The BLM helicase facilitates RNA polymerase I-mediated ribosomal RNA
Dissertation
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
Patrick Michael Grierson, B.S.
Graduate Program in Integrated Biomedical Science
The Ohio State University
2012
Dissertation Committee:
Joanna Groden PhD, Advisor
Samir Acharya PhD
Denis Guttridge PhD
Debbie Parris PhD
Larry Schlesinger MD
Copyright by
Patrick Michael Grierson
2012
Abstract
The BLM helicase is a DNA repair protein mutated in the hereditary condition Bloom’s syndrome (BS). BLM is best known for its roles in regulating homologous recombination-mediated DNA repair and telomere maintenance. A more limited body of work suggests that BLM regulates stability of the nucleolar ribosomal DNA. Previous work in our laboratory suggested that BLM might function in nucleolar ribosome biogenesis.
In the present study, we investigated a role for BLM in nucleolar RNA polymerase I-mediated ribosomal RNA (rRNA) transcription. We determined that the nucleolar localization of BLM is sensitive to inhibition of RNA polymerase I- mediated transcription by actinomycin D and 4-nitroquinoline-1-oxide, drugs that inhibit RNA polymerase I. We demonstrated that BLM facilitates 45S rRNA transcription using pulse-chase and nuclear run-on techniques and identified
RNA polymerase I and DNA topoisomerase I as nucleolar protein partners. DNA topoisomerase I directly interacts with BLM, an interaction mediated predominantly by the C-terminus of BLM. The interaction of BLM with DNA topoisomerase I is functionally significant as it stimulates BLM unwinding activity in vitro using a RNA:DNA hybrid oligomer substrate that models rRNA transcription from rDNA. We established the biochemical requirement of a 3’
ii overhang of single-stranded DNA for BLM to unwind nucleolar-relevant
RNA:DNA hybrids. Additionally, we discovered a physical interaction between
BLM and the nucleolar protein nucleophosmin (NPM) both in nucleoli and nucleoplasm, suggesting a mechanism by which nucleolar trafficking of BLM may be mediated. Overall, this work demonstrates that BLM functions in a pathway of nucleolar ribosome biogenesis and suggests a mechanism by which it may do so. These findings may impact our ability to inhibit or promote cell growth in appropriate clinical settings.
iii
Acknowledgments
There are many people whom I must acknowledge. First and most importantly is my dissertation advisor, Joanna Groden, who has been incredibly supportive.
She will always stand as a model of scientific excellence. I am extremely grateful to Samir Acharya for his scientific guidance, mentoring and friendship. His endless energy and enthusiasm for science have truly been an inspiration. I thank the remainder of my dissertation committee, Denis Guttridge, Larry
Schlesinger and Debbie Parris for insightful suggestions and helpful discussions through the years. Many thanks to Jeremy Keirsey for scientific discussions and friendship throughout my entire time in the lab. Cathy Ebert’s scientific insight and technical wisdom proved invaluable for several of my early studies. I thank current and former members of the Groden Lab for discussions, advice and friendship.
Finally, I gratefully acknowledge my parents and brothers. They have been endlessly supportive and inspiring in my pursuit of higher education.
iv
Vita
June 2005...... B.S. Biochemistry, University of
...... Wisconsin-Madison
2005 to present ...... MD/PhD student, The Ohio State
University
Publications
Russell, B., Bhattacharyya, S., Keirsey, J., Sandy, A., Grierson, P., Perchiniak,
E., Kavecansky, J., Acharya, S., and Groden, J. (2011) Chromosome breakage is regulated by the interaction of the BLM helicase and topoisomerase IIα.
Cancer Res., 71, 561-571.
Grierson, P., Lillard, K., Behbehani, G. K., Combs, K. A., Bhattacharyya, S.,
Acharya, S., and Groden, J. (2012) BLM helicase facilitates RNA polymerase I- mediated ribosomal RNA transcription. Hum. Mol. Genet., 21, 1172-1183.
Fields of Study
Major Field: Integrated Biomedical Science
v
Table of Contents
Abstract………………………………………………………...…………………………ii
Acknowledgments……………………………………………...……………………….iv
Vita………………………………………………………………...…………………..….v
Table of Contents…………………………………………...……..……………….…..vi
List of Figures……………………………………………….………..…………….……x
Chapter 1: Literature review………………………………………….………………..1
The clinical presentation and genetics of Bloom’s syndrome………………1 Identification of the BLM gene, BLM protein structure and function….……7 Role for BLM in DSB repair……………………………………….….11 Role for BLM in DNA replication…………………………….…….....14 Role for BLM in telomere maintenance……………………………..14 Role for BLM in mitotic chromosome decatenation………………..16 Role for BLM in nucleolar metabolism………………….……...... 16 Nucleolar metabolism……………………………………………….………...18 Ribosomal RNA transcription………………………………………...20 Ribosomal DNA replication………………………………………..….26 Ribosomal DNA recombination…………………………………...….28 Nucleolar trafficking……………………………………………………30 Ribosome biogenesis and cell growth regulation…………………..32
Chapter Two. Thesis Rationale and Research Objectives………………..………35
vi Chapter Three. The BLM helicase facilitates RNA polymerase I-mediated ribosomal RNA transcription……………………………………………...... 38
I.
Introduction………………………………………………………………………….....38
II. Materials and Methods…………….……………..……………………………..…41
Cell lines………………………………………………………………………..41
Transfection, immunofluorescence and antibodies……………………..…42
Protein co-immunoprecipitation………………………………………………43
Pulse-chase assays………………………………………………………...…43
Biotin-labeled nuclear run-on assay…………………………………...…….44
Nuclear run-on/ dot blot assay……………………………………………….45
Protein purification…………………………………………………………..…46
Helicase assays and electrophoretic mobility-shift assays (EMSA)……...46
III. Results……………………………………………………………….……………...47
BLM re-localizes within the nucleus following inhibition of RNA
polymerase I-mediated rRNA transcription……………………...………….47
BLM interacts with the RNA polymerase I-specific subunit RPA194…….56
BLM-deficient cells display a slower rate of RNA polymerase I-mediated
rRNA transcription……………………………………………………..………58
BLM binds and unwinds rDNA-like GC-rich DNA20:DNA33 and
RNA20:DNA33 nucleic acid duplexes……………………………………..…..67
IV.
Discussion……………………………………………………………………………...72
vii Chapter Four. DNA topoisomerase I interacts directly with the BLM helicase and stimulates its unwinding activity of duplex templates that mimic rDNA……..…...81
I. Introduction…………………………………………………………………………..81
II. Materials and Methods………………………………………………………….....83
Cell lines……………………………………………………………….……….83
Nucleolar isolation……………………………………………………………..83
Protein co-immunoprecipitation………………………………………………84
Protein purification……………………………………………………………..84
In vitro transcription/ translation (IVTT)……………………………………...86
Helicase assays………………………………………………………………..86
III. Results……………………………………………………………………………....87
BLM interacts with DNA topoisomerase I from nuclear and nucleolar-
enriched extracts.……………………………………………………………...87
BLM and DNA topoisomerase I interact directly in vitro via the BLM C-
terminus ………………………………………………………...... 91
DNA topoisomerase I stimulates BLM helicase activity on RNA20:DNA33
substrates designed to model rRNA/rDNA hybrids …...... …94
IV. Discussion…………………………………………………………….…………....99
Chapter Five. Nucleolar trafficking of BLM and protein partnering with nucleophosmin…………………………………………………………………….….103
I. Introduction………………………………………………………………………....103
II. Materials and Methods……………………………………………………….…..104
viii Cell lines ……………………………………………………………………...104
Transfection, immunofluorescence and antibodies………………………104
Nucleolar isolation……………………………………………………………104
Protein co-immunoprecipitation and western blotting…………………....105
III. Results……………………………………………………………………………..106
BLM and NPM co-localize in nucleoli of MCF7 cells……………...……...106
BLM and NPM co-immunoprecipitate from nuclear extracts of 293T
cells…………………………………………………………………………….106
BLM and NPM co-immunoprecipitate in nucleolar and nucleoplasmic
fractions of 293T cells…………………………………………………..……107
IV. Discussion…………………………………………………………………….…..109
Chapter Six. Thesis summary………………………………………………………112
Bibliography…………………………………………………………………………..125
ix
List of Figures
Figure 1. Bloom’s syndrome patients display proportional dwarfism……………..3
Figure 2. Characteristic chromosomal structures of Bloom’s syndrome cells…...5
Figure 3. Domain structure of BLM…………………………………………………...8
Figure 4. Representative nucleic acid substrates unwound by BLM……………10
Figure 5. Pathway of homologous recombination-mediated DNA double strand break (DSB) repair…………………………………………………………………….12
Figure 6. Nucleolar localization of BLM is dependent upon ongoing RNA polymerase I transcription…………………………………………………………….51
Figure 7. BLM associates with the RNA polymerase I-specific subunit
RPA194…………………………………………………………………………………57
Figure 8. BLM deficiency slows RNA polymerase I-mediated 45S rRNA transcription rate……………………………………………………………………….60
Figure 9. BLM unwinds duplex substrates with a 3’ DNA overhang but not those with a 3’ RNA overhang……………………………………………………………….68
Figure 10. BLM binds to DNA20:DNA33 and RNA20:DNA33, and less strongly to
DNA20:RNA33 and RNA20:RNA33 duplexes………………………………………….71
Figure 11. Model for the role of BLM in rRNA transcription…………………...... 76
Figure 12. BLM and DNA topoisomerase I associate in 293T cells……………..88
x Figure 13. BLM and DNA topoisomerase I directly interact………………………90
Figure 14. DNA topoisomerase I interacts with the C-terminus of BLM………...93
Figure 15. Equimolar DNA topoisomerase I stimulates the helicase activity of
BLM on RNA20:DNA33…………………………………………………………………96
Figure 16. GFP-BLM co-localizes and interacts with NPM……………………..107
Figure 17. BLM interacts with NPM in nucleoli and nucleoplasm………………108
Figure 18. Model for the role of BLM in facilitating RNA polymerase I-mediated ribosomal RNA transcription………………………………………………………...123
xi
CHAPTER ONE
Literature Review
The clinical presentation and genetics of Bloom’s syndrome
Bloom’s syndrome (BS) is an autosomal recessive inherited disorder of chromosomal instability. Characterized predominantly by short stature, sun- sensitive facial erythema, immunodeficiency and cancer predisposition, it was initially identified by the dermatologist Dr. David Bloom in New York City in 1954.
Dr. Bloom first described BS as “congenital telangiectatic facial erythema and stunted growth.” BS is most common in individuals of Ashkenazi Jewish heritage, or those Jews who derive from eastern Europe (reviewed in German,
1968). Those affected are often identified during pregnancy at birth or due to small size. Exposure to sunlight during childhood leads to sun-sensitive facial erythema in some but not all affecteds. The three most serious complications in
BS are cancer, chronic lung disease, most likely due to chronic upper respiratory infections and diabetes mellitus (German, 1997).
As mentioned, fetuses affected by BS display intrauterine growth retardation. BS infants are born at an average gestational age of 36 weeks and weigh approximately 1.7 kg on average; males are born slightly earlier and are slightly smaller than females; these values are 2 standard deviations below the
1 mean normal values (Diaz et al., 2006). Throughout childhood, BS children exhibit failure to thrive and their growth is consistently slow; they never achieve normal stature, as adults reach an average height and weight of approximately
133 cm and 40 kg, respectively (Diaz et al., 2006) (Figure 1). Although those with BS commonly develop deregulated endocrine functioning, this does not provide an explanation for the severe growth deficiency. For example, a clinical study of BS found that, 11 of 11 individuals had deregulated insulin signaling and manifested various degrees of insulin resistance or diabetes mellitus. Five of 10 individuals tested had a degree of dyslipidemia. However, growth hormone secretion and insulin-like growth factor 1 (IGF-1) production were within normal limits, as was intestinal nutrient absorption, discounting these as likely etiologies of the growth deficiency (Diaz et al., 2006). Additionally, cell lines derived from
BS persons have slower growth rates and a diminished growth response to epidermal growth factor (EGF) (Lechner et al., 1983), recapitulating the bodily growth defect on the cellular level.
2
Figure 1: Bloom’s syndrome patients display proportional dwarfism.
Shown is Dr. David Bloom (left), Dr. James German, (top) and a full-grown young man with Bloom’s syndrome, demonstrating the severity of the proportional growth defect.
The most serious complication of BS is an increased risk of malignancy.
In the first 168 BS patients studied, 100 cancers were documented in 71 patients.
BS individuals are predisposed to those malignancies found in unaffected
3 individuals, although at a much earlier age of onset. Only two types of malignancies have never been observed in those with BS: ovarian and prostate cancer (German, personal communication). Cancers arise in BS at all ages, and the age-related incidence of specific cancers mirrors that seen in the general population. For example, acute leukemias are generally seen in the younger individuals and chronic leukemias are more frequently seen in the older individuals. Also, the rare tumors, such as medulloblastoma and Wilm’s tumor that are generally seen in the pediatric population are observed in BS patients aged 1-10 years old. Similarly, carcinomas of the stomach and colon are seen in the older BS patients, similar to that observed in the general population. Thus, the types and sites of cancers in BS very closely model that seen in the general population. The major difference is the greatly increased incidence and much earlier age of cancer onset in BS—with an average age of diagnosis of 25 years old. As expected, cancer is the leading cause of death in BS at a mean age of
24 years old (German, 1997).
Cytogenetically, cells from BS persons display a high rate of chromosomal instability including quadriradial chromosome structures, sister chromatid exchanges (SCE) and telomere fusions, consistent with an increased rate of homologous recombination (Figure 2). The initial evidence for an increased rate of homologous recombination in BS was the presence of quadriradial chromosome structures in cultured lymphocytes from patients. The predominant quadriradials in BS are symmetric—the centromeres are on opposite arms—and composed of two homologous chromosomes that have undergone chromatid
4 exchange wherein the breakpoints are in the same location on each chromatid.
This produces quadriradials where the centromeres are equidistant from the breakpoint and which will yield two monocentric chromosomes following separation at mitosis (Figure 2A). The quadriradial chromosome structures characteristic of BS may also be observed in cells from those in the normal population but occur in BS at a much greater frequency, analogous to the incidence of malignancy in BS. Thus BS is a representative model of accelerated chromosome instability and malignancy (German, 1997).
5 Figure 2: Characteristic chromosomal structures of Bloom’s syndrome cells. Shown is (A) a symmetrical quadriradial chromosome [adapted from
German et al., 1974], (B) telomeric associations [adapted from Lillard- Wetherell et al., 2004], and (C) sister-chromatid exchange [adapted from http://atlasgeneticsoncology.org].
Cells from BS persons exhibit a high number of sister chromatid exchanges or SCEs, which is diagnostic for the disorder. Lymphocytes from normal individuals have approximately 9.3 SCE per cell, while those from BS patients display approximately 94 SCE per cell, a 10-fold increase (German,
1977) (Figure 2C). Similar to the quadriradials in BS cells, the increased formation of SCE suggests excessive recombination. BS lymphocytes and fibroblasts also display a high degree of telomeric associations (TA) between homologous chromosome arms. The TAs often have a central clear region and contain telomeric repeat DNA, indicating that they are not the result of fused chromosome ends lacking telomeres, but are an entanglement of telomeric repeats from different chromosomes (Figure 2B). Similar to the occurrence of quadriradials and SCE in BS lymphocytes, the presence of TAs in BS cells is not unique, but is approximately 6-fold more common (Lillard-Wetherell et al., 2004).
Finally, analysis of the acrocentric chromosomes—those containing the highly repetitive rDNA— has identified chromosomal instability in BS cells. Some acrocentric chromosomes contain polymorphic Q-bright satellites that are stably
6 maintained over successive cell divisions. Different cells from BS individuals who are heterozygous for Q-bright satellite patterns exhibit varied patterns of acrocentric Q-bright satellites, indicating instability of these particular genomic regions (Therman et al., 1981).
Identification of the BLM gene, BLM protein structure and function
The gene responsible for BS was mapped to chromosome 15q26.1 and designated BLM following its identification by positional cloning (Ellis et al.,
1995). BLM encodes a 1417 amino acid protein homologous to the recQ-like sub-family of ATP-dependent 3’-5’ DNA helicases. BLM is a functional DNA helicase with 3’-5’ directionality (Karow et al. 1997). The recQ sub-family of helicases also includes E. coli recQ, S. cerevisiae Sgs1, and human WRN,
RecQL1, RecQL4 and RecQL5, the majority of which play roles in DNA repair and are capable of unwinding various DNA structures (reviewed in Bachrati et al.,
2003). Most BLM mutations responsible for BS introduce a premature termination codon, such as nonsense point mutations or frameshifts. Cells from
BS persons generally have no detectable BLM due to the instability of the mRNAs encoding BLM. The most common mutation, the Ashkenazi founder mutation, is a 6 base-pair deletion/7 base-pair insertion at nucleotide 2281 that leads to a frameshift and premature stop codon (Ellis et al., 1995).
7 Structural analysis of BLM indicates that it is a modular protein with several functional domains (Figure 3). The N-terminal segment is responsible for protein oligomerization, as BLM is thought to function as a hexamer (Janscak et al., 2003). The central domain contains a conserved helicase motif required for helicase activity. The carboxyl-terminal segment contains two conserved domains—the recQ C-terminal domain (RQC domain) and the Helicase and
RNase D C-terminal Domain (HRDC domain). The RQC domain binds to dsDNA, and the HRDC domain is responsible for nucleic acid substrate recognition and may function in single-stranded DNA binding. The carboxyl- terminal domain also contains the nuclear localization sequence (NLS) (reviewed in Chu and Hickson, 2009).
Figure 3: Domain structure of BLM. BLM is a modular protein containing several functional domains. The N-terminal domain is required for protein oligomerization and localization to PML bodies. The central helicase and RQC
(recQ C-terminal) domains are conserved in all recQ family members. The
HRDC (helicase and Rnase D C-terminal) domain is important for nucleic acid
8 substrate recognition. The C-terminal NLS (nuclear localization sequence) mediates BLM’s localization to nuclei. *Noted is the position of the common
Ashkenazi founder mutation (6bp deletion/ 7bp insertion) [Ellis/ Groden et al.,
1995]. **Unpublished data, as presented in this thesis.
BLM unwinds a number of different nucleic acid structures, each of which suggests a unique role for BLM in DNA metabolism. BLM requires a 3’ DNA overhang of at least 8 nucleotides to unwind traditional Watson-Crick duplex
DNA substrates. BLM binds to the 3’ overhang and subsequently translocates until encountering and unwinding a duplex region of DNA (Popuri et al., 2008).
BLM is unique in comparison to other helicases as in addition to unwinding classic Watson/Crick B-form duplex DNA, BLM is capable of unwinding unusual
DNA structures including D-loops, Holliday junctions and X-junctions; it can regress stalled replication forks, unwind G-quartet (G4) DNA and unwind
RNA:DNA hybrids (Figure 4; Popuri et al., 2008).
9
Figure 4: Representative nucleic acid substrates unwound by BLM. Shown are (A) traditional Watson-Crick B-form duplex DNA, (B) RNA:DNA hybrid duplex, (C) G quartet (G4) DNA [adapted from
10 http://genome.cshlp.org/content/16/5/644/F1.expansion], (D) X-junction, (E) displacement loop (D-loop), (F) R-loop and (G) model replication fork.
Role for BLM in DSB repair
DNA double-strand breaks (DSB) can be repaired through a pathway of non-homologous end joining (NHEJ) or recombination-mediated repair. BLM has a function in each of these two processes. Consistent with the hyper- recombination phenotype of BS cells—suggesting that BLM serves to limit homologous recombination—BLM unwinds substrates that are intermediates in the homologous recombination-mediated pathway of DNA DSB repair (Figure 5).
For example, following the induction of a DSB, exonuclease I mediates 5’ end resection of a blunt-ended DNA duplex to generate a 3’ single-stranded tail.
RAD51 recombinase binds this 3’ single-stranded tail and subsequently mediates homology-dependent strand invasion of a homologous duplex to form a displacement loop (D-loop). Notably, BLM stimulates exonuclease I to facilitate
DSB end resection, a pro-recombination function. Additionally, BLM can displace
RAD51 from the 3’ single-stranded tail as well as unwind a model D-loop in vitro.
Both of which are anti-recombinogenic functions. A double Holliday junction
(DHJ) is formed in the later stages of homologous recombination. Processing of the DHJ may occur via “dissolution” or “resolution.” DHJ resolution can lead to the formation of cross-over products such as sister chromatid exchanges, but dissolution does not lead to cross-over products. A complex of BLM, DNA
11 topoisomerase IIIα, RMI1 and RMI2 dissolves the DHJ in a pathway that does not lead to cross-over products. Thus, in the absence of BLM, DHJs are processed by resolution leading to crossing-over, seen as SCE, in BS cells
(reviewed in Chu and Hickson, 2009). Importantly, BLM can unwind an X- junction structure in vitro, which is a model of a Holliday junction (Mohaghegh et al., 2001).
Figure 5. Pathway of homologous recombination-mediated DNA double strand break (DSB) repair. Upon induction of a DSB, 5’ end resection generates a 3’ single-stranded tail. Ultimately, Rad51 binds the 3’ single- stranded tail and mediates homologous duplex invasion, yielding a D-loop. BLM
12 can catalyze D-loop unwinding. Following strand extension, a double Holliday junction (DHJ) is formed, which is dissolved in the presence of BLM. Adapted from Chu and Hickson, 2009.
BLM localizes to PML bodies, storage depots for DNA repair proteins, consistent with a role for BLM in DNA repair. PML bodies also include PML,
DNA topoisomerase III∝, p53, RAD50, MRE11 and NBS1 (Lombard and
Guarente 2000; Hu et al., 2001; Sengupta et al., 2003). Many proteins in PML bodies are covalently SUMO modified; SUMOylation regulates protein localization to and retention in PML bodies (reviewed in Muller et al., 2004).
Tetracycline-inducible pGFP-BLM deletion constructs in BS cells identified a domain within the N-terminus of BLM (amino acid residues 135-600) that mediates BLM localization to PML bodies (Yankiwski et al., 2001). BLM has
SUMO-binding sites; while SUMOylation of the N-terminus of BLM is required for its localization to PML bodies (Eladad et al., 2005). Phosphorylation also regulates BLM localization to PML bodies, as induction of a DSB leads to ATM- or ATR-mediated threonine-99 phosphorylation of BLM and BLM re-localization from PML bodies to γH2AX-marked DSB (Rao et al., 2005). Overall, BLM trafficking between PML bodies and sites of DNA damage is regulated by a dynamic interplay of SUMOylation and phosphorylation.
