig-
ure 6, C and D, the RMI2-K100A and RMI2-K121A vari-
ants were proficient in interaction with BLM, Topo III?,
and RMI1. Interestingly, RMI2-K24A, RMI2-W59A and
RMI2-W135A variants were unable to interact with the
other BTB complex components. Next, we examined
whether variants that were deficient in binding the other
BTB components could protect RMI1 and Topo III?
against proteolysis in the absence of endogenous RMI2.
We used the siRNA-based complementation system de-
scribed above to address this question. As shown in Fig-
ure 6E, unlike wild-type RMI2 and the K100A variant,
the K24A, W59A, and W135A mutants did not prevent
the degradation of RMI1 and Topo III?. However, we
cannot rule out the possibility that the lack of associa-
tion of the K24A, W59A, and W135A mutants with the
BTB complex stems from their reduced steady state level
in cells rather than an inability to interact with the BTB
Cancer predisposition is a major characteristic of BS,
which arises likely because of the elevated genomic in-
stability in the cells of patients. The BLM protein has
been shown to interact with many other proteins in-
volved in DNA damage repair and response (Brosh et al.
2000; Wang et al. 2000, 2001; Langland et al. 2001;
Meetei et al. 2003; Yin et al. 2005), and consistent with
this, BS cells are hypersensitive to DNA damaging
agents, such as MMS (German and Ellis 2002). This phe-
notype and the increased rate of SCEs in BS cells (Ger-
man 1969; German et al. 1977b) are consistent with the
notion that BLM mediates its genome preservation func-
tion via influencing DNA damage repair. In this regard, a
thorough understanding of the partner proteins of BLM is
essential for deciphering the mechanism of the BLM-
dependent genome maintenance pathway.
In this study, we used an affinity purification approach
(Ling et al. 2007) to isolate a BLM-containing protein
complex that harbors a novel protein of 18 kDa molecu-
lar mass, which we refer to as RMI2. Previous work has
shown that BLM forms a complex, termed the BTB com-
plex, with Topo III? and RMI1 (Yin et al. 2005; Raynard
et al. 2006; Wu et al. 2006). Here we presented extensive
biochemical and cytological evidence that RMI2 is also a
critical component of the BTB complex and, as discussed
that bind inefficiently to the BTB complex are unstable. (A,
lanes 2,4) Immunoblot showing that siRMI2-3?UTR specifically
depletes endogenous RMI2 and not the exogenously expressed
HF-RMI2. (Lanes 1,3) Scrambled oligos were used as a control.
(B) Immunoblot showing that RMI2 mutants K24A, W59A, and
W135A were unstable, as compared with wild-type (WT) RMI2
or the K100A or K121A mutants. GFP was used as an internal
control to show that mRNA expression is comparable in all the
cells stably expressing HF-RMI2 variants. Immunoprecipitation
coupled with immunoblotting to assay for the binding of the
various variants to the BTB complex in HeLa cells (C) and
HEK293 cells (D). (E) Immunoblot showing the levels of BTB
members in HeLa cells depleted of the endogenous RMI2. Note
that when the endogenous RMI2 was knocked down using
siRMI2-3?UTR, HF-RMI2 mutants (K24A, W59A, and W135A),
which could not bind to BTB complex, were unable to protect
Topo III? and RMI1 from degradation (lanes 4,6,10), while wild
type or the K100A could rescue Topo III? and RMI1 from deg-
radation (lanes 2,8). Asterisk (*) denotes nonspecific crossreac-
tive bands detected by anti-Topo III? and anti-RMI1 antibodies.
Point mutations in the conserved residues of RMI2
BLAP18/RMI2, a novel component of the BTB complex
GENES & DEVELOPMENT 2863
further below, influences the function of the BTB com-
plex in various ways.
RMI2 is a novel component of the BTB complex
Our studies revealed remarkable interdependence among
Topo III?, RMI2, and RMI1, such that depletion of any
one of these proteins leads to the destabilization of the
others. Interestingly, the cellular level of these proteins
is not affected by the BLM status, and their association
occurs in the absence of BLM. Furtheremore, we found
that RMI2 colocalizes with BLM in nuclear foci. The
results from our cytological and cell extract fraction ex-
periments showed that the complex of Topo III?, RMI2,
and RMI1 is required for the efficient targeting of BLM
to chromatin and to nuclear foci. These nuclear foci
form both spontaneously and in response to replication
stress induced by HU. The observation that there is
an increased interaction of BLM with the Topo III?–
RMI1–RMI2 complex in chromatin in cells treated with
a replication blocker or DNA interstrand cross-linker
strengthens the hypothesis that BLM function depends
upon the Topo III?–RMI1–RMI2 complex.
