Drosophila Melanogaster Dis3 Is a Dynamic Endo- and 3’Æ5’

DROSOPHILA MELANOGASTER DIS3 IS A DYNAMIC ENDO- AND 3’Æ5’

EXORIBONUCLEASE

MEGAN CHRISTINE MAMOLEN

Submitted in partial fulfillment of the requirements

for the degree of Doctor of Philosophy

Dissertation advisor: Erik D. Andrulis, Ph.D.

Department of Molecular Biology and Microbiology

CASE WESTERN RESERVE UNIVERSITY

August, 2010

We hereby approve the thesis/dissertation of

______Megan Mamolen______

candidate for the ______Ph.D.______degree *.

(signed)______Dr. Jonathan Karn______

(chair of the committee)

______Dr. Peter Harte______

______Dr. Alan Tartakoff______

______Dr. Erik Andrulis______

(date) ______6-15-10______

*We also certify that written approval has been obtained for any

proprietary material contained therein.

This work is dedicated to my husband and best friend, Mike Smolko. Thank you for

encouraging me to believe in myself. This is only the beginning of a wonderful journey

together.

Table of Contents

Table of Contents ...... 5

List of Tables ...... 10

List of Figures ...... 11

Acknowledgements ...... 14

List of Abbreviations ...... 16

Abstract ...... 20

Chapter I:

Introduction ...... 21

I.A RNA turnover and human health ...... 22

I.B RNA metabolism in eukaryotes ...... 23

I.B.1 RNA expression: RNA production versus RNA turnover ...... 23

I.B.2 RNA processing: RNA stability and maturation ...... 25

I.B.3 RNA turnover: RNA destabilization and degradation ...... 26

I.C Ribonucleases ...... 31

I.C.1 Physical features of RNases ...... 32

I.C.2 RNase mechanisms of action ...... 34

I.C.3 RNase product formation and substrate specificity ...... 39

I.D Dis3 ...... 45

I.D.1a Dis3 in vivo: mitosis...... 45

I.D.1b Dis3 in vivo: RNA processing and turnover ...... 46

I.D.1c Dis3 in vivo: additional features ...... 48

I.D.2 Dis3 in vitro characteristics ...... 49

I.E Summary ...... 54

I.F Hypothesis ...... 54

Chapter II:

Characterization of the Drosophila melanogaster Dis3 Ribonuclease ...... 58

II.A Abstract ...... 59

II.B Introduction ...... 60

II.C Materials and Methods ...... 62

II.C.1 Molecular cloning ...... 63

II.C.2 Expression and purification of recombinant proteins ...... 63

II.C.3 Preparation of RNA substrates ...... 64

II.C.4 Ribonuclease activity assays ...... 64

II.C.5 Quantification of RNase activity ...... 65

II.C.6 Dis3 immunoprecipitation ...... 65

II.D Results ...... 66

II.D.1 Recombinant dDis3 is active in vitro ...... 66

II.D.2 MBP-dDis3 RNase activity requires monovalent and divalent cations ...... 66

II.D.3 MBP-dDis3 activity is not affected by non-ionic reaction conditions ...... 74

II.D.4 MBP-dDis3 RNase activity is not sequence specific ...... 77

II.D.5 dDis3 likely associates with exosome proteins via ionic and hydrophobic interactions ...... 83

II.E Discussion ...... 85

II.E.1 Drosophila melanogaster Dis3 is a functional ribonuclease in vitro ...... 85

II.E.2 dDis3 Mg2+ requirements point to a metal-ion catalyzed reaction mechanism .. 87

II.E.3 dDis3 product formation may depend on substrate identity ...... 88

II.E.4 dDis3 stably associates with core exosome proteins and exosome co-factors ... 90

II.E.5 Conclusions ...... 91

II.F Funding ...... 91

Chapter III:

Drosophila melanogaster Dis3 N-terminal domains are required for ribonuclease

activities, subcellular localization, and exosome interactions ...... 92

III.A Abstract ...... 93

III.B Introduction ...... 94

III.C Materials and Methods ...... 97

III.C.1 Molecular cloning ...... 97

III.C.2 Purification of recombinant proteins ...... 97

III.C.3 Preparation of RNA substrates ...... 98

III.C.4 Ribonuclease activity assays ...... 98

III.C.5 Quantification of RNase activity ...... 99

III.C.6 Cell culture ...... 99

III.C.7 Immunofluorescence, immunoprecipitation, and western blotting ...... 99

III.C.8 Cell fractionation and isolation of mitochondria ...... 100

III.D Results ...... 101

III.D.1 MBP-dDis3 is active on linear RNAs of varying sequences ...... 101

III.D.2 The dDis3 N-terminus harbors an endoribonuclease activity ...... 101

III.D.3 dDis3 N-terminal domains are required for nuclear localization ...... 125

III.D.4 dDis3 localizes to mitochondria via an N-terminal mitochondria targeting

sequence ...... 131

III.D.5 dDis3 N-terminal domains are required for interactions with core exosome

proteins and exosome co-factors ...... 136

III.E Discussion...... 140

III.E.1 Dis3 N-terminal endoribonuclease activity is conserved in metazoans ...... 140

III.E.2 The dDis3 N-terminus is important for subcellular localization ...... 142

III.E.3 The dDis3 N-terminus is responsible for interactions with core exosome

proteins and exosome co-factors ...... 144

III.E.4 Do dDis3 N-terminal domains link three different functions? ...... 146

III.E.5 Conclusions ...... 148

III.F Funding ...... 148

Chapter IV:

General Discussion and Future Directions ...... 149

IV.A The Drosophila melanogaster Dis3 ribonuclease ...... 150

IV.B dDis3 in vitro ...... 152

IV.B.1 dDis3 ion requirements and reaction mechanism ...... 152

IV.B.2 dDis3 domain function ...... 153

IV.C dDis3 in vivo ...... 155

IV.C.1 dDis3 substrate specificity ...... 155

IV.C.2 Regulation of dDis3 RNase activities ...... 159

IV.C.2a dDis3 localization ...... 159

IV.C.2b dDis3 protein-protein interactions ...... 161

IV.D Dis3 in the bigger picture: endo-exoRNases in RNA metabolism ...... 162

Appendix A ...... 164

Bibliography ...... 165

List of Tables

Table 1. Summary of MBP-dDis3 condition-specific RNase assays ...... 80

Table 2. Putative MTS alignment ...... 160

List of Figures

Figure 1. Kinetics dictate the balance between formation of a mature RNA and its

degradation ...... 24

Figure 2. General mRNA turnover pathways in eukaryotes ...... 27

Figure 3. RNA quality control pathways in eukaryotes ...... 29

Figure 4. RNA phosphodiester bond cleavage by acid-base catalysis ...... 36

Figure 5. RNA phosphodiester bond cleavage by metal-ion catalysis ...... 38

Figure 6. EndoRNases cleave internal RNA phosphodiester bonds ...... 40

Figure 7. ExoRNase product formation: hydrolytic versus phosphorolytic cleavage ...... 43

Figure 8. S. cerevisiae Dis3 crystal structure ...... 50

Figure 9. Contacts between S. cerevisiae Dis3 amino acids and polyA RNA

nucleotides…...... 53

Figure 10. Multiple sequence alignment of putative Dis3 exoRNase active sites ...... 56

Figure 11. Multiple sequence alignment of putative Dis3 endoRNase active sites ...... 57

Figure 12. Drosophila melanogaster Dis3 has ribonuclease activity in vitro ...... 67

Figure 13. MBP-dDis3 is active in the presence of various monovalent cations ...... 68

Figure 14. MBP is not active on polyU RNA in any ionic condition tested ...... 70

Figure 15. MBP-dDis3 is activated by divalent cations ...... 72

Figure 16. MBP-dDis3 RNase activity requires divalent cations ...... 73

Figure 17. MBP-dDis3 RNase activity is sensitive to monovalent cation concentrations 75

Figure 18. Divalent cation concentrations affect MBP-dDis3 RNase activity ...... 76

Figure 19. MBP-dDis3 RNase activity is not affected by certain non-ionic conditions in

vitro ...... 78

Figure 20. Nucleotide co-factors are not required for MBP-dDis3 activity in vitro...... 79

Figure 21. MBP-dDis3 RNase activity is not sequence specific ...... 81

Figure 22. MBP-dDis3 has 3’Æ5’ exoribonuclease activity ...... 82

Figure 23. dDis3-dRrp6 and dDis3-core exosome interactions are stable in vitro ...... 84

Figure 24. Full-length MBP-dDis3 efficiently degrades multiple RNA substrates ...... 102

Figure 25. Recombinant mutant MBP-dDis3 proteins used in in vitro RNase assays ... 104

Figure 26. N-terminal domains are necessary for full-length MBP-dDis3 in vitro RNase

activity...... 105

Figure 27. N-terminally truncated MBP-dDis3 mutants retain RNase activity ...... 107

Figure 28. Point mutations to the PIN domain affect MBP-dDis3 RNase activity ...... 109

Figure 29. MBP has little background RNase activity ...... 111

Figure 30. N-terminal domains are sufficient for MBP-dDis3 in vitro RNase activity .. 112

Figure 31. Circularized RNA substrates are not cleaved by control enzymes ...... 115

Figure 32. The dDis3 N-terminus has endoribonuclease activity ...... 117

Figure 33. The N-terminus of MBP-dDis3 cleaves circular RNAs less efficiently than the

full-length protein ...... 118

Figure 34. MBP has little background RNase activity on circularized RNA substrates 119

Figure 35. MBP-dDis3 cleavage of 3’ end-labeled RNAs confirms endoribonuclease

activity...... 122

Figure 36. Full-length MBP-dDis3 cleaves 3’ end-labeled RNAs more efficiently than the

N-terminus alone ...... 123

Figure 37. MBP has little or no background activity on 3’ end-labeled RNAs ...... 124

Figure 38. The N-terminus of dDis3 contributes to its subcellular distribution in

Drosophila S2 cells ...... 126

Figure 39. Images of mutant dDis3 localization patterns ...... 128

Figure 40. dDis3 fractionates with mitochondria from Drosophila S2 cells ...... 132

Figure 41. dDis3 localizes to mitochondria in Drosophila S2 cells ...... 134

Figure 42. The N-terminus of dDis3 is sufficient for mitochondrial targeting ...... 135

Figure 43. The dDis3 N-terminus is required for interactions with core exosome

proteins… ...... 137

Figure 44. N- and C-terminally truncated dDis3 mutant polypeptides fail to

immunoprecipitate endogenous Dis3 ...... 139

Figure 45. dDis3 functional regions identified in this work ...... 147

Figure 46. Summary of Dis3 in vivo functions ...... 156

Acknowledgements

There are several people who assisted me in completing graduate school

successfully:

Erik Andrulis – Thank you for challenging me to become a better scientist and

respecting my scientific ideas. I feel I have been able to grow both personally and

professionally in your lab.

Dan Kiss – Thank you for all of your help and suggestions over the years. Thank

you most of all for your friendship and the confidence you always had in me and my

work.

Amy Graham, Alex Smith, Ed Turk, Miriam Ruiz – Thank you for your help with

experiments and manuscripts. Each of you made graduate school a more enjoyable experience.

Jonathan Karn, Peter Harte, Alan Tartakoff – Thank you for your support and

helpful suggestions regarding my work and the resulting manuscripts.

Jim Lissemore, Dave Mascotti – Thank you for providing me with a strong

science foundation and encouraging me to follow my scientific dreams.

Lisa Stempak – To my dearest friend: Thank you for understanding me and

getting me through all the ups and downs. We made it!

Mike Smolko – Not only is this dissertation dedicated to you, but you deserve an

extra thanks for always being there, and for being the most supportive spouse that anyone

could ask for.

Finally, thank you to the people who provided me with the tools necessary to do

some of the experiments: Piet de Boer for microscope use, Eckhard Jankowsky for some

of the RNA substrates, and everyone in the Molecular Biology office (Brinn, Holly, Brad,

Karen, Dorothy) for always taking care of the little things.

List of Abbreviations

Δ domain deletion

α antibody

μ microsomes

2Me 2-mercaptoethanol

AMP adenosine monophosphate

ATP adenosine trisphosphate

C cytoplasm

C3 3 cysteine residues; iron-sulfur cluster homology

CMP cytidine monophosphate

Csl4 cep 1 synthetic lethal 4

DAPI 4',6-diamidino-2-phenylindole

dATP deoxy- adenosine trisphosphate

DIC differential interference contrast microscopy

Dis3 defective in sister chromatid disjoining 3

dDis3 Drosophila melanogaster Dis3

dsRNA double-stranded RNA

E. coli Escherichia coli

EDTA ethylenediaminetetraacetic acid

EM electromagnetic imaging

endoRNase endoribonuclease

exoRNase exoribonuclease

F FLAG tag

FISH fluorescence in situ hybridization

GFP green fluorescent protein

IP immunoprecipitation

M mitochondria

MBP maltose binding protein

metalloRNase metal-ion requiring RNase

Min minutes

mitoRNA mitochondrial encoded RNA

mRNA messenger RNA

Mtnp metallothinine promoter

Mtr mRNA transport mutant

MTS mitochondria targeting sequence

N nucleus

NDP nucleotide disphosphate

NGD no-go decay

NLS nuclear localization signal

NMD nonsense-mediated decay

NMP nucleotide monophosphate

NP-40 nonidet P-40

NRD non-functional rRNA decay

NSD non-stop decay

OH hydroxyl group

OB oligonucleotide binding fold

p plasmid or phosphate

P pellet

Pi inorganic phosphate

PCR polymerase chain reaction

PIN Pil-T N-terminal domain

PNPase polynucleotide phosphorylase polyA RNA polyadenine RNA

polyC RNA polycytidine RNA

polyN RNA polynucleotide RNA

polyU RNA polyuridine RNA

PROMPTs promoter upstream transcripts

PTC premature termination codon

REMD ribosome extension mediated decay

RNA ribonucleic acid

RNAi RNA interference

RNase ribonuclease

RNase II/R RNase II and RNase R

RNB RNase II domain

RNP ribonucleoprotein complex

RNR RNase II/R family

rRNA ribosomal RNA

Rrp ribosomal RNA processing

RTD rapid tRNA decay

S supernatant

S2 Drosophila Schneider 2 cells

S. cerevisiae Saccharomyces cerevisiae

SDS sodium dodecyl sulfate

SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis

Smg suppressor affecting message stability

SnoRNA small nucleolar RNA

SnRNA small nuclear RNA

SOD superoxide dismutase

ssRNA single-stranded RNA

tRNA transfer RNA

TLC thin layer chromatography

Tryp trypsin

TX-100 triton X-100

WCE whole cell extract

Drosophila melanogaster Dis3 is a dynamic endo- and 3’Æ5’ exoribonuclease

Abstract by

MEGAN CHRISTINE MAMOLEN

Dis3 is an evolutionarily conserved and essential enzyme with functions in mitosis, and RNA processing and turnover. In this work, we employed both in vitro and

in vivo techniques to examine the biochemical and cell biological characteristics of a

metazoan Dis3 homolog. Here, for the first time, we show that Drosophila melanogaster

Dis3 (dDis3) has in vitro exo- and endoribonuclease activities. Our results suggest that

both activities employ metal-ion catalysis. Further, neither activity is substrate-specific in

vitro. Interestingly, even though the exoribonuclease active site resides in the C-terminus,

both activities require the presence of N-terminal Dis3 domains. We show that N-

terminal domains mediate additional Drosophila Dis3 functions. For example, the

Drosophila Dis3 N-terminus affects nuclear localization. The dDis3 N-terminus also

contains a mitochondrial targeting sequence. Finally, N-terminal domains are responsible

for interactions with exosome proteins and the nuclear import factor Importin-α3. This study demonstrates that Drosophila melanogaster Dis3 is a complex enzyme with multiple ribonuclease activities, localization patterns, and protein-protein interactions.

Chapter I

Introduction

I.A RNA turnover and human health

RNA turnover is an essential regulatory mechanism in the cell. Turnover of both

normal and aberrant RNAs is vitally important, and a breakdown in these processes can

have deleterious effects. For example, defects in normal RNA decay have been linked to

cancer (e.g. Hollis et al., 1988). In addition, mutated mRNAs that are not efficiently

turned over can be translated into proteins that do not localize and/or function properly.

These abnormally functioning proteins can create or exacerbate a variety of genetic

disorders (e.g. Hall and Thein, 1994). Similarly, aberrant transfer RNAs (tRNAs) and

ribosomal RNAs (rRNAs), if not degraded, can disrupt essential cellular processes such

as translation. Finally, defects in RNA decay machinery diminish the response that cells

provide against RNA viruses (e.g. Matskevich and Moelling, 2007). Thus, the

degradation of some RNA viruses by host decay enzymes is an important facet of

immunity. It is clear from clinical studies that RNA turnover is important for preserving

normal cellular functions and preventing disease states. Thus, we endeavored to

understand the underlying mechanisms of decay. To do so, we examined the catalytic

activities of a protein that is regarded as a major RNA degradation and processing

enzyme in eukaryotes; Dis3. In the following sections, a general description of the

prominent features of eukaryotic RNA metabolism is presented, with focus on RNA

turnover. A detailed analysis of Dis3 is provided in Chapters II and III.

I.B RNA metabolism in eukaryotes

I.B.1 RNA expression: RNA production versus RNA turnover

In eukaryotes, RNA levels are regulated by three broad mechanisms:

transcription, processing, and turnover. To maintain proper RNA levels in the cell, a

fixed amount of “normal” transcripts are generated from the genome. These transcripts

are then processed into mature, functional RNAs that have specialized roles in the cell.

The stability of RNAs beyond maturation depends upon their roles. For example, many mRNAs are only needed during specific cellular events, and therefore are utilized and

degraded. Conversely, the aptly named “stable RNAs,” like tRNAs and rRNAs, have

much longer half-lives, and are usually only turned over if they are non-functional or

defective (reviewed in Andersen et al., 2008). Thus, one goal of RNA metabolic pathways in a healthy, non-stressed organism is to maintain steady-state RNA levels.

There are also processes devoted to the detection and elimination of aberrant RNAs.

Collectively, these processes are known as RNA quality control, and they compete with

transcription, processing, and translation.

The balance between the production and degradation of RNA is thought to be

achieved by a kinetic competition mechanism (Figure 1; Houseley and Tollervey, 2009).

RNAs are scanned by specialized machinery during transcription and all subsequent steps

of maturation. This process is known as RNA surveillance, and it is the first step in

quality control. One way this occurs is by protein-protein interactions, where surveillance

machinery interacts with the proteins bound to RNA (reviewed in Moore et al., 2005;

Isken and Maquat, 2007). If no defects are detected by surveillance machinery, RNAs are

Figure 1. Kinetics dictate the balance between formation of a mature RNA and its degradation

The fate of an RNA is governed by two pathways: (1) normal transcription, processing, and translation events, and (2) quality control. (Top) The “correct” substrate represents an

RNA with no defects that is properly processed into a mature “product”. In this scenario, normal processes occur faster than turnover. (Bottom) The “incorrect” substrate represents an RNA with defects. Turnover occurs faster than any remaining steps of maturation. Reprinted from Cell, 136, Houseley, J., and Tollervey, D., The many pathways of RNA degradation, 763-776, copyright (2009), with permission from

Elsevier.

transcribed, processed, properly localized, and for mRNAs, translated, before they can be

degraded (Figure 1, top). Thus, normal maturation processes occur at a faster rate, and

“out-compete” turnover. Eventually, mature RNAs are turned over, but this is a function

of their stability, as mentioned above. If defects are detected, surveillance machinery will

prompt the immediate degradation of the RNA. Here, turnover occurs at a faster rate than

any remaining transcription, processing, and translation events (Figure 1, bottom). In

addition, if any of the steps involved in the generation of the mature species are too slow

or stalled, the RNA is rapidly degraded. Thus, the fate of an RNA is thought to depend on

both its composition, and the kinetic competition between RNA production and turnover

pathways (Houseley and Tollervey, 2009).