13 Role for BLM in DNA replication
The slower rate of DNA elongation in BS cells suggests a role for BLM in
DNA replication (Hand and German, 1975). Lesions in the template DNA strand can induce replication fork stalling and collapse, leading to overall slower rates of
DNA chain elongation. When lesions are encountered in template DNA strand and the replication fork stalls, the fork can be regressed and repaired to allow reformation of a functional replication fork and bypass of the lesion (reviewed in
Chu and Hickson, 2009). BLM is recruited to damaged and stalled replication forks in vivo (Davalos and Campisi, 2003) and can catalyze replication fork regression in vitro (Machwe et al., 2006). Biochemically, this function requires both the helicase and single-strand DNA annealing activities of BLM (reviewed in
Chu and Hickson, 2009). Collectively, these data suggest an in vivo role for BLM in ensuring efficient DNA replication.
Role for BLM in telomere maintenance
GC-rich regions of the genome (telomeres, immunoglobulin heavy chain switch regions and rDNA) have a propensity to form G-quartet (G4) DNA structures composed of GC-rich strands of DNA stabilized by Hoogsteen base- pairing in a planar conformation (Parkinson et al., 2002) (Figure 4C). G4 DNA can inhibit movement of the DNA replication fork (reviewed in Chu and Hickson,
2009). In vitro studies demonstrate that BLM is highly capable of unwinding G4
DNA (Sun et al., 1998). BLM is involved in the recombination-mediated pathway of telomere maintenance which may require its ability to unwind G4 DNA or
14 unwind duplex DNA and D-loop structures composed of telomeric repeats
(Lillard-Wetherell et al., 2004).
BLM involvement in telomere maintenance was first suggested by the increased frequency of telomeric associations in BS cells (Lillard-Wetherell et al.,
2004). BLM binds telomeric DNA in telomerase-positive and telomerase- negative cells (Schawalder et al., 2003)—those that use a telomerase- independent recombination-mediated pathway of telomere maintenance designated “alternative lengthening of telomeres” (ALT). The N-terminal domain of BLM is required for binding telomeric DNA (Schawalder et al., 2003). BLM co- localizes in ALT cells with the telomeric repeat proteins TRF1 and TRF2 (Lillard-
Wetherell et al., 2004). It localizes to PML bodies that co-localize with telomeric
DNA, designated ALT PML bodies (APB). BLM co-localization with TRF2 coincides with foci of DNA synthesis and increases during late S/ G2-phase when ALT occurs (Lillard-Wetherell et al., 2004). At telomeres, BLM forms a complex with TRF1, TRF2 and TEP1, proteins that regulate BLM helicase activity on telomeric repeat-containing substrates (Lillard-Wetherell et al., 2004;
Bhattacharyya et al., 2009). Specifically in ALT cells, knockdown of BLM expression leads to rapid telomere shortening (Bhattacharyya et al., 2009).
Collectively, these data suggest that BLM functions coordinately with TRF1,
TRF2 and TEP1 to regulate recombination-mediated telomere lengthening in
ALT cells.
15 Role for BLM in mitotic chromosome decatenation
Anaphase bridges—persistent entangled DNA strands between separating chromosomes during anaphase—occur with an increased incidence in cells lacking BLM. A subclass of anaphase bridges, ultra-fine anaphase bridges (UFB), contains BLM (Chan et al., 2007). DNA topoisomerase IIα functions during mitosis to decatenate entangled chromosomes, thereby preventing chromosome breakage (Holm et al., 1989). DNA topoisomerase IIα directly interacts with BLM, stimulates its helicase activity with early recombination substrates, and functions in a common pathway with BLM to prevent chromosome breakage (Russell et al., 2011).
Role for BLM in nucleolar metabolism
The ~400 rDNA repeats in a human cell are located on the p-arms of the acrocentric chromosomes and coalesce in the nucleolus. rDNA is highly GC-rich and as such is prone to the formation of G4 DNA. S. cerevisiae mutant for the recQ-like helicase and BLM homolog Sgs1 display inefficient replication solely in the rDNA, suggesting that unwinding of G4 DNA, or another inhibitory secondary structure, in this region of the genome by recQ helicases may be required for efficient replication (Versini et al., 2003).
rDNA is susceptible to the co-transcriptional formation of RNA:DNA hybrids, structures that form in association with RNA polymerases and inhibit polymerase movement. Factors that prevent the formation of RNA:DNA hybrids, such as DNA topoisomerase I, facilitate efficient transcription (Zhang et al., 1988;
16 Hraiky et al., 2000). BLM-deficient cells display slowed RNA polymerase I- mediated transcription through the rDNA (Grierson et al., 2012). In vitro studies demonstrate that BLM is capable of unwinding RNA:DNA duplexes (Popuri et al.,
2008; Grierson et al., 2012). BLM can unwind RNA:DNA hybrid duplexes in which the overhang is DNA (Figure 4B) but not RNA:DNA hybrids in which the overhang is RNA; nor can it unwind RNA:RNA duplexes. The inability to unwind substrates with a 3’ overhang of RNA is due to the inability of BLM to bind the 3’
RNA overhang (Grierson et al., 2012). This is similar to that observed for other bacterial, archaeal and eukaryotic helicases that unwind RNA:DNA duplexes: they require a 3’ overhang of DNA as they are incapable of translocating on an
RNA strand (Shin and Kelman, 2006).
Cytogenetic studies in BS cells indicate a 60-fold increased rate of recombination between the rDNA repeats, suggesting that BLM is required to maintain the stability of these regions of the genome (Therman et al., 1981).
Restriction enzyme digestion combined with Southern blotting techniques demonstrates that the rDNA of BLM-deficient cells undergoes an approximately
100-fold increased rate of recombination with every cell division cycle (Killen et al., 2009). Interestingly, cells deficient in the related recQ-like helicase WRN, as well as cells deficient in the NHEJ pathway, do not exhibit this rDNA hyper- recombination; although slight increases are seen in ATM-deficient cells. These data demonstrate that maintenance of rDNA repeat stability is a not a general consequence of DNA repair defective disorders—it is relatively unique to cells lacking BLM (Killen et al., 2009). BS lymphocytes in culture for more than one
17 year lose approximately 25% of their rDNA (Schawalder et al., 2003). BLM localizes to the nucleolus (Sanz et al., 2000) via inclusion of amino acid residues
1118-1331 within the BLM C-terminus (Yankiwski et al., 2001). The same region of BLM binds to rDNA, suggesting the hypothesis that binding to rDNA serves to retain BLM in the nucleolus. BLM binds within the 45S rRNA coding sequence as well as to regions in the rDNA intergenic spacer (IGS), suggesting BLM involvement in rDNA replication and/ or transcription (Schawalder et al., 2003).
N-terminal deletion mutants of BLM localize to nucleoli, demonstrating that distinct domains of BLM are required for its localization to PML bodies or nucleoli
(Yankiwski et al., 2001). Helicase activity is not required for BLM nucleolar localization as a GFP-BLM construct without amino acid residues 903-1115 within the helicase domain and the GFP-BLM K695T ATP-binding mutant localize to nucleoli (Yankiwski et al., 2001). These studies suggest that BLM plays a critical role in nucleolar metabolism and in maintaining nucleolar rDNA stability.
NUCLEOLAR METABOLISM
Pathologists in the 19th century first noted that malignant cells are characterized by enlarged nucleoli. Nucleoli are highly refractile non-membrane- bound sub-nuclear domains. They contain rDNA repeats and are the site of RNA polymerase I-mediated ribosomal RNA (rRNA) transcription and biogenesis of ribosomes, large macromolecules composed of rRNA and protein. Ribosome biogenesis is a crucial process as it provides the cell with the means to translate
18 protein, and thereby grow and proliferate. Rapidly growing cells are more highly engaged in ribosome biogenesis and accordingly display larger nucleoli; as such nucleolar size has prognostic value in various tumor types (reviewed in Derenzini et al., 2009).
Interphase cells can have up to 10 nucleoli, although most normal cells have 1 to 3. Nucleoli form by coalescence of the rDNA repeats on the p-arms of the acrocentric chromosomes (13, 14, 15, 21 and 22); nucleoli disassemble at the beginning of mitosis when chromosomes condense and re-assemble early in
G1-phase. Nucleoli have three distinct structural compartments. The most central region, the fibrillar center (FC), is the location of the rDNA. Surrounding the FC is the dense fibrillar component (DFC), which also contains rDNA and is the site for rRNA transcription and rRNA post-transcriptional processing and maturation (Koberna et al., 2002). Surrounding the DFC is the granular component (GC), the region for ribosome assembly (reviewed in Lamond et al.,
2007).
The nucleolus is also a site for non-ribosome biogenesis-related activity.
For example, the essential components of telomerase (TERT, hTR, dyskerin) are assembled in the nucleolus. In normal cells, telomerase is localized to the nucleolus during all but S-phase when it is released into the nucleoplasm
(reviewed in Derenzini et al., 2009). The nucleolus also plays a role in cell-cycle regulation (reviewed in Boisvert et al., 2007) and oncogene-induced ribosomal stress response. For example, oncogenic c-MYC over-expression induces the dissociation of ribosomal proteins from nucleoli into the nucleoplasm where
19 ribosomal components bind and inhibit the E3 ubiquitin ligase MDM2 to stabilize p53 (Macias et al., 2010).
Ribosomal RNA transcription
Ribosomal RNA transcription is most active during S- and G2-phases of the cell cycle and is a highly active process, accounting for 50% of all transcription in a growing cell (Klein and Grummt, 1999). RNA polymerase I mediates rRNA transcription from the highly repetitive rDNA genes. Each of the
~40 tandemly arranged rDNA repeats on each acrocentric chromosome is 44-kb in length, composed of a 13-kb 45S rRNA coding sequence and a 31-kb intergenic spacer (IGS). RNA polymerase I-mediated transcription produces the
13-kb 45S rRNA that is then processed by 2’OH methylation, pseudouridylation and endoribonuclease cleavage to yield the mature 28S, 18S and 5.8S rRNA.
The remaining 5S rRNA is transcribed by RNA polymerase III (from chromosome
1) and enters the nucleolus for assembly into ribosomes (Sorensen et al., 1991).
Enzymatic modifications of the 45S rRNA (2’OH methylation, pseudouridylation and endoribonuclease cleavage) occur at specific sites, the modifying enzymes directed by site-specific binding of small nucleolar RNA (snoRNA) to the 45S rRNA. SnoRNA are derived from the intron sequences of the RNA polymerase
II-transcribed ribosomal protein mRNAs, providing a point of coordinate regulation of rRNA maturation and ribosomal protein production. Ribosomal proteins are transported from the cytoplasm into nucleoli where they assemble with mature 28S, 18S and 5.8S rRNA to yield ribosomes, which are subsequently
20 exported out of the nucleolus to the cytoplasm where they engage in protein translation.
Formation of the RNA polymerase I transcription initiation complex is a highly regulated process and requires many protein-protein and protein-DNA interactions. RNA polymerase I, a multi-subunit enzyme, binds to hRRN3 (TIF-
1A in mouse) prior to transcription initiation. hRRN3 in turn binds selectivity factor 1 (SL1; TIF-1B in mouse) at the rDNA promoter. SL1 is a multi-subunit protein complex composed of the TATA box binding protein (TBP) and TBP- associated factors 48, 63 and 110. rDNA promoter-specific binding of SL1 is mediated in part by TBP, as well as by binding of SL1 to a dimer of upstream binding factor (UBF). UBF dimers bind rDNA and induce wrapping of ~140bp of rDNA around the dimer in a manner similar to a nucleosome. The RNA polymerase I holoenzyme, once localized to the rDNA promoter, is competent for transcription initiation (reviewed in Russell and Zomerdijk, 2005).
Growth factor addition or removal will signal cells to increase or decrease rRNA synthesis by regulating the transcriptional output per active rDNA gene.
This is accomplished by modulating the frequency of transcription initiation and thus the number of transcribing RNA polymerase I complexes per rDNA gene, which in turn can be regulated at several points. In mouse cells, growth factor signaling induces the MAPK members ERK (extra-cellular signal regulated kinase) and RSK (ribosomal S6 kinase) to phosphorylate and activate TIF-1A
(hRRN3), facilitating its interaction with RNA polymerase I to increase rDNA transcription, cell growth and proliferation (reviewed in Russell and Zomerdijk,
21 2005). Similarly, mTOR-mediated phosphorylation of TIF-1A (hRRN3) facilitates
TIF-1A/RNA polymerase I interaction to maintain TIF-1A in the nucleus (Mayer et al., 2004). P/CAF-mediated acetylation of SL1 stimulates its binding to the rDNA promoter to activate rDNA transcription. UBF is a common target of many signaling pathways that regulate rDNA transcription. Serum-induced stimulation of rDNA transcription results in UBF phosphorylation by cdk4/ cyclin D, cdk2/ cyclin E and cdk2/ cyclin A to facilitate UBF and RNA polymerase I interaction.
Serum-stimulation also leads to mTOR/RSK-mediated phosphorylation of UBF to facilitate SL1 interaction. Insulin-like growth factor 1 (IGF-1) stimulation of mouse cells leads to nucleolar translocation of insulin receptor substrate 1 (IRS-
1) and the interaction of IRS-1, PI3K and UBF. PI3K-mediated phosphorylation of UBF is associated with increased rRNA transcription. The p53 and RB tumor suppressors also localize to the nucleolus and down-regulate rRNA synthesis by binding and inhibition of SL1 and UBF, respectively (Zhai and Comai, 2000;
Cavanaugh et al., 1995). During mitosis, SL1 is phosphorylated by cyclin B/cdc2 which inhibits SL1 and UBF interaction. Mitotic phosphorylation of SL1 and UBF is associated with mitotic rDNA transcriptional silencing (Klein and Grummt,
1999). In summary, many signaling pathways converge in the nucleolus to coordinate rDNA transcription with cellular growth and proliferation.
The elongation phase of rDNA transcription is another point of rRNA regulation. Tight wrapping of rDNA on UBF—which is bound throughout the rDNA—impedes RNA polymerase I passage. Epidermal growth factor (EGF) stimulation induces MAP kinase (MAPK)-mediated phosphorylation of UBF to
22 relax rDNA wrapping of UBF and facilitate the elongation phase of rRNA transcription (Stefanovsky et al., 2001). DNA topoisomerase I associates with actively transcribed rDNA genes to relax positive supercoils that accumulate ahead of the transcription complex and relax negative supercoils that accumulate behind the transcription complex. Persistence of negative supercoils behind the transcription complex facilitates re-annealing of the nascent rRNA with the template rDNA to form an rRNA:rDNA hybrid duplex. The rRNA:rDNA hybrid presents an obstacle to subsequent RNA polymerases by slowing their rate of movement and thus the overall rate of rDNA transcription. By relieving negative supercoils behind the transcription complex, DNA topoisomerase I prevents formation of rRNA:rDNA duplexes and thus facilitates efficient rDNA transcription elongation (Hraiky et al., 2000; Hage et al., 2010). In E. coli, over expression of
RNase H—an enzyme that selectively degrades the RNA strand of an RNA:DNA duplex—rescues a DNA topoisomerase I-deficiency-induced slowing of rDNA transcription. Yeast RNA polymerase A has intrinsic RNase H activity, further supporting the hypothesis that the removal of RNA:DNA hybrids is important for transcription (Huet et al., 1976). The BLM-related recQ-like helicases WRN in humans and Sgs1 in S. cerevisiae associate with the RNA polymerase I complex and facilitate rRNA transcription by unwinding DNA secondary structures that inhibit the elongating polymerase (Shiratori et al., 2002; Lee et al., 1999).
Termination of rRNA transcription occurs at a defined region in each rDNA repeat. At the 3’ end of each rDNA gene, transcription termination factor 1 (TTF-
1) binds to the rDNA and terminates transcription when it encounters RNA
23 polymerase I. RNA polymerase I subsequently releases the rRNA transcript and dissociates from the rDNA. It then may or may not re-associate with promoter- bound SL1 and UBF to initiate an additional round of rDNA transcription (Kuhn et al., 1990).
Ribosomal DNA transcription is also regulated by controlling the number of actively transcribed rDNA genes in addition to controlling the rate of transcription from each activated gene. Cells contain an excess number of rDNA genes and tightly control those activated at any given time. In yeast, approximately 50% of the total cellular rDNA genes are transcribed, while the total number of rDNA genes can be reduced to 30% of the wild-type number without decreasing the rRNA transcriptional output or growth rate. Yeast compensate for a reduced number of rDNA genes by increasing the RNA polymerase I density per gene to maintain the same number of transcribing polymerases and subsequent rRNA output (French et al., 2003). Cellular differentiation is associated with decreased rDNA transcription mediated by rDNA gene silencing (reviewed in Reeder, 1999). Histone de-acetylation is also a major contributor to the mechanism of rDNA gene silencing (Sandmeier et al.,
2002).
In addition to the pre-rRNA gene promoter directing transcription of the
45S rRNA, the IGS of the rDNA repeats contains a promoter (IGS promoter) upstream of the pre-rRNA gene promoter with sequence similarity to the pre- rRNA gene promoter (de Winter and Moss, 1986). Transcription from both the
IGS promoter and the pre-rRNA gene promoter is mediated by RNA polymerase
24 I. IGS promoter-driven transcripts are processed into approximately 200 nucleotide RNAs (pRNA) that bind and activate the nucleolar-remodeling complex (NoRC) (Mayer et al., 2006). The activated NoRC recruits DNA methyltransferases and histone deacetylases to mediate heterochromatin formation and silencing of rDNA genes (Santoro et al., 2002). RNA polymerase
I-mediated IGS rRNA transcription occurs from the hypo-methylated rDNA repeats as is the case for pre-rRNA synthesis. Ectopic over-expression of pRNA drives CpG methylation at the pre-rRNA gene promoter, increases heterochromatic histone modifications (H3K9me3, H4K20me3) and decreases euchromatic histone modifications (H3K4me3) at the pre-rRNA gene promoter.
As predicted, this correlates with silencing of pre-rRNA expression and demonstrates that IGS rRNA-mediated transcription leads to pRNA-dependent rDNA silencing in trans (Santoro et al., 2010). The relative rate of transcription from the spacer promoter is approximately 10% of the pre-rRNA promoter rate; the steady state level of IGS rRNA is 1000-fold less than rRNA (Santoro et al.,
2010). Transcription from the spacer promoter is mediated by RNA polymerase I using the same basal RNA polymerase I transcription factors TIF-1A and SL1 bound to both promoters. TIF-1A over-expression drives transcription from both promoters (Santoro et al., 2010), although, pre-rRNA gene transcription is driven by growth factor-stimulated phosphorylation of TIF-1A (Zhao et al., 2003) and
UBF (Stefanovsky et al., 2006). Growth factor stimulation has no effect on the rate of transcription from the IGS promoter, suggesting that transcription is regulated by distinct mechanisms from each promoter. Additionally, IGS rRNA
25 transcriptional output peaks slightly later during S-phase than does pre-rRNA gene transcription, further suggesting that transcription of these two rRNA regions is differentially regulated (Santoro et al., 2010).
Ribosomal DNA replication
The large-scale organization of the rDNA, with 13kb rRNA transcription units separated by 31kb intergenic spacers (IGS), provides a unique challenge for the cell to coordinate rRNA transcription and rDNA replication during S-phase.
Given the bi-directional replication fork movement from replication origins, mechanisms must exist to prevent head-on collisions between simultaneous replication and transcription complexes. rDNA replication initiates in multiple locations within the IGS, but does not initiate in the rRNA transcription unit or in the rRNA promoter (Little et al., 1993). Two-D gel electrophoresis demonstrates that a high number of bi-directional replication forks stall at the 3’ end of the rRNA gene and imply that this region acts as a replication fork barrier (RFB).
Replication fork stalling in this region occurs when replication forks encounter the barrier from either direction or due to head-on replication fork convergence. This observation suggests that when moving down-stream, rDNA replication complexes proceed co-directionally with the RNA polymerase I transcription complex. Upstream movement or anti-directional replication initiated within the
IGS is stalled at the 3’ end of the rRNA transcription unit, the same site as RNA polymerase I transcription termination to prevent replication/transcription complex collision and interference. Therefore, the 3’ end of the rRNA transcription unit is
26 a site of repeated replication fork and transcription complex stalling. In mammalian cells, transcription termination factor 1 (TTF-1) binds at the RFB to mediate replication fork stalling (Pfleiderer et al., 1990). In yeast, the replication fork barrier (RFB) is bound by Fob1 protein that analogously induces replication fork stalling (reviewed in Eckert-Boulet and Lisby, 2009).
Approximately 50% of the rDNA repeats are actively transcribed during S- and G2-phases of the cell cycle. Actively transcribed rDNA repeats associate with UBF, a necessary basal transcription factor for RNA polymerase I; they replicate early in S-phase while the inactive rDNA replicates later in S-phase (Li et al., 2005). The combination of light microscopy and nuclear run-on assays by
Dimitrova (2011) demonstrates that early replicating (transcriptionally active) rDNA dissociates from UBF and moves to the nucleolar periphery or exterior during replication. After replication, the rDNA fraction returns to the nucleolar interior and re-associates with UBF to allow resumption of rDNA transcription.
Dimitrova (2011) detected rDNA replication initiation within the transcribed regions of rDNA genes and found that the late replicating rDNA fraction, that does not associate with UBF and is transcriptionally inactive, is replicated in the nucleolar interior. These data suggest a model in which rDNA transcription halts, rDNA is moved to the nucleolar exterior to be replicated, and rDNA is returned to the nucleolar interior for the resumption of transcription. This model proposes that minimization of interference between rDNA transcription and replication complexes occurs through the spatial re-localization of the rDNA to the nucleolar periphery and the transient cessation of rDNA transcription. It explains how rDNA
27 replication initiates within the rDNA genes without interfering with transcription
(Dimitrova 2011).
Ribosomal DNA recombination
Aging yeast display a characteristic enlargement and fragmentation of the nucleolus due to increased recombination within the rDNA. Such recombination leads to the formation of extra-chromosomal rDNA circles (ERC). Age- associated ERC selectively accumulate in the yeast mother cell and may cause aging by titrating the cellular DNA replication machinery to the point that genomic replication ceases, inducing senescence and death of the yeast mother cell
(Sinclair and Guarente, 1997). To prevent these processes in aging yeast, Sir proteins re-locate from extra-nucleolar sites to the nucleolus, inhibit rDNA recombination, slow aging and prolong lifespan. The S. cerevisiae mutant for the recQ-like helicase Sgs1 prematurely display sterility, nucleolar fragmentation and enlargement, and the accumulation of ERC, all markers of aging (Sinclair and
Guarente, 1997). Sgs1 localizes to the nucleolus (Sinclair et al., 1997) and suppresses rDNA recombination (Gangloff et al., 1994). Similarly, BLM localizes to the nucleolus (Sanz et al., 2000); BS cells display hyper-recombination within rDNA (Killen et al., 2009); and individuals with BS display various degrees of premature aging. These observations and similarities suggest the hypothesis that rDNA hyper-recombination with resultant ERC formation are associated with aging and may be a conserved mechanism that prevents or promotes aging.