The other proteins that were isolated with the RMI2
complex as determined by MS analysis and clearly vis-
ible on the silver-stained SDS-PAGE gel are BLAP300,
FANCM, and BLAP15. These proteins may regulate the
function of the BTB–RMI2 complex. Some other proteins
were not visible as a clear band on the gel, but they were
identified in the MS analysis of the purified RMI2-asso-
ciated complex. These proteins included MSH2, MSH6,
and RFC, which are part of the BASC complex (Wang et
al. 2000), as well as all the three subunits of RPA. Un-
derstanding the functional interaction of these proteins
with the BTB–RMI2 complex will be an important focus
of future studies. The presence of FANCM and its part-
ner FAAP24 in the RMI2 complex suggests that RMI2 is
also a component of the BRAFT complex that contains
BLM and components of the Fanconi anemia core com-
plex (Meetei et al. 2003).
Biochemical functions of the RMI2 protein
RMI2 is unique to higher eukaryotes and contains a pu-
tative OB-fold domain. Even though OB-folds are gener-
ally assumed to be DNA-binding modules, we previously
provided evidence that the OB-fold located within the
N-terminal region of the RMI1 protein is not responsible
for DNA binding but and may in fact be involved in
mediating the interaction of RMI1 with BLM and Topo
III?. In this study, by mutating highly conserved residues
within the putative OB-fold of RMI2, three mutants—
K24A, W59A and W135A—that fail to associate with the
BTB complex, have been identified. It thus seems likely
that the RMI2 OB-fold is involved in protein–protein in-
teractions with RMI1 and possibly Topo III?. Future
studies will address whether RMI2 has DNA-binding ac-
tivity and the relevance of its putative OB-fold in such an
We provided evidence that RMI2 enhances the dHJ
dissolvase activity of the BTB complex. Based on our
previous finding that complex formation with Topo III?
as being indispensable for the ability of RMI1 to function
as a stimulatory factor of the dHJ dissolution reaction, it
seems reasonable to suggest that the dHJ dissolvase
stimulatory attribute of RMI2 is likewise reliant on its
interactions with RMI1 and possibly with Topo III? as
Like BLM, RMI2 also undergoes mitotic phosphoryla-
tion. It will be interesting to determine whether phos-
phorylation of BLM and RMI2 is mediated by the same
protein kinase(s) and to ascertain the functional signifi-
cance of phosphorylation. In this regard, since it has been
demonstrated previously that ATM, ATR, and MPS1
phosphorylate BLM (Davies et al. 2004; Leng et al. 2006),
it will be important to test the relevance of these kinases
in RMI2 phosphorylation.
The RMI2-binding domain in RMI1
We showed previously that RMI1 interacts with BLM
and Topo III? through its N-terminal region, likely its
OB-fold (Raynard et al. 2008). In contrast, we demon-
strated that RMI1 interacts with RMI2 through its C-
terminal region. Comparison of the RMI1 protein and its
presumptive Saccharomyces cerevisiae ortholog Rmi1
protein shows that the middle and C-terminal of RMI1
are absent from Rmi1 (Raynard et al. 2008). It has been
suggested that this region has evolved in higher eukary-
otes to perform specific cellular functions and may me-
diate interaction with other DNA repair proteins (Ray-
nard et al. 2008). The observations that RMI2 is absent in
yeast and that the RMI1 C terminus mediates RMI2 in-
teraction is certainly in congruence with this particular
There is compelling evidence linking the BLM protein to
tumor suppression. First, BS patients exhibit a marked
predisposition to various cancers (German et al. 1977a;
German 1997). Second, mutation of BLM is found in a
variety of tumors in the general population (Gruber et al.
2002). Finally, heterozygosity for BLM increases the in-
cidence of cancer in mice (Goss et al. 2002). Given that
depletion of RMI2 results in DNA damage sensitivity
and chromosomal instability in the form of elevated lev-
els of chromosomal gaps and breaks, it will be particu-
larly interesting to determine whether RMI2 and the
other BTB components, including RMI1 and Topo III?,
also function as tumor suppressors.
Materials and methods
HeLa, HT1080, HEK293, and U2OS cells were cultured in
DMEM medium (Invitrogen) supplemented with 10% FBS (At-
Singh et al.
2864GENES & DEVELOPMENT
lanta Biologicals) and 1× penicillin and streptomycin solution
(Invitrogen) in a humidified atmosphere of 5% CO2at 37°C.