I.B.2 RNA processing: RNA stability and maturation

There are complex pathways which compete with one another to promote the

stability of functional RNAs and prevent the accumulation of defective ones. RNA

processing is one method utilized by eukaryotes to stabilize RNA. For mRNAs,

processing entails splicing, and acquisition of a 5’ 7-methylguanosine cap and 3’

polyadenosine (polyA) tail. While splicing transforms pre-mRNAs into more functional,

“translatable” messages, the addition of the 5’ cap and 3’ polyA tail prevents their

degradation, and also increases their translatability (reviewed in Mangus et al., 2003;

Lejeune et al., 2004). In contrast, the processing of tRNAs, rRNAs, small nucleolar

(snoRNAs) and small nuclear (snRNAs) RNAs includes degradation events. pre-RNAs

are subjected to multiple cleavages, plus “trimming” of their ends during the later steps of

processing (Mitchell et al., 1996; Mitchell et al., 1997; Allmang and Tollervey, 1998;

Allmang et al., 1999a; Allmang 2000). Thus, initially, degradation of sequences within

these RNAs actually lends to their stability. Interestingly, some of the proteins that

participate in the maturation of these RNAs also facilitate their turnover.

During the life of an RNA, specific ribonucleoprotein complexes (RNPs) also

provide stability. These complexes both physically inhibit RNA degradation and signal

for normal processes to take place, until the RNA is no longer needed in the cell

(reviewed in Moore et al., 2005). Components of RNPs also aid in the folding of RNAs

into more stable forms, and direct the subcellular localization of mature RNAs to sites

where they function, or are translated into their protein counterparts (reviewed in Moore,

2005). Defective RNAs, and RNAs no longer needed by the cell, can also form RNPs that signal for their degradation (reviewed in Moore et al., 2005). Thus, RNP formation has a

dual role; it either maintains RNA stability or prompts RNA decay, depending on which

proteins are present in the complex.

I.B.3 RNA turnover: RNA destabilization and degradation

All RNA turnover pathways culminate in the destabilization and degradation of

RNAs. As alluded to in previous sections, there are two categories of RNA turnover. The first type, which will be referred to as “general” RNA turnover, directly controls the levels of normal RNAs in the cell. Although mature mRNAs are stabilized as described above, there are three mechanisms in place to degrade them (Figure 2; Parker and Song,

2004). First, mRNAs can be cleaved internally; the biochemistry of this reaction will be discussed in later sections. Two RNA fragments are produced from the initial cleavage event (Figure 2, right). One fragment retains the 5’ cap, and is degraded from its 3’ end

Figure 2. General mRNA turnover pathways in eukaryotes

The three mechanisms of general mRNA turnover are represented by three straight

arrows at the top of the figure. These include: (1) endonucleolytic cleavage, (2)

deadenylation-dependent degradation, and (3) deadenylation-independent degradation.

Reprinted by permission from Macmillan Publishers Ltd: [Nature Struc. and Molec.

Biol.] (Parker, R., and Song. H. The enzymes and control of eukaryotic mRNA turnover,

11, 121-127), copyright (2004).

since the cap acts as a physical block to degradation. The other fragment retains the

polyA tail, and is degraded from its 5’ end. The second mechanism of general turnover is

characterized by removal of the polyA tail prior to degradation. This process is called deadenylation, and is either directly followed by degradation from the 3’ end, or removal

of the 5’ cap and then degradation from the 5’ end (Figure 2, center). The third

mechanism of general turnover is referred to as deadenylation-independent degradation.

Here, the 5’ cap of an mRNA is removed first in a process called decapping, and then the

mRNA is degraded from the 5’ end (Figure 2, left). As described, all three pathways

result in the turnover of mRNAs to maintain necessary levels in the cell.

The second type of turnover in eukaryotes is “specialized” RNA turnover. There are

multiple specialized, quality control pathways, all which aim to eliminate aberrant RNAs

from the cell (Figure 3; Doma and Parker, 2007). Although these pathways vary, there

are some common features. As mentioned previously, all defective RNAs are first

recognized by surveillance machinery. Following detection, RNAs are either degraded

immediately, or “marked” in some way, which subsequently prompts their turnover.

Often, adenines are added to the 3’ ends of defective RNAs to signal decay. It is thought

that this adenylation process provides an unstructured, single-stranded region for

enzymes to latch onto and begin degradation (Doma and Parker, 2007). These types of

quality control mechanisms have been observed for aberrant stable RNAs. For example,

tRNAs that are not processed or modified correctly are subject to turnover by one of two

pathways. During a process called rapid tRNA decay (RTD), tRNAs lacking a covalent

modification are recognized by surveillance machinery, and turned over (Alexandrov et

Figure 3. RNA quality control pathways in eukaryotes

Quality control occurs in both the nucleus and cytoplasm of eukaryotic cells. Quality

control pathways are highlighted in red. Reprinted from Cell, 131, Doma, M.K. and

Parker. R., RNA quality control in eukaryotes, 660-668, 2007, with permission from

Elsevier.

al., 2006). Alternatively, some defective tRNAs are adenylated first, and then degraded

(Kadaba et al., 2004; LaCava et al., 2005; Wyers et al., 2005; Vanacova et al., 2005;

Kadaba et al., 2006; Schneider et al., 2007). Aberrant rRNAs are turned over in a similar

manner. Mutated, non-functional, rRNAs are degraded via the non-functional RNA decay pathway (NRD; LaRiviere et al., 2006), while mis-processed rRNAs are adenylated and

then degraded (Fang et al., 2004; Kuai et al, 2004; LaCava et al., 2005; Dez et al., 2006;

Kadaba et al., 2006; Win et al., 2006; Slomovic et al., 2006).

There are additional quality control pathways associated with mRNA processing

(Figure 3). These include distinct pathways for the turnover of mis-spliced mRNAs

(Bousquet-Antonelli et al., 2000; Hilleren and Parker, 2003; Conrad et al., 2006),

mRNAs lacking a functional 5’ cap (Schwer et al., 1998), and mRNAs with abnormal

polyA tails or other 3’ end mutations (Custodio et al., 1999; Hilleren et al., 2001; Libri et

al., 2002; Das et al., 2006; Rougemaille et al., 2007). mRNA quality control is not limited

to processing, however. In recent years, several groups have uncovered pathways devoted

to the turnover of mRNAs with errors in translation. During nonsense-mediated decay

(NMD), mRNAs containing a premature termination codon (PTC) are recognized and

turned over prior to the completion of translation (Hodgkin et al., 1989; Lim and Maquat,

1992; Lim et al., 1992; Belgrader et al., 1994; Muhlrad and Parker, 1994; van Hoof and

Green, 1996; Muhlrad and Parker, 1999; Buhler et al., 2002; Mitchell and Tollervey,

2003; Amrani et al., 2004; Couttet and Grange, 2004; Gatfield and Izaurralde, 2004;

Chan et al., 2007). After they are recognized by surveillance machinery, PTC-containing

mRNAs may be degraded from either end, or cleaved internally and then degraded.

Ultimately, the method of degradation used is organism-dependent (reviewed in Isken

and Maquat, 2007).

Three additional pathways rid the cell of translationally-defective mRNAs. The first

is non-stop decay (NSD), which is responsible for the degradation of mRNAs lacking a

stop codon (Frischmeyer et al., 2002; van Hoof et al., 2002; Inada and Aiba, 2005). NSD

targets are degraded via deadenylation-independent mechanisms (reviewed in Isken and

Maquat, 2007). Related to NSD is ribosome extension-mediated decay (REMD). During

REMD, mRNAs that have been translated past their stop codons are turned over (Inada and Aiba, 2005; Kong and Liebhaber, 2007). mRNAs associated with the opposite defect, stalls in translation, are turned over via a pathway called no-go decay (NGD; Doma and

Parker, 2006). NGD targets are cleaved internally, and the resulting fragments are degraded (Doma and Parker, 2006). Although all of these quality control pathways have

been identified, information regarding known pathways is incomplete. There are many

enzymes and co-factors involved, yet their biochemical characteristics, including the mechanisms they utilize in RNA turnover, are not fully understood.

I.C Ribonucleases

Although many proteins participate in RNA turnover, decay ultimately requires ribonucleases (RNases). These enzymes are physically responsible for the degradation of

RNAs. Many RNases also function in pre-RNA processing. Thus, RNases represent a crucial and fundamental feature of all living cells, and studies of these enzymes are important for understanding general mechanisms and regulation of RNA metabolism.

I.C.1 Physical features of RNases

RNases range from simple, single-domain proteins to complex, multi-functional

enzymes with multiple interaction and localization sequences. Some RNases may also

consist of catalytic RNA alone, but for brevity, these have been excluded from the discussion here. Structurally, all protein RNases have two things in common: a sequence that confers RNA binding, and at least one catalytic site involved in the destruction of

RNA phosphodiester bonds. RNA binding is mediated by a variety of protein domains.

Some domains are specific for RNA structure. That is, they bind either single-stranded or

double-stranded RNA regions (reviewed in Messias and Sattler, 2004). Other binding domains are specific for RNA sequence, and can confer the substrate specificity of an

RNase (reviewed in Auweter et al., 2006). Additionally, there are RNA binding domains

which are not specific at all, and simply mediate electrostatic interactions between the

RNase and any RNA. In later chapters, I describe an RNase called Dis3 that contains two

cold shock domains and an S1 domain. Cold shock domains were originally found in

RNA chaperone proteins involved in the cold shock response (Murzin, 1993; Graumann

and Marahiel, 1998). The binding reactions that this type of domain participates in are

diverse, and therefore, often non-specific (Graumann and Marahiel, 1998). S1 domains

are named after the ribosomal protein S1, which was initially found to bind U-rich

mRNA sequences (Boni et al., 1991). Since then, S1 domains have also been shown to bind more variable sequences, like those found in tRNAs (Schneider et al., 2007). Both cold shock and S1 domains contain distinct oligonucleotide binding (OB) folds (Boni et al., 1991; Graumann and Marahiel, 1998). These folds consist of a series of anti-parallel

β-sheets, which partake in stacking interactions with nucleic acids (reviewed in Theobald

et al., 2003). In RNases, these domains provide an efficient way for a single protein to

both contact RNA substrates and, in some cases, render them more structurally

vulnerable for degradation.

Degradation is coordinated by amino acids within the catalytic domains of RNases.

In most cases, these amino acids are separate from RNA binding domains, but there are a few enzymes in which the catalytic domain is able to bind RNA as well (e.g. Amblar et

al., 2006; Vincent and Deutscher, 2009). It is thought that additional binding by the

catalytic domain may support entry of an RNA into the active site (e.g. Frazao et al.,

2006). The active sites themselves contain conserved residues with polar, charged side

chains. Thus, lysines, arginines, histidines, aspartates, and glutamates are prevalent. Upon substrate binding, RNase active site residues form unique spatial arrangements that enable them to better coordinate catalysis. It is the type of activity mediated by the catalytic domain, however, that usually gives the domain its name. For example, Dis3 contains two catalytic domains; a PIN domain and an RNB domain. PIN domains were discovered in proteins involved in bacterial pilli biogenesis, hence the name of the Pil-T

N-terminal domain (Wall and Kaiser, 1999). RNB domains, conversely, were found in conserved proteins containing an RNase II-like activity (Zuo and Deutscher, 2001).

In addition to binding and activity domains, RNases may contain other sequences that aid in RNA degradation. Localization signals guide RNases to the subcellular sites where they function. For example, RNases that degrade mitochondrial RNAs

(mitoRNAs) may contain a defined mitochondrial targeting sequence (MTS). Many

RNases also have protein-interacting domains. Interactions with other proteins, or with other copies of the same protein, can modulate RNase activity (e.g. Dziembowski et al.,

2007; Lui et al., 2006). Finally, some RNases have specialized domains that render RNA

substrates more susceptible to degradation. An example is a helicase domain, which

unwinds the secondary structures of RNAs, thereby creating single-stranded regions

which are more easily degraded (e.g. Awano et al., 2010). The organization of RNA

binding, catalytic, and “specialized” domains within the enzyme is specific to each family

of RNases. Ultimately, it is the domain content of each RNase that makes it a highly

efficient degradation machine in the cell.

I.C.2 RNase mechanisms of action

All ribonucleolytic mechanisms consist of three steps: binding of the RNA

substrate, coordination of RNase active site residues, and cleavage of the target RNA by

some type of nucleophilic attack. Binding occurs by hydrogen bonding, nonpolar

stacking/packing interactions, and electrostatic interactions between RNA nucleotides

and RNase amino acid side chain groups (reviewed in Auweter et al., 2006). Binding of

the RNA usually elicits a conformational change in both the RNA and the RNase

(reviewed in Williamson, 2000). This change in structure helps align RNA

phosphodiester bonds within the active site of the RNase. In addition, it may position

active site residues closer to the target phosphodiester bond, which is favorable for

catalysis. Once a phosphodiester bond is properly positioned in the RNase active site,

cleavage can take place. For most RNases, this occurs by general acid-base or metal-ion

catalysis.

RNases that utilize general-acid base chemistry require a specific environmental pH

to function properly (e.g. Raines, 1998). Depending on pH, amino acid side chain groups

may either be protonated or deprotonated. Based on their protonation states, these groups

may act as general acids and/or bases during catalysis. For example, a deprotonated

amino acid group acts as a general base by abstracting a proton from the 2’ hydroxyl

group (2’ OH) of an RNA nucleotide (Figure 4, orange box; Cochrane and Strobel,

2008). This proton abstraction activates the 2’ OH, allowing it to become a nucleophile and to “attack” the phosphate group within the RNA phosphodiester bond. An additional protonated active site residue, acting as a general acid, donates a proton to the 5’ oxygen of what will be the “leaving” nucleotide (Figure 4, green box; Cochrane and Strobel,

2008). This stabilizes the leaving nucleotide, and results in destruction of the phosphodiester bond. Thus, in acid-base catalysis, the actions of acidic and basic residues within the RNase itself catalyze RNA cleavage.

Although general acid-base catalysis is effective, metal-ion catalysis is considered a more efficient method of RNA cleavage. RNases that utilize metal-ion chemistry

(metalloRNases) require a specific type and concentration of metal ions to function

(reviewed in Yang, 2008). The ions commonly used by metalloRNases are magnesium, manganese, calcium, or zinc (reviewed in Yang, 2008). It is not known why one type of metal ion is preferred over another, but it has been suggested that ion specificity is related to size of the RNase active site (Page and DiCera, 2006; Yang et al., 2006). Regardless of the metal-ion used, the chemistry is thought to be the same. Currently, it is thought that most protein metalloRNases utilize two metal ions during catalysis, although some

RNases may only use one (reviewed in Steitz and Steitz, 2003; Yang et al., 2006).

Figure 4. RNA phosphodiester bond cleavage by acid-base catalysis

Here, in a classic example of acid-base chemistry, RNase A histidine residues (orange

and green boxes) act as proton acceptors and proton donors to facilitate phosphodiester

bond scission. The lysine residue (blue box) is thought to stabilize the charged

intermediate. Reprinted with permission from Cochrane, J.C., and Stroble, S.A. Acc.

Chem. Res. Catalytic strategies of self-cleaving ribozymes, 41, 1027-1035. Copyright

2008 American Chemical Society.

In the RNase active site, both metal ions are coordinated by conserved amino acids. In

contrast to general acid-base chemistry, where amino acid side chain groups activate the

nucleophile, metal ions activate the nucleophile in this reaction. As described by Yang

and colleagues, one metal ion assists in the deprotonation of a water molecule (Figure

5A; Yang et al., 2006). The activated water/hydroxide then attacks the phosphate group of the target RNA phosphodiester bond (Figure 5B; Yang et al., 2006). A second metal ion stabilizes the intermediate that is formed prior to phosphodiester bond scission

(Figure 5B; Yang et al., 2006). Finally, the 3’ oxygen of the leaving nucleotide is protonated by another water molecule, which effectively destroys the phosphodiester bond (Figure 5C; Yang et al., 2006). Although water is used as the nucleophile here, some metalloRNases use metal ions to activate other nucleophiles such as inorganic phosphate (Pi), or even a 2’OH from the target RNA (reviewed in Symmons et al., 2002;

Yang et al., 2006).

Ions are not only used to coordinate cleavage in RNase reactions. In vivo, RNAs are coated with cations, which helps neutralize their negatively charged phosphate groups. Initially, this makes the RNA more stable, but this also allows proteins to bind, since cations disrupt the repulsion between negatively charged amino acid side chain groups and RNA phosphates (reviewed in Dupureur, 2008). Thus, RNases sometimes require the presence of cations to merely bind RNAs. In addition, metal ions may coordinate RNase substrate specificity (Dupureur, 2008). This suggests that cations can be important for all aspects of RNase activity.

Figure 5. RNA phosphodiester bond cleavage by metal-ion catalysis

(A), (B) A nucleophile (water, RNA 2’OH, Pi), which is activated by a metal ion in the

active site, attacks the phosphate group of an RNA phosphodiester bond. (C) The 3’

hydroxyl of the leaving nucleotide is protonated by another water molecule. Reprinted

from Molecular Cell, 22, W. Yang, J.Y., Lee, and M. Nowotny. Making and breaking

nucleic acids: Two-Mg2+-ion catalysis and substrate specificity, 5-13., Copyright (2006),

with permission from Elsevier.

I.C.3 RNase product formation and substrate specificity

Regardless of the active site chemistry used, RNases ultimately coordinate the

breakdown of RNA substrates into nucleotides or small oligonucleotide products. Based

on the position of the bonds they cleave, and the reaction products that they generate,

RNases are classified as either endo- or exoribonucleolytic. Endoribonucleases

(endoRNases) only cleave RNAs internally; they do not degrade every bond of a target

RNA. Thus, endoRNase cleavage reactions produce oligonucleotide fragments (Figure 6;

Yang et al., 2006). It is not known how endoRNases exclusively cleave the bodies of

RNAs. However, many endoRNases are sequence or structure specific. For example,

there are endoRNases which target A-U rich regions (AREs) (e.g. Claverie-Martin et al.,

1997). Other endoRNases target stem-loop structures (e.g. Korennykh, et al., 2007).

Thus, endoRNases can recognize distinct RNA elements, and cleave phosphodiester bonds at those sites.

EndoRNases also possess the unique ability to cleave single-stranded and/or double-stranded RNAs. In that regard, they are quite versatile, and function in a variety of RNA decay pathways. This includes general mRNA turnover (reviewed in Li et al.,

2010). One way that mRNA levels are controlled is by RNA interference (RNAi) (Fire et al., 1998; Montgomery et al., 1998). During RNAi, small double-stranded RNAs target complementary mRNAs for degradation (Montgomery et al., 1998). EndoRNases are responsible for creation of the double-stranded RNAs (Bernstein et al., 2001), and cleavage of the mRNA targets during RNAi (Liu et al., 2004). Other sequence-specific

Figure 6. EndoRNases cleave internal RNA phosphodiester bonds

Here, an endoRNase uses water as a nucleophile to attack internal bonds. Products

include one RNA fragment with a 3’ hydroxyl group and one fragment with a 5’

phosphate group. EndoRNases also utilize other reaction mechanisms to cleave RNAs,

which produces RNA fragments with different terminal groups (reviewed in Li et al.,

2010). Reprinted from Molecular Cell, 22, W. Yang, J.Y., Lee, and M. Nowotny. Making

and breaking nucleic acids: Two-Mg2+-ion catalysis and substrate specificity, 5-13.,

endoRNases directly recognize and cleave mRNAs in a deadenylation-independent

manner (schematized in Figure 2; reviewed in Li et al., 2010). Currently, there are 7

known endoRNases that target specific mRNAs for degradation in eukaryotes (reviewed

in Li et al., 2010). EndoRNases also participate in aberrant mRNA turnover. For example, the endoRNase SMG6 cleaves PTC-containing mRNAs targeted by NMD

(Huntzinger et al., 2008; Eberle et al., 2009). SMG6, like the Dis3 endoRNase that is the focus of this study, contains a PIN domain that mediates its activity (Glavan et al., 2006).

EndoRNases cleave mRNAs in the NGD pathway as well (Doma and Parker, 2006).

In addition to decay, endoRNases function in RNA processing pathways. One example is Dis3, which is thought to cleave 7S rRNA pre-cursors into functional 5.8S rRNA units (Lebreton et al., 2008; Schneider et al., 2009). Other endoRNases function in rRNA processing and tRNA processing pathways as well (reviewed in Granneman and

Baserga, 2004; Hopper and Phizicky, 2003). Although endoRNase activity is limited to the cleavage of one or a few phosphodiester bonds per RNA, it is clear these enzymes are important for the initial steps of RNA degradation and for the calculated processing of pre-RNAs.