28 Recombinational repair of DSB within the rDNA is tightly controlled. In the general response to a DSB, blunt ends of a DSB are initially recognized by the
MRN complex (MRX in yeast), followed by 5’ end-resection to generate a 3’ overhang. The 3’ overhang is bound by RPA, followed by the recruitment of
Rad51 and Rad52. Rad51 catalyzes homologous duplex invasion by the 3’ single-strand end. Following strand invasion and extension, the invading stand is removed from the invaded duplex and DSB repair is completed (Figure 5)
(reviewed in Eckert-Boulet and Lisby, 2009). The induction of a DSB in the rDNA is followed by initial DSB recognition and processing in the nucleolus. The rDNA
DSB then moves to the nucleoplasm where it associates with Rad52, largely excluded from nucleoli. Repaired rDNA DSB then returns to the nucleolar interior.
Exclusion of Rad52 from the nucleolus, largely regulated by SUMOylation, may limit rDNA recombination. The SMC5-SMC6 SUMO E3 ligase-containing complex is abundant in the nucleolus and regulates Rad52 nucleolar accumulation (reviewed in Eckert-Boulet and Lisby, 2009). A Rad52 mutant with point mutations that prevent SUMOylation accumulates in nucleoli at much greater levels than wt-Rad52. Similarly, smc5-smc6-/- S. cerevisiae display increased Rad52 accumulation in nucleoli, suggesting that SMC5-SMC6- mediated SUMOylation prevents nucleolar accumulation of Rad52. Importantly,
SUMO-mutant Rad52 or smc5-smc6-/- S. cerevisiae display hyper-recombination within the rDNA (reviewed in Eckert-Boulet and Lisby, 2009).
29 Stalled replication forks throughout the genome can often be resolved by homologous recombination. As replication forks repeatedly stall at the RFB at the 3’ end of rDNA genes, homologous recombination may also be the favored mechanism to repair rDNA. In human rDNA, there is great variability in the number of tandemly repeated ~700bp sequence elements within the 3’ regions, hypothesized to be due to unequal homologous recombination (Little et al.,
1993). In S. cerevisiae, the Fob1 protein binds at the RFB and is required to induce replication fork stalling. Fob1-/- yeast defective in replication fork stalling display decreased levels of ERC (Defossez et al., 1999); rrm3-/- yeast that do not remove Fob1 from the RFB during replication fork passage display elevated levels of ERC. These data strongly suggest that replication fork stalling at the
RFB leads to homologous recombination-mediated repair and consequent ERC formation (reviewed in Eckert-Boulet and Lisby, 2009).
Nucleolar trafficking
Nucleoli are non-membrane-bound organelles. However, access to nucleoli is restricted as not all nuclear proteins are found in nucleoli. Nucleolar localization is mediated by various mechanisms including direct protein-protein interaction with resident nucleolar proteins, direct binding to rDNA or rRNA or binding to nucleolar non-coding RNA (ncRNA). Nucleolar targeting of proteins is a dynamic process, as proteins continually move between nucleoli as well as between nucleoli and the nucleoplasm. A prevailing theory is that some nucleolar proteins function as a hub, and that protein-protein interactions with a
30 hub serve to retain other proteins in the nucleolus. For example, nucleophosmin
(NPM) is required for maintenance of nucleolar structure. It interacts with many nucleolar proteins in the nucleolus and the nucleoplasm, and perhaps allows subsequent nucleolar import (reviewed in Emmott and Hiscox, 2009). NPM is a relatively small protein with many disordered domains, both characteristics of hub proteins. In contrast to the well-defined nuclear localization sequence (NLS) that mediates nuclear localization, a minority of nucleolar proteins have nucleolar localization sequences (NoLS). NoLSs are not as highly conserved as the NLS and are largely defined by function: a domain required for nucleolar localization and a frequent involvement in nucleolar protein-protein interactions. Other proteins directly involved in ribosome biogenesis are retained in nucleoli solely by an ability to bind rRNA (reviewed in Emmott and Hiscox, 2009).
Nucleolar localization of proteins with a role in ribosome biogenesis is sensitive to changes in RNA polymerase I activity. The RNA polymerase I- selective inhibitor actinomycin D (AMD) intercalates into GC-rich rDNA and disrupts RNA polymerase I-mediated transcription (Fetherston et al., 1984).
Following exposure to AMD, RNA polymerase I and multiple ribosome biogenesis-related proteins dissociate from the nucleolus and diffuse into the nucleoplasm (reviewed in Drygin et al., 2010). These observations suggest that
RNA polymerase I may function as a hub protein. Nucleolar proteins without a role in ribosome biogenesis, such as the telomeric protein TRF2, do not behave similarly following AMD treatment (Zhang et al., 2004).
31 Some nucleolar proteins permanently reside in nucleoli and include RNA polymerase I and other proteins involved in ribosome biogenesis. Other proteins localize to nucleoli only under certain conditions, such as heat-shock inducing the re-localization of HSP70 to nucleoli, acidosis re-localizing VHL to nucleoli, and ribosomal stress re-localizing MDM2 to nucleoli. These specific stimuli induce the expression of RNA transcripts from the rDNA IGS that are transcribed by
RNA polymerase I and are not polyadenylated, similar to rRNA, but are processed to ~300bp non-coding RNA (ncRNA). These ncRNA selectively bind and retain proteins with a nucleolar detention sequence (NoDS). There are multiple ncRNA loci in the IGS that are independently expressed under certain stress conditions (again including heat shock, acidosis and ribosomal stress) to cause their nucleolar retention (Audas et al., 2012).
Ribosome biogenesis and cell growth regulation
Conditions that limit the rate of ribosome biogenesis limit the cellular growth rate. Depletion of small ribosomal subunit processing components in yeast leads to defective ribosomal subunit synthesis and G1-phase cell cycle arrest. This occurs prior to any detectable change in the total translational capacity of the cell or the total number of ribosomes. Such a temporal pattern suggests that the rate of new ribosome biogenesis is sensed by the cell. G1 arrest induced by ribosome biogenesis inhibition coincides with nuclear accumulation of Whi5, the yeast RB homolog, also suggesting that cell cycle progression and proliferation are coordinated with nucleolar ribosome biogenesis
32 (Bernstein et al., 2007). Similar observations have been made in bacteria
(Hraiky et al., 2000) and mammalian cells. For example, MEFs with mutations in
TIF-1A are characterized by slower rates of rRNA transcription and cell growth rates; the mouse model, although embryonic lethal, yields small embryos (Yuan et al., 2005). Mice lacking the non-essential ribosomal subunit, Rpl29, are approximately half the size of wild-type mice in utero, and do not exhibit catch-up growth after birth. As expected, Rpl29-/- MEFs have slower rates of protein synthesis and proliferation (Kirn-Safran et al., 2007).
Growth factors stimulate cell growth and up-regulate rRNA transcription
(Stefanovsky et al., 2001). For example, nutrient availability up-regulates rRNA synthesis via mTOR signaling (Mayer et al., 2004). Serum, FGF and EGF stimulate rRNA synthesis and cell growth by activating the MAPK cascade, leading to phosphorylation and activation of TIF-1A (Zhao et al., 2003).
Epidermal growth factor (EGF) signaling through the MAPK pathway also leads to ERK-dependent phosphorylation of UBF, required for EGF-mediated stimulation of rRNA transcription. ERK-dependent phosphorylation of UBF relaxes rDNA wrapping around UBF, facilitating rRNA transcriptional elongation and rRNA transcriptional output (Stefanovsky et al., 2001; reviewed in Moss et al., 2006). These studies provide nucleolar-related mechanisms by which growth factors stimulate cell growth. Given the decreased responsiveness of BS cells to
EGF (Lechner et al., 1983) and the involvement of BLM in rDNA transcription
(Grierson et al., 2012), it is conceivable that limited rDNA transcription in the absence of BLM at least partially underlies resistance to EGF.
33 In humans, several disorders known as “ribosomopathies” feature small stature and result from limitation of ribosome biogenesis. For example,
Diamond-Blackfan Anemia (DBA) is caused by mutations in various ribosomal subunit proteins and is defined clinically by growth stunting, anemia and cancer risk. Mutations in the RMRP gene, encoding the RNA component of Rnase MRP required for 45S rRNA processing, lead to cartilage hair hypoplasia, metaphyseal dysplasia without hypotrichosis and anauxetic dysplasia—all are syndromes including significant growth stunting. Many of the human ribosomopathies are also characterized by an increased risk of malignancy, similar to that seen in BS
(reviewed in Narla and Ebert, 2010). Werner syndrome, a disorder characterized by malignancy and growth stunting, is caused by lack of expression of the recQ- like helicase WRN. WRN localizes to nucleoli and facilitates rDNA transcription.
Collectively, these disorders highlight important connections between growth potential, nucleolar metabolism and cancer susceptibility.
34
CHAPTER TWO
Thesis Rationale and Research Objectives
Bloom’s syndrome (BS) is an autosomal recessive inherited chromosomal instability disorder characterized invariably by proportional growth deficiency, an extremely high incidence of cancer and immunodeficiency (German, 1968;
German, 1997). The presentation of growth deficiency, as well as the concurrent affliction with malignancy, is similar to that observed in several human disorders of disrupted nucleolar metabolism (reviewed in Narla and Ebert, 2010). BS is caused by mutation of the BLM gene which encodes a 1417 amino acid DNA helicase with homology to the recQ family of DNA helicases (Ellis et al., 1995).
Cells lacking BLM display chromosomal instability suggestive of an increased rate of recombination (German et al., 1977). BLM is required for maintenance of genomic stability most likely through its roles in the regulation of both non- homologous end joining and homologous recombination-mediated DNA double- strand break repair, and in pathways of recombination-mediated telomere maintenance (reviewed in
Chu and Hickson, 2009; reviewed in Bhattacharyya et al., 2010).
A limited number of studies have suggested a role for BLM in nucleolar metabolism. An early cytogenetic study demonstrated that cells from BS patients
35 display variability of ribosomal DNA (rDNA) satellite repeats on the acrocentric chromosomes (Therman et al., 1981). A molecular study using restriction enzyme digestion of rDNA demonstrated that BLM deficiency dramatically increased recombination between the rDNA repeats on acrocentric chromosomes (Killen et al., 2009). Immunocytochemistry demonstrated that
BLM localizes to the nucleolus predominantly during S- and G2-phase of the cell cycle, the time during which rDNA replication and rRNA transcription are occurring (Yankiwski et al., 2000). A region within the C-terminus of BLM is required for nucleolar localization and directly binds the rDNA (Yankiwski et al.,
2001; Schawalder et al., 2003). A specific function for nucleolar BLM, however, has not yet been established.
This work tests the hypothesis that the BLM helicase facilitates the unwinding of rDNA and rRNA during replication and transcription of rDNA in the nucleolus. Previous preliminary work in our lab suggested that BLM nucleolar localization may be related to ongoing RNA polymerase I-mediated rRNA transcription as BS lymphocytes displayed a slower rate of rRNA transcription.
The goal of this work was to characterize the functions of BLM in nucleolar metabolism. We first measured the rate of RNA polymerase I-mediated rRNA transcription in BLM-proficient versus BLM-deficient cells. We then studied the effects of several cell-stressing agents on nucleolar retention of BLM, as well as identified RNA polymerase I and DNA topoisomerase I as protein partners of
BLM in the nucleolus. We defined the biochemical requirements for BLM unwinding and binding of GC-rich DNA:DNA and RNA:DNA duplexes. Finally,
36 we demonstrated a functional interaction between BLM and DNA topoisomerase
I in the unwinding of nucleic acid structures.
37
CHAPTER THREE
The BLM helicase facilitates RNA polymerase I-mediated ribosomal RNA
transcription.
I. Introduction
The approximately 400 ribosomal DNA (rDNA) genes within human cells are distributed tandemly on the p-arms of the five acrocentric chromosomes, 13,
14, 15, 21 and 22. These rDNA repeats, along with RNA polymerase I and numerous other proteins, are localized in interphase cells in a nuclear structure known as the nucleolus. The predominant function of nucleoli is transcription of ribosomal RNA (rRNA) from rDNA, a process mediated by RNA polymerase I that occurs most prominently during S- and G2-phases of the cell cycle
(Schwarzacher et al., 1993; Ayrault et al., 2006). Ribosomal RNA transcription is a major determinant of ribosome biogenesis, which in turn drives protein translation, cellular growth and proliferation (Birch et al., 2008). Animal models show that mutation of RNA polymerase I transcription factors inhibits rRNA transcription and impairs growth (Yuan et al., 2005), while human syndromes caused by defects within the ribosome biogenesis pathway similarly display growth impairment (Narla et al., 2010).
38 The nucleolus contains three distinct sub-structural components, the fibrillar center (FC), dense fibrillar component (DFC), and the granular component (GC) (reviewed in Huang et al., 2002). The FC and the DFC contain rDNA and RNA polymerase I; the DFC also contains factors required for rRNA processing (Huang et al., 2002; Koberna et al., 2002). RNA polymerase I transcription most likely occurs at the FC/DFC interface, or entirely within the
DFC (Koberna et al., 2002). The GC is the outermost region of the nucleolus and contains factors necessary for ribosomal assembly (Huang et al., 2002).
Proteomic analysis reveals a large number of putative RNA and DNA helicases, particularly those belonging to the DEAD-box family of RNA-dependent
ATPases, that localize to all nucleolar regions and suggest a necessity for diverse helicases in ribosomal RNA synthesis, processing and assembly into ribosomes (Sutherland et al., 2001; Andersen et al., 2002).
Bloom’s syndrome (BS) is a rare autosomal recessive disorder characterized by a high predisposition to cancer and severe growth retardation
(German, 1969). Cells from BS persons grow poorly in culture and have a decreased response to growth factors (Lechner et al., 1983). Affected individuals invariably display intra-uterine growth retardation (IUGR) with a mean birth weight of approximately 1.7 kg, and proportional dwarfism that persists throughout life with a mean adult height of 133 cm. The etiology of the BS growth defect remains unknown despite extensive clinical investigation (Diaz et al., 2006).
39 BLM, the protein absent in BS, belongs to the conserved recQ subfamily of ATP-dependent 3’-5’ DNA helicases (Ellis/ Groden et al., 1995). The BLM helicase localizes to PML bodies and nucleoli, most prominently during S-phase
(Yankiwski et al., 2000). The N-terminus of BLM is required for its accumulation to PML bodies, while nucleolar localization of BLM requires the C-terminal region that also directly binds rDNA repeats (Yankiwski et al., 2001). Within rDNA sequences, BLM specifically associates with the 18S-coding region and Alu- repeat regions, upstream of the region where replication is initiated (Schawalder et al., 2003). Furthermore, clonally selected BS cells have less rDNA than BLM- proficient cells, suggesting the hypothesis that nucleolar BLM, by binding to rDNA, is necessary to maintain the stability of rDNA (Yankiwski et al., 2001;
Schawalder et al., 2003). The related recQ-like WRN helicase localizes to nucleoli in some human cell types and accelerates RNA polymerase I transcription (Gray et al., 1998; Shiratori et al., 2002). The Saccharomyces cerevisiae BLM ortholog Sgs1 facilitates rDNA replication and maintains the stability of rDNA repeats (Versini et al., 2003; Heo et al., 1999). Sgs1 is also essential for RNA polymerase I transcription in the absence of the Srs2 helicase, suggesting the possibility of a similar function for BLM in rDNA replication and transcription (Lee et al., 1999).
Here, we report that treatment of human cells with the RNA polymerase I inhibitor actinomycin D (AMD) results in redistribution of BLM from nucleoli to the nucleoplasm and nucleolar periphery, consistent with an association of BLM with the RNA polymerase I transcription complex. In vivo protein co-
40 immunoprecipitation demonstrates a physical interaction between BLM and the
RNA polymerase I-specific subunit RPA194. 3H-uridine pulse-chase assays demonstrate a decreased production of the 45S rRNA transcript in BLM-deficient cells compared to wild-type cells, indicating a slower rate of RNA polymerase I transcription in the absence of BLM. In vitro, BLM binds and unwinds GC-rich rDNA-like DNA20:DNA33 and RNA20:DNA33 duplexes predicted to form during rRNA transcription, but not DNA20:RNA33 or RNA20:RNA33 duplexes. We propose that BLM is part of an RNA polymerase I transcription complex in the nucleolus and modulates rDNA to remove secondary structures that, if left unresolved, stall
RNA polymerase I transcription and increase recombination within rDNA repeats.
These data may help to understand the instability of rDNA repeats in BS cells
(Therman et al., 1981), as well as the documented cellular (Lechner et al., 1983) and whole-body growth defect in BS (Diaz et al., 2006; German, 1969).
II. Materials and Methods
Cell lines. MCF7 and 293T cells were obtained from ATCC and cultured in
Dulbecco’s Modified Eagle Medium (Invitrogen) containing 10% heat-inactivated
FBS (Hyclone). GM08505 cells were obtained from Coriell Cell Repository and cultured in Minimal Essential Medium (MEM) (Invitrogen) containing 10% heat- inactivated FBS (Hyclone). GM01806, GM03403, GM11973 and GM16375 lymphoblastoid cells were obtained from Coriell Cell Repository and cultured in
41 RPMI (Invitrogen) containing 15% heat-inactivated FBS (Hyclone). All cells were cultured at 37°C and 5% CO2.
Transfection, immunofluorescence and antibodies. pGFP-BLM was generated by cloning BLM cDNA into pEGFP-C1 (Clontech). MCF7 cells were transfected with the pGFP-BLM expression vector using Effectene Transfection
Reagent (Qiagen) according to manufacturer’s instructions. Forty-eight hours after pGFP-BLM transfection, cells were treated with either 5µg/ml actinomycin D or 30µg/ml α-amanitin for 1 hour at 37°C and then processed for immunofluorescence. MCF7 cells were treated with 4-nitroquinoline-1 oxide
(4NQO) according to published protocols (Gray et al., 1998). Forty-eight hours after pGFP-BLM transfection, cells were treated with 0.8µg/ml 4NQO in DMSO
(or mock-treated with DMSO) for 1 hour at 37°C, then into fresh media for 2 hours and processed for immunofluorescence. Finally, for hydroxyurea (HU) treatment, 48 hours after pGFP-BLM transfection MCF7 cells were treated with
2mM HU for 16 hours, and then processed for immunofluorescence. Following drug treatments, cells were washed in phosphate-buffered saline, fixed with 10% formamide (Sigma), stained with anti-RPA194 (Santa Cruz, sc-48385), anti- nucleophosmin (NPM, Abcam FC82291) or anti-PML (Santa Cruz, sc-966), mounted on glass slides and subsequently scored for localization of GFP-BLM using a Zeiss Axiovert 200M microscope and Axiovision 4.5 software. Western blotting was performed according to standard procedures using anti-BLM (Bethyl
Laboratories, A300-110A), anti-Lamin B (Santa Cruz Biotech, sc-6217), anti-
42 RPA194 (Santa Cruz Biotech, sc-48385), anti-RNA polymerase II (Abcam, ab817) or anti-WRN (Abcam, ab66606). GM08505 cells were transfected with pGFP-BLM, pGFP-BLM-D795A or pGFP-empty using FuGENE HD Transfection
Reagent (Roche) according to manufacturer’s instructions. 293T cells were transfected with either BLM Silencer pre-designed siRNA (Ambion) (5’-
GGAAGUUGUAUGCACUACCTT-3’) or Silencer negative control siRNA
(Ambion) using Lipofectamine 2000 Transfection Reagent (Invitrogen) according to manufacturer’s instructions.
Protein co-immunoprecipitation. Protein co-immunoprecipitations used 293T nuclear lysates prepared according to published protocols (Russell et al., 2011).
Antibodies used in co-immunoprecipitation included anti-BLM (Santa Cruz
Biotech, sc-7790), anti-RPA194 (Santa Cruz Biotech, sc-48385) and anti-RNA polymerase II (Abcam, ab817). Protein-antibody complexes were captured with
Dynabeads Protein G (Invitrogen, 100-04D), washed, eluted and separated by
8% SDS-PAGE.
Pulse-chase assays. Pulse-chase analysis was performed as previously described (Schlosser et al., 2003). Cells were pulse-labeled for 30 minutes in medium containing 2.5µCi/ml 3H-uridine and chased in medium containing
0.5mM uridine for the indicated amount of time. RNA was isolated using an
RNeasy Midi Kit (Qiagen) or TRI-Reagent (MRC), separated by electrophoresis in a 1% formaldehyde-MOPS agarose gel, transferred to a nylon membrane and
43 exposed to either Kodak BioMax MS film with BioMax TranScreen LE
Intensifying Screen or placed in a Tritium Storage Phosphor Screen (Amersham
Biosciences) for several days. Methylene blue or ethidium bromide staining demonstrated equal RNA loading. Band intensities were determined by densitometry analysis with ImageQuant software.
Biotin-labeled nuclear run-on assay. In vivo biotin-labeled nuclear run-on assays were performed as previously described (Patrone et al., 2000) with minor modifications. After siRNA transfection, cells were collected, washed in PBS and lysed in lysis buffer (final 13.3mM Tris, pH 7.5, 340mM sucrose, 13.3mM NaCl,
53mM KCl, 2mM EDTA, 0.5mM spermidine, 0.13mM spermine, 0.1% Triton X-
100, 2mM MgCl2) for 5 minutes on ice. Lysates were added to a sucrose cushion solution (final 13.3mM Tris, pH 7.5, 1.2M sucrose, 13.3mM NaCl, 53mM
KCl, 2mM EDTA, 0.5mM EGTA, 0.5mM spermidine, 0.13mM spermine, 2mM
MgCl2) and nuclei were collected by centrifugation at 2400g for 30 minutes at
4°C. Nuclei were re-suspended in freezing buffer (50mM Tris, pH 8.3, 40% glycerol, 5mM MgCl2, 0.1mM EDTA) and isolated nuclei incubated in transcription buffer (2X: 200mM KCl, 20mM Tris, pH 8.0, 5mM MgCl2, 4mM DTT,
4mM ATP, GTP, CTP, 200mM sucrose, 20% glycerol) with 1mM biotin-16-UTP
(Roche) for 1 hour at 30ºC. Transcription was terminated by passage through a
25G needle and DNase I (Roche) treatment according to manufacturer’s instructions. Total RNA was isolated using Tri-Reagent (MRC). Eight micrograms of total RNA were bound to Dynabeads M-280 streptavidin
44 (Invitrogen) for 1 hour according to manufacturer’s instructions. Beads were washed 4 times with 500ul of 2X SSC, 15% formamide, 0.2% Tween-20 for 5 minutes with gentle rotation, then once in 1ml of 2X SSC. RNA was eluted in
H2O by heating to 95ºC for 1 minute, and used in cDNA synthesis using
ThermoScript RT-PCR system (Invitrogen) according to manufacturer’s instructions. cDNA was amplified using primers specific for GAPDH (forward primer: GACATCAAGAAGGTGGTGAAG, reverse primer:
CCAGGAAATGAGCTTGACAAAG) and the 5’ external transcribed spacer (5’
ETS) of 45S rDNA (forward primer: GCCGGGTCCGAGCCGCGACGG, reverse primer: GCGGCGGGCGGGACGGCGAGG) using Taq DNA polymerase
(Roche), separated by agarose gel electrophoresis and analyzed using
ImageQuant software.