MMS was purchased from Sigma. A stock solution was diluted
to 2000 µM with DMSO. Nocodazole and taxol (Sigma) were
resuspended in DMSO to a stock concentration of 1 mg/mL and
10 mM, respectively. HU (Sigma) was resuspended in water, to
a stock concentration of 1 M. Mitomycin C (MMC; Sigma) was
dissolved in 70% ethanol (Sigma) to a stock concentration of
Protein knockdown by siRNA treatment
All siRNA oligos were purchased from Dharmacon. For the
knockdown of RMI2, we designed an siRNA in the 3?UTR
(3?UTR1C: 5?-UGUUGGAACUGUCGUUAAAUU-3?), and we
also used ON-TARGETplus SMARTpool siRNA (catalog no.
L-015684-01, human C16orf75, NM_152308). The sequence of
the siRNA for RMI1 is described in Yin et al (2005). ON-TAR-
GETplus SMARTpool siRNAs were also used for the knock-
downof BLM(catalog no.
NM_000057) and Topo III? (catalog no. L-005279-00, human
Topo III?, NM_004618). A nonspecific control siRNA (catalog
no. D-001210-01) was used in all experiments. Cells were trans-
fected with siRNA using lipofectamine2000 for 5 h in reduced
serum OptiMEM medium, as recommended by the manufac-
turer (Invitrogen). After 5 h, OptiMEM was removed and re-
placed by complete DMEM medium. Cells were harvested 4 d
post-transfection and analyzed by immunoblotting.
Different orthologs of RMI2 were identified by searching pro-
tein sequence databases available at NCBI using the BLAST
program. Multiple sequence analysis of the various orthologs of
RMI2 was carried out using the multiple sequence alignment
program ClustalW 1.8 available at the BCM search launcher
(http://searchlauncher.bcm.tmc.edu). Shading of the multiple
alignments was carried out using the BOXSHADE 3.21 program
(EMBnet server; http://www.ch.embnet.org/software/BOX_form.
html). Putative domains were identified by searching the Inter-
Pro database using the InterProScan tool (http://www.ebi.ac.uk/
Tools/InterProScan). Secondary structure (?-sheets) characteris-
tic of OB-fold domains, was identified using PSIPRED, a highly
accurate method for protein secondary structure prediction
Two rabbit RMI2 polyclonal antibodies were raised against fu-
sion proteins containing full-length RMI2 with either MBP (An-
tibody A) or glutathione-S-transferase (GST) (Antibody B) and
affinity-purified. These fusion proteins were expressed and pu-
rified from E. coli in accordance with the manufacturer’s pro-
tocols. A polyclonal antibody against BLM (69D) has been de-
scribed elsewhere (Meetei et al. 2003), and Topo III? and RMI1
antibodies were described elsewhere (Yin et al. 2005). A goat
polyclonal anti-?-actin antibody and a rabbit anti-GFP antibody
were from Santa Cruz Biotechnologies. Anti-histone H2A anti-
body was purchased from Millipore.
Construction of double-tagged BLM and RMI2 plasmids
The pMIEG3 retroviral vector described in this study is a kind
gift from Dr. David Williams (Cincinnati Children’s Hospital,
Cincinnati, OH), and has been described earlier (Chandra et al.
2005; Ling et al. 2007). In order to generate a double-tagged BLM
construct, we obtained a BLM cDNA tagged with a (His)6affin-
ity epitope at its C terminus (Raynard et al. 2006). N-terminal
Flag was introduced by site-directed mutagenesis using PCR
and cloned into pMIEG3 vector to generate pMIEG3-F-BLM-H.
Human RMI2 cDNA was obtained from Open Biosystems.
RMI2 ORF was PCR-amplified with high-fidelity pfu Polymer-
ase (Stratagene) with an N-terminal (His)6-Flag tag. This product
was cloned into EcoR1 and Xho1 sites of the pMIEG3 vector to
generate pMIEG3-HF-RMI2. Next, amphotropic retroviruses
were created and used to infect the target cells. To prepare tran-
sient virus stocks, 1.5 × 106293T cells were plated in 10-cm
dishes. The next day, cells were cotransfected using lipofect-
amine2000 with the retroviral expression vectors as described
above, together with the appropriate helper plasmid (gag-pol and
RD114). The medium was changed 5 h post-transfection, and
retrovirus-containing medium was harvested in 12-h incre-
ments at 24 h post-transfection.
Generation of stable cell lines
Cells were seeded in six-well plates at the density of 5 × 104
cells per well in 3 mL of complete medium. The next day, cells
were transduced in the presence of 8 ng/mL polybrene (Sigma
Aldrich). The plates were then spun at 1000g for 1 h, transferred
to a humidified incubator (5% CO2), and cultured for 17 hat
37°C. Cells were washed twice to remove polybrene and resus-
pended in complete DMEM medium (GIBCO-BRL). After cul-
ture for 48–72 h, the EGFP-positive cells were sorted using a
Becton Dickinson FACS Vantage SE instrument (Becton Dick-
inson Immunocytometry Systems).