Exoribonucleases (exoRNases) are quite different in terms of RNA cleavage and product formation. These enzymes start at one end of a target RNA, and cleave every subsequent phosphodiester bond in a linear manner, until they can no longer bind the

RNA. exoRNases are described as either 5’Æ3’ or 3’Æ5’, depending on the direction in which they cleave an RNA. Nucleotides are the primary products of exoRNase degradation reactions (reviewed in Ibrahim et al., 2008). The type of nucleotides produced depends on the nucleophile used during catalysis. Hydrolytic exoRNases,

which use activated water for catalysis, produce nucleotide monophosphates (NMPs)

(Figure 7, left). Phosphorolytic exoRNases use activated inorganic phosphates, and generate nucleotide diphosphates (NDPs; Figure 7, right).

ExoRNase reaction products are not limited to NMPs and NDPs, however.

Typically, exoRNases degrade an RNA up to the last few nucleotides. These RNA

fragments are then released as an end-product of the reaction because most exoRNases

cannot remain bound to small oligonucleotides. On average, exoRNases require a stretch

of 4-7 nucleotides to remain bound and continue degradation (e.g. Liu et al., 2006). The

amount of variability in oligonucleotide end-product length depends on the efficiency of

the exoRNase (reviewed in Symmons et al., 2002). Most exoRNases bind to an RNA

target once, and remain bound until they complete degradation. These exoRNases are

highly efficient, and their activity is described as “processive.” Processive exoRNases

generate a mix of NMPs or NDPs, and oligonucleotides of one defined length.

Conversely, “distributive” exoRNases bind and degrade small stretches of RNAs, and

then fall off the RNA (Symmons et al., 2002). This process is repeated multiple times; a

distributive exoRNase can re-bind and continue cleaving its original target, or can

degrade a new RNA. This type of activity results in the generation of both NMPs or

NDPs, and oligonucleotide fragments of various sizes.

Unlike endoRNases, exoRNases are characteristically non-specific (Ibrahim et al.,

2008). In that regard, they can recognize any free end of an RNA, and degrade it. Single-

stranded RNA regions are usually the substrate of choice (Ibrahim et al., 2008). However,

some exoRNases, like Dis3, are also able to degrade through regions with secondary

structure, without the aid of other proteins (Liu et al., 2006; Lorentzen et al., 2008). The

Figure 7. ExoRNase product formation: hydrolytic versus phosphorolytic cleavage

exoRNases use activated water or phosphate nucleophiles during catalysis to generate

NMP or NDP products, respectively. The reactions of 3’Æ5’ exoRNases are displayed

here. Image courtesy of R. Rajimakers. Reprinted from

double-stranded regions, however, are usually preceded by a single-stranded region that

the exoRNase can bind to and start degradation (e.g. Liu et al., 2006).

Because of their ability to completely degrade RNAs, exoRNases are involved in

nearly every RNA turnover pathway. A group of 3’Æ5’ exoRNases called the exosome,

which will be discussed in more detail later, degrades normal mRNAs that have been

deadenylated (Anderson and Parker, 1998). This same group of exoRNases degrades

RNA fragments produced by endonucleolytic cleavage in RNAi (Orban and Izzauralde,

2005). ExoRNases have also been implicated in the degradation of aberrant RNAs or

RNA fragments in almost every specialized pathway discussed in section I.B.3 (e.g.

Muhlrad and Parker, 1994; Mitchell and Tollervey, 2003 (NMD); van Hoof et al., 2005

(NSD); Kadaba et al., 2004; Schneider et al., 2007 (hypomodified tRNAs); Dez et al.,

2006 (defective rRNAs)).

Like endoRNases, exoRNases also participate in stable RNA processing. For

example, the exosome “trims” the 3’ ends of rRNAs, snRNAs, and snoRNAs during

maturation (Mitchell et al., 1996; Mitchell et al., 1997; Allmang et al., 1999a). The

exosome also degrades RNA spacer regions that are discarded during rRNA processing

(Allmang et al., 1999a). Finally, exoRNases are thought to participate in the trimming of

tRNAs during maturation as well (Papadimitriou and Gross, 1996). All of these studies

suggest that exoRNases are powerhouses of RNA processing and degradation in the

eukaryotic cell.

There is one additional category of RNases that has recently emerged. These

RNases possess both endo- and exoribonucleolytic activities. Because there are only a

few examples of these enzymes in eukaryotes, their biochemical characteristics have not

been completely established (e.g. Yang et al., 2009; Lebreton et al., 2008). It is assumed

that endo-exoRNases utilize the same active site chemistries that RNases with one

activity use. Product formation and substrate specificity are quite puzzling because the

interplay between the two active sites in endo-exoRNases has not been determined. Thus,

it is not known if each activity targets distinct substrates, or if both activities work on one

substrate. Additionally, it is not know how both activities are regulated. Like many

enzymes, endo-exoRNases could be regulated by interactions with other proteins, post-

translational modifications, subcellular localization, allostery, or a combination of these things. Work on one enzyme, Dis3, has shed some light on the function of endo- exoRNases, but many questions remain.

I.D Dis3

I.D.1a Dis3 in vivo: mitosis

Within the last 15 years, Dis3 has emerged as one of the major RNA turnover and

processing enzymes in eukaryotes. Dis3 was originally identified as a mitotic gene in

Schizosaccharomyces pombe. In 1988, Yanagida and colleagues showed that cold-

sensitive mutants of Dis3 were defective in sister chromatid disjoining (dis3-54; Ohkura

et al., 1988). Interestingly, this work also revealed that Dis3 mutants are sensitive to

caffeine (Ohkura et al., 1988). The authors concluded that caffeine only affects Dis3

expression (Ohkura et al., 1988). In a follow-up study, the same group showed that Dis3

is an ~110 kDa protein that is essential for viability (Kinoshita et al., 1991); Dis3 mutants

do not divide during mitosis. Additional analyses with Dis3 cold-sensitive mutants

showed that Dis3 constructs truncated at either the N- or C-terminus do not complement

the mutant phenotype (Kinoshita et al., 1991). This suggests that N- and C-terminal

sequences may be important for Dis3 essential function. In terms of in vivo function, it is

clear that Dis3 is needed for mitosis, but its exact role has not been described to date. In

recent years, the original cold-sensitive mutation has been identified as a change from a

proline to leucine (P509L) in the C-terminus of Dis3. Yanagida and colleagues showed

that recombinant proteins with this mutation have a reduced ability to degrade total RNA

in vitro (Murakami et al., 2007). From this, they concluded that Dis3 RNase activity is likely needed for progression through mitosis (Murakami et al, 2007). The connection

between Dis3 RNase activity and its function in mitosis has not been examined further.

I.D.1b Dis3 in vivo: RNA processing and turnover

Although Dis3’s role in mitosis remains elusive, its functions in RNA metabolism

are better defined. Six years after its original discovery, Dis3 was identified in a screen

for proteins involved in mRNA transport in Saccharomyces cerevisiae (mtr-17;

Kadowaki et al., 1994). The Dis3 mutant in this study was also shown to have rRNA and

tRNA processing defects (Kadowaki et al., 1994). Dis3’s involvement in rRNA

processing was later confirmed by Tollervey and colleagues. Dis3 (called Rrp44 in this

study), was isolated from S. cerevisiae in a complex of 3’Æ5’ exoRNases called the

exosome (Mitchell et al., 1997). The exosome complex contains 9-11 proteins, depending

on the organism (reviewed in Schmid and Jensen, 2008). This includes 6 proteins with homology to RNase PH (Rrp41/Ski6, Rrp42, Rrp43, Rrp45, Rrp46), 3 S1 domain- containing proteins (Rrp4, Rrp40, Csl4), 1 protein with homology to RNase D (Rrp6), and Dis3. In the 1997 study, strains depleted of Dis3 or other exosome proteins

accumulated an extended form of the 5.8S rRNA (Mitchell et al., 1997). Thus, it was

concluded that Dis3 and other exosome proteins trim the 3’ ends of S. cerevisiae 5.8S

rRNA precursors during maturation (Mitchell et al., 1997). Additional studies have suggested that Dis3, as part of the exosome, participates in many aspects of rRNA processing, including 3’ end trimming of 35S, 27S, and 7S pre-rRNA, and degradation of the 5’ external transcribed spacer (5’ ETS), a pre-rRNA sequence that is discarded during processing (Allmang et al., 1999a; Allmang et al., 2000; Suzuki et al., 2001; Houseley and Tollervey, 2006; Dziembowski et al., 2007; Lebreton et al., 2008; Schaeffer et al.,

2009; Schneider et al., 2009; Suzuki et al., 2001). Dis3 is also required for snRNA and snoRNA maturation, as several studies have shown that Dis3 mutants accumulate 3’

extended forms of these RNAs (Allmang et al., 1999a; Houalla et al., 2006; Schneider et

al., 2009). Together, these studies demonstrate that Dis3 activity is an important element

of RNA processing.

Following the analyses described above, Dis3 was found to be involved in

multiple RNA turnover pathways. As mentioned earlier, mRNAs with defects in splicing

are recognized and degraded by quality control machinery. This occurs in the nucleus,

and turnover of the mis-spliced mRNAs requires Dis3 (Bousquet-Antonelli et al., 2000).

mRNAs with defects in RNP formation are also recognized, and then degraded

exonucleolytically by Dis3 (Assenholt et al., 2008). Dis3 activities are not limited to

aberrant mRNA decay. In several studies, Dis3 has been shown to directly bind to and

degrade a hypomodified, aberrant tRNA (Kadaba et al., 2004; Schneider et al., 2007). It

was suggested that degradation of the tRNA by Dis3 was enhanced by adenylation of the

tRNA 3’ end (Schneider et al., 2007). Thus, polyA sequences may be a target of Dis3

exoRNase activity. In an unrelated study, Dis3 was shown to be involved in degradation

of an unstable class of RNAs in human cells (Preker et al., 2008). These unstable RNAs

are known as promoter upstream transcripts, or PROMPTS, and they are thought to be

synthesized and degraded, as a normal part of transcription (Preker et al., 2008).

In addition to aberrant RNA decay, Dis3 is thought to function in general mRNA

turnover. A recent study showed that elimination of Dis3 3’Æ5’ exoRNase activity

results in increased levels of a reporter mRNA in S. cerevisiae cells (Dziembowski et al.,

2007). It is not known if Dis3 targets specific endogenous mRNAs. However, from all of

the studies in yeast and human cells discussed to this point, one can conclude that Dis3 is

a significant component in both RNA turnover and maturation.

I.D.1c Dis3 in vivo: additional features

Although Dis3 is involved in many RNA metabolic pathways, Dis3 was isolated

as a member of the exosome, and thus it was thought for many years that Dis3 was yet

another enzyme in a larger complex of exoRNases. Initially, all Dis3 activities were

credited to the exosome. However, a study by Seraphin and colleagues thrust Dis3 into

the spotlight. In their 2007 work, they introduced a point mutation to the C-terminus of S.

cerevisiae Dis3, which eliminated its 3’Æ5’ exoRNase activity (D551N; Dziembowski et

al., 2007). Surprisingly, this mutation also abolished the exoRNase activity of the entire

S. cerevisiae exosome complex in vitro, and it also affected exosome rRNA processing

activity in vivo (Dziembowski et al., 2007). This result was quite unexpected because

other exosome proteins were shown to possess exoRNase activity in a previous study

(Mitchell et al., 1997). Although unexpected, this study has become the cornerstone for a

paradigm in which Dis3 is the only active 3’Æ5’ exoRNase activity in the S. cerevisiae

exosome, with the remaining exosome proteins serving as supporting or regulatory

factors.

Our group has shown that Dis3 may not function solely in context of the exosome

complex (Graham et al., 2009a; Kiss and Andrulis, 2009). There are several reports that

support this idea. As described above, Dis3 is active in mitosis (Ohkura et al., 1988;

Kinoshita et al., 1991; Murakami et al., 2007). Only one other exosome protein, Rrp6,

has similar functions (Graham et al., 2009b). In addition, Dis3 is found in complexes

lacking exosome proteins in both yeast and humans (Noguchi et al., 1996; Shiomi et al.,

1998). In these complexes, Dis3 binds to Ran, a protein involved in nucleocytoplasmic transport (Noguchi et al., 1996; Shiomi et al., 1998). It is possible that Dis3 is involved in

transport, in addition to mitosis, RNA turnover, and RNA processing. It is apparent from these analyses that we are only beginning to understand the many functions of Dis3.

I.D.2 Dis3 in vitro characteristics

Since Dis3 was labeled as the sole RNase of the exosome complex, several

RNase activity studies have been performed with Dis3 in vitro. The findings of these

experiments are described in the introduction sections of Chapters II and III. However,

most known in vitro biochemical characteristics can be summarized by a crystal

structure. In 2008, Conti and colleagues published a structural representation of an N-

terminally truncated form of S. cerevisiae Dis3 (Figure 8; Lorentzen et al., 2008). The

structure lacks the first 241 amino acids, which includes the PIN domain, a sequence later

found to contain an endoRNase active site (Lebreton et al., 2008; Schaeffer et al., 2009;

Figure 8. S. cerevisiae Dis3 crystal structure

An N-terminally truncated Dis3 construct bound to RNA is shown. Dis3 exoRNase

activity is inactivated here by the D551N mutation to the RNB domain. Dis3 domains are

in color. N-terminal cold-shock RNA binding domains are in yellow and orange. The C-

terminal exoribonucleolytic RNB domain is in blue, and the C-terminal S1 RNA binding

domain is in red. The bound RNA is in black. Additionally, a purple magnesium ion is

shown in the active site. Reprinted from Cell, 29, Lorentzen, E., Basquin, J., Tomecki, R.,

Dziembowski, A., and Conti, E. Structure of the active subunit of the yeast exosome core,

Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family, 717-728,

Schneider et al., 2009). Thus, the representation of the enzyme’s folds may be skewed.

Nevertheless, the structure shows the association of Dis3 with a 13 nucleotide polyA

RNA (Figure 8). It appears that the C-terminal exoribonucleolytic RNB domain makes

most of the contacts to the bound RNA (Figure 8; blue domain). The N-terminal cold-

shock domains also bind to the 5’ end (Figure 8; yellow and orange domains).

Interestingly, the protein appears to fold into a clamp (Lorentzen et al., 2008). It has been

suggested that this formation allows ssRNA substrates to be threaded into the exoRNase

active site (Lorentzen et al., 2008).

A more detailed schematic of S. cerevisiae Dis3-RNA contacts revealed a putative

reaction mechanism for exoRNase activity (Figure 9; Lorentzen et al., 2008). As shown

in Figure 9, a magnesium ion was found in the active site of the crystal. The authors speculated that an additional Mg2+ is present in the active site, but did not appear in the

crystal structure due to mutation of a critical residue (D551N; Lorentzen et al., 2008).

Both ions are coordinated by aspartic acids in the active site (Lorentzen et al., 2008).

Thus, it appears that S. cerevisiae Dis3 most likely uses two metal-ion catalysis for exonucleolytic cleavage of RNA substrates. Although not displayed in the schematic, the nucleophile is most likely a water molecule, as S. cerevisiae Dis3 does not require phosphate for RNA cleavage (Dziembowski et al., 2007).

Although a fairly detailed analysis of S. cerevisiae Dis3 exoRNase activity was provided with the crystal structure, endoRNase activity was not discussed in the work. As mentioned above, the N-terminal PIN domain of S. cerevisiae Dis3 was eventually found to mediate an endoRNase activity in vitro, but this was not reported until almost a full year after publication of the crystal structure (Lebreton et al., 2008; Schaeffer et al.,

Figure 9. Contacts between S. cerevisiae Dis3 amino acids and polyA RNA nucleotides

RNB domain residues are shown in blue, and an N-terminal cold-shock domain is in

yellow. The magnesium ion found in the crystal is shown in purple; a second putative ion

is circled. Dis3 contacts the RNA by both polar (dotted lines) and stacking (solid lines)

interactions. Reprinted from Cell, 29, Lorentzen, E., Basquin, J., Tomecki, R.,

Dziembowski, A., and Conti, E. Structure of the active subunit of the yeast exosome core,

Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family, 717-728,

Schneider et al., 2009). In fact, a complete analysis of S. cerevisiae Dis3 endoRNase

activity is still lacking. The exact requirements for this activity have not been established.

Substrate specificity is unknown. It is also unclear how exo- and endoRNase activities are

related, thus there is room for future analyses.

I.E Summary

Dis3 is a unique eukaryotic enzyme, possessing both endo- and

exoribonucleolytic activities. Because Dis3 functions in multiple RNA processing and

turnover pathways, studying this enzyme will allow us to define conserved mechanisms

of RNA cleavage and degradation. Although some detailed information regarding Dis3

RNase activity has been presented, the in vitro analyses are limited. To date, all Dis3 work has been performed with the yeast enzyme. Thus, it is not known if Dis3 RNase activity is conserved in multicellular eukaryotes. There are also many unanswered questions regarding both Dis3 activities, even in yeast. Further experiments are warranted

to formulate a complete analysis of Dis3 activities, both in vitro and in vivo.

I.F Hypothesis

Our group has endeavored to understand the functions of a previously

uncharacterized homolog of Dis3. Specifically, we wanted to determine the roles of Dis3

in the metazoan, Drosophila melanogaster. Prior to this study, work done by our group showed that Dis3 is associated with exosome proteins in Drosophila (Andrulis et al.,

2002; Graham et al., 2006; Graham et al., 2009a). Thus, we hypothesized that Dis3 may

function as an RNase in this organism as well. Indeed, sequence alignments of D.

melanogaster and S. cerevisiae Dis3 homologs have identified similarities between the

two proteins. Aspartic acid residues that have been implicated in S. cerevisiae exoRNase

function are conserved in the Drosophila homolog (Figure 10, purple). Similarly,

sequence alignments showed that catalytic residues in the PIN domain are also present in

Drosophila Dis3 (Figure 11, arrows). Thus, it appears that endoRNase activity may be

conserved in the Drosophila homolog as well. Based on this information, we developed

an in vitro system to assay the putative RNase activities of Drosophila Dis3. Using a

systematic approach, we were able to confirm conservation of Dis3 endo- and exoRNase

activities. Further, we were able to uncover other biochemical and cell biological

characteristics that were previously unidentified. These results are presented in the

following chapters.

Figure 10. Multiple sequence alignment of putative Dis3 exoRNase active sites

Conserved residues are highlighted in grey. Putative active site residues are highlighted

in purple. D551 is the amino acid in S. cerevisiae that has been mutated to knock out exoRNase activity. Reprinted from Cell, 29, Lorentzen, E., Basquin, J., Tomecki, R.,

Dziembowski, A., and Conti, E. Structure of the active subunit of the yeast exosome core,

Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family, 717-728,

Figure 11. Multiple sequence alignment of putative Dis3 endoRNase active sites

Conserved aspartic acids in the PIN domain endoRNase active site are designated by the

arrows. Additional conserved residues are in color. Reprinted by permission from

Macmillan Publishers Ltd: [Nature] (Lebreton, A., Tomecki, R., Dziembowski, A., and

Seraphin, B. Endonucleolytic RNA cleavage by a eukaryotic exosome, 456, 993-997),

Chapter II

Characterization of the Drosophila melanogaster Dis3 Ribonuclease

Mamolen, M., and Andrulis, E.D. (2009) Biochem and Biophys Res Commun 390, 529-

534

II.A Abstract

The Dis3 ribonuclease is a member of the exonucleolytic RNase II/RNase R

(RNR) family of enzymes. Although much progress has been made in understanding the

structure, function, and enzymatic activities of prokaryotic and yeast RNR proteins, there

are no functional studies of a metazoan Dis3 homolog. Here, we characterize the RNase

activity of Drosophila melanogaster Dis3 (dDis3). We find that dDis3 is active in the

presence of various ions, and requires divalent cations for activity. dDis3 hydrolyzes

compositionally distinct RNAs, yet releases different products depending upon the substrate. Reaction products included NMPs, which suggests that dDis3 has exoRNase

activity. A study of dDis3 interactions with dRrp6 and core exosome subunits in extracts

revealed sensitivity to higher divalent cation concentrations and detergent, suggesting the

presence of both ionic and hydrophobic interactions in dDis3-exosome complexes. Our

study thus broadens our mechanistic understanding of the functions of Dis3 and RNR

family members.