Nuclear run-on/ dot blot assay. Nuclei were isolated from 293T cells and used in nuclear run-on assays according to that described above, which the exception that α32P-UTP (Perkin-Elmer) was used in place of bitoin-16-UTP. After termination of transcription, total RNA was isolated using TRI-Reagent (MRC).
Total RNA was hybridized to a membrane made as follows: beta-actin encoding plasmid (Addgene, plasmid 27123) was spotted onto a membrane (Zeta-Probe
Blotting Membrane; Bio-Rad cat# 162-0153) using a Bio-Rad Microfiltration
Apparatus (cat# 170-6545) according to manufacturer’s instructions. The 18S rDNA coding sequence was PCR amplified and cloned into pYES2 (Invitrogen), maxi-preped using DH5α and spotted onto the membrane as described above for
45 beta-actin encoding plasmid. Total RNA was hybridized to the membrane in hybridization buffer (final: 5X Denhardt’s, 6X SSC, 0.1M Na2HPO4, 1% SDS,
0.2mg/ml sheared-sonicated salmon sperm DNA) at 50ºC overnight. The membrane was washed in 6XSSC 0.5% SDS, in 1X SSC 0.5% SDS and in
0.1XSSC 0.1% SDS at 50ºC. The membrane was subsequently placed in a
Storage Phosphor Screen (Amersham Biosciences) and analyzed using
ImageQuant software.
Protein purification. I. Hickson (University of Oxford, Oxford, UK) provided the pYES-BLM expression vector (pJK1). BLM was purified as previously described
(Russell et al., 2011). Briefly, hexa-histidine (6X-His)-tagged BLM was over- expressed in Saccharomyces cerevisiae. Yeast were lysed at 20k psi using a
French Press Cell Disrupter (Thermo) and lysates were separated by ultracentrifugation at 65,000g for 1hr at 4°C. The cleared lysate was purified by
FPLC using Ni-NTA Superflow (Qiagen), followed by Q-Sepharose (Sigma). The purity of the resultant BLM was determined by 8% SDS-PAGE and staining of the gel with SYPRO Ruby Protein Gel Stain (Sigma) and analysis using ImageQuant software as previously described.
Helicase assays and electrophoretic mobility-shift assays (EMSA).
Oligonucleotides were purchased from Invitrogen. Oligonucleotide sequences
(5’-3’ orientation) are as follows: DNA20- CGCTAGCAATATTCTGCAGC, DNA33-
GCTGCAGAATATTGCTAGCGGGAATTCGGCGCG, RNA20-
46 CGCUAGCAAUAUUCUGCAGC, RNA33-
GCUGCAGAAUAUUGCUAGCGGGAAUUCGGCGCG, DNA46-
GCGCGGAAGCTTGGCTGCAGAATATTGCTAGCGGGAATTCGGCGCG and
RNA46-
GCGCGGAAGCUUGGCUGCAGAAUAUUGCUAGCGGGAAUUCGGCGCG.
32 RNA20, DNA20, DNA46 and RNA46 were P end-labeled using polynucleotide kinase (NEB) according to manufacturer’s instructions. Substrates were duplexed by heating to 95°C for 5 minutes and slowly cooling. Helicase assays were performed as previously described (Lillard-Wetherell et al., 2004), with the exception that reactions were performed with 20ul final volume using 2 fmol of substrate per reaction. Helicase products were separated on 12% non- denaturing polyacrylamide gels. For EMSA, reactions were set up identically to those in helicase assays but with ATP omitted. Binding products were separated using 4% polyacrylamide, 5% glycerol, 1X TBE gels electrophoresed at 4ºC.
EMSA reactions using single-stranded substrate were resolved on 1% agarose
TBE gels electrophoresed at room temperature. Helicase and binding assays were analyzed using ImageQuant software.
III. Results
BLM re-localizes within the nucleus following inhibition of RNA polymerase
I-mediated rRNA transcription.
47 BLM localizes to the nucleolus (Yankiwski et al., 2000), the site of RNA polymerase I-mediated rRNA transcription. To investigate a role for nucleolar
BLM in rRNA transcription, we examined whether BLM localization is dependent upon RNA polymerase I-mediated rRNA transcription using the RNA polymerase
I inhibitor actinomycin D (AMD). Actinomycin D binds selectively to GC-rich DNA and strongly inhibits RNA polymerase I transcription (Drygin et al., 2010) as rRNA genes have a high GC-content. RNA polymerase I and its associated factors distribute in a characteristic fashion following AMD treatment (Jordan et al., 1996; Svarcova et al., 2008). A low concentration of AMD treatment changes the localization of RNA polymerase I from the center of nucleoli to the nucleolar periphery (Jordan et al., 1996). A high concentration of AMD results in the dispersal of RNA polymerase I throughout the nucleoplasm (Svarcova et al.,
2008). In order to visualize the localization of BLM during AMD treatment, a
GFP-tagged BLM construct was used in transfections of the breast cancer cell line MCF7. MCF7 cells were chosen to minimize artifacts of BLM over- expression as they express a relatively low level of endogenous BLM and an effective αBLM antibody for immunofluorescence is not available. pGFP-BLM- transfected MCF7 cells were treated with AMD and the sub-nuclear localization of GFP-BLM analyzed by immunofluorescence microscopy. Nucleolar localization of GFP-BLM was observed in untreated cells using BLM co- localization with the RNA polymerase I-specific subunit RPA194 (Figure 6A) and the nucleolar protein nucleophosmin (NPM/ B23) (Figure 6B); cells transfected with the pGFP control vector showed a diffuse nuclear staining of GFP (Figure
48 6A). A short treatment with AMD results in a dramatic redistribution of NPM/
B23, RPA194 and GFP-BLM from the nucleolus to the nucleoplasm (Figure 6A and Figure 6B). Exposure to AMD significantly decreased the nucleolar localization of GFP-BLM (Figure 6C; 40% decrease); phase contrast microscopy confirmed that nucleoli remained intact after AMD treatment (Figure 6B). Such redistribution is consistent with that observed for RNA polymerase I after similar
AMD treatments (Svarcova et al., 2008). We also tested the effect of the RNA polymerase II inhibitor α-amanitin (αAMT) on BLM localization. Our results indicated that GFP-BLM is retained in the nucleolus following treatment of cells with αAMT and demonstrate that although RNA polymerase II is inhibited at the
AMD concentration used (Svarcova et al., 2008), the effect upon GFP-BLM nucleolar localization is specific to inhibition of RNA polymerase I and independent of RNA polymerase II.
Finally, cells were treated with the DNA polymerase-stalling drug hydroxyurea (HU) to test whether the effect of AMD on BLM localization is a general response to nuclear stress. Nucleolar GFP-BLM was undisturbed
(Figure 6A, C). Similar results were obtained using the human embryonic kidney
293T cell line and the BS fibroblast cell line GM08505 (Figure 6D). The recQ-like
Werner syndrome helicase (WRN) functions in nucleolar rRNA transcription and localizes to nucleoli in a 4-nitroquinoline-1 oxide (4NQO)-sensitive manner (Gray et al., 1998). 4NQO induces DNA lesions usually corrected by nucleotide excision repair. We analyzed nucleolar localization of GFP-BLM in MCF7 cells following treatment with 4NQO and found, similarly to WRN, nucleolar
49 dissociation of GFP-BLM (Figure 6A), suggesting that these two related helicases may also function similarly in rRNA transcription (Shiratori et al., 2002).
MCF7 cells transiently transfected with pGFP-BLM and treated with either AMD or 4NQO were stained with anti-PML to demonstrate that co-localization of PML and BLM is not disturbed following AMD or 4NQO treatment. These results show that the effect of AMD and 4NQO on GFP-BLM is specific to nucleolar BLM
(Figure 6E). These results indicate that localization of BLM within the nucleolus is associated with ongoing RNA polymerase I-mediated rRNA transcription.
50 Figure 6. Nucleolar localization of BLM is dependent upon ongoing RNA polymerase I transcription. A) The breast cancer cell line MCF7 was transiently transfected with pGFP-BLM and stained with αRPA194, a nucleolar protein, to demonstrate co-localization of GFP-BLM to nucleoli. MCF7 cells were treated with either actinomycin D (AMD), 4-nitroquinoline-1 oxide (4NQO), α- amanitin, hydroxyurea (HU), DMSO (negative control for AMD and 4NQO) or
H2O (negative control for α-amanitin and HU), followed by staining with αRPA194 and visualization of transiently expressed GFP-BLM and RPA194. MCF7 cells were transfected with pGFP control vector and similarly treated to demonstrate lack of an effect on GFP. B) MCF7 cells were treated as in A but stained with
αNPM, a nucleolar marker, to demonstrate co-localization of GFP-BLM with NPM and thus localization of BLM to nucleoli. Cells were treated with either actinomycin D (AMD) or DMSO (AMD mock treatment), followed by visualization of transiently expressed GFP-BLM or staining with αNPM. DAPI images show nuclear staining; fluorescence images show either GFP-BLM or NPM; phase- contrast images show location of nucleoli. C) The results of scoring transiently transfected MCF7 cells for GFP-BLM localization following the indicated treatments are shown. Averages + standard deviation were calculated for a minimum of 60 cells per treatment. D) The 293T cell line and the BS fibroblast cell line GM08505 were transiently transfected with pGFP-BLM, stained with
αRPA194 and treated as in A. E) MCF7 cells were transfected and treated as in
A but stained with αPML to analyze the co-localization of GFP-BLM and PML when BLM dissociates from the nucleolus. 51 A
continued
52
Figure 6 continued
B
continued
53
Figure 6 continued
C
continued
54 Figure 6 continued
D
E
55
BLM interacts with the RNA polymerase I-specific subunit RPA194.
Further investigation of the role of BLM in RNA polymerase I-mediated rRNA transcription used protein co-immunoprecipitation experiments to demonstrate the interaction of BLM and at least one subunit of RNA polymerase
I. RNA polymerase I is a nucleolar-specific polymerase solely dedicated to rDNA transcription. It is a multi-subunit enzyme, sharing some subunits with RNA polymerases II and III, although its largest subunit, RPA194, is not shared
(Seither et al., 1997). αBLM and αRPA194 antibodies were used in co- immunoprecipitation experiments with nuclear lysates from 293T and MCF7 cells to demonstrate that BLM and RPA194 interact (Figure 7A and B). Co- immunoprecipitation of BLM and RPA194 was observed regardless of which antibody was used for immunoprecipitation. An antibody specific to RNA polymerase II was unable to co-immunoprecipitate BLM and RNA polymerase II to control for the possibility that BLM non-specifically interacts with RNA polymerases (Figure 7C). The interaction of BLM and RNA polymerase I supports a function for BLM in the modulation of rDNA structures in association with RNA polymerase I to facilitate rRNA transcription.
56
Figure 7. BLM associates with the RNA polymerase I-specific subunit
RPA194. A, B) Co-immunoprecipitations were performed with nuclear extracts from 293T and MCF7 cells using either αBLM or αRPA194 antibodies for immunoprecipitation (IP). Proteins were separated using 8% SDS-PAGE, blotted and analyzed with αBLM and αRPA194 antibodies; goat IgG is an isotype- matched negative control for αBLM, mouse IgG is an isotype-matched negative control for αRPA194. C) Co-immunoprecipitations were performed as in A, B but 57 αRPB1 (RNA polymerase II subunit) was used; mouse IgG is an isotype- matched negative control for anti-RNA polymerase II. In panel C we were unable to detect an interaction between BLM and RNA polymerase II, demonstrating the specificity of its interaction with RPA194 in A, B.
BLM-deficient cells display a slower rate of RNA polymerase I-mediated rRNA transcription.
As nucleolar localization of GFP-BLM is associated with ongoing RNA polymerase I-mediated rRNA transcription, we asked whether BLM plays a role in rRNA expression. We used 293T cells transfected with an αBLM-directed siRNA to knockdown BLM expression or scrambled control siRNA for 3H-uridine pulse-chase assays to measure the rRNA transcription rate. Figure 3A shows that the abundance of radio-labeled 45S rRNA transcript diminishes following knockdown of BLM expression in 293T cells (Figure 8B). αBLM-directed siRNA does not affect the level of the related recQ-like helicase WRN (Figure 8C).
Additionally, we performed 3H-uridine pulse-chase assays using the BS fibroblast line GM08505 transiently transfected with pGFP-BLM or pGFP control vector.
Figure 8D shows that the abundance of radio-labeled 45S rRNA transcript increases upon re-expression of BLM via GFP-BLM in GM08505 cells (Figure
8B). In contrast, the helicase-dead GFP-BLM-D795A was unable to rescue the
45S rRNA transcript defect of GM08505 cells (Figure 8E). Our results suggest that BLM facilitates RNA polymerase I-mediated rRNA transcription and
58 emphasize the requirement for BLM helicase activity. To support these data,
293T cells were used in in vivo biotin-labeled nuclear run-on assays to measure the 45S rRNA transcription rate, as RNA polymerase I-mediated rRNA transcription initially produces the 45S rRNA transcript (reviewed in Drygin et al.,
2010). GAPDH, an RNA polymerase II-transcribed gene, served as normalization for RNA polymerase I-mediated transcription (Figure 8F). Figure
8G shows that siRNA-mediated knockdown of BLM expression (Figure 8F) decreases the GAPDH-normalized 45S rRNA transcription rate. The p-value (p =
0.03) was determined with the Wilcoxon signed-rank test as the experiment was conducted in a pair-wise manner (Figure 8H). 293T cells treated with AMD or mock-treated with DMSO were used in in vivo biotin-labeled nuclear run-on assays to verify the inhibitory effect of AMD on RNA polymerase I-mediated transcription (Figure 8I). The results of the in vivo biotin-labeled nuclear run-on assays were confirmed in 293T cells using traditional 32P-labeled nuclear run-on/ dot blot assays. Figure 8J shows that BLM knockdown (Figure 8K) slows the rRNA transcription rate by this technique as well. Furthermore, 3H-uridine pulse- chase analyses in age- and sex-matched BS (GM03403 and GM16375) versus wild-type lymphoblastoid cells (GM01806 and GM11973) similarly show that BLM is required to maintain a normal rRNA transcription rate (Figure 8L). Our findings demonstrate that BLM expression and helicase activity are required for efficient rRNA transcription.
59 Figure 8. BLM deficiency slows RNA polymerase I-mediated 45S rRNA transcription rate. A) 293T cells transfected with either an αBLM directed siRNA or scrambled control siRNA were pulse-labeled with 3H-uridine for 30 minutes (P) and chased for 1 hour (C) with cold uridine. Isolated RNAs were separated on a 1% MOPS-formaldehyde agarose gel, transferred to a nylon membrane and analyzed by autoradiography. 45S, 28S and 18S rRNA species are indicated next to autoradiograph (top panel). Ethidium bromide staining demonstrates equal loading of RNA (middle panel). Western blot demonstrates efficiency of BLM knockdown; lamin B serves as a protein loading control (bottom panel). B) The pulse-chase analyses in 293T and GM08505 cells were analyzed using ImageQuant software to measure the 45S rRNA transcript abundance in the BLM-proficient cells (293T scrambled siRNA transfected or GM08505 pGFP-
BLM transfected) compared to the BLM deficient cells (293T αBLM directed siRNA transfected or GM08505 pGFP transfected). C) 293T cells transfected with either αBLM directed siRNA or scrambled control siRNA were used in western analysis and probed with αBLM or αWRN to demonstrate specificity of the αBLM directed siRNA; lamin B serves as a protein loading control. D) 3H- uridine pulse-chase analysis in BS fibroblasts (GM08505) transfected with either pGFP-BLM or pGFP; figure is labeled as in A. E) GM08505 cells transfected with pGFP-BLM-D795A or pGFP control vector were used in pulse-chase assays as in D. BLM-D795A is unable to rescue the rRNA transcriptional rate in
GM08505 cells. F) Representative 45S rRNA and GAPDH mRNA PCR readouts of in vivo biotin-labeled nuclear run-on assays from 293T cells transfected with 60 scrambled control siRNA or BLM-directed siRNA to knockdown BLM expression.
Representative western blot shows the efficiency of BLM knockdown in BLM- directed siRNA-transfected 293T cells. Lamin B serves as a protein loading control. G) Graph shows quantification of the RNA polymerase I-mediated 45S rRNA transcription rate normalized to the RNA polymerase II-mediated GAPDH mRNA transcription rate as determined by in vivo biotin-labeled nuclear run-on assays; the populations are significantly different as determined by the Wilcoxon signed-rank test (p=0.03). Horizontal lines indicate the average value for each population. H) Table shows the pair-wise relationship of the data from in vivo biotin-labeled nuclear run-on assays. I) 293T cells treated with AMD (0.05ug/ml) or mock-treated with DMSO were used in in vivo biotin-labeled nuclear run-on assays to confirm the effect of AMD on RNA polymerase I-mediated 45S rRNA transcription in this assay. J) 293T cells transfected with scrambled control siRNA or BLM-directed siRNA to knockdown BLM expression were used in 32P- labeled nuclear run-on/ dot blot assays. Total RNA was hybridized to 18S rDNA coding sequence as a read-out of RNA polymerase I-mediated transcription, and to beta-actin coding sequence to provide normalization to RNA polymerase II- mediated transcription. K) The efficiency of BLM knockdown is shown. L) BS
(GM03403 and GM16375) and wild-type lymphoblastoid cells (GM01806 and
GM11973) were pulse-labeled with [5-3H]-uridine for 30 minutes (P) and chased for 1 hour (C1) and 2 hours (C2) with cold uridine. RNA was isolated and analyzed as in A. 45S, 32S, 28S and 18S rRNA species are indicated next to
61 autoradiograph. Methylene blue staining demonstrates rRNA integrity and equal loading of RNA (lower panel).
continued
62 Figure 8 continued
continued
63 Figure 8 continued
continued
64 Figure 8 continued
continued
65 Figure 8 continued
66 BLM binds and unwinds rDNA-like GC-rich DNA20:DNA33 and RNA20:DNA33 nucleic acid duplexes.
Altered localization of nucleolar BLM following AMD treatment, slower rates of RNA polymerase I-mediated rRNA transcription in the absence of BLM, and the association of BLM and RNA polymerase I suggest that BLM helicase functions may be required during rRNA transcription. As rRNA/rDNA duplexes can form in vivo (Aguilera, 2002) and inhibit movement of transcription complexes (Hraiky et al., 2000; Hage et al., 2010), we investigated the activity of
BLM on RNA/DNA nucleic acid substrates. Recombinant BLM was expressed in yeast and FPLC-purified (Figure 9A) for in vitro helicase assays and electrophoretic mobility shift assays (EMSA) using nucleic acid substrates with
20bp of duplexed sequence and a 13-nucleotide 3’ single-stranded overhang.
Previous in vitro studies have determined that BLM requires a 3’ overhang of at least 8 nucleotides for unwinding of standard duplex DNA (Popuri et al., 2008;
Mohaghegh et al., 2001). Substrates in our experiments had either a 3’ overhang of DNA (DNA20:DNA33 and RNA20:DNA33) or RNA (DNA20:RNA33 and
RNA20:RNA33). Substrates were incubated with BLM and products resolved using non-denaturing polyacrylamide gels (Figure 9B). Figure 9C shows that
BLM unwinds DNA overhang substrates (DNA20:DNA33 and RNA20:DNA33) following Michaelis-Menten kinetics with maximum unwinding, Umax: 95.26 +
1.14% (T1/2 = 0.28 min) and 103.84 + 2.80% (T1/2 = 3.24 min), respectively.
BLM displays very low activity unwinding RNA overhang substrates,
DNA20:RNA33 (Umax: 9.46 + 1.57%, T1/2 = 10.41 min) and RNA20:RNA33 (Umax:
67 1.77 + 0.17%, T1/2 = 6.16 min). Helicase assays using duplexes containing a
26nt 3’ overhang of either scrambled sequence, poly-T or poly-U sequences to control for secondary structure, show that BLM is unable to unwind duplexes with a 3’ RNA overhang regardless of its potential to form secondary structure (data not shown). These in vitro assays demonstrate the ability of BLM to unwind the
RNA/DNA hybrid duplexes predicted to form during RNA polymerase I-mediated transcription, and extend previously published unwinding assays (Popuri et al.,
2008).
68 Figure 9. BLM unwinds duplex substrates with a 3’ DNA overhang but not those with a 3’ RNA overhang. A) The purity of BLM protein was determined using electrophoresis with 8% SDS-PAGE gels and staining with SYPRO Ruby
Protein Gel Stain (Sigma). B) Autoradiographs of representative gels illustrate unwinding activities. BLM (3.8nM) was incubated with substrate for 0, 2.5, 5, 10,
20 or 30 minutes at 37°C as described in Methods. Products were resolved using 12% non-denaturing acrylamide gels. Unwinding is demonstrated by conversion of duplexed substrate to faster migrating single-stranded oligonucleotide. HD is heat-denatured substrate produced by heating to 95°C for
5 minutes. C) Kinetics of BLM unwinding of RNA- and DNA-containing substrates. BLM unwinds DNA20:DNA33 and RNA20:DNA33 but does not appreciably unwind DNA20:RNA33 or RNA20:RNA33. Unwinding of each duplex substrate was calculated by comparing the amount of single-stranded substrate produced to the total amount of substrate in the reaction with correction for any un-annealed substrate in zero-time controls. Percent unwinding is graphed as a function of time.
We next performed EMSAs to investigate whether the hybrid duplex unwinding activity of BLM reflected different binding affinities for the substrates.
EMSAs (Figure 10A, B) show that BLM binding to DNA20:DNA33 and
RNA20:DNA33 follows Michaelis-Menten kinetics with Kd values of 60.46 +
7.45nM BLM and 58.92 + 12.18nM BLM, respectively. BLM binding to
69 DNA20:RNA33 and RNA20:RNA33 is less favorable and does not follow Michaelis-
Menten kinetics. The lower binding affinity of BLM for DNA20:RNA33 and
RNA20:RNA33 may partially explain the reduced unwinding activity of BLM on these substrates. To address the possibility that the DNA20:RNA33 and
RNA20:RNA33 substrates are not bound or unwound because of insufficient RNA overhang length, EMSAs were performed with single-stranded DNA46 and single- stranded RNA46. Figure 10C shows that while BLM binds DNA46, it does not efficiently bind RNA46. These data and the duplex binding data suggest that BLM is unable to unwind DNA20:RNA33 and RNA20:RNA33 due to an inability to bind the single-stranded RNA overhang in these duplexes. Overall, these in vitro studies support a function for BLM at the interface of RNA and DNA metabolism in the nucleolus.