Purification of BLM and RMI2 complexes
BLM and RMI2 complexes were isolated from HT1080-F-
BLM-H and HeLa S3-HF-RMI2 nuclear extracts, respectively, by
using a modified two-step affinity chromatography protocol de-
scribed previously (Ling et al. 2007). Briefly, 50 150-mm-diam-
eter plates of cells expressing F-BLM-H or HF-RMI2 were grown
to 85%–95% confluency. Cells were washed with phosphate-
buffered saline and collected as a pellet. Nuclear extracts were
prepared as described (Meetei et al. 2003). The first purification
step was performed with anti-Flag M2 agarose beads (Sigma).
Multiple tubes containing 1.5 mL each of the nuclear extracts
were incubated with 25 µL anti-Flag M2 agarose beads over-
night, washed four times with 250-lysis buffer (20 mM HEPES
at pH 7.9, 250 mM NaCl, 0.2 DTT, 0.5 mM sodium orthovana-
date, 50 mM sodium fluoride, protease inhibitor cocktail, 2 mM
PMSF, 1% Triton X-100, 10% glycerol) with 10 min rotation at
4°C, and eluted with 3XFlag peptide (Sigma) for 1 h on ice. The
second purification step was performed by incubating the eluate
of the first step with 30 µL Talon metal affinity Resin (BD) for
3 h at 4°C in the presence of 3 mM imidazole. Then the beads
were washed four times with 250-lysis buffer containing 10 mM
imidazole, and eluted with 50 mM EDTA. The purified protein
complex was resolved on an 8%–16% SDS-PAGE gel and
stained by either SilverQuest silver or colloidal blue staining kit
according to the manufacturer’s instructions (Invitrogen). Pro-
tein identification by MS was carried out by using either excised
gel pieces or the entire gel, as described previously (Meetei et al.
2003). Actual MS data are not shown but are available upon
For immunodepletion, the total cell lysate of HeLa S3 cells
stably expressing HF-RMI2 was incubated with 50 µL of anti-
Flag M2 agarose beads overnight, then washed four times with
BLAP18/RMI2, a novel component of the BTB complex
GENES & DEVELOPMENT2865
250-lysis buffer and eluted with 3XFlag peptide (Sigma) for 1 h
on ice. CoIP experiments using anti BLAP-18 and anti-BLM an-
tibodies were carried out as described (Meetei et al. 2003).
Protein phosphatase treatment
We treated HeLa cells expressing HF-RMI2 with taxol or noco-
dazol for 16 h or left the cells untreated. Cells were prepared
with lysis buffer, then HF-RMI2 was immunoprecipitated with
anti-M2 agarose. Flag immunoprecipitates were incubated at
30°C with 400 U of ?-protein phosphatase, either in the pres-
ence or absence of phosphatase inhibitors, for 60 min prior to
immunoblot analysis according to the manufacturer’s instruc-
tions (New England Biolabs). For phosphatase inhibition, a com-
bination of 10 mM sodium orthovanadate and 50 mM sodium
fluoride was added to the protein samples prior to the addition
of the ?-protein phosphatase.
Preparation of cellular fractions
Cells were treated with HU (1.5 mM) for 16 h Then the cells
were trypsinized and washed with cold phosphate-buffered sa-
line (PBS). The pellets were resuspended in cold buffer A (10
mM PIPES at pH 7.0, 100 mM NaCl, 1 mM EGTA, 300 mM
sucrose, 0.5 mM sodium orthovanadate, 50 mM sodium fluo-
ride, protease inhibitor cocktail, 1 mM PMSF, 0.5% Triton
X-100) at five times the volume of the cell pellet and were
incubated for 2 min at room temperature to permeabilize the
cells. The suspension was then centrifuged at 200g for 3 min,
and the supernatant (S100; detergent-soluble nuclear proteins)
was collected. Pellets were washed with cold buffer A. The pel-
let was then extracted with buffer B (buffer A with the NaCl
concentration increased to 300 mM) for 5 min. The supernatant
was collected (S300). The pellet was washed once with buffer A
and then boiled with 2× SDS buffer (P300).
Purification of the RMI2/RMI1 complex
All purification steps were conducted at 0 to 4°C, and the elu-
tion of the RMI2/RMI1 protein complex was monitored by 12%
SDS-PAGE and Coomassie Blue staining. For purification, E.
coli BL21:DE3 Rosetta cells were transformed with pRSF-Duet
vector (Invitrogen) carrying Flag-RMI2 cDNA and pMAL-p2X
cDNA. Cells were grown at 37°C to OD600= 0.8, and the ex-
pression of proteins was induced by the addition of 0.1 mM
IPTG at 16°C for 16 h. Bacterial pellets (30 g) were suspended in
100 mL of Buffer K (20 mM KH2PO4, 10% glycerol, 0.5 mM
EDTA, 0.01% IGEPAL, 1 mM DTT, 1 mM phenylmethylsulfo-
nyl fluoride, 0.5 mM benzamidine, and 5 µg/mL each of aproti-
nin, chymostatin, leupeptin, pepstatin) containing 150 mM
KCl. The cells were disrupted by sonication, and the crude ly-
sate was subjected to ultracentrifugation (100,000g for 90 min).