II.B Introduction

Dis3 is an essential enzyme (Kinoshita et al., 2001), with critical roles in RNA

metabolism. Although Dis3 is present in the majority of multicellular eukaryotes (Zuo

and Deutscher, 2001), biochemical data has only been collected for yeast homologs.

From these analyses, it appears that S. cerevisiae Dis3 acts as both an exo- and

endoRNase (Dziembowski et al., 2007; Schneider et al., 2007; Lebreton et al., 2008;

Schaeffer et al., 2009; Schneider et al., 2009). These activities only require divalent

cations (Dziembowski et al., 2007; Lebreton et al., 2008; Schaeffer et al., 2009;

Schneider et al., 2009). However, exact reaction mechanisms for this enzyme are

unknown. From studies of yeast homologs, it has been proposed that Dis3 exoRNase

activity consists of a series of processive hydrolytic cleavage events (Lorentzen et al.,

2008). Subsequently, it has been suggested that magnesium ions coordinate these

degradation reactions (Dziembowski et al., 2007; Lorentzen et al., 2008; Schneider et al.,

2007). Dis3 endoRNase activity is much less understood, as this activity was only

recently discovered in S. cerevisiae. It has been suggested that manganese increases S.

cerevisiae Dis3 endoRNase activity (Lebreton et al., 2008; Schaeffer et al., 2009;

Schneider et al., 2009), but the cause for manganese specificity has not been examined

further. Thus, many things remain to be studied about Dis3 function at the most basic

biochemical level.

The substrate specificity of Dis3 is an additional feature of the enzyme that

warrants further investigation. Since S. cerevisiae Dis3 cleaves RNAs regardless of

sequence or structure, it has been suggested that the enzyme is not a sequence specific

RNase (Mitchell et al., 1997; Dziembowski et al., 2007; Liu et al., 2006; Schneider et al.,

2007). However, reaction products do vary depending upon the substrate. For example, S.

cerevisiae Dis3 degrades pre-tRNAs, releasing products that are 3 nucleotides long

(Schneider et al., 2007). Conversely, degradation of single-stranded RNAs results in 4-5

nucleotide fragments (Liu et al., 2006). Yeast and bacterial Dis3 homologs have also

been shown to produce a mix of NMPs and 2-5 nucleotide fragments from degradation of

various RNAs (Cheng and Deutscher, 2002; Dziembowski et al., 2007). These studies

suggest that the RNase reaction mechanism utilized by Dis3 depends on substrate

identity. Consistent with this, substrate sequences influence the efficiency of Dis3

activity. For example, bacterial homolog RNase R has been shown to degrade polyA

RNAs better than polyC or polyU RNAs (Cheng and Deutscher, 2002). Additionally,

studies of S. cerevisiae Dis3 show it degrades polyA RNAs less efficiently than AU-rich

RNAs, or RNAs consisting of all four nucleotides (Liu et al., 2006). These in vitro

studies clearly demonstrate that Dis3 is able to degrade a variety of RNAs. However,

based on the observed fluctuations in degradation efficiency, it is possible that Dis3 has a

preference for specific RNA types. Further, as observed for the yeast and bacterial

systems, this putative substrate preference could vary between homologs.

There is also a lack of knowledge regarding Dis3 protein-protein interactions. It is

known that Dis3 directly binds to core exosome subunits, exosome co-factors, and

nuclear import proteins (Noguchi et al., 1996; Mitchell et al., 1997; Dziembowski et al.,

2007; Graham et al., 2009a). However, studies have suggested that Dis3 has different

interaction profiles depending on the organism (Allmang et al., 1999b; Andrulis et al.,

2002; Graham et al., 2009a). The manner in which these proteins associate is not entirely

clear either. Immunoprecipitation analyses with S. cerevisiae proteins suggest that Dis3

binds to core exosome subunits via ionic interactions (Allmang et al., 1999b;

Dziembowski et al., 2007). Yeast protein purification and electromagnetic imaging (EM), studies have suggested the relative arrangement of Dis3 in the largest known yeast exosome complex (Gavin et al., 2002; Gavin et al., 2006; Wang et al., 2007).

Additionally, crystallization studies have shown S. cerevisiae Dis3 associations with exosome subunits Rrp41 and Rrp45 in a small sub-complex (Bonneau et al., 2009). The specific types of interactions within these complexes have not been examined thoroughly, nor has the position of Dis3 in other protein-protein complexes been reported. As it has been widely speculated that Dis3-core exosome binding influences Dis3 functions, further analyses of these interactions may uncover the basis for regulation of this enzyme both in yeast and other organisms.

Whether metazoan Dis3 homologs have similar activities, ionic requirements, substrate specificity, and interactions has not been explored. To address this, and to clarify some of the results from the yeast studies, we characterized the Drosophila melanogaster Dis3 enzyme (dDis3). Specifically, we assessed requirements for in vitro

RNase activity, the ability of the enzyme to degrade different RNA substrates, the products of the ribonucleolytic reactions, and characteristics of dDis3-core exosome interactions. As this study represents the first analysis of a metazoan Dis3, our findings help build a more complete picture of the general features of Dis3 functions.

II.C Materials and Methods

II.C.1 Molecular cloning

All plasmids were constructed using basic molecular cloning techniques. All constructs

were screened by digestion with restriction endonucleases and sequenced to confirm the

absence of errors. For dDis3-FLAG constructs used in immunoprecipitation assays,

dDis3 genes were PCR amplified from the full-length ORF using primers shown in Table

1, Appendix A. The 5’ primer has a unique BglII site and the 3’ primer has an in-frame

FLAG (DYKDDDK) tag followed by a stop codon and a unique SalI site. This PCR

product was digested with BglII and SalI and cloned into the BamHI and SalI sites of

pRmHa3 to obtain metallothionein (Mtn) promoter driven dDis3 genes. These constructs

were transiently transfected into S2 cells using CELLFectin (Invitrogen), tested for

copper-inducible expression, and then established as stable cell lines as described

previously (Graham et al., 2009a). For in vitro experiments, the wild-type dDis3 gene

was cloned into pMal c2 to create MBP fusions used in RNase assays.

II.C.2 Expression and purification of recombinant proteins

MBP or MBP-dDis3 was transformed into Escherichia coli strain DH5α. Over-

expression of proteins was induced with the addition of 1 mM IPTG (Denville Scientific)

to 500 mL cultures overnight at 20˚C. Cells were lysed by treatment with 1 mg/mL

lysozyme (Sigma), and by sonication in buffer containing 2 mM tris-HCl, pH 7.5, 100

mM NaCl, 0.1 mM PMSF, and 1x EDTA-free protease inhibitor cocktail (Roche).

Lysates were loaded onto 1 mL amylose resin (New England BioLabs), and washed with

80 mL of buffer (20 mM tris-HCl, pH 7.5, 100 mM NaCl, 0.1 mM PMSF). Proteins were

eluted in buffer containing 20 mM tris-HCl, pH 7.5, 100 mM NaCl, and 50 mM maltose.

Proteins were also dialyzed into buffer containing 20 mM tris-HCl, pH 7.5, 100 mM

NaCl, and 10% glycerol. Dialyzed proteins were visualized by 12% SDS-PAGE and

Coomassie staining. Protein concentrations were determined from Coomassie stained gels

by comparison to a BSA standard curve using Quantity One Software.

II.C.3 Preparation of RNA substrates

RNA substrates polyU (32 nucleotides), polyA (30 nucleotides; Dharmacon), polyC (30

nucleotides, Dharmacon), and polyN (31 nucleotides;

5’GCGUCUUUACGGUGCUUAAAACAAAACAAAA3’) were used in RNase assays.

RNA concentrations were determined by spectrophotometric analysis. RNAs were

radiolabeled at the 5’ end with γ32ATP (Perkin-Elmer) using T4 polynucleotide kinase

(Promega). Unincorporated radiolabeled nucleotides were removed using NucAwayTM

spin columns (Ambion) per manufacturer’s recommendations.

II.C.4 Ribonuclease activity assays

Assays were adapted from previously published protocols (Dziembowski et al., 2007).

Radiolabeled RNA was incubated alone or with recombinant proteins at 37˚C in buffer

containing 10 mM tris-HCl, pH 8.0, 75 mM KCl, and 40 μM MgCl2, unless otherwise

indicated. In all reactions, RNA concentration was 120 nM, and protein concentration

was 60 nM, except in Figure 12 experiments where MBP-dDis3 concentration varied. At

the time points indicated in each figure, 10 uL of RNase reactions were taken out and

stopped with 10 uL of formamide loading buffer (10 mM EDTA, 0.1% bromophenol

blue, 0.1% xylene cyanol, 95% formamide). Reaction products were separated on 12.5%

acrylamide, 8M urea denaturing gels or TLC plates (developed in KH2PO4 buffer), and

visualized by autoradiography.

II.C.5 Quantification of RNase activity

The following method was used to create all graphical representations of RNase activity.

The amount of RNA remaining at each experimental time point was quantified by densitometry using ImageQuant Software. For the majority of experiments, % polyU remaining was determined as the ratio of full-length RNA remaining at a particular time point to the amount of RNA present at time zero. For experiments in Figures 14 and 15,

% polyU was calculated as the ratio of product to full-length RNA for each reaction since there was only one time point. Data was graphed using GraphPad Prism Software.

II.C.6 Dis3 immunoprecipitation

S2 stable cell lines were constructed and maintained as described previously (Graham et al., 2009a). FLAG immunoprecipitations were performed as before with the following exceptions (Graham et al., 2009a). Wash buffer (10 mM tris-HCl, pH 7.4, 150 mM NaCl,

3 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 1% Triton X-100, 10% glycerol, protease

inhibitor cocktail (Invitrogen)) was supplemented with NaCl to 1.0 or 1.25 M, Triton X-

100 to 5%, or MgCl2 to 0.1 M as designated in Figure 16A. For experiments in Figure

16B, NaCl concentration was increased to 0.2 or 1.0 M, and MgCl2 concentration was

0.01, 0.2, 0.4, 0.6, 0.8, 1.0, or 1.2 M as designated. Western blotting of immunoprecipitation reactions was performed as described previously with antibodies to

dDis3F (dDis3-FLAG) or endogenous exosome subunits (Graham et al., 2006; Graham et

al., 2009a).

II.D Results

II.D.1 Recombinant dDis3 is active in vitro

Although the RNase activity of S. cerevisiae Dis3 has been characterized, there is no evidence a metazoan Dis3 homolog functions in a similar manner. To address this, we purified full-length recombinant Drosophila melanogaster Dis3 as maltose binding protein fusions (MBP-dDis3; Figure 12A), and established a system to assess RNase activity. MBP-dDis3 completely degraded a 5’end-labeled polyU RNA within 60 minutes

(Figure 12B). Further, MBP-dDis3 activity was robust, as polyU RNA was also completely degraded when the enzyme to substrate ratio was 1:120 (Figure 12B, lane 8).

II.D.2 MBP-dDis3 RNase activity requires monovalent and divalent cations

Optimal ionic conditions for RNase activity can vary from one enzyme to another, including bacterial homologs of Dis3 (Cheng and Deutscher, 2002). We tested multiple assay buffers to determine which ionic environments promote Dis3 enzymatic activity in

vitro. First, we assessed the activity of MBP-dDis3 on polyU RNA in reaction buffer

containing no added monovalent or divalent cations. As expected, little activity was

observed under these conditions during the 60 minute reaction period, indicating that

dDis3 requires the presence of cations to function properly (Figure 13, “tris”). Upon

addition of any monovalent cation tested, MBP-dDis3 activity was enhanced (Figure

13A). Graphical analysis of the assays shows that different monovalent cations have

Figure 12. Drosophila melanogaster Dis3 has ribonuclease activity in vitro

(A) Recombinant MBP and MBP-dDis3. Proteins and prestained protein marker (New

England BioLabs) were loaded onto the gel. Molecular weight standards are labeled on the left side of the gel. (B) MBP-dDis3 is active at multiple concentrations in vitro. PolyU

RNA was incubated alone (buffer) or with recombinant protein for 60 minutes. Full-length

RNA is marked by (*) and the smallest degradation product is marked by (♦); these symbols are used throughout, unless otherwise noted. Image is representative of at least two independent experiments.

Figure 13. MBP-dDis3 is active in the presence of various monovalent cations

(A) MBP-dDis3 degrades polyU RNA in any monovalent cationic condition tested.

Reaction buffer contained 10 mM tris-HCl, pH 8.0 alone, or tris-HCl and 75 mM of the

monovalent cation designated on the left side of each gel. Additionally, 1 mM 2-

mercaptoethanol was present in one of the experiments (K+ + 2me). (B) MBP-dDis3 is

equivalently active in all monovalent cationic conditions tested. RNase activity was

quantified as described in Materials and Methods. The MBP control line (dashed line, ●)

represents data averaged for all of the reaction conditions. Control data has also been

separated for each reaction condition, and can be viewed in Figure 11. The remaining

lines on the graph are as follows: Tris ; KCl S; KCl + 2me T; NaCl ¡; NH4Cl c;

CsCl Δ; LiCl ⁬. These represent data averaged for at least two independent experiments.

Figure 14. MBP is not active on polyU RNA in any ionic condition tested

(A) Graphical representation of MBP activity in monovalent cation-containing buffers.

Ionic conditions are represented as follows: Tris ; KCl S; KCl + 2-Me T; NaCl ¡,

NH4Cl c; CsCl Δ; LiCl ⁬. (B) MBP activity in divalent cation-containing buffers. Ionic

conditions are represented as follows: MnCl2 + KCl ●; MnCl2 ; MgCl2 S. Data shown represents averages of two independent experiments for each condition.

approximately the same affect on MBP-dDis3 activity, suggesting this activity is not

monovalent cation-specific (Figure 13B). As a control, MBP alone had negligible activity

in any ionic reaction condition (Figure 14).

We also examined the effects of divalent cations on RNase activity, as several

reports have stated that Dis3 utilizes either magnesium or manganese for catalysis.

Divalent cations alone were able to elicit MBP-dDis3 activity (Figure 15A, Mn2+ and

Mg2+), although activity was not optimal. Addition of potassium to the reaction buffer was sufficient to increase degradation of polyU RNA (Figure 15, Mn2+/K+). Graphical

presentation of RNase activity time courses demonstrates that the t1/2 (defined here as the

time at which 50% of polyU is remaining) for Mn2+/K+ is 2 minutes (Figure 15B). In

2+ 2+ comparison, the t1/2 for both Mn and Mg alone is 20 minutes. This represents a 10- fold increase in MBP-dDis3 activity when monovalent and divalent cations are in the reaction buffer. These data suggest that monovalent cations are required for efficient dDis3 activity.

Although the assays above revealed a possible requirement for monovalent cations, it was not clear whether divalent cations are necessary for activity. The experiments in

Figure 13 were performed using reaction buffers lacking divalent cations, and MBP- dDis3 was still active. However, it is possible that divalent cations were imported into the reactions along with the proteins or RNAs. To resolve this issue, we added a chelating agent to our reactions; EDTA completely inhibited degradation of polyU RNA (Figure

16). We interpret this to mean that the RNase activity of MBP-dDis3 requires divalent cations in vitro.

Figure 15. MBP-dDis3 is activated by divalent cations

(A) MBP-dDis3 degrades polyU RNA in the presence of divalent cations. Reaction buffers contained 10 mM tris-HCl, pH 8.0, 1 mM 2me, and 75 mM KCl + 40 μM MnCl2

(top panel), 40 μM MnCl2 (middle panel), or 40 μM MgCl2 (bottom panel). (B) MBP- dDis3 is efficiently activated by a combination of monovalent and divalent cations in

vitro. % polyU remaining for the MBP control represents data averaged from all reaction

conditions (dashed line, ●). Separated control data is presented in Figure 14. Ionic

conditions are graphed using the following symbols: MnCl2 + KCl ●; MnCl2 ; MgCl2

S. Data shown represents averages of at least two independent experiments for each

condition.

Figure 16. MBP-dDis3 RNase activity requires divalent cations

MBP-dDis3 was incubated with polyU RNA for 10 minutes in buffer with (right panel) or without (left panel) 5 mM EDTA. Images shown are representative of at least two

independent experiments.

To complete our study of dDis3 ionic requirements, we performed two experiments in which ion concentrations were increased and/or decreased in a series of 10 minute reactions. We reasoned that dDis3 activity likely requires specific ionic concentrations for catalysis, as this is apparent for other Dis3 homologs (Cheng and Deutscher, 2002;

Dziembowski et al., 2007; Lebreton et al., 2009). In the first experiment, magnesium concentration was constant (either no magnesium, or 40 μM ), and potassium concentration was varied. Decreases in potassium concentration below 75 mM resulted in nearly complete inhibition of MBP-dDis3 activity, either in the absence or presence of magnesium (Figure 17A). The sensitivity to potassium concentration is manifest by a shift from 0% full-length polyU remaining at 75 mM K+ to 70% remaining with 7.5 mM

K+ in the reaction buffer (Figure 17B). This suggests that monovalent cation concentrations play a role in the efficiency of MBP-dDis3 activity. In the second experiment, magnesium concentration differed, and potassium concentration remained the same (either no potassium, or 7.5 mM). We observed a peak of MBP-dDis3 activity at

4 mM magnesium (Figure 18). Increases or decreases in magnesium concentration beyond that point led to a reduction in MBP-dDis3 activity. This suggests that specific concentrations of magnesium ions are likely required for efficient catalysis as well.

II.D.3 MBP-dDis3 activity is not affected by non-ionic reaction conditions

Mechanistically, other conditions are often needed by RNases for catalysis, including co-factors, specific pH, and denaturing environments. To determine if dDis3 activity is enhanced by any of these factors, we performed several additional assays in buffers containing various concentrations of reducing agents, acids, or nucleotides. MBP- dDis3 activity was only modestly affected by increases in 2-mercaptoethanol

Figure 17. MBP-dDis3 RNase activity is sensitive to monovalent cation concentrations

(A) Optimal MBP-dDis3 RNase activity requires greater concentrations of

monovalent cations. Gels depict reaction products that accumulated after 10 minute

incubations of protein with polyU RNA. Each lane represents independent reactions with

differing concentrations of KCl, as specified at the top of the gels. Two experiments are

displayed; one set of reactions contained no additional magnesium in the reaction buffer

(top panel), the other set contained 40 μM MgCl2 (bottom panel). Control reactions were

assayed using only one buffer condition. (B) Monovalent cations prompt more efficient

MBP-dDis3 RNase activity in vitro. Ionic conditions were graphed as follows: -MgCl2 line ●; + MgCl2 line . Data was averaged from two independent experiments.

Figure 18. Divalent cation concentrations affect MBP-dDis3 RNase activity

(A) MBP-dDis3 RNase activity varies with divalent cation concentrations. Assays were performed as described in Figure 17, except MgCl2 concentration varied, and KCl concentration was constant. Either KCl was not added to reaction buffers (top panel) or was present at 7.5 mM (bottom panel). Arrows point to uneven gel fronts caused by the salt content of reaction buffers. (B) MBP-dDis3 RNase activity is stimulated by specific magnesium concentrations. Experiments are represented as follows: -KCl line ●; +KCl line . Data was averaged from two independent experiments.

concentration (Figure 19A) or pH (Figure 19B). Additionally, activity did not require nor

was enhanced by nucleotides, as MBP-dDis3 efficiently degraded polyU RNA in the absence or presence of ATP or dATP (Figure 20). Thus, it appears that out of the conditions tested, dDis3 only utilizes ions for degradation of RNAs in vitro. A summary of all dDis3 condition-specific data is displayed in Table 1.