70
71 Figure 10. BLM binds to DNA20:DNA33 and RNA20:DNA33, and less strongly to DNA20:RNA33 and RNA20:RNA33 duplexes. A) Purified BLM (0, 11, 15, 23,
30, 38, 56, 75, 110nM) was incubated with 32P-labeled substrates as described in
Methods. Reactions were separated using acrylamide gel electrophoresis and analyzed using ImageQuant software. Duplexes bound by BLM migrate more slowly (°) than unbound duplexes (*). B) Binding of each duplex was calculated by comparing the amount of bound complex to the total amount of duplex in the reaction. Percent binding is graphed as a function of BLM concentration. C)
Purified BLM (0, 13.3, 29, 40, 55nM) was incubated with 32P-labeled single- stranded DNA46 or RNA46 as described in Methods. Reactions were separated using 1% agarose gels that were dried and analyzed. Substrate bound by BLM migrates more slowly (°) than unbound substrate (*).
IV. Discussion
Cellular and animal models demonstrate that defects in RNA polymerase
I-mediated rRNA transcription or ribosome function negatively impact growth
(Yuan et al., 2005; Chester et al., 1998). In cell culture, deletion of the RNA polymerase I transcription factor Tif-1a leads to cell cycle arrest and apoptosis while the Tif-1a-/- mouse model, although an embryonic lethal, produces small embryos (Yuan et al., 2005). Deletion of the non-essential Rpl29 ribosomal subunit slows protein synthesis and proliferation in murine fibroblasts and
72 produces viable yet proportionally small mice (Kirn-Safran et al., 2007). Similar to these models of perturbed ribosome biogenesis and function, cultured BS cells and persons with BS invariably display a growth defect (German, 1969; Lechner et al., 1983; Diaz et al., 2006), and although Blm-/- leads to embryonic lethality, before death and at all stages of embryogenesis the embryos are very small
(Chester et al., 1998). Importantly, BLM localizes to nucleoli (Yankiwski et al.,
2000), associates with rDNA in vivo and is required to maintain rDNA integrity
(Schawalder et al., 2003). As ribosome biogenesis from rRNA transcription to ribosome subunit assembly occurs predominantly in the nucleolus, we investigated a role for BLM in RNA polymerase I-mediated transcription.
Actinomycin D (AMD), a selective RNA polymerase I inhibitor, disrupts the nucleolar localization of RNA polymerase I and its related factors (Jordan et al.,
1996). We demonstrated that treatment of cells in culture re-localizes BLM from the nucleolus to the nucleoplasm and nucleolar periphery. This nucleolar redistribution pattern of BLM is consistent with an interaction of BLM with the
RNA polymerase I transcription machinery and the interaction of BLM and rDNA
(Schawalder et al., 2003). Using protein co-immunoprecipitation, we demonstrated that BLM physically interacts with the RNA polymerase I-specific subunit RPA194. The nucleolar re-localization of BLM in response to AMD as well and the physical association of BLM with RPA194, an RNA polymerase I subunit, suggests a role for BLM in rRNA transcription. Accordingly, the 45S rRNA transcription rate was measured in cells with either an innate or experimentally induced BLM deficiency, as 45S rRNA is the initial transcript
73 produced by RNA polymerase I. Cells lacking BLM display a slower RNA polymerase I-mediated 45S rRNA transcription rate suggesting that BLM facilitates rRNA transcription. Published work demonstrates that BS lymphoblastoid cells in long-term culture (one year or more) have approximately
25% less rDNA than wild-type cells (Schawalder et al., 2003). Our observations of slowed rRNA transcription rates following short-term siRNA-mediated BLM depletion support a mechanism not explained by a decrease of template rDNA but rather one with the persistence of an impeding DNA structure. Our finding in
BS cells that transient re-expression of BLM via GFP-BLM rescues the decreased rRNA transcription rate further supports this point. Similar data were reported in cells in the absence of the related recQ-like WRN helicase and suggest that BLM and WRN have the same or similar functions in RNA polymerase I-mediated transcription (Shiratori et al., 2002).
The formation of RNA/DNA hybrids during rRNA transcription and their ability to impede RNA polymerase I transcription complexes have been demonstrated in bacterial and yeast systems (Hraiky et al., 2000; Hage et al.,
2010). When unwinding fully duplexed DNA, BLM requires a 3’ overhang of at least 8 nucleotides and it is upon this strand that BLM translocates as it unwinds the duplex (Karow et al., 1997; Popuri et al., 2008). BLM can unwind hybrid
RNA/DNA substrates with a DNA overhang in vitro (Popuri et al., 2008). Thus, we asked whether the unwinding ability of BLM using a hybrid duplex would proceed with a 3’ overhang of RNA. Our results show that BLM binds and unwinds DNA overhang substrates DNA20:DNA33 and RNA20:DNA33, but not RNA
74 overhang substrates DNA20:RNA33 or RNA20:RNA33. The similarities between substrate unwinding and binding ability of BLM suggest that substrate binding may be the significant determinant of unwinding. This is consistent with previous studies of BLM using G4 DNA and Holliday substrates, in which G4 DNA is both better bound and unwound (Huber et al., 2002). Our results are also similar to those found for bacterial, archaeal and eukaryotic replicative helicases that unwind RNA/DNA hybrid duplexes only when binding to and translocating on the
DNA strand (Shin and Kelman, 2006). Our in vitro data support a role for BLM in rRNA transcription, rather than rRNA post-transcriptional processing or maturation, as helicases involved in these latter processes possess the ability to unwind RNA/RNA duplexes (Kressler et al., 2010).
These experiments suggest a novel role for the BLM helicase in RNA polymerase I-mediated transcription. We propose a model by which nucleolar
BLM maintains rDNA stability and promotes RNA polymerase I transcription
(Figure 11). The GC-rich nature of rDNA and rRNA transcripts may promote the formation of RNA/DNA hybrids, generating R-loops (Salazar et al., 1996). R-loop formation within rDNA, and its inhibition of rRNA transcription, occurs in E. coli and S. cerevisiae (Hraiky et al., 2000; Hage et al., 2010). Thus, the propensity of actively transcribed rDNA to form RNA/DNA hybrids is conserved. BLM readily unwinds RNA/DNA hybrids and R-loops (Popuri et al., 2008); our data support these suggested functions of BLM while emphasizing a requirement for a 3’ DNA overhang. BLM may disrupt RNA/DNA hybrids and R-loops formed during RNA polymerase I transcription, as these structures otherwise inhibit progression of
75 the transcription complex (Hraiky et al., 2000). Additionally, co-transcriptionally formed stable RNA/DNA hybrids, and thus R-loops, may induce genomic instability and recombination (Huertas and Aguilera, 2003), a phenomenon termed transcription-associated recombination (TAR) (reviewed in Aguilera,
2002). Excessive variability in the number of Q-bright satellites in BS lymphocytes indicates a high frequency of recombination between rDNA gene clusters (Therman et al., 1981). Recent work using BS cells and various DNA- repair defective cells demonstrates that the recombination within the rDNA is relatively unique to cells lacking BLM (Killen et al., 2009). Therefore, this model may explain both the slower rate of RNA polymerase I transcription and the well- known high frequency of recombination and rDNA instability in BS cells
(Schawalder et al., 2003; Therman et al., 1981).
76 Figure 11. Model for the role of BLM in rRNA transcription. During RNA polymerase I-mediated 45S rRNA transcription, RNA/DNA hybrids may form between the nascent rRNA transcript and the template rDNA. We propose that
BLM unwinds the RNA/DNA hybrid so that transcription and replication proceed unaffected. In the absence of BLM, the RNA/DNA hybrid may remain unresolved resulting in retardation of further rRNA transcription, stalling of transcription and replication forks, and the induction of recombination within the rDNA.
Severe proportional dwarfism was among the first characteristics documented in BS, and one for which a satisfactory explanation is still lacking
(Diaz et al., 2006). Our data suggest the hypothesis that deficient rRNA transcription may contribute to the growth deficiency typical of BS. All BS neonates exhibit intra-uterine growth retardation (IUGR), with typical birth weights of 1.5 kg and 1.8 kg for males and females respectively, 2.6 and 3.2 kg
SD below the mean considering average gestational ages of 35 and 38 weeks, respectively (Diaz et al., 2006). In the general population, approximately 92% of babies born with IUGR eventually exhibit catch-up growth during childhood
(Albertsson-Wikland et al., 1993). This contrasts with the observation that none of the BS infants exhibit catch-up growth. Proportional dwarfism is invariably present in BS with a typical adult height and weight of approximately 133 cm and
40 kg (Diaz et al., 2006). These observations suggest that the etiology of the growth defect in BS is fundamentally different from that observed in many other
77 common causes of growth stunting, such as growth hormone deficiency, thyroid hormone deficiency or malnutrition. Accordingly, these clinical parameters have been evaluated in BS persons, ranging from 9 months to 28 years of age.
Growth hormone secretion, thyroid hormone levels and intestinal absorption are found within normal range, excluding these as likely etiological factors in growth retardation (Diaz et al., 2006).
In contrast, the growth characteristics of BS persons are similar to those observed in some rare disorders designated as “ribosomopathies.” Namely,
Diamond-Blackfan anemia includes anemia as well as growth stunting, and is due to mutations in various genes encoding ribosomal proteins that in turn decrease ribosome biogenesis (Narla and Ebert, 2010). Approximately 20% of those affected by Diamond-Blackfan anemia exhibit IUGR, of which only 40% will exhibit catch-up growth during childhood, not quite analogous to that seen in BS yet distinctly different from that typical of IUGR in the general population (Chen et al., 2005). Short stature is also a prominent feature of Shwachman-Diamond syndrome, a disorder caused by mutations in the SBDS gene. The specific function of the SBDS gene product is unknown, but it associates with the 60S ribosomal subunit; its mutation leads to defective ribosome biogenesis (Narla and
Ebert, 2010; Makitie et al., 2004). Cartilage hair hypoplasia, its variant metaphyseal dysplasia without hypotrichosis, and anauxetic dysplasia are congenital syndromes displaying significant growth retardation; all are due to various mutations in the RMRP gene that encodes the untranslated RNA component of RNase MRP (Ridanpaa et al., 2001; Bonafe et al., 2002; Thiel et
78 al., 2005). RNase MRP is an endoribonuclease required for processing the precursor 45S rRNA into mature rRNA and, when deficient, leads to defective rRNA processing and impaired ribosome biogenesis (Narla and Ebert, 2010).
Mutation of the recQ-like helicase WRN is responsible for Werner syndrome, a disorder characterized by premature aging, an increased incidence of malignancy, as well as growth stunting and short stature. Similar to BLM, the
WRN helicase localizes to nucleoli and facilitates rRNA transcription (Shiratori et al., 2002). These disorders clearly suggest links between impaired ribosome biogenesis and growth deficiency.
Many BS persons have deregulated insulin signaling, observed as either insulin resistance in younger children, or as insulin-resistant diabetes mellitus in young adults in their twenties (Diaz et al., 2006). The presentation of insulin resistance is unique in BS, as it begins in early childhood and BS children are quite thin; this contrasts with the typical presentation of insulin-resistant diabetes mellitus in older overweight adults. Whereas postnatal growth hormone is a major mitogen, insulin is the predominant mitogen in utero. A familiar illustration is the macrosomia characteristic of neonates born to poorly controlled diabetic mothers whose fetuses are exposed to high insulin levels in utero (Riskin and
Garcia-Prats, 2011). Importantly, growth hormone and insulin both signal through the insulin receptor substrate-1 (IRS-1) pathway (Myers et al., 1994). In addition to the classical roles of IRS-1 in the insulin-signaling pathway, IRS-1 localizes to nucleoli, physically interacts with the RNA polymerase I transcription factor UBF and promotes rRNA transcription (Tu et al., 2002), suggesting that the
79 growth-promoting effect of insulin is mediated in part through its effects on ribosome biogenesis. Furthermore, Irs1-/- mice are insulin-resistant and have prenatal and postnatal growth retardation (Tamemoto et al., 1994) similar to that from impaired ribosome biogenesis. Thus, the absence of nucleolar BLM may limit the full potential of insulin- and growth hormone-stimulated rRNA transcription, also limiting growth in BS. Our data may help to understand the relationship between the cellular and metabolic abnormalities as well as the growth deficiency observed in BS.
80
CHAPTER FOUR
DNA topoisomerase I interacts directly with the BLM helicase and
stimulates its unwinding activity of duplex templates that mimic rDNA.
I. Introduction
Bloom’s syndrome (BS) is a rare autosomal recessive disorder characterized by a high predisposition to cancer and severe growth retardation
(German, 1969). The protein that is mutated in BS, BLM, belongs to the conserved recQ subfamily of ATP-dependent 3’-5’ helicases (Ellis/ Groden et al.,
1995; Karow et al., 1997) and localizes to the nucleolus (Sanz et al., 2000). The
C-terminus is required for nucleolar retention and rDNA binding within the 18S- coding region and the intergenic spacers (IGS) (Yankiwski et al., 2001;
Schawalder et al., 2003). BLM deficiency leads to hyper-recombination within rDNA (Therman et al., 1984; Killen et al., 2009) and a reduction of overall rDNA repeat numbers in comparison to wild-type cells (Yankiwski et al., 2001;
Schawalder et al., 2003). These data support the hypothesis that nucleolar BLM maintains the stability of rDNA via its direct binding to rDNA.
Human cells in interphase contain several nucleoli, sub-nuclear structures that contain the highly repetitive ribosomal DNA (rDNA) genes on the p-arms of the acrocentric chromosomes. Nucleolar rDNA associates with nucleolar-
81 dedicated RNA polymerase I and a multitude of other proteins required for ribosome biogenesis. The predominant function of nucleoli is transcription of ribosomal RNA (rRNA) from rDNA, occurring during S- and G2-phases of the cell cycle (Schwarzacher and Wachtler, 1993; Ayrault et al., 2006).
Nascent rRNAs generated during RNA polymerase I-mediated rRNA transcription have a tendency to re-associate with template rDNA and form rRNA:rDNA hybrids that can inhibit rRNA transcription and facilitate rDNA recombination (reviewed in Aguilera, 2002). DNA topoisomerase I, a component of the RNA polymerase I transcription complex, relaxes the negative supercoiling associated with rRNA transcription and prevents the formation of inhibitory rRNA:rDNA hybrids (Hraiky et al., 2000; Hage et al., 2010). BLM is also a component of the RNA polymerase I transcription complex and unwinds
RNA:DNA hybrids with 3’ overhangs of DNA (Grierson et al., 2012). Hyper- recombination within rDNA generates extra-chromosomal rDNA circles (ERC), the accumulation of which cause aging in S. cerevisiae (Sinclair and Guarente,
1997). BLM-deficient cells display rDNA hyper-recombination (Therman et al.,
1984; Killen et al., 2009), while some of the clinical characteristics of BS are associated with premature aging. These studies suggest that BLM and DNA topoisomerase I may cooperatively function to limit the accumulation of rRNA:rDNA hybrids in the nucleolus.
Here, we report that BLM interacts directly with DNA topoisomerase I.
Protein co-immunoprecipitation from nuclear extracts and sub-fractionated nuclei from cultured cells demonstrate that their interaction occurs in nucleoli. Purified
82 recombinant proteins co-immunoprecipitate in vitro, while in vitro transcription/translation (IVTT) coupled to immunoprecipitation demonstrates that the interaction is mediated by a small domain within the C-terminus of BLM.
Finally, we show using helicase assays that DNA topoisomerase I stimulates
BLM helicase activity 1.3-fold using an rDNA-like RNA20:DNA33 duplex substrate that models a co-transcriptionally formed rRNA:rDNA hybrid. Our data suggest that BLM and DNA topoisomerase I interact and cooperate to promote efficient rRNA transcription.
II. Materials and Methods
Cell lines. MCF7 and HEK 293T cells were obtained from ATCC and cultured in
Dulbecco’s Modified Eagle Medium (Invitrogen) containing 10% heat-inactivated
FBS (Hyclone). All cells were cultured at 37°C and 5% CO2.
Nucleolar isolation. Nucleoli were isolated from 293T cells according to the protocol of the Lamond Lab (www.lamondlab.com). Briefly, proliferating 293T cells were harvested by trypsinization, washed in PBS, re-suspended in buffer A
(10mM HEPES, pH7.9, 10mM KCl, 1.5mM MgCl2, 0.5mM DTT) and incubated on ice for 5 min. Cell suspensions were homogenized until approximately 90% of the cells were disrupted to produce intact nuclei. Lysis was monitored by light microscopy. Homogenized suspensions were centrifuged at 218g for 5 min at
4°C, nuclear pellets re-suspended in 3ml of S1 solution (0.25M sucrose, 10mM
83 MgCl2), layered over 3ml of S2 solution (0.38M sucrose, 0.5mM MgCl2), and centrifuged at 1430g for 5 min at 4°C. Resultant nuclear pellets were re- suspended in 3ml of S2 solution and sonicated at 4°C (Fisher Scientific Sonic
Dismembrator model 500). Liberation of nucleoli was monitored by light microscopy. Resultant nucleolar suspensions were layered over 3ml of S3 solution (0.88M sucrose, 0.5mM MgCl2), centrifuged at 3000g for 10 min at 4°C and re-suspended in 500ul of S2 solution.
Protein co-immunoprecipitation. Protein co-immunoprecipitations used 293T nuclear lysates prepared according to published protocols (Russell et al., 2011) or nucleolar and nucleoplasmic lysates prepared as described above. Antibodies used in co-immunoprecipitation included αBLM (Santa Cruz Biotech, sc-7790) and αDNA topoisomerase I (Bethyl, A302-589A). Protein-antibody complexes were captured with Dynabeads Protein G (Invitrogen, 100-04D), washed, eluted and separated by 8% SDS-PAGE and detected using standard western blotting procedures using αBLM (Bethyl Laboratories, A300-110A), αRPA194 (Santa
Cruz Biotech, sc-48385) and αRNA polymerase II (Abcam, ab817).
Protein purification. I. Hickson (Karow et al., 1997) provided the pYES-BLM expression vector (pJK1). BLM was purified as previously described (Russell et al., 2011), with an additional heparin-sepharose purification step. Briefly, hexa- histidine (6X-His)-tagged BLM was over-expressed in Saccharomyces cerevisiae. Yeast were lysed at 20k psi using a French Press Cell Disrupter 84 (Thermo) and lysates were separated by ultracentrifugation at 65,000g for 1hr at
4°C. Cleared lysates were purified by FPLC using a 1ml column of Ni-NTA
Superflow (Qiagen) with buffer A (50mM Tris, pH 7.5 at room temperature, 10% glycerol, 500mM NaCl, 1:1000 PMSF and 1:500 mammalian protease inhibitor cocktail [Sigma]) and buffer B (50mM Tris, pH 7.5 at room temperature, 10% glycerol, 500mM NaCl, 250mM imidazole, 1:1000 PMSF and 1:500 mammalian protease inhibitor cocktail [Sigma]) over a 50ml gradient from 14% buffer B to
100% buffer B. BLM eluted between 145mM and 250mM imidazole. The eluted
BLM was dialyzed for 4hrs at 4°C in 4L of 50mM Tris, pH 7.5 at room temperature, 10% glycerol, 1mM DTT, 1mM EDTA, 200mM NaCl, 1:10,000
PMSF and 1:10,000 mammalian protease inhibitor cocktail (Sigma). The resultant BLM was then purified using a 1ml column of Heparin-Sepharose 6
Fast Flow (Amersham Biosciences) with buffer A (50mM Tris, pH 7.5 at room temperature, 10% glycerol, 1mM DTT, 1mM EDTA, 1:10,000 PMSF and
1:10,000 mammalian protease inhibitor cocktail [Sigma]) and buffer B (50mM
Tris, pH 7.5 at room temperature, 10% glycerol, 1mM DTT, 1mM EDTA, 1M
NaCl, 1:10,000 PMSF and 1:10,000 mammalian protease inhibitor cocktail
[Sigma]) over a 30ml gradient from 20% buffer B to 100% buffer B. BLM eluted between 600mM NaCl and 920mM NaCl. The eluted BLM was dialyzed for 4hrs at 4°C in 4L of 50mM Tris, pH 7.5 at room temperature, 10% glycerol, 1mM DTT,
1mM EDTA, 200mM NaCl, 1:10,000 PMSF and 1:10,000 mammalian protease inhibitor cocktail (Sigma). The resultant BLM was then purified using a 1ml column of Q-Sepharose (Sigma) with buffer A (50mM Tris, pH 7.5 at room
85 temperature, 10% glycerol, 1mM DTT, 1mM EDTA, 1:10,000 PMSF and
1:10,000 mammalian protease inhibitor cocktail [Sigma]) and buffer B (50mM
Tris, pH 7.5 at room temperature, 10% glycerol, 1mM DTT, 1mM EDTA, 1M
NaCl, 1:10,000 PMSF and 1:10,000 mammalian protease inhibitor cocktail
[Sigma]) over a 10ml gradient from 20% buffer B to 70% buffer B. BLM eluted between 450mM NaCl and 550mM NaCl. Purity of the resultant BLM was determined by 8% SDS-PAGE, staining of gels with SYPRO Ruby Protein Gel
Stain (Sigma) and analysis using ImageQuant software.
In vitro transcription/ translation (IVTT). IVTT reactions were performed according to Lillard-Wetherell et al., 2004. Briefly, pET24D-BLM-N, pET24D-
BLM-H and pET24D-BLM-C were used in IVTT according to manufacturer’s instructions (TNT Rabbit Reticulocyte Lysate kit, Promega). IVTT products were mixed with full-length wild-type human DNA topoisomerase I (Topogen) according to published protocols (Langland et al., 2001) and incubated for 2hrs at
4°C with rotation. Subsequently, αDNA topoisomerase I (Bethyl, A302-589A) was added with an additional 2hr incubation at 4°C. Finally, Dynabeads Protein
G (Invitrogen, 100-04D) were added for 2hrs at 4°C, washed 5 times with binding buffer, eluted with 1X SDS-PAGE sample buffer, separated on 10% SDS-PAGE, dried and imaged using ImageQuant software.
Helicase assays. Oligonucleotides were purchased from Invitrogen.
Oligonucleotide sequences (5’-3’ orientation) are as follows: DNA20- 86 CGCTAGCAATATTCTGCAGC, DNA33-
GCTGCAGAATATTGCTAGCGGGAATTCGGCGCG and RNA20-
32 CGCUAGCAAUAUUCUGCAGC. RNA20 and DNA20 were P end-labeled using polynucleotide kinase (PNK; NEB) according to manufacturer’s instructions.
Duplex substrates were generated by heating to 95°C for 5 min and slow cooling.
Helicase assays were performed as previously described (Grierson et al., 2012).