The clarified lysate was incubated with 5 mL of anti-Flag M2
agarose (Sigma) for 4 h with constant agitation. The matrix was
poured into a column with an internal diameter of 1 cm and was
washed with 100 mL of Buffer K with 150 mM KCl. The bound
proteins were eluted in 5 mL of PBS containing 250 ng/µL of
Flag peptide (Sigma). The elutions were repeated five times, 30
min each with constant agitation. Fractions containing RMI2/
RMI1 were incubated with 3 mL Amylose beads (New England
Biolabs) for 3 h with constant agitation. The matrix was washed
as above, and proteins were eluted with 15 mL of 10 mM malt-
ose in Buffer K with 150 mM KCl. Fractions containing RMI2/
RMI1 complexes were combined and concentrated in an Ami-
con Ultra microconcentrator (Millipore) to 600 µL and then sub-
jected to gel filtra-
tion in a 35 mL Sephacryl S200 column in Buffer K with 150
mM KCl. The size of collected fractions was 1 mL. Fractions
containing the RMI2/RMI1 peak were combined and concen-
trated as before to 6 mg/mL (RMI1) and stored in small portions
at −80°C. The concentration of the proteins was determined by
densitometric comparison of multiple loadings of the purified
proteins against known amounts of bovine serum albumin in a
Coomassie Blue-stained polyacrylamide gel. The identities of
the proteins were determined by Western blotting with ?-Flag
(Sigma) and ?-MBP (New England Biolabs) antibodies.
Purification of RMI2/RMI1(C) complex
MBP-RMI1(N)-(His)6, MBP-RMI1(M)-(His)6, and MBP-RMI1(C)-
(His)6(Raynard et al. 2008) were each expressed alone or coex-
pressed with Flag-RMI2 in the E. coli cells and the cell lysates
were prepared and incubated with anti-Flag M2 agarose as
above. A small portion of anti-Flag M2 agarose (25 µL bead bed)
was taken from each sample and bound proteins were eluted
with 2% SDS and analyzed by 12% SDS-PAGE and Coomassie
Blue staining. Only MBP-RMI1(C)-(His)6was found in the com-
plex with RMI2 and the complex was purified as above.
Purification of RMI1 and other proteins
MBP-RMI1-(His)6was expressed from a modified pMAL-p2X
vector (New England Biolabs), as described above, and purified
using our published procedure devised for GST-RMI1-(His)6
(Raynard et al. 2006) with the following modifications: amylose
resin (New England Biolabs) was substituted for glutathione-
Sepharose (GE Healthcare) and elution of proteins bound to the
resin was done with 10 mM maltose in K buffer containing 150
mM KCl. Recombinant (His)6-tagged BLM and (His)6-tagged
Topo III? were expressed in yeast and E. coli, respectively, and
purified following our published procedures (Bussen et al. 2007).
The concentration of these proteins was determined as above.
The dHJ substrate was prepared by hybridizing and ligating two
partially complementary oligonucleotides, as described (Fu et
al. 1994; Wu and Hickson 2003; Raynard et al. 2006).
dHJ dissolution assay
BLM (7.5 nM) was incubated with Topo III? (120 nM) and RMI1
or RMI1/RMI2 (50 nM) for 10 min on ice in 24 µL of reaction
buffer (50 mM Tris-HCl at pH 7.8, 1 mM DTT, 0.8 mM MgCl2,
200 µg/mL bovine serum albumin, 2 mM ATP, 80 mM KCl [or
amount as indicated], an ATP regenerating system consisting of
10 mM creatine phosphate and 50 µg/mL creatine kinase), fol-
lowed by the addition of dHJ substrate (1.2 nM) in 1 µL of water.
The reactions were incubated at 37°C, and after 5 min, or time
as indicated, 8-µL aliquots of the reaction mixtures were re-
moved and mixed with 2 µL of 1% SDS and 0.5 µL of proteinase
K (10 µg/µL stock), followed by a 3-min incubation at 37°C. The
deproteinized reactions were mixed with an equal volume of
sample loading buffer (20 mM Tris-HCl at pH 7.5, 50% glycerol,
and 0.08% Orange G) containing 50% urea, incubated at 95°C
for 3 min, and resolved in an 8% polyacrylamide gel containing
20% formamide and 8 M urea in TAE buffer at 55°C.