II.D.4 MBP-dDis3 RNase activity is not sequence specific

We next wished to examine both the nature and specificity of dDis3 RNase activity. To ensure that the activity of MBP-dDis3 is not specific to the polyU substrate utilized in all of the assays above, we tested its activity on additional single-stranded

RNAs. As expected for sequence non-specific RNases, MBP-dDis3 degraded polyA, polyC, and an RNA containing all four nucleotides (polyN) within 1 hour (Figure 21, top). Within 3 hours, MBP-dDis3 degraded some of the reaction intermediates present at the 1 hour time point (Figure 21, bottom). Reaction products also appear to be different between the distinct substrates. MBP-dDis3 activity on polyA, polyC, and polyU

liberated similar ladders of products. However, MBP-dDis3 activity on polyN yielded

mainly the larger and smaller products, with a bias against mid-size products. This

suggests that although activity is not sequence specific in general (dDis3 cleaved all

RNAs), RNA sequence may still influence the types of reaction products that are

generated by dDis3.

To determine the identity of the smallest reaction products, we performed thin layer chromatography (TLC; Figure 22). TLC analysis of products from MBP-dDis3 degradation of polyA or polyC revealed spots that migrated at the locations of AMP and

CMP, respectively (Figure 22). These data show that MBP-dDis3 liberates nucleotide

Figure 19. MBP-dDis3 RNase activity is not affected by certain non-ionic conditions

in vitro

(A) MBP-dDis3 is active in a range of reducing conditions. RNase activity was assayed on a polyU RNA in buffers containing different 2-mercaptoethanol concentrations, as

designated. Reactions products were analyzed after 60 minutes. (B) MBP-dDis3 is active

in a range of pH conditions. Reaction buffer pH values are listed. Data shown is

representative of at least two independent experiments.

Figure 20. Nucleotide co-factors are not required for MBP-dDis3 activity in vitro

RNase activity was assessed on a polyU RNA in buffers supplemented with ATP or

dATP. 1.2 μM ATP (lanes 2, 5, 8) or dATP (lanes 3, 6, 9) was added to the RNase

reactions. Data shown is representative of two independent experiments in which reaction

products were analyzed after 60 minutes.

Table 1. Summary of MBP-dDis3 condition-specific RNase assays

Construct Substrate Buffer Time (min) Result 1-982 U tris only ≤ 60 activity impaired 1-982 U K + ≤ 60 active 1-982 U K+, 2Me ≤ 60 active 1-982 U Na + ≤ 60 active 1-982 U NH 4+ ≤ 60 active 1-982 U Cs + ≤ 60 active 1-982 U Li + ≤ 60 active 1-982 U Mn 2+, K + ≤ 60 active 2+ 80 1-982 U Mn alone ≤ 60 activity impaired 1-982 U Mg 2+ alone ≤ 60 activity impaired 1-982 U Mg 2+, K +, ±EDTA 10 activity impaired by EDTA 1-982 U 2Me titration 60 active; slight inhibition at higher 2Me concentrations 1-982 U pH titration 60 active; slight inhibition at more basic pH values 1-982 U ATP addition 60 active

Figure 21. MBP-dDis3 RNase activity is not sequence specific

Activity was assessed on four different RNA substrates, as designated. Nucleotide composition of the RNA substrates is listed in Materials and Methods. Data shown is representative of at least two independent experiments for each substrate.

Figure 22. MBP-dDis3 has 3’Æ5’ exoribonuclease activity

Reaction products that accumulated after one hour incubations of MBP-dDis3 with polyA or polyC RNAs were separated on TLC plates. Products were identified by comparison to non-labeled nucleotide standards run on the same TLC plates.

monophosphates as a product of catalysis. Together, these experiments suggest that

dDis3, like S. cerevisiae Dis3, is an exoRNase.

II.D.5 dDis3 likely associates with exosome proteins via ionic and hydrophobic interactions

To complete our basic analysis of dDis3 function, we investigated dDis3 protein- protein interactions. Specifically, we looked at how dDis3 associates with exosome subunits. Previous work in yeast has shown that interactions between S. cerevisiae Dis3 and the exosome core are sensitive to divalent cation concentrations, with S. cerevisiae

Dis3 disassociating at 200 to 800 mM of magnesium (Allmang et al., 1999b). This

suggests these interactions are ionic. As the enzymatic activity of dDis3 is sensitive to

changes in monovalent and divalent cation concentration, we suspected that varying salt

conditions could change functionally relevant protein-protein interactions also. We first

assessed whether monovalent cations affect dDis3-dRrp6-core exosome interactions. We

performed FLAG immunoprecipitation experiments with dDis3-FLAG isolated from S2

whole cell extracts, and measured the recovery of dRrp6, the nuclear co-factor dRrp47,

and core exosome subunits dRrp46, dRrp4, and dCsl4. The control for these experiments

was extracts from cells expressing the vector alone (Mtn). We saw no effect on the ability

of dDis3-FLAG to recover these proteins at concentrations up to 1.25 M Na+, and ~8-fold

increase over normal wash buffer conditions (Figure 23A, lanes 3-10). The recovery was

also stable in the presence of 100 mM Mg2+. Interestingly, we only observed a difference

at high salt concentrations when 5% Triton X-100 was also present. The addition of TX-

100 to binding buffer resulted in a 2-fold reduction in co-immunoprecipitation efficiency

Figure 23. dDis3-dRrp6 and dDis3-core exosome interactions are stable in vitro

(A) dDis3 associates with exosome subunits through hydrophobic interactions. dDis3-FLAG interactions with dRrp6

and core exosome subunits were analyzed in buffer enriched with NaCl and Triton X-100 (TX). (B) dDis3 also

associates with dRrp6 and core exosome subunits through ionic interactions. dDis3 interactions were analyzed in

buffers containing varying concentrations of NaCl and MgCl2, as specified. Note that recovery of dDis3-FLAG resin is

impeded at MgCl2 concentrations higher than 0.6 M. Data courtesy of Erik Andrulis.

(Figure 23A, lane 11). The sensitivity of the interactions to detergent suggests exosome

proteins may associate with dDis3 via hydrophobic interactions.

We also tested the effect of divalent cations on dDis3-dRrp6-core exosome

interactions. Divalent cation concentration varied in these assays, and monovalent

concentrations were constant (either 0.2 or 1.0 M Na+). At low Mg2+ concentrations, we

observed quantitative recovery of the assessed proteins (Figure 23B, lanes 1 and 15). A

20-fold increase in Mg2+ concentration elicited a modest effect on dDis3-FLAG binding to the FLAG resin, yet caused a 2-5 fold reduction in dDis3-FLAG-mediated recovery of

dRrp4, dRrp46, dCsl4, and dRrp47, with a lesser effect on dRrp6 recovery. Increasing the

Mg2+ concentration to 0.4 M led to a significant reduction of dDis3-FLAG binding to the

resin, and a consequential loss of binding to the other proteins as well (Figure 23B, lanes

5 and 19). Additional findings are complicated as higher Mg2+ concentration (0.6-1.2 M)

reduced or eliminated dDis3-FLAG interaction with the resin. Together, these data

suggest that dDis3 likely associates with exosome proteins through both ionic and

hydrophobic interactions, where divalent cations mediate the ionic interactions.

II.E Discussion

In this work we have developed an experimental system to characterize the

activity of a metazoan Dis3 homolog. We have uncovered several fundamental features

regarding the in vitro RNase activity, and exosome interactions of Drosophila

melanogaster Dis3. These analyses have provided us with a basic understanding of dDis3 biochemistry, and have facilitated more advanced studies of this enzyme.

II.E.1 Drosophila melanogaster Dis3 is a functional ribonuclease in vitro

We have shown for the first time that full-length recombinant dDis3 is able to

degrade RNA in vitro. This activity only requires cations, and is not influenced by

changes in pH, denaturing conditions, or nucleotide co-factors. Although divalent cations

likely drive the dDis3 reaction (explained in detail in II.E.2), our studies uncovered that

dDis3 RNase activity requires monovalent cations as well. However, monovalent cation identity doesn’t appear to be important. This requirement is similar, but not identical to, that of dDis3 bacterial and yeast homologs. Prior to our study, S. cerevisiae Dis3 was

shown to be similarly active in solutions containing K+, and Na+; no other monovalent

ions were tested (Dziembowski et al., 2007). By comparison, bacterial homolog RNase II

has been shown to vary in activity depending on the monovalent ion. RNase II is

+ + + + optimally activated by K and NH4 , whereas Li varies in its contributions, and Na weakly stimulates activity (Spahr and Schlessinger, 1963; Spahr, 1964; Singer and

Tolbert, 1964; Gupta et al., 1977). This divergence in monovalent ion requirements between eukaryotic and bacterial Dis3 homologs could point to differences in reaction mechanism. However, monovalent cations have never been shown to directly participate in Dis3 catalysis. Consistent with this, they have not been observed in the active sites of

Dis3 crystals (Lorentzen et al., 2008; Bonneau et al., 2009). We propose that monovalent cations are important to Dis3 RNase activity because they mediate Dis3-RNA interactions. Thus, decreases in monovalent cation concentrations in our assays likely perturbed associations between enzyme and substrate, resulting in loss of RNA degradation. This could also explain why dDis3 activity does not vary in different monovalent ionic environments. The charges of these ions alone may facilitate large scale interactions, but do not participate in site specific coordination by active site amino acids.

If dDis3 activity was monovalent cation-specific, we would expect that different cations

would produce variations in reaction efficiency as they are different in size, and would presumably interact differently with dDis3 active site residues. Further experiments are warranted to identify the exact roles monovalent cations play in activating Dis3.

II.E.2 dDis3 Mg2+ requirements point to a metal-ion catalyzed reaction

mechanism

Our in vitro data confirms that dDis3 RNase activity requires divalent cations,

since chelation of magnesium resulted in complete loss of dDis3 activity. As dDis3

requires Mg2+ at specific concentrations to function in vitro, it is likely that dDis3 utilizes

metal-ion catalyzed chemistry to cleave phosphodiester bonds. Further, we show that

dDis3 activity is not greatly affected by changes in pH, which suggests this enzyme does

not employ a general acid-base catalytic mechanism. The nucleophile in the reaction is

most likely water, as dDis3 does not require phosphate for cleavage. Thus, we envision a

hydrolytic metal-ion catalyzed reaction mechanism for dDis3, similar to the mechanism

depicted in Figure 5. Our observations are consistent with analyses of S. cerevisiae Dis3

crystal structures. As mentioned previously, these structure show that magnesium ions

are associated with a C-terminal RNB domain active site (Lorentzen et al., 2008). Other

functional assays have suggested that these magnesium ions coordinate Dis3 3’Æ5’

exoRNase activity, and they are not simply required for substrate binding (Schneider et

al., 2007). Our data shows that full-length dDis3 is able to release NMP products from

cleavage of polynucleotide RNAs in the presence of magnesium. Thus, our data confirms

that magnesium-activated exoRNase activity is conserved among Dis3 homologs.

Notably, we also observed that full-length dDis3 RNase activity is stimulated by manganese. This is consistent with reports that dDis3 uses manganese to cleave RNAs

(Lebreton et al., 2008; Schaeffer et al., 2009; Schneider et al., 2009; Mn2+ stimulates

endoRNase activity in S. cerevisiae Dis3). However, it is not clear from our studies and

others whether manganese plays a role in degradation, substrate binding, or both.

Interestingly, wild-type S. cerevisiae Dis3 degradation of some RNAs is inhibited by

manganese (Schneider et al., 2009). This could suggest that divalent ions also play a role in Dis3 substrate specificity. Further analysis will be required to determine if manganese directly participates in catalysis, or if it plays an auxiliary role in substrate binding and/or regulation of activity.

II.E.3 dDis3 product formation may depend on substrate identity

In our study, dDis3 degraded multiple RNAs consisting of varying sequences.

This indicates that dDis3 activity is not sequence specific in vitro. However, reaction

products did vary depending on substrate. Oligonucleotide fragments generated from

cleavage of polyN RNAs were different than those produced from degradation of polyA,

polyC, and polyU. This could point to variations in reaction mechanisms used by dDis3

to cleave diverse pools of RNAs. We envision several putative reaction mechanisms

based upon our data. First, to generate the reaction products we observed for all the

RNAs, dDis3 may have liberated NMPs via a processive exoRNase activity until

reaching the very 5’ end of the RNA. On some RNAs, dDis3 may have been unable to

bind the RNA any longer, and would have released the last few nucleotides as larger

RNA fragments. The lengths of the oligonucleotide fragments varied in our assays,

perhaps due to structural differences among the RNAs. This mechanism of action has

been proposed for other RNR family members (Frazao et al., 2006; Zuo et al., 2006).

This particular model is based upon the observation that these enzymes require an ~7

nucleotide region for tight binding to RNA, and/or for the RNA to reach the active site

(Vincent and Deutscher, 2006). Alternatively, although we have not identified an

endoRNase activity in dDis3 to this point, a combination of exo- and endoRNase

activities may have be used by dDis3 to generate the observed products. EndoRNase

activity may have produced the fragments, and exoRNase activity may have produced the

NMPs. It is possible that dDis3 utilizes two activities to degrade a single substrate.

Indeed, it has been suggested that S. cerevisiae Dis3 uses both endo- and exoRNase

activities during rRNA processing (Lebreton et al., 2008).

Neither of these reaction mechanisms, however, account for the differences we

saw between degradation of homopolymer and polyN RNAs. It is possible that dDis3

may have toggled between mechanisms depending on the substrate. As mentioned before,

activity on all of the homopolymers produced similar sized products, but fragment

products generate from polyN degradation varied in size. This could be indicative of a switch between processive and distributive activity. Perhaps dDis3 remains bound to a homopolymer RNA and processively degrades through long stretches of the same nucleotide. Thus, similar products would be generated from degradation of all homopolymers, as we observed here. However, when dDis3 encounters different types of nucleotides, like in the polyN substrate, dDis3 may become distributive, degrading short stretches of the RNA, and then falling off. This reaction mechanism usually generates unequal ladders of products, like we observed here for polyN. This type of distributive product pattern has also been observed for the exosome subunit Rrp4 (Mitchell et al.,

1997). Although it is difficult to determine which, if any, of these reaction mechanisms

was used by dDis3 to degrade each substrate tested in our assays, our results demonstrate

that dDis3 cleavage of distinct RNA sequences is not entirely uniform.

II.E.4 dDis3 stably associates with core exosome proteins and exosome co-factors

Our immunoprecipitation study shows that interactions between dDis3 and core

exosome subunits remain stable at varying monovalent cation (Na+) concentrations.

Conversely, increasing concentrations of divalent cations (Mg2+) result in dissolution of these interactions. This is in agreement with observations of S. cerevisiae Dis3 interactions, and suggests that Dis3 associates with exosome proteins through ionic interactions. Interestingly, our experiments show that exosome subunits dRrp4, dRrp46, dCsl4, and dRrp47 lose interactions with dDis3 at lower magnesium concentrations than dRrp6, suggesting that the dDis3-dRrp6 interaction is more stable. This data supports previous findings that dDis3 and dRrp6 form a complex independent of the core exosome

(Graham et al., 2009a). It is possible that formation of this complex is affected by magnesium concentrations in vivo. No studies to date have determined how assembly of

exosome complexes is regulated in the cell.

Our data also shows that dDis3 interactions with core exosome subunits are sensitive

to the detergent TX-100. This suggests that not only ionic, but hydrophobic interactions

within the dDis3-exosome complex lend to its stability. As exosome complexes can

consist of associations between several different proteins, it is not surprising that various

types of interactions occur. Additional studies will better define the relationships between

Dis3 and other exosome proteins.

II.E.5 Conclusions

As this represents the first study of a Dis3 ribonuclease from a multicellular

organism, our observations highlight the conservation of activity among Dis3 homologs.

Our study provides the foundation for further investigation of Dis3 properties, including

domain function, regulation of activities, and in vivo substrate specificity. Further analysis will reveal general and conserved properties of Dis3 family members, all which are essential proteins involved in many aspects of RNA metabolism.

II.F Funding

This work was supported by grants GM072820 to E.D.A. and T32HD007104 to

M.M. from the National Institutes of Health.

Chapter III

Drosophila melanogaster Dis3 N-terminal domains are required for ribonuclease

activities, subcellular localization, and exosome interactions

Excerpts from:

Mamolen, M., Smith, A., and Andrulis, E.D. (2010) Nuc Acids Res in press

Mamolen, M., and Andrulis, E.D. (2009) Biochem and Biophys Res Commun 390, 529-

534

Turk, E., Mamolen, M., Graham, A., Smith, S., Kiss, D., and Andrulis, E.D. (2010) in

preparation

III.A Abstract

The conserved RNase, Dis3, plays important roles in several RNA metabolic

pathways. Despite much progress in understanding general characteristics of the Dis3

enzyme in vitro and in vivo, much less is known about the contributions of Dis3 domains

to its activities, subcellular localization, and protein-protein interactions. To address this,

we constructed a set of Drosophila Dis3 mutants and assessed their enzymatic activities

in vitro and their localizations and interactions in S2 tissue culture cells. We show that

the dDis3 N-terminus is sufficient for endoRNase activity in vitro. Additional mutational

analyses showed that dDis3 contains a second independent C-terminal active site, which

requires proper N-terminal domain structure to function. We find that the dDis3 N-

terminus also contributes to its subcellular distribution, and specifically contains a

sequence that directs dDis3 to mitochondria. dDis3 N-terminal domains are necessary

and sufficient for interactions with core exosome proteins as well. Finally, we found that

dDis3 interaction with dRrp6 and dImportin-α3 is independent of core interactions and

occurs through two different regions. Taken together, our data suggest that the dDis3 N-

terminus is a dynamic and complex set of domains that orchestrate RNA metabolic

functions and exosome interactions.

III.B Introduction

Dis3 is a highly conserved RNase, with significant structural and functional

similarities to its bacterial homologs, E. coli RNase II and RNase R (Zuo and Deutscher,

2001). All Dis3 homologs belong to the RNR family of enzymes, which are proteins that share a C-terminal RNB domain (Figure 24A). This domain houses the 3’Æ5’ exoRNase active site of RNase II/R and S. cerevisiae Dis3 (Amblar et al., 2005; Dziembowski et al.,

2007; Schneider et al., 2007); its function has not been confirmed in other organisms. In addition to the RNB domain, eukaryotic and bacterial Dis3 enzymes harbor two N- terminal oligonucleotide binding (OB) fold domains (also known as cold-shock domains), and a C-terminal S1 RNA binding domain (Figure 24A). Most Dis3 homologs also have C-terminal extensions that contain additional, uncharacterized variant sequences, and in the Drosophila homolog, a nuclear localization signal (NLS, (Graham

et al., 2009a)). The major structural difference between RNase II/R and Dis3 is an ~300

amino acid N-terminal extension that contains multiple bioinformatically identified

domains ((Zuo and Deutscher, 2001), Figure 24A). These include a conserved set of three

cysteine residues that resemble an iron-sulfur cluster motif (referred to as C3, (Graham et

al., 2009a; Schaeffer et al., 2009)), a PIN endoRNase domain (Lebreton et al., 2008;

Schaeffer et al., 2009; Schneider et al., 2009), and, in the Drosophila protein, a motif

with homology to the cohesin protein STAG (Graham et al., 2009a). Few of these

domains have been unequivocally demonstrated to have functional relevance to Dis3

RNase activity, localization, or interactions.

As mentioned previously, analyses regarding Dis3 structure and function have

been primarily done in yeast cells and/or using recombinant yeast polypeptides. Thus, it

is not known if the majority of observed domain functions are conserved in multicellular

eukaryotes. Studies of S. cerevisiae Dis3 have revealed that mutations to the N-terminal

C3 domain impede cell growth, but for unknown reasons (Schaeffer et al., 2009).

Mutations to conserved residues in the PIN or RNB domains result in loss of endoRNase

or exoRNase activities, respectively (Dziembowski et al., 2007; Schneider et al., 2007;

Lebreton et al., 2008; Schaeffer et al., 2009; Schneider et al., 2009). These analyses

indicate the PIN and RNB domains of S. cerevisiae Dis3 contain RNase active sites,

although it is not known how other domains in the protein contribute to these activities.

Finally, N-terminal domains appear to be important for localization in Drosophila Dis3

(Graham et al., 2009a), but the relationship of the Dis3 N-terminus to its subcellular

distribution is largely unknown.

Although Dis3 has been shown to function independently in vitro, it was initially

co-purified in a group of exoRNases called the exosome, as mentioned previously

(Mitchell et al., 1997). Exosome proteins are thought to assemble into a core complex,

and function in both the nucleus and cytoplasm (reviewed in Schmid and Jensen, 2008).