Helicase products were separated on 12% non-denaturing polyacrylamide gels, dried and analyzed using ImageQuant software.
III. Results
BLM interacts with DNA topoisomerase I from nuclear and nucleolar- enriched extracts.
BLM localizes to the nucleolus (Sanz et al., 2000), interacts with RNA polymerase I and facilitates RNA polymerase I-mediated rRNA transcription
(Grierson et al., 2012). DNA topoisomerase I functions as a component of the
RNA polymerase I transcription complex and is required for efficient rRNA transcription in bacteria, yeast and human cells (Hraiky et al., 2000; Hage et al.,
2010; Zhang et al., 1988). Therefore, we asked whether BLM and DNA topoisomerase I co-immunoprecipitate from nuclear extracts from MCF7 and
293T cells using αBLM and αDNA topoisomerase I antibodies. Figure 12A shows that BLM and DNA topoisomerase I co-immunoprecipitate suggesting a common RNA polymerase I-associated complex.
87 As BLM and DNA topoisomerase I co-localize in the nucleolus and the nucleoplasm, we asked whether the BLM-DNA topoisomerase I interaction also occurs in the nucleolus. Nuclei were fractionated into nucleoli and nucleoplasm
(Figure 12B) and the resultant sub-nuclear fractions used in protein co- immunoprecipitation with αBLM and αDNA topoisomerase I. The RNA polymerase I-specific subunit RPA194 was used as a nucleolar marker, RNA polymerase II as a nucleoplasmic marker and beta-actin as a cytoplasmic marker. Figure 12C shows the enhanced interactions of BLM and DNA topoisomerase I in nucleoli, suggesting that their function may be specific to nucleolar metabolism.
88
Figure 12. BLM and DNA topoisomerase I associate in 293T cells. (A) Co- immunoprecipitations used MCF7 and 293T nuclear extracts, and either αBLM or
αDNA topoisomerase I. Immunoprecipitated proteins were separated using 8%
SDS-PAGE and blotted using αBLM or αDNA topoisomerase I. Goat IgG was chosen as an isotype-matched negative control for αBLM; rabbit IgG was an isotype-matched negative control for αDNA topoisomerase I. (B) 293T nuclei were sub-fractionated into nucleoli and nucleoplasm. The purity of the sub- cellular fractions is shown using the RNA polymerase I-specific subunit RPA194 as a nucleolar marker; RNA polymerase II as a nucleoplasmic marker (indicated with an *); and beta-actin as a cytoplasmic marker. (C) Nucleolar and nucleoplasmic fractions were used in protein co-immunoprecipitation with αBLM.
Immunoprecipitated proteins were separated using 8% SDS-PAGE and blotted using αBLM, αDNA topoisomerase I, αRPA194 and αRNA polymerase II. Goat
IgG was used as an isotype-matched negative control for αBLM.
To determine whether the interaction between BLM and DNA topoisomerase I is direct or indirect, purified full-length human proteins were used for in vitro protein co-immunoprecipitations. Recombinant BLM was expressed in S. cerevisiae and FPLC-purified as previously described; DNA topoisomerase I was purchased from Topogen (cat # TG2005H-RC1). Proteins are shown in Figure 13A. Immunoprecipitation with either αBLM or αDNA 89 topoisomerase I recovered both proteins demonstrating a direct protein-protein interaction (Figure 13B).
90 Figure 13. BLM and DNA topoisomerase I directly interact. (A) The purity of recombinant FPLC-purified BLM and purchased DNA topoisomerase I (Topogen) was determined using 8% SDS-PAGE and staining with Sypro Ruby Protein Gel
Stain (Sigma). DNA topoisomerase I is supplied with 50ng/ul of bovine serum albumin (BSA), as indicated. (B) BLM and DNA topoisomerase I were used in in vitro protein co-immunoprecipitation using αBLM or αDNA topoisomerase I.
Immunoprecipitated proteins were separated using 8% SDS-PAGE and blotted using αBLM and αDNA topoisomerase I. Goat IgG was used as an isotype- matched negative control for αBLM; rabbit IgG was used as an isotype-matched negative control for αDNA topoisomerase I.
BLM and DNA topoisomerase I interact directly in vitro via the BLM C- terminus.
BLM is a modular protein with distinct domains dictating localization to specific sub-nuclear regions. The N-terminus of BLM is required for its accumulation in PML bodies, while the C-terminus of BLM is required for localization to nucleoli (Yankiwski et al., 2001). Expression of pGFP-BLM deletion constructs that retain the BLM NLS sequence in BS cells demonstrates that deletion of amino acids 1118-1164 or amino acids 1166-1331, although permissive for nuclear localization and co-localization with PML, prevents nucleolar accumulation of BLM. These data demonstrate that nucleolar accumulation of BLM is specified by its C-terminal region within amino acids 91 1118-1331. Importantly, this altered sub-nuclear localization is not an artifact of impaired helicase activity, as the helicase-dead GFP-BLM K695T still localizes to nucleoli while GFP-BLM deletion constructs without amino acids 1118-1164 or
1166-1331 remain helicase-proficient (Yankiwski et al., 2001; Schawalder et al.,
2003). Full-length wild-type GFP-BLM directly binds nucleolar rDNA, both in the rDNA transcription unit as well as in the intergenic spacer region (IGS)
(Schawalder et al., 2003). GFP-BLM deletion constructs without amino acids
1118-1164 or 1166-1331 that fail to localize to the nucleolus also fail to bind rDNA, suggesting that BLM binding to rDNA mediates nucleolar retention of BLM
(Schawalder et al., 2003).
In vitro transcription/translation (IVTT) of BLM using 35S-methionine was used to generate the N-terminal (N: amino acids 25-612), helicase (H: amino acids 613-1025) and C-terminal (C: amino acids 1021-1417) segments of BLM.
Protein co-immunoprecipitation of these segments with full-length human DNA topoisomerase I showed that the C-terminal region of BLM selectively interacts with DNA topoisomerase I (Figure 14). These data suggest that the C-terminal region of BLM may not only mediate BLM nucleolar localization and rDNA binding (Yankiwski et al., 2001; Schawalder et al., 2003), but also mediate direct protein-protein interaction with other known nucleolar proteins.
92
Figure 14. DNA topoisomerase I interacts with the C-terminus of BLM. In vitro transcription/translation (IVTT) was performed using the TNT Rabbit
Reticulocyte Lysate kit (Promega) using pET24D-BLM-N, pET24D-BLM-H and pET24D-BLM-C. The resultant 35S-methionine-labeled input BLM fragments are shown on the left. N-, H- and C-fragments were used in co-immunoprecipitation with purified DNA topoisomerase I protein (Topogen) and αDNA topoisomerase I.
Following capture with Protein G Dynabeads (Invitrogen), immunoprecipitated proteins were separated using 8% SDS-PAGE; western blotting identified DNA topoisomerase I; autoradiography identified BLM fragments. Negative control
93 immunoprecipitations were carried out by omitting DNA topoisomerase I to control for non-specific binding of BLM fragments to αDNA topoisomerase I or to the beads.
DNA topoisomerase I stimulates BLM helicase activity on RNA20:DNA33 substrates designed to model rRNA/rDNA hybrids.
Given the direct interactions of BLM and DNA topoisomerase I, their independent functions as components of an RNA polymerase I-associated transcription complex to prevent rRNA:rDNA hybrid formation (Hraiky et al.,
2000) or resolution of rRNA:rDNA hybrids (Grierson et al., 2012), we asked whether BLM and DNA topoisomerase I cooperate in vitro to unwind substrates presumed to form in the nucleolus during rRNA transcription and rDNA replication. In vitro helicase assays used FPLC-purified BLM and equimolar human DNA topoisomerase I (Topogen) with the RNA20:DNA33 duplex substrate to model a co-transcriptionally formed rRNA:rDNA hybrid. Helicase assays evaluated BLM concentrations ranging from 0.15nM-0.3nM at time points from 15 sec-6 min (Figure 15A). BLM by itself or BLM with equimolar DNA topoisomerase I unwinds a GC-rich RNA20:DNA33 duplex following Michaelis-
Menten kinetics (Figure 15B). Equimolar DNA topoisomerase I stimulates BLM unwinding activity at all protein concentrations studied (Figure 15C-F). The specific activity of BLM unwinding RNA20:DNA33 is 3.11 fmol unwound/ min/ nM
BLM, which increases to 4.17 fmol unwound/ min/ nM BLM in the presence of
94 equimolar DNA topoisomerase I. DNA topoisomerase I did not display unwinding activity by itself. Our data are analogous to that of Hu et al. (2001),
Bhattacharyya et al., (2009) and Russell et al. (2011), which demonstrate that
DNA topoisomerase IIIα and DNA topoisomerase IIα, respectively, modulate
BLM helicase activity.
95 Figure 15. Equimolar DNA topoisomerase I stimulates the helicase activity of BLM on RNA20:DNA33. The purity of BLM and DNA topoisomerase I is shown in Figure 2A. (A) Representative gels demonstrate the unwinding activity of
BLM. BLM (0.15-0.3nM) was incubated with RNA20:DNA33 alone or in the presence of equimolar DNA topoisomerase I for 15 sec, 30 sec, 45 sec, 1 min,
1.5 min, 2 min, 3 min, 4 min and 6 min as described in Materials and Methods.
Reaction products were separated using 12% non-denaturing acrylamide gels, dried and imaged using ImageQuant software. Unwinding is evaluated by conversion of the double-stranded substrate to single-stranded substrate. Heat- denatured substrate (HD) is generated by heating to 95ºC for 5 min followed by rapid cooling. DNA topoisomerase I (topo I) does not display unwinding activity by itself. (B) Kinetics of BLM unwinding of RNA20:DNA33 alone or with equimolar
DNA topoisomerase I. Unwinding was calculated by quantifying the amount of single-stranded substrate relative to the total amount of substrate in the reaction.
All unwinding reactions follow Michaelis-Menten kinetics. (C-F) Shown are the component curves (from B) of BLM alone or BLM with DNA topoisomerase I at each concentration. (G) The specific activity of BLM alone (3.11 fmol unwound/ min/ nM BLM) or BLM with DNA topoisomerase I (4.17 fmol unwound/ min/ nM
BLM) on RNA20:DNA33 was calculated using the initial unwinding rates up to 45 sec.
96
continued
97 Figure 15 continued
continued
98 Figure 15 continued
IV. Discussion
Regulation of rDNA transcription occurs at both transcription initiation and elongation. Transcriptionally formed rRNA:rDNA hybrids inhibit transcription elongation and slow the overall rRNA transcriptional output. In bacteria, yeast and human cells, DNA topoisomerase I functions as a component of the RNA
99 polymerase I transcription complex to relax rDNA template negative supercoiling to disfavor the formation of RNA:DNA hybrids and facilitate rRNA transcription.
The role of BLM in facilitating rRNA transcription and its ability to unwind
RNA:DNA hybrids suggested that BLM and DNA topoisomerase I may interact directly and cooperate to unwind these rRNA transcription hybrids.
These studies demonstrate that BLM and DNA topoisomerase I interact in nucleolar-enriched extracts of MCF7 and 293T cells, consistent with both proteins being components of the RNA polymerase I transcription complex. This direct interaction is mediated by the C-terminal domain of BLM, the same region required for BLM nucleolar localization. Previously characterized interactions of
BLM with other DNA topoisomerases, for example IIIα and IIα, occur primarily via the N-terminal region of BLM (Russell et al., 2011; Hu et al., 2001). Our studies show that more than one region of BLM mediates interaction with all topoisomerases, suggesting that specific interactions modulate specific functions of BLM and that these interactions can be targeted independently.
BLM interactions with DNA topoisomerase IIα and DNA topoisomerase
IIIα modify BLM helicase activity on unique nucleic acid substrates (Russell et al.,
2011; Bhattacharyya et al., 2009; Hu et al., 2001). We asked whether the interaction between BLM and DNA topoisomerase I similarly influences BLM activity. In vitro helicase assays demonstrated that equimolar DNA topoisomerase I stimulates BLM unwinding activity on a GC-rich RNA20:DNA33 substrate, modeling a co-transcriptionally formed rRNA:rDNA hybrid. Our in vitro data suggest that BLM and DNA topoisomerase I may function cooperatively to
100 modulate rRNA:rDNA hybrid formation in the nucleolar rDNA. This newly identified functional interaction between BLM and DNA topoisomerase I further strengthens the view of conserved interactions between helicases and topoisomerases.
These experiments suggest a model in which BLM and DNA topoisomerase I are components of the RNA polymerase I transcription complex and serve to exclude rRNA:rDNA hybrids from the rDNA. Prevention of
RNA:DNA hybrid formation by DNA topoisomerase I is imperfect, as some
RNA:DNA hybrids form in DNA topoisomerase I-proficient cells (Hage et al.,
2010). The presence of BLM within the RNA polymerase I transcription complex may allow erroneously formed RNA:DNA hybrids to be removed, thereby preventing interference with transcription and replication. The ability of DNA topoisomerase I to stimulate this ability of BLM may further ensure exclusion of
RNA:DNA hybrids from the rDNA.
The coordinate functioning of BLM and DNA topoisomerase I to promote efficient rRNA transcription and thus cell growth has therapeutic implications for cancer therapeutics. Cells deficient in either BLM or DNA topoisomerase I display slower growth rates (Lechner et al., 1983; Hraiky et al., 2000). Drugs targeting DNA topoisomerase I are used chemotherapeutically, such as topotecan to treat small cell lung cancer and irinotecan to treat colon cancer
(UpToDate online). Understanding the nucleolar mechanisms by which BLM functions would provide the opportunity to inhibit selectively its growth-promoting activity in nucleolar metabolism while preserving its role in maintaining extra-
101 nucleolar genomic stability. Thus, chemotherapeutic dual targeting of DNA topoisomerase I and BLM nucleolar function may synergistically inhibit tumor cell growth.
102
CHAPTER FIVE
Nucleolar trafficking of BLM and protein partnering with nucleophosmin
I. Introduction
The BLM helicase localizes to PML bodies, at sites of DNA double strand break repair and at telomeres. BLM also localizes to the nucleolus where it regulates rRNA transcription and rDNA recombination (Sanz et al., 2000;
Grierson et al., 2012; Killen et al., 2009). The C-terminal domain of BLM is required for nucleolar retention and rDNA binding (Yankiwski et al., 2001;
Schawalder et al., 2003), however the mechanism by which BLM is trafficked to nucleoli is unknown. The protein nucleophosmin (NPM/B23) localizes to the nucleoplasm but predominantly to nucleoli where it participates in ribosome biogenesis. NPM mediates nucleoplasmic/nucleolar trafficking of several proteins via direct protein-protein interaction. For example, the tumor suppressor p19ARF physically interacts with and is trafficked to nucleoli by NPM (Colombo et al., 2005), as is the RNA polymerase I transcription factor TTF-1 (Lessard et al., 2010). In these experiments, we tested whether NPM mediates nucleolar trafficking of BLM.
These studies show that BLM and NPM co-localize in nucleoli and co- immunoprecipitate from nuclear lysates of HEK 293T and MCF7 cells. Sub-
103 nuclear fractionation and co-immunoprecipitation demonstrate that BLM and
NPM interact in nucleoli and nucleoplasm. These data suggest that NPM may mediate the trafficking of BLM from the nucleoplasm into nucleoli.
II. Materials and Methods
Cell lines. HEK 293T and MCF7 cells were obtained from ATCC and cultured in
Dulbecco’s Modified Eagle Medium at 37°C and 10% fetal bovine serum
(Hyclone). All cells were cultured at 37°C and 5% CO2.
Transfection, immunofluorescence and antibodies. pGFP-BLM was generated by cloning BLM cDNA into pEGFP-C1 (Clontech). MCF7 cells were transfected with the pGFP-BLM expression vector using Effectene Transfection
Reagent (Qiagen) according to manufacturer’s instructions. Forty-eight hours after transfection, cells were washed in phosphate-buffered saline, fixed with
10% formamide (Sigma), stained with α-nucleophosmin (NPM, Abcam FC82291) mounted on glass slides and subsequently scored for localization of GFP-BLM using a Zeiss Axiovert 200M microscope and Axiovision 4.5 software.
Nucleolar isolation. Nucleoli were isolated from 293T cells according to the protocol of the Lamond Lab (www.lamondlab.com). Briefly, proliferating 293T cells were harvested by trypsinization, washed in PBS, re-suspended in buffer A
(10mM HEPES, pH7.9, 10mM KCl, 1.5mM MgCl2, 0.5mM DTT) and incubated on 104 ice for 5 min. Cell suspensions were homogenized until approximately 90% of the cells were disrupted to produce intact nuclei. Lysis was monitored by light microscopy. Homogenized suspensions were centrifuged at 218g for 5 min at
4°C, nuclear pellets re-suspended in 3ml of S1 solution (0.25M sucrose, 10mM
MgCl2), layered over 3ml of S2 solution (0.38M sucrose, 0.5mM MgCl2), and centrifuged at 1430g for 5 min at 4°C. Resultant nuclear pellets were re- suspended in 3ml of S2 solution and sonicated at 4°C (Fisher Scientific Sonic
Dismembrator model 500). Liberation of nucleoli was monitored by light microscopy. Resultant nucleolar suspensions were layered over 3ml of S3 solution (0.88M sucrose, 0.5mM MgCl2), centrifuged at 3000g for 10 min at 4°C and re-suspended in 500ul of S2 solution.
Protein co-immunoprecipitation and western blotting. Protein co- immunoprecipitations using 293T nuclear lysates were prepared according to
Jiang et al., 2003. Nucleolar lysates or nucleoplasmic lysates were used in co- immunoprecipitation prepared as previously described. Antibodies for co- immunoprecipitation included α-BLM (Santa Cruz Biotech, sc-7790) and α-NPM
(Santa Cruz Biotech, sc-47725). Protein-antibody complexes were captured with
Dynabeads Protein G (Invitrogen, 100-04D), washed, eluted and separated by
8% SDS-PAGE. Western blotting was performed according to standard procedures using α-BLM (Bethyl Laboratories, A300-110A), α-beta-actin
(Abcam, ab8226-25),α-RPA194 (Santa Cruz Biotech, sc-48385), α-RNA
105 polymerase II (Abcam, ab817) and α-NPM (Santa Cruz Biotech, sc-6013-R or
Santa Cruz Biotech, sc-47725).
III. Results
BLM and NPM co-localize in nucleoli of MCF7 cells.
MCF7 cells were transiently transfected with pGFP-BLM, processed for immunofluorescence and stained with an antibody specific for the nucleolar marker NPM. Figure 16A demonstrates that BLM and NPM co-localize strongly in nucleoli. As shown earlier in Chapter 3, BLM and NPM both dissociate from nucleoli following treatment with actinomycin D, consistent with their involvement in nucleolar ribosome biogenesis (Grierson et al., 2012). These observations led us to investigate whether the two proteins interact in order to test the hypothesis that NPM mediates nucleolar trafficking of BLM.
BLM and NPM co-immunoprecipitate from nuclear extracts of 293T cells.
Proteins trafficked to nucleoli by NPM must physically interact with NPM
(Colombo et al., 2005). Experiments to test whether BLM and NPM interact used nuclear lysates from 293T cells for protein co-immunoprecipitation with α-
BLM and α-NPM antibodies. Figure 16B shows IP-westerns of BLM and NPM co-immunoprecipitated from 293T nuclear extracts, confirming an interaction between the two proteins.
106
Figure 16. GFP-BLM co-localizes and interacts with NPM. A) MCF7 cells were transiently transfected with pGFP-BLM expression vector and stained with
α-NPM to demonstrate co-localization in nucleoli. B) Nuclear lysates from 293T cells were used in protein co-immunoprecipitation using α-BLM and α-NPM antibodies. Immunoprecipitated proteins were separated using 8% SDS-PAGE and blotted with α-BLM and α-NPM; goat IgG is isotype-matched negative control for α-BLM; mouse IgG is isotype-matched negative control for α-NPM.
BLM and NPM co-immunoprecipitate in nucleolar and nucleoplasmic fractions of 293T cells.
If NPM mediates nucleolar trafficking of BLM, both proteins would be expected to interact in the nucleoli and in the nucleoplasm. Lysates from 293T 107 cells were sub-nuclear fractionated to isolate nucleoli and nucleoplasm. Western blotting was used to control for the purity of the sub-cellular fractions with RNA polymerase I-specific subunit RPA194 as a nucleolar marker, RNA polymerase II as a nucleoplasmic marker and beta-actin as a cytoplasmic marker (Figure 17A).
Nucleolar and nucleoplasmic fractions were used in co-immunoprecipitation experiments with α-BLM and α-NPM antibodies. Figure 17B shows that BLM and NPM interact in both nucleoli and nucleoplasm. Detection of the interaction between BLM and RPA194 was used as a positive control in nucleoli. These data suggest that NPM may mediate nucleolar trafficking of BLM.
108 Figure 17. BLM interacts with NPM in nucleoli and nucleoplasm. A) nucleoli were isolated from HEK 293T cells according to published protocols (Lamond
Lab; www.lamondlab.com). The purity of nucleolar, nucleoplasmic and cytoplasmic fractions was monitored using 8% SDS-PAGE and blotting with α-
RPA194 (a nucleolar marker), α-RNA polymerase II (a nucleoplasmic marker; the
RNA pol II specific band is denoted with an *) and α-beta-actin (a cytoplasmic marker). B) The nucleolar (or nucleoplasmic) protein fraction was used in protein co-immunoprecipitation with α-BLM or α-NPM, or goat IgG or mouse IgG as negative controls, respectively, separated using 8% SDS-PAGE and blotted with
α-BLM, α-NPM or α-RPA194.
IV. Discussion
Nucleoli are non-membrane-bound sub-nuclear organelles that contain many but not all nuclear proteins. Trafficking of proteins to the nucleolus is mediated by various mechanisms including direct protein-protein interaction with resident nucleolar proteins, interactions with a trafficking protein in the nucleoplasm, binding to nucleolar rDNA or rRNA or binding to nucleolar non- coding RNA (ncRNA). Nucleolar targeting is a dynamic process, as proteins continually move between nucleoli as well as between nucleoli and the nucleoplasm.
109 Nucleophosmin (NPM) localizes to the granular component (GC) of the nucleolus and participates in ribosome assembly (Ochs et al., 1983). NPM is required for maintenance of nucleolar structure and interacts with many nucleolar proteins, both in the nucleolus and in the nucleoplasm, where it subsequently traffics other proteins to nucleoli. For example, nucleolar localization of the tumor suppressor p19ARF depends on NPM expression and the interaction of the two proteins (Colombo et al., 2005). The RNA polymerase I transcription termination factor 1 (TTF-1) is trafficked to nucleoli by NPM (Lessard et al., 2010).