Affinity pull-down assays
For pull-down assays, 5 µg of BLM or Topo III? were incubated
in 30 µL of Buffer K containing 200 mM KCl with or without
RMI2/RMI1 complex (5 µg of RMI1) for 30 min at 4°C. The
Singh et al.
2866 GENES & DEVELOPMENT
reactions were mixed with 15 µL of Amylose beads (which rec-
ognize the MBP tag at the N terminus of RMI1) (Fig 5B) or 20 µL
Flag agarose (which binds Flag epitope at the N terminus of
RMI2) (Fig 5C) for 30 min at 4°C. After washing the beads twice
with 200 µL of the same buffer, bound proteins were eluted with
25 µL of 2% SDS. Fifteen percent of total supernatant (S) and
elution (E) fractions, and 2% of total wash (W) fraction were
analyzed by 4%–15% gradient SDS-PAGE and Coomassie Blue
Adherent cells were grown on 12-mm-diameter glass coverslips,
which were coated with poly-D-lysine, for a minimum of 24 h
prior to treatment or fixation. Cells were left untreated or were
exposed to 2 mM HU for 24 h. For depletion, RMI2 cells were
transfected with siRMI2 or the scrambled control and treated
with HU beginning at 72 h after transfection.
Cells were fixed with 2% paraformaldelyde for 20 min at
37°C. Cells were then washed with PBS, permeabilized 3 min
with 0.2% Triton X-100 in PBS, and washed again with PBS.
Cells were incubated 1 h at 37°C with primary antibodies di-
luted in antibody buffer (PBS/3% bovine serum albumin/0.05%
Tween 20/0.04% sodium azide). Primary antibodies included
mouse anti-Flag (M2; 1:100; Sigma) and goat anti-BLM (1:200;
Santa Cruz Biotechnologies) antibodies. After incubation with
primary antibodies, cells were washed with PBS (3×, 5 min each)
and then incubated with the appropriate secondary antibody
(Jackson Immunoresearch; 1:500) for 30 min at 37°C. Secondary
antibodies included FITC-conjugated donkey anti-goat IgG and
Rhodamine B-conjugated donkey anti-mouse IgG diluted in an-
tibody buffer. Cells were washed three times with PBS and
mounted over Vectashield containing DAPI (Vector Laborato-
ries) to stain DNA, sealed with nail polish, and stored at −20°C
Labeled cells were observed with a Zeiss Axiovert 200M mi-
croscope, and images were collected with a Hamamatsu Camera
using Openlab software (Improvision). Images were processed
into figures using Photoshop (Adobe).
MMS sensitivity assay
HEK293 cells were transfected with siRNA targeting RMI2,
RMI1, BLM, or control oligos, as described above. At 3 d post-
transfection, 200 cells were seeded per 10-cm dish containing
the indicated concentration of MMS in DMEM medium. After
10 d cells were fixed, stained, and visible colonies were counted.
Detection of chromosome gaps/breaks
Analysis of chromosome gaps/breaks was carried out at the cy-
togenetic facility at Cincinnati Children’s Hospital, using a
standard protocol. Three days post-transfection with siRNA,
HEK 293 cells were cultured in colcemid for 2.5 h. The cells
were harvested and treated with hypotonic solution containing
0.35% Sodium citrate and 0.28% KCl for 30 min at room tem-
perature. Cells were then fixed with methanol/acetic acid (3:1)
for 60 min. The cell suspension was dropped onto ice-cold wet
glass slides and air dried. The cells were stained with Giemsa
solution for 1.5 min and examined by light microscopy.
We thank the Mass Spectrometry and Proteomics Facility at
Ohio State University and the Taplin Biological Mass Spectrom-
etry Facility at Harvard Medical School for MS analysis. We
thank Drs. Weidong Wang and David A Williams for reagents.
We thank Dr. Teresa Smolarek, Audra Birri, and Jenny Coffman
of the Cytogenetics Laboratory at Cincinnati Children’s Hospi-
tal Medical Center for chromosome gaps/breakage analyses. We
also thank Wahengbam Kebola Devi for technical assistance.
This work was supported by National Institutes of Health Re-
search Grants RO1 HL085587 (to P.R.A.), ES015632 (to P.S.),
and R01 HL084082 (to A.R.M.), and an American Society of
Hematology Junior Faculty Award (to A.R.M). V.B. was sup-
ported by Ruth L. Kirschtein post-doctoral fellowship F32
Bischof, O., Galande, S., Farzaneh, F., Kohwi-Shigematsu, T.,
and Campisi, J. 2001. Selective cleavage of BLM, the Bloom
syndrome protein, during apoptotic cell death. J. Biol. Chem.