In S. cerevisiae, the N-terminal PIN domain of Dis3 is responsible for interactions with

these proteins (Schneider et al., 2007). N-terminal domains also appear to be important

for Drosophila Dis3 interactions with the exosome (Graham et al., 2009a), although the

exact interacting domain has not been identified.

The exosome core associates with two additional proteins, Rrp6 and Rrp47.

However, these proteins also function independently of the exosome core as well

(Mitchell et al., 2003; Callahan and Butler, 2008; Graham et al., 2009b; Callahan and

Butler, 2010). In this regard, there is significant biochemical evidence that exosome

proteins assemble into multiple, distinct complexes, several of which lack many of the

“traditional” subunits, including Dis3 (Chen et al., 2001; Estevez et al., 2003). As

mentioned before, Dis3 itself is found in complexes independent of the exosome

(Noguchi et al, 1996; Shiomi et al., 1998; Gavin et al., 2002; Gavin et al., 2006).

Additionally, proteomic analyses in S. cerevisiae have suggested that Dis3 localizes to

the mitochondria, but the rest of the exosome proteins do not (Sickmann et al., 2003;

Prokisch et al., 2004; Reinders et al., 2006; Vogtle et al., 2009). Although this putative

mitochondrial localization has not been confirmed in any organism, this data also

suggests Dis3 functions independently of the exosome. The exact number of Dis3

complexes, the site(s) of their assembly and disassembly, and the identification of

domains of Dis3 that mediate specific protein-protein interactions remain largely

unknown.

In this work, we focus on the contributions of Drosophila Dis3 N-terminal

domains to functions in vitro and in vivo. First, we examine if the N-terminal endoRNase

activity reported for S. cerevisiae Dis3 is conserved in the Drosophila Dis3 enzyme. We

also use truncation and point mutants to explore the contributions of the dDis3 N-

terminus to its subcellular distribution, and interactions with core exosome subunits,

dRrp6, dRrp47, and dImportin-α3. This study reveals novel features of the dDis3 N-

terminus that are functionally relevant and hence of general and broad importance to

exosome-mediated RNA metabolic pathways and mechanisms.

III.C Materials and Methods

III.C.1 Molecular cloning

All MBP- and FLAG-tagged dDis3 constructs for in vitro and cell based analyses were

made as detailed in Chapter II. Mutant dDis3 genes were PCR amplified from DNA

templates as described previously, using the primers listed in Table 1, Appendix A

(Andrulis et al., 2002; Graham et al., 2006). Constructs utilized in the mitochondrial

studies, Mtnp-dDis31-35-GFP, Mtnp-MTS4A-dDis31-35-GFP, and Mtnp-myc-dDis31-35-

GFP, were also constructed by standard molecular cloning. Briefly, dDis3 DNA corresponding to the first 35 amino acids of dDis3 was PCR amplified from a full-length dDis3 open reading frame using oligonucleotides described in Table 2, Appendix A.

These PCR products were digested with BglII and BamHI and then cloned into pRM-

Ha3-GFP (Lis et al., 2000).

III.C.2 Purification of recombinant proteins

Wild-type and mutant MBP-dDis3 proteins were expressed as described before, except for internal domain deletion proteins (MBP-dDis3C3Δ, MBP-dDis3PINΔ, MBP-dDis3STAGΔ,

MBP-dDis3OB1Δ) which were induced for two hours at 37˚C. Purification was performed

as before with the following exceptions. Cells harboring MBP-dDis31-394 were further

lysed in 1% triton. For internal domain deletion proteins and some full-length protein

preparations, lysis, wash, and dialysis buffers contained 1 mM EDTA.

III.C.3 Preparation of RNA substrates

5’ end-labeled RNAs were generated as before. Circular RNA substrates were created by

incubating 5’ end-labeled RNAs with T4 RNA ligase (New England BioLabs) for 15

minutes at 37˚C. Ligation was terminated by boiling the reactions for two minutes. To

ensure ligated RNAs were actually circularized, RNAs were incubated with calf intestine

alkaline phosphatase (Ambion), Exonuclease T (New England BioLabs), and Xrn1 (New

England BioLabs) in separate reactions per manufacturer’s recommendations. Reaction products were visualized as described before.

For 3’ end-labeling, 3’CMP was first incubated with γ32ATP (Perkin-Elmer) and T4

polynucleotide kinase (New England BioLabs) to generate radiolabeled pCp. RNAs and

[32P] pCp were then incubated with T4 RNA ligase (New England BioLabs) for 5 hours

at room temperature to generate 3’ end-labeled RNAs. Unincorporated nucleotides were

removed by NucAwayTM spin columns (Ambion).

III.C.4 Ribonuclease activity assays

General RNase activity and endonuclease assays were developed based on previously

published protocols (Dziembowski et al., 2007; Lebreton et al., 2008; Schaeffer et al.,

2009). Experiments were performed as before, except all reaction buffers contained 10

mM tris-HCl, pH 8.0, 75 mM KCl, 1 mM 2me, and 40 μM MgCl2 or MnCl2

(endonuclease assays). RNA concentrations were 120 nM for all experiments, except those in Figure 32, where RNAs were at 20 nM. Protein concentrations were as follows:

MBP-dDis31-982 and MBP-dDis31-394 were 60 nM in all experiments. MBP-dDis3C3Δ,

MBP-dDis3PINΔ, MBP-dDis3STAGΔ, and MBP-dDis3OB1Δ were 5 nM. MBP-dDis3189-982

was 60 nM, MBP-dDis362-982 was 43 nM, and MBP-dDis329-982 was 42 nM. The

concentration of MBP matched that of the highest concentrated MBP-dDis3 protein in

each experiment.

III.C.5 Quantification of RNase activity

Activity was quantified as detailed in Chapter II. For assays in which both circular and

linear RNAs were present, each form of the RNA was quantified separately. For example,

% circular polyA remaining was calculated as the ratio of circular polyA at a particular

time point, to circular polyA present at time zero for the same reaction.

III.C.6 Cell culture

Drosophila melanogaster S2 cell culture, including transient transfections and stable cell

lines were established and maintained as previously described (Graham et al., 2006;

Graham et al., 2009a; Graham et al., 2009b).

III.C.7 Immunofluorescence, immunoprecipitation, and western blotting

All immunofluorescence, immunoprecipitations, and western blotting with anti-FLAG,

and anti-dDis3, -dRrp6, -dImportin-α3, and exosome antibodies were performed as

described previously (Graham et al., 2009a). IP experiments contained RNase A and/or

ethidium bromide to confirm that interactions were independent of nucleic acids. For

immunofluorescence experiments where mitochondria were visualized, S2 cells were

incubated with 250 nM MitoTracker (Invitrogen) for 2 hours at 27˚C prior to staining.

III.C.8 Cell fractionation and isolation of mitochondria

Mitochondria were isolated from Drosophila S2 cells using previously published protocols (Morrow et al., 2000) or a commercial kit (Qiagen). For both methods, cells were grown to 80-100% confluence in 100 mm2 petri dishes in Hyclone CCM3TM media

(Hyclone) at 27˚C. For crude fractionations (Figure 38A-B), cells were washed twice

with 1 mL of wash buffer (10 mM tris-HCl, pH 7.5, 140 mM NaCl). Cells were lysed by

hypotonic shock in low-salt buffer (10 mM tris-HCl, pH 6.7, 10 mM KCl, 0.15 mM

MgCl2, 1x protease inhibitor cocktail (Roche)) for 5 minutes on ice. Sucrose was added to the lysates, and the mixture was centrifuged at 1000 x g for 10 minutes at 4˚C.

Mitochondria-containing supernatants were centrifuged at 8100 x g for 10 minutes at

4˚C. Cytoplasmic supernatants were removed and reserved, and pelleted mitochondria were washed twice with 200 uL low-salt buffer. Untreated supernatant and pellet fractions were prepared for western blotting. For NP-40 treatment, isolated mitochondria were lysed in buffer (300 mM NaCl, 10 mM CaCl2, 100 mM tris-HCl, pH 8.5, 0.5% NP-

40, 1 mM PMSF) for 1 hour on ice. Lysates were centrifuged at 14000 rpm for 5 minutes at 4˚C to obtain supernatant and pellet fractions analyzed by western blotting. For RNase

A treatment, isolated mitochondria fractions were incubated in low-salt buffer containing

100 μg/μL RNase A for 1 hour at 37˚C. Finally, for SDS and trypsin analysis, isolated mitochondria were incubated in buffer (0.35 M sucrose, 100 mM EDTA, 10 mM tris-

HCl, pH 7.5), with or without 0.1% SDS and/or 50 μg/μL trypsin, for 1 hour on ice. For

fractionation using a commercial kit (Qiagen; Figure 40C), S2 cells were washed with

0.9% NaCl solution, and mitochondria was isolated per manufacturer’s

recommendations. All fractions were examined using standard western blotting

100

techniques. dDis3 was detected using an antibody previously described (Graham et al.,

2006). Commercially available antibodies were used to detect α-Tubulin (Sigma) and

Superoxide Dismutase (Abcam).

III.D Results

III.D.1 MBP-dDis3 is active on linear RNAs of varying sequences

We have shown before that MBP-dDis3 cleaves RNAs regardless of sequence

(Chapter II). To establish a baseline for wild-type activity before testing N-terminal

mutants, we utilized the same substrates in time course assays. Full-length MBP-dDis3

(MBP-dDis31-982) completely degraded polyA, polyC, polyU, and polyN within 10 minutes (Figure 24B). Interestingly, all of the substrates were cleaved with approximately the same efficiency (Figure 24C). As a control, MBP itself did not degrade any RNAs in any assay. This data depicts the robust activity of MBP-dDis31-982 on linear substrates of

varying sequences.

III.D.2 The dDis3 N-terminus harbors an endoribonuclease activity

To assess the contributions of N-terminal domains to dDis3 RNase activity, we purified N-terminal mutants as MBP-dDis3 fusions (Figure 25), and tested their activities in vitro. We first examined a set of proteins in which N-terminal domains were individually deleted. These included MBP-dDis3C3Δ, MBP-dDis3PINΔ, MBP-dDis3STAGΔ,

and MBP-dDis3OB1Δ (Figure 26A). As shown in Figure 26B, every N-terminal internal

domain deletion mutant was inactive. Further, this effect was not substrate-specific. This

101

102

Figure 24. Full-length MBP-dDis3 efficiently degrades multiple RNA substrates

(A) Schematic of full-length Drosophila melanogaster Dis3. (B) MBP-dDis31-982 degrades 5’ end-labeled RNA substrates. Composition of the RNAs is depicted on the left side of each gel ((*) represents the position of the 32P label); this notation is used

hereafter. The smallest reaction product is marked on the right side of each gel. (C) MBP- dDis3 degrades compositionally distinct RNAs at similar rates. MBP (dashed line, ●) and

MBP-dDis31-982 () activity was averaged from two independent experiments for each

substrate and graphed.

103

Figure 25. Recombinant mutant MBP-dDis3 proteins used in in vitro RNase assays ~200 ng of full-length proteins were loaded onto the gel. Molecular weight standards

(labeled on the left side of each gel) were obtained by running prestained protein marker on each gel (not shown here).

104

105

Figure 26. N-terminal domains are necessary for full-length MBP-dDis3 in vitro RNase activity

(A) Schematics of N-terminal internal domain deletion mutant polypeptides. (B) MBP-dDis3 mutant proteins lacking

specific N-terminal domains are inactive. As a comparison, full-length MBP-dDis3 at the same concentration (5 nM) is

active (Figure 12, data not shown).

data suggests that N-terminal domains are important for maintaining the RNase activity

of full-length dDis3. This result was puzzling at first because all of these mutants retain

the RNB domain, which is suspected to mediate the 3’Æ5’ exoRNase activity of Dis3

proteins (Dziembowski et al., 2007; Schneider et al., 2007). We reasoned that the mutant

N-termini could be mis-folding in such a manner that they inactivated or blocked the

RNB catalytic sites in the domain deletion mutants. To test this idea, we truncated dDis3

at the N-terminus and observed activity. These mutants lacked the first 28, 61, or 188

amino acids. As depicted in Figure 27A, the missing residues correspond to an

uncharacterized N-terminal region, the C3 domain, and the PIN domain. Interestingly,

MBP-dDis329-982, MBP-dDis362-982, and MBP-dDis3189-982 completely degraded polyU

RNA within 60 minutes (Figure 27B). Two conclusions can be drawn from this data.

First, it appears that the dDis3 C-terminus itself is active, which is consistent with reports

of an independent RNB catalytic site (Lorentzen et al., 2008). Second, N-terminal

domains may interfere with this active site when they are not properly ordered and/or

mis-folded. Hence why N-terminal internal domain deletion mutants are inactive, yet N-

terminally truncated mutants retain activity.

Although our data suggests that N-terminal domains may be structurally significant to Dis3 RNase activity overall, recent studies have reported that the N- terminus of Dis3 is also important for enzymatic function because it contains an

additional endoribonucleolytic active site (Clissold and Pontig, 2000; Lebreton et al.,

2008; Schaeffer et al., 2009; Schneider et al., 2009). To determine if the dDis3 N-

terminus has activity, we first made single mutations to conserved residues in the PIN

106

Figure 27. N-terminally truncated MBP-dDis3 mutants retain RNase activity

(A) Schematics of N-terminal truncation mutants. (B) The RNase activity of N-terminally truncated mutants is approximately equivalent to full-length MBP-dDis3 RNase activity.

Enzymatic activity was assessed on a polyU substrate.

107

domain (Figure 28A). These mutants, MBP-dDis3D67K and MBP-dDis3D183L, degraded all

RNAs tested (Figure 28B-C), but their activities were significantly less efficient than wild-type (Figure 28D). For example, MBP-dDis31-982 degradation of polyA resulted in

less than 40% RNA remaining at 2 minutes, whereas MBP-dDis3D67K and MBP- dDis3D183L degradation left approximately 90% and 65% poly A RNA remaining at the

same time point, respectively (Figure 28D). MBP displayed negligible activity in these

assays (Figure 28, Figure 29). This reduction in dDis3 RNase activity suggested the

presence of an additional catalytic site that is directly perturbed by point mutations.

However, it was difficult to definitively conclude from this experiment that we had

discovered an N-terminal active site. An alternative explanation for our results is that the

N-terminal point mutations somehow altered the RNB active site, thereby reducing the

activity of dDis3.

To directly test if the dDis3 N-terminus alone harbors any RNase activity, we

purified a recombinant polypeptide in which MBP is fused to the first 394 amino acids of

dDis3, and assessed its ability to degrade various single-stranded RNAs. As depicted in

Figure 30A, this mutant lacked the entire RNB domain as well as additional C-terminal

sequences. MBP-dDis31-394 cleaved all RNAs tested, albeit much slower than full-length

MBP-dDis3 (Figure 30B-C). These assays show that the dDis3 N-terminus contains an

independent RNase activity. Our results are also consistent with previous studies showing

that an N-terminal fragment of S. cerevisiae Dis3 is active (Lebreton et al., 2008;

Schaeffer et al., 2009).

To determine the type of RNase activity harbored by the dDis3 N-terminus, we

engineered substrates for endonucleolytic cleavage: circular polyA, polyC, polyU, and

108

109

Figure 28. Point mutations to the PIN domain affect MBP-dDis3 RNase activity

(A) Schematic of MBP-dDis3 point mutations. Arrows point to relative locations of point mutations. (B),(C) MBP-dDis3 point mutants retain some ability to degrade 5’ end- labeled RNAs. (D) PIN domain mutant MBP-dDis3 polypeptides have reduced RNase activities compared to wild-type. Reactions were graphed as follows: MBP ●; MBP- dDis31-982 ; MBP-dDis3D67K ¡; MBP-dDis3D183L . For ease of graphing MBP control reactions, data was combined from both point mutant experiments and averaged for each substrate. Control data that is separated by experiment can be viewed in Figure 29. The

MBP-dDis31-982 line is graphed for comparison purposes; this data was originally shown in Figure 24. For point mutant lines, data was averaged from at least two independent experiments for each substrate.

110

Figure 29. MBP has little background RNase activity

Graph displays MBP control data for each point mutant experiment on every substrate.

For MBP-dDis3D67K control experiments, symbols are: MBP + polyA ●; MBP + polyC

S; MBP + polyU ¡; MBP + polyN ⁬. For MBP-dDis3D183L control experiments, symbols are: MBP + polyA ; MBP + polyC S; MBP + polyU c; MBP + polyN.

111

112

Figure 30. N-terminal domains are sufficient for MBP-dDis3 in vitro RNase activity

(A) Schematic of C-terminally truncated MBP-dDis3 mutant. (B) The N-terminus of

MBP-dDis3 alone is ribonucleolytically active. (C) MBP-dDis31-394 is less active than

full-length MBP-dDis3. Experiments were graphed as follows: MBP ●; MBP-dDis31-982

; MBP-dDis31-394 S. Data was averaged from at least two independent experiments for

each substrate.

113

polyN RNAs. As a control, RNAs marked as circles were not able to be

dephosphorylated by calf intestine alkaline phosphatase, nor cleaved by the

3’Æ5’exoribonuclease RNase T, or by the 5’Æ3’ exoribonuclease Xrn1 (Figure 31). In

contrast, MBP-dDis31-394 cleaved each of the circular substrates (Figure 32A), suggesting

that the dDis3 N-terminal activity is endoribonucleolytic. Similar to S. cerevisiae Dis3

(Lebreton et al., 2008; Schaeffer et al., 2009), the ability to cleave circular substrates was

retained in full-length dDis3 (Figure 32B). Quantification of the assays is displayed in

Figure 33. Quantification of MBP control reactions for this set of assays is displayed in

Figure 34. For clarity, degradation of different RNA types is presented as separate lines

for each protein. As shown by the graphical analyses, full-length MBP-dDis3 is more

active than the N-terminus alone in every assay, when comparing the same RNA types.

For example, MBP-dDis31-982 degraded circular polyU to less than 20% remaining at 10

minutes, whereas there was ~70% circular polyU remaining at 10 minutes for MBP- dDis31-394. Comparison of all of the graphs shows that MBP-dDis31-982 and MBP-dDis31-

394 cleaved linear and circular polyA and polyC RNAs at approximately the same rate.

For polyU and polyN, however, both proteins cleaved linear RNA more efficiently.

Together, these results demonstrate that although degradation rates vary, the N-terminal

active site facilitates cleavage of circular RNAs regardless of sequence.

We next examined the activity of MBP-dDis3 proteins on RNAs that were

radioactively labeled at the 3’ end. We also sought to determine which activity, exo- or

endoribonucleolytic, full-length dDis3 uses to cleave linear RNAs in certain in vitro

environments. We used ionic conditions suggested to, but not confirmed to, promote

endoRNase activity (Lebreton et al., 2008; Schaeffer et al., 2009; Schneider et al., 2009).

114

Figure 31. Circularized RNA substrates are not cleaved by control enzymes

115

Figure 31. Circularized RNA substrates are not cleaved by control enzymes

RNAs were incubated with calf intestine alkaline phosphatase (CIAP) (A), the 3’Æ5’ exonuclease ExoT (B), or the 5’Æ3’ exonuclease Xrn1 (C). For each experiment, lanes 1-

2 represent degradation of a 5’ linear RNA by the control enzymes to ensure control enzymes functioned properly in the assay (positive control). Lanes 3-10 display cleavage assays with circularized RNAs. Note, two bands appear for the RNAs. One band is linear

RNA remaining from inefficient ligation reactions during preparation of the circular substrates (top band), the other is circularized RNA (bottom band). None of the control enzymes cleave the bottom circularized RNA bands.

116

Figure 32. The dDis3 N-terminus has endoribonuclease activity

MBP-dDis31-394 (A) and MBP-dDis31-982 (B) cleaved circularized RNA substrates. Full- length linear RNA is marked as (—) and RNA circles as (○). These symbols are used throughout. Data shown is representative of at least two independent experiments.

117

Figure 33. The N-terminus of MBP-dDis3 cleaves circular RNAs less efficiently than

the full-length protein

Quantification of the activity of MBP-dDis31-394 and MBP-dDis31-982 on linear and

circular polyA (A), polyC (B), polyU (C) and polyN (D) is displayed. This data

accompanies Figure 32 experiments. For all graphs, experiments are as follows: MBP-

dDis31-394 + linear RNA S; MBP-dDis31-394 + circular RNA ●; MBP-dDis31-982 + linear

RNA ; MBP-dDis31-982 + circular RNA c. % RNA remaining increased at the 60

minute time point in some assays due to overloading. Data was averaged from at least

two independent experiments for every substrate.