Furthermore, NPM may act as a hub protein, as it is a relatively small protein and has many disordered domains, both characteristics of hub proteins. In contrast to the well-defined nuclear localization sequence (NLS) that mediates nuclear localization, a minority of nucleolar proteins has a nucleolar localization sequence (NoLS). Nucleolar localization sequences, when present, are generally 7-20 amino acids long and rich in lysine and arginine residues
(reviewed in Emmott and Hiscox, 2009). The NoLSs are not as highly conserved as the NLS and are largely identified on a functional basis—that is, a domain required for nucleolar localization, often by its involvement in nucleolar protein- protein interactions (reviewed in Emmott and Hiscox, 2009).
Those nucleolar proteins with a role in ribosome biogenesis are sensitive to changes in the activity of RNA polymerase I, including inhibition with actinomycin D (AMD). Exposure to AMD causes NPM, BLM and RNA polymerase I to dissociate from the nucleolus and diffusely distribute into the nucleoplasm. This is consistent with the established roles of these proteins in
110 ribosome biogenesis (Ochs et al., 1983; Grierson et al., 2012) as well as supportive of the role of NPM in trafficking BLM to the nucleolus.
111
CHAPTER SIX
Thesis Summary
The BLM gene encodes the 1417 amino acid BLM helicase that is mutated in the growth-defective, cancer-prone disorder Bloom’s syndrome (BS).
BLM expression is required for genomic stability, as the occurrence of abnormal chromosome structures increases in its absence. BLM localizes to PML bodies with other DNA repair proteins and participates in homologous recombination- mediated and non-homologous end-joining DNA double-strand break repair pathways. BLM also localizes to telomeres and functions in the recombination- mediated telomere maintenance pathway designated alternative lengthening of telomeres (ALT) (reviewed in Chu and Hickson, 2009; Langland et al., 2001).
Lastly, BLM localizes to nucleoli. It is this function of BLM that is the topic of this dissertation research.
BS cells display instability of the nucleolar ribosomal DNA repeats
(Therman et al., 1981). BLM expression is cell-cycle regulated and localizes to the nucleolus during rRNA transcription (Sanz et al., 2000; Klein and Grummt,
1999). A mechanistic role for nucleolar BLM had not been previously established, although studies of G-quartet (G4) DNA have demonstrated the proficiency of BLM to unwind this structure, known to occur during rDNA
112 replication (Sun et al., 1998). Preliminary data from our lab suggested that BLM may function in nucleolar rRNA metabolism. This thesis tests the hypothesis that nucleolar BLM facilitates rRNA and rDNA metabolism.
The RNA polymerase I inhibitor actinomycin D (AMD) is commonly used to test for the involvement of a nucleolar-localized protein in various steps of ribosome biogenesis. Following AMD treatment and subsequent inhibition of
RNA polymerase I, those proteins involved in ribosome biogenesis diffuse from the nucleoli to the nucleoplasm (reviewed in Drygin et al., 2010). We first performed such immunofluorescence experiments in two cancer cell lines to determine the nucleolar localization of BLM and to observe its response to AMD treatment. Cell lines were transfected with a pGFP-BLM expression construct and treated with AMD, the RNA polymerase II inhibitor alpha-amanitin, or various
DNA-damaging agents. BLM nucleolar localization was sensitive to AMD and the DNA-damaging agent 4NQO, both inhibitors of nucleolar rRNA transcription
(reviewed in Drygin et al., 2010; Gray et al., 1998). BLM nucleolar localization was insensitive to the RNA polymerase II inhibitor alpha-amanitin and the replication fork stalling drug hydroxyurea. These data demonstrate that nucleolar localization of BLM depends on RNA polymerase I activity and implies that its primary function, unlike that in other regions of the genome, may not be to resolve stalled replication forks. Our study is the first to analyze the sensitivity of
BLM nucleolar localization.
BLM is a non-processive helicase that is proficient in the unwinding of unusual nucleic acid structures (Popuri et al., 2008) such as G4 DNA and
113 RNA:DNA hybrids, both of which form in the nucleolus. We hypothesized that
BLM functions in rRNA transcription. Transcription assays were designed to measure the dependence of rRNA transcription rate on BLM expression in HEK
293T cells knocked down for BLM with siRNA: in vivo biotin-labeled nuclear run- on assays, 32P-labeled nuclear run-on/dot blot assays and 3H-uridine pulse- chase assays. As a complementary approach, we also performed 3H-uridine pulse-chase assays in GM08505 Bloom’s syndrome fibroblasts re-expressing
BLM. These data demonstrated that BLM-deficient cells display an approximately 2-fold slower rate of rRNA transcription in comparison to matched cells expressing BLM. BLM expression does not affect rRNA processing, as the mature 18S rRNA accumulates with similar kinetics during the pulse-phase of 3H- uridine pulse-chase experiments regardless of BLM expression. Our findings are similar to those of Shiratori et al. (2002) who demonstrated that the related recQ- like helicase WRN, mutated in the chromosome breakage and cancer-prone disorder Werner’s syndrome facilitates rRNA transcription.
The co-transcriptional formation of RNA:DNA hybrids in the rDNA slows movement of the RNA polymerase I transcription complex and thus the overall rRNA transcription rate. This is an evolutionarily conserved process, occurring in multiple organisms (Hraiky et al., 2000; Hage et al., 2010). As BLM is a helicase capable of unwinding unusual nucleic acid structures, we next asked whether
BLM function in rRNA transcription is to unwind inhibitory RNA:DNA hybrids. In vitro helicase assays were carried out using DNA:DNA and RNA:DNA substrates to define the structural requirements that are permissive for BLM unwinding
114 (Popuri et al., 2008). BLM is capable of unwinding DNA:DNA and RNA:DNA duplexes with a DNA overhang, but is unable to unwind duplexes with an RNA overhang of similar length. We then tested whether the inability of BLM to unwind RNA overhang duplexes is due to an inability to bind these substrates.
Electrophoretic mobility shift assays (EMSA) were performed and demonstrated that BLM is unable to bind ssRNA, potentially explaining the lack of unwinding of these substrates. This finding is consistent with that reported for replicative helicases from other organisms, which similarly unwind RNA:DNA hybrids specifically by binding to and translocating on the DNA strand (Shin and Kelman,
2006). It may be that the binding conditions used in our assays were not ideal for
RNA binding, or that the length of the experimental substrate overhang was of insufficient length (n=13) for BLM binding. Insufficient overhang length is an unlikely explanation, as an 8-nucleotide overhang is sufficient for unwinding duplex DNA (Popuri et al., 2008). Alternatively, the active site of BLM that binds single-strand DNA may not accommodate the additional 2’OH that defines RNA.
Further experiments could be designed to show more definitively that
RNA:DNA hybrids are the source of decreased rRNA production in cells without
BLM. Anti-RNA:DNA hybrid antibodies (Hage et al., 2010) could be used to quantify RNA:DNA hybrid abundance following BLM depletion. Additionally, as
RNA:DNA hybrids induce transcription complex stalling and “pile up” on the rDNA, the ability of Rnase H to rescue any RNA polymerase I transcription complex pile up in BLM-deficient cells would also support the involvement of this structure (Hage et al., 2010).
115 RNA:DNA hybrid formation and their persistence in the absence of BLM may not be the only structure that inhibits rRNA transcription in the absence of
BLM. rDNA has a high propensity to form G4 DNA, similarly capable of stalling replication and transcription complexes (reviewed in Chu and Hickson, 2009).
The hypothesis that persistent G4 DNA in the absence of BLM leads to transcriptional slowing may be tested using the drug NMM (N-methyl mesoporphyrin IX)—a G4 DNA stabilizer that inhibits its unwinding by BLM—in transcriptional assays (Huber et al., 2002). Specifically, an absence of rRNA transcriptional recovery following BLM re-expression in the presence of NMM would implicate G4 DNA as a transcriptional block in BS cells.
Having established a role for BLM in rRNA transcription, we then determined some of the components of the nucleolar protein complex in which
BLM functions. HEK 293T nuclear extracts were used for protein co- immunoprecipitation with antibodies specific for BLM and the RNA polymerase I- specific subunit RPA194. RPA194 was chosen as many of the RNA polymerase
I subunits are shared with RNA polymerase II or III and RPA194 is not. We determined that RPA194 and BLM interact indirectly or directly, as each antibody was used inversely for immunoprecipitation and western analysis. The interaction of BLM with the RNA polymerase I complex is specific, as we were unable to identify an interaction between BLM and components of RNA polymerase II using protein co-immunoprecipitation. Experiments were repeated in the breast cancer cell line MCF7, confirming that BLM is a component of the
116 RNA polymerase I transcription complex and suggesting that BLM functions co- transcriptionally to modulate RNA:DNA hybrid formation.
DNA topoisomerase I is a component of the RNA polymerase I transcription complex and relaxes negative supercoils induced during rRNA transcription to inhibit rRNA:rDNA hybrid formation and facilitate transcription
(Hage et al., 2010). Therefore, we investigated an interaction between BLM and
DNA topoisomerase I to test the hypothesis that they function in an analogous way to promote rRNA transcription. Protein co-immunoprecipitation identified an interaction between the two proteins using extracts from two cell lines. Nuclei were sub-fractionated to confirm the specificity of co-immunoprecipitation to the nucleolar fraction. In vitro immunoprecipitations using purified proteins demonstrated that the interaction is direct. Subsequently, the DNA topoisomerase I interaction region was mapped to the C-terminus of BLM. The
C-terminus is also required for nucleolar retention and rDNA binding, suggesting that the BLM C-terminus may be important for many of its nucleolar functions
(Yankiwski et al., 2001; Schawalder et al., 2003). Preliminary informatic analyses of the BLM protein sequence identify a number of candidate phosphorylation sites in this region suggesting that phosphorylation could also regulate BLM/DNA topoisomerase I interactions. These sites identified by
KinasePhos 2.0 include: S1209, S1296, S1342, S1352, S1355 and S1386
(Groden, unpublished data).
In addition to BLM interactions with DNA topoisomerase I, BLM interacts with DNA topoisomerase IIα (Bhattacharyya et al., 2009; Russell et al., 2011)
117 and DNA topoisomerase IIIα (Hu et al., 2001). These three studies also confirmed the effects of topoisomerase interactions on BLM unwinding proficiency with a number of DNA duplex substrates. We investigated the consequences of the interaction between BLM and DNA topoisomerase I using in vitro helicase assays with the RNA20:DNA33 substrate to model co- transcriptionally-formed rRNA:rDNA hybrids. BLM was used with equimolar DNA topoisomerase I, a condition suggested to be optimal based on published work with DNA topoisomerase IIα (Bhattacharyya et al., 2009; Russell et al., 2011).
Equimolar DNA topoisomerase I stimulates BLM helicase activity using the
RNA20:DNA33 duplex. These data support a model in which DNA topoisomerase
I and BLM cooperate to prevent formation of or unwind, respectively, rRNA:rDNA hybrids that form during rRNA transcription.
Future experiments could use varying BLM:DNA topoisomerase I molar ratios to validate that the 1:1 molar ratio is optimal. Studies using DNA20:DNA33 would also demonstrate that DNA topoisomerase I effects on unwinding might be specific for the RNA:DNA substrate only. Other substrates could also be studied, particularly a supercoiled plasmid-based R-loop and G4 DNA. A supercoiled plasmid-based R-loop would more closely model the RNA:DNA hybrid in the context of transcription. Helicase assays could also be performed with BLM phosphorylation site mutants within the region of interaction in the C-terminus of
BLM.
At present, the mechanism by which DNA topoisomerase I modulates
BLM activity is unknown. Mechanisms could include BLM processivity or
118 alteration of substrate-binding affinity. Functional studies investigating the interaction between BLM and RPA194, or other components of the RNA polymerase I transcription machinery, would also be informative as they may regulate or modify BLM function. DNA topoisomerase I-mediated plasmid relaxation assays could be performed to investigate a reciprocal effect of BLM in modulating DNA topoisomerase I activity. Fine-mapping of the BLM C-terminal sub-domain with which DNA topoisomerase I interacts will allow generation of a
BLM deletion mutant that localizes to nucleoli but that doesn’t interact with DNA topoisomerase I. Such a BLM deletion mutant in cellular transcriptional assays would permit investigation of the role of the BLM/ DNA topoisomerase I interaction in facilitating rRNA transcription.
Nucleolar trafficking of proteins is a regulated process, driven by protein- protein interaction or protein-nucleic acid interaction (reviewed in Emmott and
Hiscox, 2009). The C-terminal domain of BLM is required for nucleolar retention
(Schawalder et al., 2001), although the mechanism for nucleolar trafficking of
BLM has not been described. The similar responses of BLM and NPM to RNA polymerase I inhibition and the established roles of NPM in mediating nucleolar trafficking (reviewed in Emmott and Hiscox, 2009) suggested a role for NPM in
BLM nucleolar trafficking. These proteins co-localize and interact in nuclear extracts from HEK 293T cells. Sub-fractionated nuclei were isolated and used for co-immunoprecipitation. NPM and BLM co-immunoprecipitate in nucleoli and in the nucleoplasm, a predicted result if one mediates the nucleolar trafficking of the other. Future studies will determine whether NPM is responsible for nucleolar
119 trafficking of BLM. Knock down of NPM with siRNA to assess nucleolar accumulation of BLM may be problematic, as NPM is required for nucleolar integrity (reviewed in Emmott and Hiscox, 2009). Another approach may be to map the domain of BLM that interacts with NPM, delete it in the pGFP-BLM expression vector, express in cells and assay for nucleolar accumulation of GFP-
BLM using immunofluorescence or by sub-nuclear fractionation and western blotting. Both phosphorylation and SUMOylation regulate BLM localization to
PML bodies. Thus the impact of these post-translational modifications on BLM nucleolar trafficking should be studied further.
Recent work by Audas et al. (2012) has identified a novel mechanism regulating nucleolar retention of proteins. Specific cellular stresses (such as heat-shock, acidosis and ribosomal stress) induce the expression of specific non- coding RNA (ncRNA) from rDNA intergenic spacer promoters. The stress- specific ncRNAs are bound by specific proteins and retained in the nucleolus.
Importantly, ncRNA binding is sequence-specific, as expression of random sequence RNA does not lead to protein retention (Audas et al., 2012). As BLM responds to replication stress, DNA damage and ribosomal stress, it is possible that BLM nucleolar trafficking is partially regulated by stress-induced nucleolar ncRNA expression. The absence of detectable BLM binding to single-stranded
RNA in Chapter 3, Figure 10 does not preclude this possibility, as an RNA of random sequence was used. Future experiments investigating any Rnase- sensitive nucleolar accumulation of BLM may be informative.
120 Disruption of nucleolar metabolism leads to a nucleolar stress response.
Nucleolar stress causes release of various ribosomal proteins (RP) (such as
RPL5, RPL11 and RPL23) from nucleoli that subsequently bind MDM2, inhibiting its E3 ubiquitin ligase activity, and leading to the stabilization of p53 (Macias et al., 2010). p53 stabilization coincident with increased RP/MDM2 binding provides a signature for nucleolar stress-induced p53 induction. This response occurs following actinomycin D treatment. Knock-down of BLM and measurement of p53 levels coincident with increased RP/MDM2 binding may elucidate a nucleolar BLM stress pathway.
There is an apparent paradox in Bloom’s syndrome: an anti- growth/proliferative phenotype (small stature at all ages, reduced proliferation of non-transformed cells) and a pro-growth/proliferation phenotype in tumor formation (increased incidence of virtually all malignancies) (German, 1997;
Lechner et al., 1983). The mutations that arise in the context of BLM-deficiency evidently allow unrestrained cell growth and proliferation that we see clinically as cancer. Importantly, nucleolar metabolism is a central regulator of cell growth and cell cycle control (reviewed in Boisvert et al., 2007). Bloom’s syndrome clinical phenotypes are analogous to those that result from the disease-causing
RMRP mutations in cartilage hair hypoplasia that leads to loss of Rnase MRP function. Loss of Rnase MRP function results in the failure of efficient rRNA processing and the slowing of ribosome biogenesis. Rnase MRP is also responsible for degrading cyclinB mRNA; cyclinB accumulation in Rnase MRP- deficient cells leads to chromosomal instability and cancer. Overall, loss of
121 Rnase MRP function leads to slowed ribosome biogenesis (anti- growth/proliferation) simultaneous with increased cyclin B (pro-growth/ proliferation), and ultimately leads to both growth defects and increased cancer incidence (Thiel et al., 2005). The similarities between the two conditions suggest the possibility that altered nucleolar metabolism may be a contributor to the cellular growth and clinical characteristics of BS.
This thesis as a whole describes a role for the BLM helicase in a pathway of cell growth control regulation and suggests ways in which it may be manipulated to control the growth of cancer cells. Therapeutics that slow rRNA transcription slow cellular growth rates and cause tumor regression (Drygin et al.,
2011). Strategies that lead to persistence of rRNA:rDNA hybrids are similarly expected to slow growth. This may be achieved by selective targeting of the nucleolar role of BLM, as general inhibition of BLM activity leads to genomic instability and promotes tumorigenesis. Thus, selective inhibition of BLM partnering with RNA polymerase I, partnering with DNA topoisomerase I, or of its ability to resolve RNA:DNA hybrids is expected to induce nucleolar stress and slow cellular growth without compromising stability of the extra-nucleolar genome. Additionally, the hypothesis that NPM mediates BLM nucleolar trafficking suggests strategies to disrupt BLM/NPM interaction and thereby impair
BLM nucleolar trafficking to slow cellular growth.
122
123 Figure 18. Model for the role of BLM in facilitating RNA polymerase I- mediated ribosomal RNA transcription. Ribosomal RNA (rRNA) transcription in the nucleolus leads to the formation of rRNA:rDNA hybrids that impede progression of subsequent transcription complexes. DNA topoisomerase I (topo
I) functions in association with RNA polymerase I (RNA pol I) to prevent the formation of such hybrids. We propose that BLM functions coordinately with
DNA topoisomerase I and RNA polymerase I to unwind aberrantly formed rRNA:rDNA hybrids to facilitate rRNA transcription. The nucleolar trafficking protein nucleophosmin (NPM/B23) may mediate the nucleolar/ nucleoplasmic shuttling of BLM. (Symbols: circle is BLM; triangle is RNA pol I; square is DNA topo I; diamond is NPM/ B23).
124
Bibliography
Aguilera, A. (2002) The connection between transcription and genomic instability. EMBO J., 21, 195-201.
Albertsson-Wikland, K., Wennergren, G., Wennergren, M., Vilbergsson, G. and Rosberg, S. (1993) Longitudinal follow-up of growth in children born small for gestational age. Acta Paediatr., 82, 438-443.
Andersen, J.S., Lyon, C.E., Fox, A.H., Leung, A.K., Lam, Y.W., Steen, H., Mann, M. and Lamond, A.I. (2002) Directed proteomic analysis of the human nucleolus. Curr. Biol., 12, 1 – 11.
Audas, T. E., Jacob, M. D., and Lee, S. (2012) Immobilization of proteins in the nucleolus by ribosomal intergenic spacer noncoding RNA. Mol. Cell, 45, 147- 157.
Ayrault, O., Andrique, L., Fauvin, D., Eymin, B., Gazzeri, S. and Seite, P. (2006) Human tumor suppressor p14ARF negatively regulates rRNA transcription and inhibits UBF1 transcription factor phosphorylation. Oncogene, 25, 7577-7586.
Bachrati, C. Z., and Hickson I. D. (2003) RecQ helicases: guardian angels of the DNA replication fork. Chromosoma, 117, 219-233.
Bernstein, K. A., Bleichert, F., Bean, J. M., Cross, F. R., and Baserga, S. J. (2007) Ribossome biogenesis is sensed at the start cell cycle checkpoint. Mol. Biol. Cell, 18, 953-964.
Bhattacharyya, S., Keirsey, J., Russell, B., Kavecansky, J., Lillard-Wetherell, K., Tahmaseb, K., Turchi, J. J., and Groden, J. (2009) Telomerase-associated protein 1, HSP90, and topoisomerase IIalpha associate directly with the BLM helicase in immortalized cells using ALT and modulate its helicase activity using telomeric DNA substrates. J. Biol. Chem., 284, 14966-14977.
Birch, J.L. and Zomerdijk, J. (2008) Structure and function of ribosomal RNA gene chromatin. Biochem. Soc. Trans., 36, 619–624.
125 Boisvert, F. M., van Koningsbruggen, S., Navascues, J., and Lamond, A. I. (2009) The multifunctional nucleolus. Nat. Rev. Mol. Cell Biol., 8, 574-585.
Bonafe, L., Schmitt, K., Eich, G., Giedion, A. and Superti-Furga, A. (2002) RMRP gene sequence analysis confirms a cartilage-hair hypoplasia variant with only skeletal manifestations and reveals a high density of single-nucleotide polymorphisms. Clin. Genet., 61, 146 – 151.
Cavanaugh, A. H., Hempel, W. M., Taylor, L. J., Rogalsky, V., Todorov, G., and Rothblum, L. I. (1995) Activity of RNA polymerase I transcription factor UBF blocked by Rb gene product. Nature, 374, 177-180.
Chan, K. L., North, P. S., and Hickson, I. D. (2007) BLM is required for faithful chromosome segregation and its localization defines a class of ultrafine anaphase bridges. EMBO J., 26, 3397-3409.
Chen, S., Warszawski, J., Bader-Meunier, B., Tchernia, G., Da Costa, L., Marie, I. and Dommergues, J.P. (2005) Diamond-Blackfan anemia and growth status: the French registry. J. Pediatr., 147, 669 – 673.
Chester, N., Kuo, F., Kozak, C., O’Hara, C.D. and Leder, P. (1998) Stage- specific apoptosis, developmental delay, and embryonic lethality in mice homozygous for a targeted disruption in the murine Bloom’s syndrome gene. Genes Dev., 12, 3382-3393.
Chu, W. K., and Hickson I. D. (2009) RecQ helicases: multifunctional genome caretakers. Nat. Rev. Cancer, 9, 644-654.
Colombo, E., Bonetti, P., Lazzerini, D. E., Martinelli, P., Zamponi, R., Marine, J. C., Helin, K., Falini, B., and Pelicci, P. G. (2005) Nucleophosmin is required for DNA integrity and p19Arf protein stability. Mol. Cell Biol., 25, 8874-8886.
Davalos, A. R., and Campisi J. (2003) Bloom syndrome cells undergo p53- dependent apoptosis and delayed assembly of BRCA1 and NBS1 repair complexes at stalled replication forks. J. Cell Biol., 162, 1197-1209.
Defossez P. A., Prusty R., Kaeberlein M., Lin S. J., Ferrigno P., Silver P. A., Keil R. L., and Guarente L. (1999). Elimination of the replication fork block protein Fob1 extends the life span of yeast mother cells. Mol. Cell, 3; 447-455.
Derenzini, M., Montanaro, L., and Trere, D. (2009) What the nucleolus says to a tumor pathologist. Histopathology, 54, 753-762.
De Winter, R. F., and Moss, T. (1986) Spacer promoters are essential for efficient enhancement of X. laevis ribosomal transcription. Cell, 44, 313-318.