Brosh Jr., R.M., Li, J.L., Kenny, M.K., Karow, J.K., Cooper, M.P.,
Kureekattil, R.P., Hickson, I.D., and Bohr, V.A. 2000. Repli-
cation protein A physically interacts with the Bloom’s syn-
drome protein and stimulates its helicase activity. J. Biol.
Chem. 275: 23500–23508.
Bussen, W., Raynard, S., Busygina, V., Singh, A.K., and Sung, P.
2007. Holliday junction processing activity of the BLM–
Topo III?–BLAP75 complex. J. Biol. Chem. 282: 31484–
Chandra, S., Levran, O., Jurickova, I., Maas, C., Kapur, R., Schin-
dler, D., Henry, R., Milton, K., Batish, S.D., Cancelas, J.A., et
al. 2005. A rapid method for retrovirus-mediated identifica-
tion of complementation groups in Fanconi anemia patients.
Mol. Ther. 12: 976–984.
Davalos, A.R. and Campisi, J. 2003. Bloom syndrome cells un-
dergo p53-dependent apoptosis and delayed assembly of
BRCA1 and NBS1 repair complexes at stalled replication
forks. J. Cell Biol. 162: 1197–1209.
Davies, S.L., North, P.S., Dart, A., Lakin, N.D., and Hickson,
I.D. 2004. Phosphorylation of the Bloom’s syndrome heli-
case and its role in recovery from S-phase arrest. Mol. Cell.
Biol. 24: 1279–1291.
Dutertre, S., Ababou, M., Onclercq, R., Delic, J., Chatton, B.,
Jaulin, C., and Amor-Gueret, M. 2000. Cell cycle regulation
of the endogenous wild type Bloom’s syndrome DNA heli-
case. Oncogene 19: 2731–2738.
Ellis, N.A., Groden, J., Ye, T.Z., Straughen, J., Lennon, D.J.,
Ciocci, S., Proytcheva, M., and German, J. 1995. The
Bloom’s syndrome gene product is homologous to RecQ he-
licases. Cell 83: 655–666.
Fu, T.J., Tse-Dinh, Y.C., and Seeman, N.C. 1994. Holliday junc-
tion crossover topology. J. Mol. Biol. 236: 91–105.
German, J. 1969. Bloom’s syndrome. I. Genetical and clinical
observations in the first twenty-seven patients. Am. J. Hum.
Genet. 21: 196–227.
German, J. 1997. Bloom’s syndrome. XX. The first 100 cancers.
Cancer Genet. Cytogenet. 93: 100–106.
German, J. and Ellis, N.A. 2002. Bloom syndrome. In The ge-
netic basis of human cancer (eds. B. Vogelstein and K.W.
Kinzler), pp. 267–288. McGraw-Hill, New York.
German, J., Bloom, D., and Passarge, E. 1977a. Bloom’s syn-
drome. V. Surveillance for cancer in affected families. Clin.
Genet. 12: 162–168.
German, J., Bloom, D., Passarge, E., Fried, K., Goodman, R.M.,
Katzenellenbogen, I., Laron, Z., Legum, C., Levin, S., and
Wahrman, J. 1977b. Bloom’s syndrome. VI. The disorder in
BLAP18/RMI2, a novel component of the BTB complex
GENES & DEVELOPMENT2867
Israel and an estimation of the gene frequency in the Ash- Download full-text
kenazim. Am. J. Hum. Genet. 29: 553–562.
German, J., Schonberg, S., Louie, E., and Chaganti, R.S. 1977c.
Bloom’s syndrome. IV. Sister-chromatid exchanges in lym-
phocytes. Am. J. Hum. Genet. 29: 248–255.
Goss, K.H., Risinger, M.A., Kordich, J.J., Sanz, M.M., Straughen,
J.E., Slovek, L.E., Capobianco, A.J., German, J., Boivin, G.P.,
and Groden, J. 2002. Enhanced tumor formation in mice het-
erozygous for Blm mutation. Science 297: 2051–2053.
Gruber, S.B., Ellis, N.A., Scott, K.K., Almog, R., Kolachana, P.,
Bonner, J.D., Kirchhoff, T., Tomsho, L.P., Nafa, K., Pierce,
H., et al. 2002. BLM heterozygosity and the risk of colorectal
cancer. Science 297: 2013.
Hanada, K. and Hickson, I.D. 2007. Molecular genetics of RecQ
helicase disorders. Cell. Mol. Life Sci. 64: 2306–2322.
Hu, Y., Raynard, S., Sehorn, M.G., Lu, X., Bussen, W., Zheng, L.,
Stark, J.M., Barnes, E.L., Chi, P., Janscak, P., et al. 2007.
RECQL5/Recql5 helicase regulates homologous recombina-
tion and suppresses tumor formation via disruption of Rad51
presynaptic filaments. Genes & Dev. 21: 3073–3084.