118

Figure 34. MBP has little background RNase activity on circularized RNA substrates

119

Figure 34. MBP has little background RNase activity on circularized RNA

substrates

MBP activity on all substrates utilized in Figure 32 experiments is graphed. Degradation of both the linear (A) and circular (B) forms of each RNA is presented. For MBP-dDis31-

394 control experiments, symbols are as follows for both graphs: MBP + polyA ; MBP

+ polyC T; MBP + polyU c; MBP + polyN Δ. For MBP-dDis31-982 control experiments,

symbols are as follows: MBP + polyA ●; MBP + polyC S; MBP + polyU ¡; MBP + polyN . Data was averaged from at least two independent experiments for every substrate.

120

Specifically, manganese, not magnesium, was added to the reaction buffers. MBP-dDis31-

394 cleavage of the 3’ end-labeled RNAs produced multiple RNA fragments, which is

again suggestive of an endoRNase activity (Figure 35A). For MBP-dDis31-982, we

anticipated we may or may not observe accumulation of product fragments. Rather, if the

full-length enzyme utilized its RNB domain-mediated 3’Æ5’ exoRNase activity to

degrade the 3’ end-labeled substrates, we expected accumulation of NMP products that

would not be visible on our gels. Interestingly, full-length dDis3 cleavage of these RNAs

resulted in accumulation of RNA fragments of various sizes (Figure 35B), suggesting that

full-length dDis3 does use endoRNase activity to cleave 3’ end-labeled RNAs in these

assay conditions. Graphical analyses show that MBP-dDis31-982 is more active on 3’ end-

labeled RNAs than MBP-dDis31-394, consistent with previous results (Figure 36). In

contrast to the assays in Figure 33, linear and circular polyA, polyC, and polyN RNAs 60

minutes, there was ~20% of linear polyU RNA remaining for MBP-dDis31-982, and twice

as much circular polyU remaining (40%). Likewise, there was ~20% linear and 60%

circular polyU remaining for MBP-dDis31-394 at 60 minutes. The graphs also show that

MBP-dDis31-982 degradation of all 3’ end-labeled RNAs plateaued after two minutes, yet the reactions did not go to completion. MBP again had little activity on any substrate

(Figure 37). This may suggest that protein concentration was limiting for these particular assays. Collectively, this data confirms that the N-terminus of dDis3 contains an independent active site that facilitates cleavage of 3’ end-labeled RNAs in such a manner

as to suggest endoRNase activity. Further, observations of polyU degradation suggest

dDis3 endoRNase activity may be more efficient at cleaving particular RNA structures.

121

Figure 35. MBP-dDis3 cleavage of 3’ end-labeled RNAs confirms endoribonuclease

activity

MBP-dDis31-394 (A) and MBP-dDis31-982 (B) cleave 3’ end-labeled polyA, polyC, polyU, and polyN RNA substrates. A portion of the RNAs used in these assays is circular as a result of the ligation reaction during the 3’ end-labeling procedure. Note that reaction products for both proteins are RNA fragments. Data shown is representative of at least two independent experiments.

122

Figure 36. Full-length MBP-dDis3 cleaves 3’ end-labeled RNAs more efficiently than the N-terminus alone

RNase activities of MBP-dDis31-394 and MBP-dDis31-982 on 3’ end-labeled linear and

circular polyA (A), polyC (B), polyU (C), and polyN (D) were quantified. For all graphs,

experiments are as follows: MBP-dDis31-394 + linear RNA S; MBP-dDis31-394 + circular

RNA ●; MBP-dDis31-982 + linear RNA ; MBP-dDis31-982 + circular RNA c. Data was

averaged from at least two independent experiments for every substrate.

123

Figure 37. MBP has little or no background activity on 3’ end-labeled RNAs

MBP control experiments from Figure 35 are graphed. For MBP-dDis31-394 control

experiments, symbols are as follows for both graphs: MBP + polyA ; MBP + polyC T;

MBP + polyU c; MBP + polyN Δ. For MBP-dDis31-982 control experiments, symbols are as follows: MBP + polyA ●; MBP + polyC S; MBP + polyU ¡; MBP + polyN .

Data was averaged from at least two independent experiments for every substrate.

124

III.D.3 dDis3 N-terminal domains are required for nuclear localization

Upon completion of our enzymatic studies, we next wanted to determine the

physiological relevance of dDis3 N-terminal domains. To accomplish this, we analyzed

the subcellular distribution and protein-protein interactions (section III.D.5) of multiple

FLAG-tagged dDis3 mutant proteins. For cytological analyses, cells were scored for

immunofluorescence staining in the nucleus, cytoplasm, or throughout the entire cell. We

also observed two minor types of staining that were classified as large structures or small

foci, both which appear in the cytoplasm (Figure 38A). As a control, we first examined

the subcellular distribution of a set of C-terminally truncated polypeptides. These

mutants, schematized in Figure 38B, were designed to remove the NLS as well as

potentially functional domains. Localization patterns of each mutant were quantified and

are presented as graphs (Figure 38C) and fluorescence images (Figure 39A). Consistent

with previous reports (Graham et al., 2009a), full-length dDis3 (dDis31-982) was

predominantly nuclear (>90%). Conversely, all C-terminally truncated dDis3 mutants

were mostly cytoplasmic (Figure 38C). The dDis31-188 mutant was unique in that it accumulated within small cytoplasmic foci in ~80% of cells. In the other 20%, dDis31-188 was found in large cytoplasmic structures. These data confirm that loss of the dDis3 C- terminal NLS ablates nuclear accumulation of the protein.

To expand upon this observation and improve our understanding of dDis3 domain contributions to nucleocytoplasmic localization, we constructed a set of dDis3 N-terminal truncations (Figure 38D). All of these mutants retain the C-terminal NLS and hence are predicted to be nuclear. We observed no significant difference in the localization of

125

Figure 38. The N-terminus of dDis3 contributes to its subcellular distribution in

Drosophila S2 cells

126

Figure 38. The N-terminus of dDis3 contributes to its subcellular distribution in

Drosophila S2 cells (A) Examples images of major localization patterns of dDis3 mutant polypeptides. (B) Schematic of dDis3 C-terminally truncated mutants utilized as controls in this study. (C) dDis3 mutants lacking an NLS are predominantly cytoplasmic. Graph represents quantification of the distributions of full-length and mutant dDis3 polypeptides in S2 cells. For simplicity, localization patterns were grouped into the five categories listed (the nuclear category represents staining in the nucleus alone; cytoplasmic is general staining throughout the cytoplasm or at the plasma membrane; entire cell is staining in the nucleus and cytoplasm; large structures is staining in distinct large structures alone or staining in large structures and diffuse cytoplasmic staining; small foci is staining in distinct foci alone, or staining in foci and diffuse cytoplasmic staining). This method was used for all additional graphs. The “normal” dDis3 distribution pattern is represented by the dDis31-982 bar, where full-length dDis3 is >90% nuclear. (D)

Schematic of dDis3 N-terminally truncated mutants. (E) N-terminally truncated dDis3 mutants, despite harboring an NLS, are predominantly cytoplasmic. (F) Schematic of dDis3 N-terminal point mutants. (G) Point mutations to the N-terminus of dDis3 perturb its normal subcellular distribution pattern. Data shown was collected and averaged from two (point mutants) or three (truncations) independent experiments where at least 88 expressing cells were counted. Error bars are excluded for clarity. Data in (A), (C), and

(E) courtesy of Alexandra Smith.

127

Figure 39. Images of mutant dDis3 localization patterns

128

Figure 39. Images of mutant dDis3 localization patterns

(A) C-terminally truncated proteins are predominantly cytoplasmic, as expected with

removal of the NLS. (B) N-terminally truncated proteins exhibit multiple localization patterns. Images are representative of three independent localization experiments, and

accompany data in Figure 38. Data courtesy of Alexandra Smith.

129

dDis329-982 as compared to full-length dDis3 (Figure 38E, 39B). In contrast, a polypeptide lacking the C3 domain (dDis362-982) was predominantly (60%) cytoplasmic; less than

10% of cells had nuclear staining. When the first 188 amino acids were removed

(dDis3189-982), we observed an unexpected reversal of localization, with 60% of the

protein being found in the nucleus. The remaining constructs (dDis3253-982, dDis3395-982,

and dDis3731-982) accumulated mainly in cytoplasmic foci and large structures. Thus, in

addition to the C-terminal NLS, N-terminal domains appear to contribute to dDis3

subcellular localization.

To further investigate dDis3 localization, we engineered point mutations in dDis3 N- terminal domains and assessed their effects on subcellular distribution patterns (38F-G).

We mutated two cysteine residues in the C3 domain (C31A, C36A), two active site

aspartates in the PIN domain (D67K, D183L), a conserved residue in the OB1 domain

(D291N), and we changed a residue to recapitulate the S. cerevisiae P463L mutation in

the OB2 domain, described previously ((Suzuki et al., 2001), P419L). Despite the

presence of the NLS in all of these proteins, both the dDis3C31A and dDis3C36A mutants

were only ~20-30% nuclear (Figure 38G). Disruption of the first active site residue in the

PIN domain, D67K, elicited a complete loss of nuclear localization. In contrast, nuclear

accumulation of dDis3D183L (~60%) and dDis3D291N (~80%) mutants was comparable to

wild-type levels (Figure 38G). Finally, we observed complete loss of nuclear staining

with the dDis3P419L mutant (Figure 38G). These data show that specific changes to

individual amino acids in the dDis3 N-terminus render the polypeptide incapable of

maintaining its normal nucleocytoplasmic distribution ratio. Given the complex

localization patterns of these mutants, it is possible that multiple, distinct N-terminal

130

regions of dDis3 help regulate the nuclear targeting and/or retention of this enzyme in the cell.

III.D.4 dDis3 localizes to mitochondria via an N-terminal mitochondria targeting sequence

Since our observations above suggested the N-terminus may participate in coordination of dDis3 localization, we completed bioinformatic analyses to identify putative signaling sequences in that region. From these analyses, we learned that the dDis3 N-terminus is predicted to contain a MTS, and additional signals that are discussed in Chapter IV. We performed several experiments to determine if dDis3 localizes to mitochondria. First, we fractionated Drosophila S2 cells, and used western blotting to see if endogenous dDis3 is contained within a membranous subcellular compartment. In untreated cells, dDis3 was present in both the cytoplasmic supernatant (S) and pelleted organellar (P) fractions (Figure 40A, lanes 3-4). When these fractions were treated with detergent to break apart or otherwise disrupt organelles, dDis3 was found only in the supernatant fraction, (Figure 40A, lanes 5-6, 11-12). Next, we treated samples with trypsin, a proteolytic agent, and saw that dDis3 remained intact in organellar fractions, whereas there was no protein in supernatant fractions (Figure 40A, lanes 7-8). This suggests dDis3 was inaccessible to digestion while in the pellet. Treatment with a combination of detergent and protease resulted in complete loss of dDis3 in both fractions (Figure 40A, lanes 9-10). Finally, when samples were treated with RNase A, dDis3 remained intact in the organellar pellet, suggesting the protein does not associate with, or is not directed to, organelles via RNA binding (Figure 40A, lanes 13-14).

131

Figure 40. dDis3 fractionates with mitochondria from Drosophila S2 cells (A) dDis3 proteins contained within organellar fractions are inaccessible to protease treatment. dDis3 proteins in cytoplasmic (S) or pelleted microsomal fractions (P) were detected by western blotting (lanes 3-4). Some fractions were treated with SDS, NP-40 detergent, RNase A, and/or trypsin (lanes 5-14). Whole cell extract (WCE) and dDis3-

FLAG are also shown (lanes 1-2). (B) dDis3 co-fractionates with the mitochondrial protein Superoxide Dismutase. Tubulin was used as a cytoplasmic control. (C) dDis3 is present in S2 cell mitochondrial fractions. S2 cells were separated into nuclear (N), cytoplasmic (C), microsomal (μ), and mitochondrial (M) components. As a control, no proteins were detected in buffer used to wash isolated mitochondria (lanes 4-6). Images shown are representative of at least three independent experiments. Data in (A) courtesy of Erik Andrulis.

132

Together, these data suggest that dDis3 is contained within a membrane-bound compartment in S2 cells.

We performed western blotting with additional antibodies to identify the specific organellar location of dDis3. Our analysis shows that dDis3 co-localizes with superoxide dismutase (SOD), a mitochondrial protein, in pelleted organellar fractions (Figure 40B, lane 2). As a control, the cytoplasmic protein α-tubulin was not present in these fractions

(Figure 40B, lane 2). In agreement with these results, dDis3 was also found in a pure mitochondrial fraction isolated from S2 cells (Figure 40C, lane 7). As additional confirmation of dDis3 mitochondrial localization, we performed indirect immunofluorescence experiments. dDis3-FLAG was overexpressed in Drosophila S2 and

Kc cells to show localization is not cell type specific. As shown in Figure 41A, dDis3-

FLAG staining overlapped with MitoTracker, a mitochondrial marker, in both cell types.

However, only a small fraction of dDis3 proteins was found in mitochondria in these experiments. To ensure that this observation wasn’t an effect of protein overexpression, we isolated mitochondria from S2 cells and examined endogenous dDis3 expression. dDis3 was present in the mitochondria, and localization coincided with MitoTracker staining, although co- localization was not 100% (Figure 41B).

Finally, we combined mutational and immunofluorescence analysis to determine if the dDis3 MTS detected by our bioinformatics is functional. This sequence is displayed in Figure 42A. We fused the first 35 amino acids of dDis3 to GFP and expressed this construct in S2 cells (Figure 42B). The N-terminal dDis3 sequence, dDis31-35-GFP, showed a high level of co-localization with MitoTracker (Figure 42C). In contrast, GFP alone was mainly in the nucleus (Figure 42C). Mutations to the putative MTS (red amino

133

Figure 41. dDis3 localizes to mitochondria in Drosophila S2 cells

(A) dDis3-FLAG mitochondrial localization is not cell-type specific. Cells were pre-

treated with Mitotracker to mark mitochondria. (B) Endogenous dDis3 is located within

isolated Drosophila mitochondria. Immunofluoresence analyses were performed on

mitochondrial fractions separated from S2 cells. Data in (A) courtesy of Amy Graham.

134

Figure 42. The N-terminus of dDis3 is sufficient for mitochondrial targeting

(A) Schematic of a putative dDis3 mitochondria targeting sequence. (B) dDis3 N-

terminal mutants are detected by Western analysis. The extreme N-terminus of dDis3,

containing a putative sequence mitochondria targeting signal (MTS; amino acids 1-35),

was overexpressed in S2 cells. A mutant version of this sequence (MTS4A) was also

expressed. Mutated residues are shown in red in (A); residues were changed to alanines.

Data courtesy of Amy Graham and Erik Andrulis.

135

acids in 42A were changed to alanines) resulted in a loss of dDis31-35-GFP co-localization with MitoTracker. Together, these results suggest that dDis3 does localize to mitochondria, and its N-terminus is responsible for targeting the protein there.

III.D.5 dDis3 N-terminal domains are required for interactions with core exosome proteins and exosome co-factors

To ascertain the contributions of dDis3 domains to its protein-protein interactions, we used FLAG immunoprecipitation to recover full-length and mutant dDis3 polypeptides from whole cell extracts (Figure 43). We specifically examined the requirements for dDis3 interactions with core exosome proteins, exosome co-factors, and the nuclear import protein dImportin-α3, all proteins previously shown to interact with dDis3 (Graham et al., 2009a). The full-length dDis3 protein (dDis31-982) served as a positive control, and was able to co-precipitate core exosome proteins dRrp42, dRrp45, dRrp41, dRrp4, dRrp46, dCsl4, and dRrp40. dDis31-982 also co-precipitated the nuclear exosome cofactor dRrp47, as well as dRrp6 and dImportin-α3 (Figure 43A, lane 14;

Figure 43B, lane 9). The negative control, beads incubated with extract from vector- harboring S2 cells, showed little or no background binding using any of these proteins

(Figure 43A, lane 26; Figure 43B, lane 16). This confirms prior observations that Dis3 not only interacts with core exosome proteins and exosome cofactors, but also interacts with proteins involved in nucleocytoplasmic transport (Mitchell et al., 1997; Noguchi et al, 1996; Shiomi et al., 1998; Graham et al., 2009a).

136

Figure 43. The dDis3 N-terminus is required for interactions with core exosome proteins

(A) dDis3 amino acids 1-252 contain the core exosome interacting region. Western blot analysis of dDis3 co-immunoprecipitation experiments are presented. Input, 2.5%; immunoprecipitate (IP), 5%. Asterisks represent the heavy and light chains from the α-

FLAG resin used to immunoprecipitate dDis3-FLAG constructs. Note, amino acids 1-188 retain interactions with dRrp47, dRrp6, and dImportin-α3, but not core proteins. (B) The dDis3 C-terminus interacts with dRrp6 and dImportin-α3, independently of core exosome proteins. Here, dDis3 amino acids 731-982 only co-precipitate dRrp6 and dImportin-α3.

Schematics of all truncated polypeptides are shown in Figure 38. Data courtesy of Erik

Andrulis.

137

The majority of dDis3 C-terminally truncated polypeptides (dDis31-928, dDis31-855,

dDis31-798, dDis31-757, dDis31-732, dDis31-631, dDis31-479, dDis31-394, dDis31-319, and dDis31-

252) also co-precipitated core exosome proteins, dRrp6, and dImportin-α3 (Figure 43A,

lanes 15-24). However, the 1-188 fragment of dDis3 showed reduced binding to dRrp40,

was severely compromised and/or deficient in co-precipitating dRrp45, dRrp41, dRrp4,

dRrp46, and dCsl4, and did not co-precipitate dRrp42 (Figure 43A, lane 25). Despite this

loss or reduction in core exosome binding, dDis31-188 still co-immunoprecipitated with

dRrp6 and dRrp47. dImportin-α3 binding was observed as well, but modestly reduced.

None of the C-terminally truncated mutants co-precipitated endogenous dDis3 (Figure

44A). Based on these observations, the first 252 amino acids of the dDis3 N-terminus interacts with the exosome core, and the first 188 amino acids interact with dRrp6, dRrp47, and dImportin-α3.

Co-immunoprecipitation experiments with N-terminally truncated dDis3 mutants revealed that removal of the first 28 amino acids alone was sufficient to ablate dDis3 interaction with core subunits and dRrp47 (Figure 43B, lane 10). All additional N- terminal truncations (dDis362-982, dDis3189-982, dDis3253-982, dDis3395-982, and dDis3731-982)

were unable to co-precipitate these proteins above background binding, confirming that the dDis3 N-terminus is required for these interactions (Figure 43B, cf lanes 11-15 with lane 16). These mutants also did not recover endogenous dDis3 (Figure 44B).

By comparison, all N-terminally truncated dDis3 polypeptides recovered dRrp6 and dImportin-α3, consistent with previous data supporting an interaction between the

dDis3 C-terminus and these proteins (Graham et al., 2009a). dDis3253-982 recovered

qualitatively less dRrp6 and dImportin-α3, likely a consequence of its low level of

138

Figure 44. N- and C-terminally truncated dDis3 mutant polypeptides fail to immunoprecipitate endogenous Dis3

Western blots of (A) C-terminally truncated dDis3 proteins and (B) N-terminally truncated dDis3 proteins that were detected by α-dDis3 antibody. dDis3395-982 and dDis3731-982 are not detected by α-dDis3 antibody as the epitope for interaction is in the dDis3 N-terminus. Westerns accompany the immunoprecipitation experiments in Figure

43. Data courtesy of Erik Andrulis.

139

expression (Figure 43B, FLAG panel, lanes 5 and 13). Thus, our data show that dRrp6 and dImportin-3 interactions with dDis3 are independent, as these proteins, but not core exosome proteins, bind to the dDis3 C-terminus.