126
Diaz, A., Vogiatzi, M. G., Sanz, M. M., and German, J. (2006) Evaluation of short stature, carbohydrate metabolism and other endocrinopathies in Bloom’s syndrome. Horm. Res., 66, 111-117.
Dimitrova, D. S. (2011) DNA replication initiation patterns and spatial dynamics of the human ribosomal RNA gene loci. J. Cell Sci., 15, 2743-2752.
Drygin, D., Lin, A., Bliesath, J., Ho. C. B., O’Brien, S. E., Proffitt, C., Omori, M., Haddach, M., Schwaebe, M. K., Siddiqui-Jain, A., Streiner, N., Quin, J. E., Sanij, E., Bywater, M. J., Hannan, R. D., Ryckman, D., Anderes, K., and Rice, W. G. (2011) Targeting RNA polymerase I with an oral small molecule CX-5461 inhibits ribosomal RNA synthesis and solid tumor growth. Cancer Res., 71, 1418-1430.
Drygin D., Rice W. G., and Grummt I. (2010). The RNA polymerase I transcription machinery: an emerging target for the treatment of cancer. Annu. Rev. Pharmacol. Toxicol., 50; 131-156.
Eckert-Boulet, N., and Lisby, M. (2009) Regulation of rDNA stability by sumoylation. DNA Repair, 8, 507-516.
Eladad, S., Ye, T. Z., Hu, P., Leversha, M., Beresten, S., Matunis, M. J., and Ellis, N. A. (2005) Intra-nuclear trafficking of the BLM helicase to DNA damage- induced foci is regulated by SUMO modification. Hum. Mol. Genet., 14, 1351- 1365.
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 helicases. Cell, 83, 655-666.
Emmott E. and Hiscox J. A. (2009). Nucleolar targeting: the hub of the matter. EMBO Reports, 10; 231-238.
Fetherston J., Werner E., and Patterson R. (1984). Processing of the external transcribed spacer of murine rRNA and site of action of actinomycin D. Nucleic Acids Res., 12; 7187-7198.
French, S. L., Osheim, Y, N., Cioci, F., Nomura, M., and Beyer, A. L. (2003) In exponentially growing Saccharomyces cerevisiae cells, rRNA synthesis is determined by the summed RNA polymerase I loading rate rather than by the number of active genes. Mol. Cell Biol., 23, 1558-1568.
Gangloff S., McDonald J. P., Bendixen C., Arthur L., and Rothstein R. (1994). The yeast type I topoisomerase Top3 interacts with Sgs1, a DNA helicase homolog: a potential eukaryotic reverse gyrase. Mol. Cell Biol., 14; 8391-8398.
127
German J. (1968). Bloom’s syndrome I: genetical and clinical observations in the first twenty-seven patients.
German J. (1997). Bloom’s syndrome. XX. The first 100 cancers. Cancer Genet. Cytogenet., 93; 100-106.
Gray, M.D., Wang, L., Youssoufian, H., Martin, G.M. and Oshima, J. (1998) Werner helicase is localized to transcriptionally active nucleoli of cycling cells. Exp. Cell Res., 242, 487-494.
Grierson, P., Lillard, K., Behbehani, G. K., Combs, K. A., Bhattacharyya, S., Acharya, S., and Groden, J. (2012) BLM helicase facilitates RNA polymerase I- mediated ribosomal RNA transcription. Hum. Mol. Genet., 21, 1172-1183.
Hand, R., and German J. (1975) A retarded rate of DNA chain growth in bloom’s syndrome. Proc. Nat. Acad. Sci., 72, 758-762.
Hage, A. E., French, S. L., Beyer, A. L., and Tollervey, D. (2010) Loss of topoisomerase I leads to R-loop-mediated transcriptional blocks during ribosomal RNA synthesis. Genes Dev., 24, 1546-1558.
Heo, S.J., Tatebayashi, K., Ohsugi, I., Shimamoto, A., Furuichi, Y. and Ikeda, H. (1999) Bloom’s syndrome gene suppresses premature ageing caused by Sgs1 deficiency in yeast. Genes Cells, 4, 619-625.
Hickson I. D. (2003) RecQ helicases: caretakers of the genome. Nat. Rev. Cancer, 3, 169-178.
Holm, C., Stearns, T., and Botstein, D. (1989) DNA topoisomerase II must act at mitosis to prevent nondisjunction and chromosome breakage. Mol. Cell Biol., 9, 159-168.
Hraiky, C., Raymond, M. A., and Drolet, M. (2000) RNase H overproduction corrects a defect at the level of transcription elongation during rRNA synthesis in the absence of DNA topoisomerase I in Escherichia coli. J. Biol. Chem., 275, 11257-11263.
Hu, P., Beresten, S. F., van Brabant, A. J., Pandolfi, P. P., Johnson, F. B., Guarente, L., and Ellis, N. A. (2001) Evidence for BLM and topoisomerase IIIalpha interaction in genomic stability. Hum. Mol. Genet., 10, 1287-1298.
Huang, S. (2002) Building an efficient factory: where is pre-rRNA synthesized in the nucleolus? J. Cell Biol., 157, 739 – 741.
128 Huber, M.D., Lee, D. C., and Maizels, N. (2002) G4 DNA unwinding by BLM and Sgs1p: substrate specificity and substrate-specific inhibition. Nucleic Acids Res., 30, 3954-3961.
Huertas, P. and Aguilera, A. (2003) Cotranscriptionally formed stable DNA:RNA hybrids mediate transcription elongation impairment and transcription-associated recombination. Mol. Cell, 12, 711 – 721.
Huet, J., Wyers, F., Buhler, J. M., Sentenac, A., and Fromageot, P. (1976) Association of RNase H activity with yeast RNA polymerase A. Nature, 261, 431-433.
Janscak, P., Garcia, P. L., Hamburger, F., Makuta, Y., Shiraishi, K., Imai, Y., Ikeda, H., and Bickle T. A. (2003). Characterization and mutational analysis of the RecQ core of the bloom syndrome protein. J. Mol. Biol., 300, 29-42.
Jiang, K., Pereira, E., Maxfield, M., Russell, B., Goudelock, D.M. and Sanchez, Y. (2003) Regulation of Chk1 includes chromatin association and 14– 3–3 binding following phosphorylation on Ser-345. J. Biol. Chem., 278, 25207– 25227.
Jordan, P., Mannervik, M., Tora, L. and Carmo-Fonseca, M. (1996) In vivo evidence that TATA-binding protein/ SL1 colocalizes with UBF and RNA polymerase I when rRNA synthesis is either active or inactive. J. Cell. Biol., 133, 225-234.
Karow, J. K., Chakraverty, R. K., and Hickson, I. D. (1997) The Bloom’s syndrome gene product is a 3’-5’ DNA helicase. J. Biol. Chem., 272, 30611- 30614.
Killen, M.W., Stults, D.M., Adachi, N., Hanakahi, L., and Pierce, A.J. (2009) Loss of Bloom syndrome protein destabilizes human gene cluster architecture. Hum. Mol. Genet., 18, 3417-3428.
Kirn-Safran C. B., Oristian D. S., Focht R. J., Parker S. G., Vivian J. L., and Carson D. D. (2007). Global growth deficiencies in mice lacking the ribosomal protein HIP/RPL29. Dev. Dyn., 236; 447-460.
Klein, J., and Grummt, I. (1999) Cell cycle-dependent regulation of RNA polymerase I transcription: the nucleolar transcription factor UBF is inactive in mitosis and early G1. Proc. Natl. Acad. Sci., 96, 6096-6101.
Koberna, K., Malinsky, J., Pliss, A., Masata, M., Vecerova, J., Fialova, M., Bednar, J. and Raska, I. (2002) Ribosomal genes in focus: new transcripts label
129 the dense fibrillar components and form clusters indicative of ‘Christmas trees’ in situ. J. Cell Biol., 157, 743 – 748.
Kressler, D., Hurt, E. and Babler, J. (2010) Driving ribosome assembly. Biochim. Biophys. Acta., 1803, 673-683.
Kuhn, A., Bartsch, I., and Grummt, I. (1990) Specific interaction of the murine transcription termination factor TTF I with class-I RNA polymerases. Nature, 344, 559-562.
Langland, G., Kordich, J., Creaney, J., Goss, K. H., Lillard-Wetherell, K., Behbehani, 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, 30031-30035.
Lechner, J. F., Kaighn, M. E., Jetten, A. M., Groden, J., and German, J. (1983) Bloom’s syndrome cells have an abnormal serum growth response. Exp. Cell Res., 145, 381-388.
Lee, S. K., Johnson, R. E., Yu, S. L., Prakash, L., and Prakash, S. (1999) Requirement of yeast SGS1 and SRS2 genes for replication and transcription. Science, 286, 2339-2342.
Lessard, F., Morin, F., Ivanchuk, S., Langlois, F., Stefanovsky, V., Rutka, J., and Moss, T. (2010) The ARF tumor suppressor controls ribosome biogenesis by regulating the RNA polymerase I transcription factor TTF-1. Mol. Cell, 38, 539- 550.
Li, J., Santoro, R., Koberna, K., and Grummt, I. (2005) The chromatin remodeling complex NoRC controls replication timing of rRNA genes. EMBO J., 24, 120-127.
Lillard-Wetherell, K., Machwe, A., Langland, G.T., Combs, K.A., Behbehani, G.K., Schonberg, S.A., German, J., Turchi, J.J., Orren, D.K., and Groden, J. (2004) Association and regulation of the BLM helicase by the telomere proteins TRF1 and TRF2. Hum. Mol. Genet., 13, 1919-1932.
Little, R. D., Platt, T. H., and Schildkraut, C. L. (1993) Initiation and termination of DNA replication in human rRNA genes. Mol. Cell Biol., 13, 6600-6613.
Lombard, D. B., and Guarente, L. (2000) Nijmegen breakage syndrome disease protein and MRE11 at PML nuclear bodies and meiotic telomeres. Cancer Res., 60, 2331-2334.
130 Macias, E., Jin, A., Deisenroth, C., Bhat, K., Mao, H., Lindstrom, M. S., and Zhang, Y. (2010) An ARF-independent c-MYC-activated tumor suppression pathway mediated by ribosomal protein-Mdm2 interaction. Cancer Cell, 18, 231- 243.
Machwe, A., Xiao, L., Groden, J., and Orren D. K. (2006) The Werner and bloom syndrome proteins catalyze regression of a model replication fork. Biochemistry, 45, 13939—13946.
Makitie, O., Ellis, L., Durie, P.R., Morrison, J.A., Sochett, E.B., Rommens, J.M. and Cole, W.G. (2004) Skeletal phenotype in patients with Shwachman-Diamond syndrome and mutations in SBDS. Clin. Genet., 65, 101 – 112.
Mayer, C., Schmitz, K. M., Li, J., Grummt, I., and Santoro, R. (2006) Intergenic transcripts regulate the epigenetic state of rRNA genes. Mol. Cell, 22, 351-361.
Mayer C., Zhao J., Yuan X., and Grummt I. (2004). mTOR-dependent activation of the transcription factor TIF-1A links rRNA synthesis to nutrient availability. Genes Dev., 18; 423-434.
Mohaghegh, P., Karow, J.K., Brosh, R.M., Bohr, V.A., and Hickson, I.D. (2001) The Bloom’s and Werner’s syndrome proteins are DNA structure-specific helicases. Nucleic Acids Res., 29, 2843-2849.
Moss T., Stefanovsky V., Langlois F., and Kugler-Gagnon T. (2006). A new paradigm for the regulation of the mammalian ribosomal RNA genes. Biochem. Soc. Trans., 34; 1079-1081.
Muller, S., Ledl, A., and Schmidt, D. (2004) SUMO: a regulator of gene expression and genome integrity. Oncogene, 23, 1998-2008.
Myers, M.G. Jr., Grammer, T.C., Wang, L.M., Sun, X.J., Pierce, J.H., Blenis, J. and White, M.F. (1994) Insulin receptor substrate-1 mediates phosphatidylinositol 3′-kinase and p70S6K signaling during insulin, insulin-like growth factor-1, and interleukin-4 stimulation. J. Biol. Chem., 269, 28783-28789.
Narla, A., and Ebert, B. L. (2010). Ribosomopathies: human disorders of ribosome dysfunction. Blood, 115, 3196-3205.
Ochs, R., Lischwe, M., O’Leary, P., and Busch, H. (1983) Localization of nucleolar phosphoproteins B23 and C23 during mitosis. Exp. Cell Res., 146, 139-149.
Parkinson, G. N., Lee, M. P., and Neidle S. (2002) Crystal structure of parallel quadruplexes from human telomeric DNA. Nature, 417, 876-880.
131
Patrone, G., Puppo, F., Cusano, R., Scaranari, M., Ceccherini, I., Puliti, A. and Ravazzola, R. (2000) Nuclear run-on assay using biotin labeling, magnetic bead capture and analysis by fluorescence-based RT-PCR. Biotechniques, 29, 1012- 1014, 1016-1017.
Pfleiderer, C., Smid, A., Bartsch, I., and Grummt, I. (1990) An undecamer DNA sequence directs termination of human ribosomal gene transcription. Nucleic Acids Res., 18, 4727-4736.
Popuri, V., Bachrati, C. Z., Muzzolini, L., Mosedale, G., Costantini, S., Giacomini, E., Hickson, I. D., and Vindigni, A. (2008) The human recQ helicases, BLM and RECQ1, display distinct DNA substrate specificities. J. Biol. Chem., 283, 17766- 17776.
Rao, V. A., Fan, A. M., Meng, L, Doe, C. F., North, P. S., Hickson, I. D., and Pommier, Y. (2005). Phosphorylation of BLM, dissociation from topoisomerase IIIα, and colocalization with γ-H2AX after topoisomerase I-induced replication damage. Mol. Cell. Biol., 25, 8925-8937.
Reeder, R. H. (1999) Regulation of RNA polymerase I transcription in yeast and vertebrates. Prog. Nucleic Acid Res. Mol. Biol., 62, 293-327.
Ridanpaa, M., van Eenennaarn, H., Pelin, K., Chadwick, R., Johnson, C., Yuan, B., vanVenrooij, W., Pruijn, G., Salmela, R., Rockas, S. et al. (2001) Mutations in the RNA component of RNase MRP cause a pleiotropic human disease, cartilage-hair hypoplasia. Cell, 104, 195 – 203.
Riskin, A. and Garcia-Prats, J.A. (2011) Infant of a diabetic mother. In Basow, D.S., (ed). Up To Date, Waltham, MA.
Russell, B., Bhattacharyya, S., Keirsey, J., Sandy, A., Grierson, P., Perchiniak, E., Kavecansky, J., Acharya, S., and Groden, J. (2011) Chromosome breakage is regulated by the interaction of the BLM helicase and topoisomerase IIα. Cancer Res., 71, 561-571.
Russell, J., and Zomerdijk, J. C. (2005) RNA-polymerase-I-directed rDNA transcription, life and works. Trends Biochem. Sci., 30, 87-96.
Salazar, M., Thompson, B.D., Kerwin, S.M. and Hurley, L.H. (1996) Thermally induced DNA.RNA hybrid to G-quadruplex transitions: possible implications for telomere synthesis by telomerase. Biochemistry, 35, 16110- 16115.
132 Sandmeier, J. J., French, S., Osheim, Y., Cheung, W. L., Gallo, C. M., Beyer, A. L., and Smith, J. S. (2002) RPD3 is required for the inactivation of yeast ribosomal DNA genes in stationary phase. EMBO J., 21, 4959-4968.
Santoro, R., Li, J., and Grummt, I. (2002) The nucleolar remodeling complex NoRC mediates heterochromatin formation and silencing of ribosomal gene transcription. Nat. Genet., 32, 393-393.
Santoro, R., Schmitz, K. M., Sandoval, J., and Grummt, I. (2010) Intergenic transcripts originating from a subclass of ribosomal DNA repeats silence ribosomal RNA genes in trans. EMBO Rep., 11, 52-58.
Sanz M.M., Proytcheva M., Ellis N. A., Holloman W. K., and German J. (2000). BLM, the Bloom’s syndrome protein, varies during the cell cycle in its amount, distribution, and co-localization with other nuclear proteins. Cytogenet Cell Genet., 91; 217-223.
Schawalder, J., Paric, E., and Neff, N. F. (2003) Telomere and ribosomal DNA repeats are chromosomal targets of the Bloom syndrome DNA helicase. BMC Cell Biol., 4(1), 15.
Schlosser, I., Holzel, M., Murnseer, M., Burtcher, H., Weidle, U.H. and Eick, D. (2003) A role for c-MYC in the regulation of ribosomal RNA processing. Nucleic Acids Res., 31, 6148-6156.
Schwarzacher, H.G. and Wachtler, F. (1993) The nucleolus. Anat. Embryol., 188, 515–536.
Seither, P., Coy, J.F., Pouska, A. and Grummt, I. (1997) Molecular cloning and characterization of the cDNA encoding the largest subunit of mouse RNA polymerase I. Mol. Gen. Genet., 255, 180 – 186.
Sengupta, S., Linke, S. P., Pedeux, R., Yang, Q., Farnsworth, J., Garfield, S. H., Valerie, K., Shay, J. W., Ellis, N. A., Wasylyk, B,. and Harris, C. C. (2003) BLM helicase-dependent transport of p53 to sites of stalled DNA replication forks modulates homologous recombination. EMBO J., 22, 1210-1222.
Shin, J. H., and Kelman, Z. (2006) The replicative helicases of bacteria, archaea, and eukarya can unwind RNA-DNA hybrid substrates. J. Biol. Chem., 281, 26914-26921.
Shiratori, M., Suzuki, T., Itoh, C., Goto, M., Furuichi, Y., and Matsumoto, T. (2002) WRN helicase accelerates the transcription of ribosomal RNA as a component of an RNA polymerase I-associated complex. Oncogene, 21, 2447- 2454.
133
Sinclair D. A. and Guarente L. (1997). Extrachromosomal rDNA circles—a cause of aging in yeast. Cell, 91; 1033-1042.
Sinclair D. A., Mills K., and Guarente L. (1997). Accelerated aging and nucleolar fragmentation in yeast sgs1 mutants. Science, 29; 1313-1316.
Sorensen, P. D., Lomholt, B., Frederiksen, S,. and Tommerup, N. (1991) Fine mapping of human 5S rRNA genes to chromosome 1q42.11-q42.13. Cytogenet. Cell Genet., 57, 26-29.
Stefanovsky, V., Langlois, F., Gagnon-Kugler, T., Rothblum, L. I., and Moss, T. (2006) Growth factor signaling regulates elongation of RNA polymerase I transcription in mammals via UBF phosphorylation and r-chromatin remodeling. Mol. Cell, 21, f629-639.
Stefanovsky V. Y., Pelletier G., Hannan R., Gagnon-Kugler T., Rothblum L. I., and Moss T. (2001). An immediate response of ribosomal transcription to growth factor stimulation in mammals is mediated by ERK phosphorylation of UBF. Mol. Cell, 8; 1063-1073.
Sun, H., Karow, J. K., Hickson, I. D., and Maizels, N. (1998) The bloom’s syndrome helicase unwinds G4 DNA. J. Biol. Chem., 273, 27587-27592.
Sutherland, H.G., Mumford, G.K., Newton, K., Ford, L.V., Farrall, R., Dellaire, G., Caceres, J.F. and Bickmore, W.A. (2001) Large-scale identification of mammalian proteins localized to nuclear sub-compartments. Hum. Mol. Genet., 10, 1995-2011.
Svarcova, O., Strejcek, F., Petrovicova, I., Avery, B., Pedersen, H.G., Lucas- Hahn, A., Niemann, H., Laurincik, J. and Maddox-Hyttel, P. (2008) The role of RNA polymerase I transcription and embryonic genome activation in nucleolar development in bovine preimplantation embryos. Mol. Reprod. Dev., 75, 1095- 1103.
Tamemoto, H., Kadowaki, T, Tobe, K., Yagi, T., Sakura, H., Hayakawa, T., Terauchi, Y., Ueki, K., Kaburagi, Y., Satoh, S. et al. (1994) Insulin resistance and growth retardation in mice lacking insulin receptor substrate-1. Nature, 372, 182–186.
Therman, E., Otto, P. G., and Shahidi, N. T. (1981) Mitotic recombination and segregation of satellites in Bloom’s syndrome. Chromosoma, 82, 627-636.
Thiel, C.T., Horn, D., Zabel, B., Ekici, A.B., Salinas, K., Gebhart, E., Ruschendorf, F., Sticht, H., Spranger, J., Muller, D. et al. (2005) Severely
134 incapacitating mutations in patients with extreme short stature indentify RNA- processing endoribonuclease RMRP as an essential cell growth regulator. Am. J. Hum. Genet., 77, 795-806.
Tu, X., Batta, P., Innocent, N., Prisco, M., Casaburi, I., Belletti, B. and Baserga, R. (2002) Nuclear translocation of insulin receptor substrate-1 by oncogenes and Igf-1. Effect on ribosomal RNA synthesis. J. Biol. Chem., 277, 44357-44365.
UpToDate Online. www.uptodate.com
Versini, G., Comet, I., Wu, M., Hoopes, L., Schwob, E., and Pasero, P. (2003) The yeast Sgs1 helicase is differentially required for genomic and ribosomal DNA replication. EMBO J., 22, 1939-1949.
Yankiwski, V., Marciniak, R.A., Guarente, L. and Neff, N.F. (2000) Nuclear structure in normal and Bloom syndrome cells. Proc. Natl Acad. Sci., 97, 5214- 5219.
Yankiwski, V., Noonan, J. P., and Neff, N. F. (2001) The C-terminal domain of the Bloom syndrome DNA helicase is essential for genomic stability. BMC Cell Biol., 2, 11.
Yuan X., Zhou Y., Casanova E., Chai M., Kiss E., Grone H. J., Schutz G., and Grummt I. (2005). Genetic inactivation of the transription factor TIF-1A leads to nucleolar disruption, cell cycle arrest, and p53-mediated apoptosis. Mol. Cell, 19; 77-87.
Zhai, W., and Comai, L. (2000) Repression of RNA polymerase I transcription by the tumor suppressor p53. Mol. Cell Biol., 20, 5930-5938.
Zhang S., Hemmerich P., and Grosse, F. (2004). Nucleolar localization of the human telomeric repeat binding factor 2 (TRF2). J. Cell Sci., 117; 3935-3945.
Zhang, H., Wang, J. C., and Liu L. F. (1988) Involvement of DNA topoisomerase I in transcription of human ribosomal RNA genes. Proc. Nat. Acad. Sci., 85, 1060-1064.
Zhao, J., Yuan, X., Frodin, M., and Grummt, I. (2003) ERK-dependent phosphorylation of the transcription initiation factor TIF-1A is required for RNA polymerase I transcription and cell growth. Mol. Cell, 11, 405-413.
135