Karmakar, P., Seki, M., Kanamori, M., Hashiguchi, K., Ohtsuki,
M., Murata, E., Inoue, E., Tada, S., Lan, L., Yasui, A., et al.
2006. BLM is an early responder to DNA double-strand
breaks. Biochem. Biophys. Res. Commun. 348: 62–69.
Langland, G., Kordich, J., Creaney, J., Goss, K.H., Lillard-Weth-
erell, K., Bebenek, K., Kunkel, T.A., and Groden, J. 2001. The
Bloom’s syndrome protein (BLM) interacts with MLH1 but is
not required for DNA mismatch repair. J. Biol. Chem. 276:
Leng, M., Chan, D.W., Luo, H., Zhu, C., Qin, J., and Wang, Y.
2006. MPS1-dependent mitotic BLM phosphorylation is im-
portant for chromosome stability. Proc. Natl. Acad. Sci. 103:
Ling, C., Ishiai, M., Ali, A.M., Medhurst, A.L., Neveling, K.,
Kalb, R., Yan, Z., Xue, Y., Oostra, A.B., Auerbach, A.D., et al.
2007. FAAP100 is essential for activation of the Fanconi ane-
mia-associated DNA damage response pathway. EMBO J. 26:
Meetei, A.R., Sechi, S., Wallisch, M., Yang, D., Young, M.K.,
Joenje, H., Hoatlin, M.E., and Wang, W. 2003. A multipro-
tein nuclear complex connects Fanconi anemia and Bloom
syndrome. Mol. Cell. Biol. 23: 3417–3426.
Raynard, S., Bussen, W., and Sung, P. 2006. A double Holliday
junction dissolvasome comprising BLM, topoisomerase III?,
and BLAP75. J. Biol. Chem. 281: 13861–13864.
Raynard, S., Zhao, W., Bussen, W., Lu, L., Ding, Y.Y., Busygina,
V., Meetei, A.R., and Sung, P. 2008. Functional role of
BLAP75 in BLM–topoisomerase III?-dependent holliday
junction processing. J. Biol. Chem. 283: 15701–15708.
Seki, M., Nakagawa, T., Seki, T., Kato, G., Tada, S., Takahashi,
Y., Yoshimura, A., Kobayashi, T., Aoki, A., Otsuki, M., et al.
2006. Bloom helicase and DNA topoisomerase III? are in-
volved in the dissolution of sister chromatids. Mol. Cell.
Biol. 26: 6299–6307.
So, S., Adachi, N., Lieber, M.R., and Koyama, H. 2004. Genetic
interactions between BLM and DNA ligase IV in human
cells. J. Biol. Chem. 279: 55433–55442.
Sung, P. and Klein, H. 2006. Mechanism of homologous recom-
bination: Mediators and helicases take on regulatory func-
tions. Nat. Rev. Mol. Cell Biol. 7: 739–750.
Theobald, D.L., Mitton-Fry, R.M., and Wuttke, D.S. 2003.
Nucleic acid recognition by OB-fold proteins. Annu. Rev.
Biophys. Biomol. Struct. 32: 115–133.
Wang, Y., Cortez, D., Yazdi, P., Neff, N., Elledge, S.J., and Qin,
J. 2000. BASC, a super complex of BRCA1-associated pro-
teins involved in the recognition and repair of aberrant DNA
structures. Genes & Dev. 14: 927–939.
Wang, X.W., Tseng, A., Ellis, N.A., Spillare, E.A., Linke, S.P.,
Robles, A.I., Seker, H., Yang, Q., Hu, P., Beresten, S., et al.
2001. Functional interaction of p53 and BLM DNA helicase
in apoptosis. J. Biol. Chem. 276: 32948–32955.
Wu, L. and Hickson, I.D. 2003. The Bloom’s syndrome helicase
suppresses crossing over during homologous recombination.
Nature 426: 870–874.
Wu, L. and Hickson, I.D. 2006. DNA helicases required for ho-
mologous recombination and repair of damaged replication
forks. Annu. Rev. Genet. 40: 279–306.
Wu, L., Bachrati, C.Z., Ou, J., Xu, C., Yin, J., Chang, M., Wang,
W., Li, L., Brown, G.W., and Hickson, I.D. 2006. BLAP75/
RMI1 promotes the BLM-dependent dissolution of homolo-
gous recombination intermediates. Proc. Natl. Acad. Sci.
Yin, J., Sobeck, A., Xu, C., Meetei, A.R., Hoatlin, M., Li, L., and
Wang, W. 2005. BLAP75, an essential component of Bloom’s
syndrome protein complexes that maintain genome integ-
rity. EMBO J. 24: 1465–1476.
Singh et al.
2868 GENES & DEVELOPMENT