III.E Discussion

In this work, we have performed a structure-function analysis of Drosophila melanogaster Dis3. We show that N-terminal protein domains contribute to dDis3 enzymatic activities, subcellular compartmentalization, and interactions with the exosome core, dRrp6, and dImportin-α3. Thus, this study confirms and extends upon our initial findings regarding dDis3, as well as previous observations of S. cerevisiae Dis3 activity.

These analyses continue to build a framework for understanding the conserved roles of

Dis3 in RNA metabolism. Further, as Dis3 is a conserved endo- and exoribonuclease, these studies help us gain a better understanding of how RNases function in general.

III.E.1 Dis3 N-terminal endoribonuclease activity is conserved in metazoans

We show for the first time that the N-terminus of Drosophila melanogaster Dis3 has an RNase activity. This is the second independent active site that we have uncovered, as the dDis3 C-terminus alone was also found to have activity in this study. Since the dDis3 N-terminus can cleave circular RNA substrates, we conclude that this activity is endoribonucleolytic. However, the N-terminus is less efficient at cleaving RNAs than the full-length enzyme. This is in contrast to studies of S. cerevisiae Dis3, as its N-terminus alone has very robust endoRNase activity (Lebreton et al., 2008; Schaeffer et al., 2009).

The difference may lie in the composition of the N-terminal fragment we use here compared to those used in the yeast studies. Where our construct includes the C3, PIN,

140

and OB1 domains, the yeast construct lacks OB1. Interestingly, it has been suggested that

the OB1 domain may regulate PIN activity (Schaeffer et al., 2009). This is consistent

with our observations of MBP-dDis31-394; OB1 may reduce endoRNase activity.

In our in vitro study, we also found that dDis3 cleaved all circular RNAs tested.

Thus dDis3 endoRNase activity, like its exoRNase activity (Chapter II), is not sequence-

specific in vitro. The majority of RNAs, whether linear or circular, were cleaved by

dDis3 at approximately the same rates. polyU was an exception. This was most apparent

in experiments with 3’ end-labeled substrates. The linear version of polyU was cleaved

much more efficiently than 3’ circular polyU by both MBP-dDis31-394 and MBP-dDis31-

982. This could be directly related to RNA structure; perhaps it is easier for dDis3 to bind

to and degrade single-stranded polyU than a circle of polyU. It is also unlikely that dDis3

would naturally encounter a circular polyU in the cell, thus it’s possible that dDis3 does

not efficiently recognize it as a legitimate substrate.

Because we observe these variations, it will be important to determine the actual

targets of dDis3 activities in vivo. To date, S. cerevisiae Dis3 endoRNase activity has

only been linked to rRNA processing (Lebreton et al., 2008; Schaeffer et al., 2009;

Schneider et al., 2009); other functions have not been tested for any Dis3 homolog.

However, PIN endoRNase activities are not specific to rRNA processing pathways in

multicellular eukaryotes. Within recent years, the PIN domains of fly and human SMG6

have been linked to NMD (Glavan et al, 2006; Eberle et al., 2009). Thus, it is possible

that PIN-mediated endoRNase activity is a conserved mechanism utilized in the turnover

and processing of different classes of RNAs. Future analyses of Dis3 enzymatic activity

in vivo will likely uncover additional RNA metabolic functions.

141

Our in vitro analyses have shown that the dDis3 N-terminus not only possesses

endoRNase activity, but N-terminal domains are important for exoRNase activity in the

full-length enzyme. Deletion of any N-terminal domain resulted in a complete loss of

activity of full-length dDis3. It is possible that N-terminal domains are needed to

maintain the stability of the protein, and loss of these domains results in an unstable,

inactive protein that would be rapidly turned over in the cell. However, this is unlikely, as proteins lacking these domains, when overexpressed in Drosophila S2 cells, are visible

by western blotting (Figure 43), and hence are stable in vivo. Further, these particular domain deletion mutants still bind dRrp6 (Graham et al., 2009a), so they have some normal protein-protein interactions. A more plausible explanation is that N-terminal domains are necessary structural elements that maintain the ribonucleolytically active conformation of the wild-type protein. When these domains are deleted, remaining N- terminal sequences could take on a dominant effect on the RNase active sites, causing them to mis-fold or be blocked to substrate entry. Consistent with this, we observed that recombinant MBP-dDis3 mutants truncated at the N-terminus (mutants 29-982, 62-982, and 189-982) retain RNB-mediated exoRNase activity. Similarly, S. cerevisiae Dis3

mutants lacking the first 241 amino acids are active in vitro (Lorentzen et al., 2008).

Thus, N-terminal domains appear to play multiple roles related to dDis3 enzymatic

activity.

III.E.2 The dDis3 N-terminus is important for subcellular localization

We have shown that dDis3 localization is a consequence of a sensitive balance

between N- and C-terminal sequences. In this regard, dDis3 N-terminal mutants

containing a C-terminal NLS are not nuclear. This suggests several possibilities for the

142

function of N-terminal domains in dDis3 localization. First, N-terminal domains could maintain the proper structure of dDis3, such that the NLS is in a functional conformation.

Consistent with this idea, N-terminal domains are also necessary, probably at the structural level, for RNase activity of the entire enzyme. Alternatively, the N-terminus could contain an additional signaling or regulatory sequence that directs dDis3 localization, in conjunction with the NLS, to various subcellular compartments. Our bioinformatic, fractionation, and immunofluorescence data suggests that dDis3 does have at least one N-terminal targeting sequence, which directs the protein to mitochondria.

This is the first example of an exosome protein localizing to mitochondria. Mitochondrial

RNA metabolism in higher eukaryotes has been primarily attributed to polynucleotide phosphorylase (PNPase), a phosphorolytic exoRNase that is structurally similar to the exosome core Piwowarski et al., 2004). However, recently it has been suggested this protein is confined to the intermembrane space (Chen et al., 2006). Thus, a major mitochondrial RNase is yet to be identified.

Perhaps dDis3 shuttles in and out of the nucleus in an effort to degrade distinct classes of RNAs, including mitoRNAs. Although not yet connected to mitochondria, S. cerevisiae Dis3 has been shown to target RNAs in different subcellular compartments.

For example, S. cerevisiae Dis3 participates in the processing of rRNAs in the nucleus

(Mitchell et al., 1997; Suzuki et al., 2001; Dziembowki et al., 2007; Schaeffer et al.,

2009; Schneider et al., 2009), as well as the degradation of mRNAs in the cytoplasm

(Dziembowski et al., 2007). It is unknown whether distinct pools of Dis3 proteins degrade these targets in each compartment, or if a single, shuttling pool of Dis3 proteins is responsible for the processing and turnover of both targets.

143

The presence of one N-terminal localization signal does not explain all of the

results we observe with N-terminal mutants. For example, mutations to either the C3 or

OB2 domains disrupt proper subcellular distribution of dDis3. However, mutations to the

PIN and OB1 domains have little effect on normal dDis3 localization. Based on these observations, we suspect that the dDis3 N-terminus contains both the MTS and additional unidentified signaling sequences which may or may not be subject to several levels of regulation. It will be important to decipher the specific mechanisms by which dDis3 localization is directed in order to understand how and when Dis3 can function in different cellular locations.

III.E.3 The dDis3 N-terminus is responsible for interactions with core exosome proteins and exosome co-factors

Our immunoprecipitation studies are consistent with work demonstrating that

Dis3 N-terminal domains are responsible for core exosome interactions (Graham et al.,

2009a; Schneider et al., 2009). We found that the association between the core and dDis3 is reduced to amino acids 1-252. Moreover, dDis329-982 does not interact with the core, indicating that the extreme N-terminal 28 amino acids are required for these interactions.

Together, these data show that associations with core proteins occur through a dDis3 region containing the C3, PIN, and STAG domains. In contrast, the PIN domain alone is sufficient for S. cerevisiae Dis3 interactions with core proteins (Schneider et al., 2009).

This could point to organismal differences in how Dis3 associates with exosome subunits. However, it appears that the S. cerevisiae full-length enzyme binds these proteins better than the PIN domain alone (Schneider et al., 2009). It is reasonable that

144

the domains surrounding PIN lend stability to the interactions in both organisms.

Consistent with this, structural analysis of a S. cerevisiae Dis3 sub-complex containing

Rrp41 and Rrp45 shows that regions of Dis3 outside of the PIN domain contact the core proteins (Bonneau et al.,. 2009). Although the interacting region of Dis3 differs slightly depending on the organism, it is clear from our studies and others that the N-terminus is important for these interactions.

There are several other noteworthy observations gained from the IP data. First, dRrp40, dRrp47, dRrp6, and dImportin-α3 continue to interact with dDis31-188 despite the loss of core exosome binding. This could suggest these proteins associate with dDis3 in a complex that is independent of the remaining subunits. Formation of this complex is consistent with previously observed interactions between S. cerevisiae Rrp6 and Rrp47

(Mitchell et al., 2003; Hieronymus et al., 2009). Notably, we also show that amino acids

731-982 of dDis3 interact with dRrp6 and dImportin-α3 only. Thus, dRrp6 itself does not elicit indirect binding of the core proteins when it binds to either the dDis3 C-terminus, or the N-terminal region 1-188. This suggests that when core exosome proteins do bind dDis3, the interaction is direct, even in the presence of dRrp6. Direct interactions have been observed in the crystal structure of the S. cerevisiae Dis3-Rrp41-Rrp45 sub-complex

(Bonneau et al., 2009). We speculate that these interaction profiles reflect different dDis3 complexes and/or different assemblages of polypeptides on distinct portions of the full- length protein. Moreover, this promotes a model that exosome subunits assemble into protein complexes that are independent of the core (Callahan and Butler, 2008; Graham et al., 2006; Graham et al., 2009a; Graham et al., 2009b; Kiss and Andrulis, 2009;

Callahan and Butler, 2010).

145

III.E.4 Do dDis3 N-terminal domains link three different functions?

Our biochemical findings place additional importance on the roles of the Dis3 N-

terminus. An examination of the schematic in Figure 45 shows that the N-terminus is a

hub for endoRNase activity and interactions with core exosome proteins, dRrp6, dRrp47,

and dImportin-α3. Additionally, the N-terminus is important for localization, and

contains an MTS. As all of these functions are located at the N-terminus, they could be

linked, providing a unique way to regulate dDis3 activity. For example, RNase activities

could be regulated via interactions with exosome proteins. In S. cerevisiae, it has already

been shown that Dis3 interaction with core proteins results in an increase or reduction in

its RNase activity depending on the substrate (Liu et al., 2006; Dziembowski et al., 2007;

Bonneau et al., 2009). Differential regulation of exo- and/or endoRNase activity could

also occur with dependence on subcellular localization.

It is also possible that the N-terminus simply evolved in Dis3 proteins to mediate

additional functions not possessed by its bacterial homologs, like endoRNase activity or

mitochondrial localization. However, these N-terminal activities do not have to be

interdependent. As an example, a comparison of our in vivo data sets suggests that proper

dDis3 localization does not necessarily depend on interactions with core exosome

proteins. dDis329-982, which does not interact with the core exosome, is predominantly

nuclear, consistent with wild-type localization. Conversely, dDis31-252, which does interact with core exosome proteins, is predominantly cytoplasmic, and hence does not retain its normal subcellular distribution pattern. Further analysis will be required to determine exactly how the many functions of Dis3 are linked.

146

Cell biological features

contributes to nuclear localization MTS/mitochondria localization NLS/nuclear localization C3 PIN STAG OB (x2) RNB S1

1 29 62 188 252 394 855 928 982

dRrp6/dImpα3 dRrp6/dImpα3 interaction interaction

core exosome interaction

endoribonuclease/contributes to RNB exoribonuclease activity

Biochemical features

Figure 45. dDis3 functional regions identified in this work

(Top) dDis3 cell biological features include one N-terminal region that contributes to nuclear localization, and an N-terminal mitochondrial targeting sequence (MTS). The

NLS was previously identified (Graham et al., 2009a). (Bottom) dDis3 biochemical features include an N-terminal endoribonucleolytic active site, N-terminal domains that contribute to enzymatic activity overall, and a C-terminal exoribonucleolytic active site.

The N-terminus also interacts with core exosome proteins, dRrp47, dRrp6, and dImportin-α3. The dDis3 C-terminus contains a region for dRrp6 and dImportin-α3 interactions as well.

147

III.E.5 Conclusions

In sum, we have provided a characterization of dDis3 N-terminal domain

functions. This work adds to a larger group of studies aimed to advance knowledge of

Dis3 enzymology, core exosome function, and general RNA metabolic pathways. The

goals now are to understand the mechanisms behind Dis3 complex assembly,

disassembly, targeting, localization, and substrate specificity.

III.F Funding

This work was supported by grants GM072820 to E.D.A. and T32HD007104 to M.M.

from the National Institutes of Health.

148

Chapter IV

General Discussion and Future Directions

149

IV.A The Drosophila melanogaster Dis3 ribonuclease

As we are the first group to examine the biochemical and cell biological characteristics of a metazoan Dis3, we have made seminal contributions to the field. In this study, we showed that dDis3 has both endo- and 3’Æ5’ exoRNase activities. Thus, we have demonstrated that these activities are conserved in multicellular eukaryotic RNR family members. We also revealed novel features of the Dis3 RNase in our study. With our immunofluorescence and fractionation experiments, we uncovered two putative N- terminal localization sequences; a mitochondrial targeting sequence, and an additional sequence that appears to contribute to nuclear localization. With immunoprecipitation experiments, we also confirmed that dDis3 N-terminal domains are responsible for interactions with core exosome proteins, as has been shown for S. cerevisiae Dis3. These experiments also showed that N-terminal domains mediate Dis3 interactions with an additional exosome subunit, Rrp6, an exosome co-factor, Rrp47, and the nuclear import protein Importin-α3. The interacting domains for these proteins were previously unidentified. Thus our experiments have uncovered additional functions for Dis3 domains.

Interestingly, we found that the localization sequences, protein-protein interacting regions, and the endoRNase active site are all located within an N-terminal dDis3 region that is not present within prokaryotic RNR family members. Dis3 bacterial homologs

RNase II and RNase R do not contain the N-terminal C3, PIN, and STAG domains (Zuo and Deutscher, 2001). Consistent with this lack of N-terminal domains, these proteins do not have an endoRNase activity, nor are they known to localize to multiple subcellular compartments or associate with other RNA metabolic proteins. In fact, no other exosome

150

proteins are present in eubacteria. As mentioned previously, this could suggest that the N-

terminus of eukaryotic RNR proteins evolved for the purpose of mediating additional functions that are not required for bacterial RNA metabolism. However, two studies published recently have identified the presence of two Dis3 homologs in human cells

(Staals et al., 2010; Tomecki et al., 2010). Both of the proteins contain N-terminal domains like other eukaryotic Dis3 homologs. However, only one of the proteins, hDis3, has characteristics similar to the Drosophila protein, including endoRNase activity,

exosome binding, and distinct subcellular localization patterns (Staals et al., 2010;

Tomecki et al., 2010). The other protein, hDis3l, lacks endoRNase activity, even though

it contains the N-terminal domains thought to mediate this function (Staals et al., 2010;

Tomecki et al., 2010). Additionally, hDis3l was found to be exclusively cytoplasmic,

which differs from the localization patterns of other eukaryotic Dis3 homologs (Staals et

al., 2010; Tomecki et al., 2010). Together, these studies suggest that the N-terminal

domains in eukaryotic RNR proteins may not always convey additional functions.

Perhaps the N-terminus evolved as a central location for regulation of eukaryotic Dis3

functions instead. Although our study has uncovered both conserved and novel

characteristics of Dis3, analysis of this enzyme, including how it is regulated, is

incomplete. Hence, several lines of in vitro and in vivo investigation may follow from our

work here.

151

IV.B dDis3 in vitro

IV.B.1 dDis3 ion requirements and reaction mechanism

Our in vitro studies focused on requirements for dDis3 RNase activity and the reaction mechanisms of this enzyme. Our data show that the activity of full-length dDis3

requires metal ions in vitro. However, it is not completely clear which dDis3 activity,

endo- or exoRNase, requires these ions. We found that dDis3 is endo- and

exoribonucleolytically active in the presence of both Mg2+ and Mn2+ ions (data not

shown). Thus, it is likely that dDis3 employs metal-ion catalysis for both activities, as has

been suggested for the S. cerevisiae homolog (Lebreton et al., 2008; Schaeffer et al.,

2009; Schneider et al., 2009), but additional experiments are needed to verify these reaction mechanisms. In order to clarify the role of metal ions in dDis3 activity, experiments employing mutational analyses could be performed. Either the endoRNase or exoRNase active site could be mutated, and then Dis3 RNase activities could be

assessed in various metal ion-containing buffers to determine which types of ions are

absolutely required for each activity. Similar experiments have been performed for S.

cerevisiae Dis3, although these studies were inconclusive because they did not specify if

metal ions are necessary for RNase activity of for RNA binding (Lebreton et al., 2008).

Thus, these experiments should be paired with an analysis of substrate binding in various

ionic conditions. It is possible that certain metal ions are required for RNA binding,

rather than active site catalysis. These types of assays will provide us with more

information regarding the reaction of dDis3 and its homologs.

152

IV.B.2 dDis3 domain function

Our structure-function analysis of dDis3 confirmed that N-terminal domains are

responsible for endoRNase activity and interactions with exosome subunits. We also

showed that the dDis3 N-terminus contributes to its subcellular localization. Additionally,

we found the C-terminal RNB domain is responsible for exoRNase activity, and C-

terminal domains mediate additional interactions with Rrp6 and Importin-α3. Although

we have defined the functions of several domains as described here, and as depicted in

Figure 45, there are multiple Dis3 domains that have not been characterized in any

eukaryotic system. These include the two N-terminal OB-fold domains and the C-

terminal S1 domain. All three regions are suggested to participate in RNA binding

(Lorentzen et al., 2008). Thus, these domains may be critical for substrate recognition,

binding, and subsequently, RNA degradation. To confirm an RNA binding functions for these domains, nucleic acid filter-binding experiments could be performed. This type of

experiment has been used successfully to characterize the RNA binding domains of the

Dis3 bacterial homolog RNase II (Amblar et al., 2006). In these assays, radiolabeled

RNA substrates could be incubated with wild-type or mutant Dis3 in the presence of

EDTA. As shown in our assays, EDTA would chelate Mg2+ ions, rendering Dis3

catalytically inactive (Amblar et al., 2006). Alternatively, point mutations could be made

to active site residues to produce the same effect. Our lab has already created several

point mutants that could be used in these assays and other mutational analyses.

Regardless of the method used, the goal would be to create constructs that do not degrade

the RNA so that binding may be easily quantified. The binding reactions would then be

filtered through a nitrocellulose filter on top of an Hybond-N+ filter as described

153

previously (Amblar et al., 2006). RNA bound to Dis3 would remain on the nitrocellulose

filter, whereas free RNA would run through and would be caught by the bottom filter

(Amblar et al., 2006). The amount of bound RNA versus free RNA could be determined

by quantification of labeled RNA on each filter (Amblar et al., 2006). Finally, the amount

of RNA complexed with wild-type Dis3 could be compared to RNA complexed with

mutant Dis3 to determine if mutants lose the ability to bind substrate. Hence, with Dis3

mutants, we could determine if the OB-fold and S1 domains are required for substrate binding.

The OB-fold domains may not only be involved in RNA binding. In a recent study, the Dis3 bacterial homolog RNase R was found to have ATP-dependent helicase activity, which is mediated by one OB-fold domain (Awano et al., 2010). Interestingly, it has been shown that Dis3 does not require additional proteins, like helicases, to degrade double- stranded structures (Liu et al., 2006). However, a helicase domain has never been discovered in Dis3. Instead, it is thought that Dis3 uses a “ratcheting” mechanism to pull apart secondary structure (Liu et al., 2006). We have shown that ATP does not influence

dDis3 activity on a single-stranded RNA. However, it is possible that ATP could elicit

greater dDis3 activity on double-stranded substrates, especially if it is required for a

helicase activity. A putative helicase function could be tested via an in vitro helicase

assay (described in Awano et al., 2010). Briefly, wild-type or mutant Dis3 would be

incubated with a double-stranded RNA in which only one strand is radiolabeled. The

reactions would also contain various concentrations of ATP, which is required for

traditional helicase activities (and the helicase activity of RNase R; Awano et al., 2010).

Additionally, the assays would require the presence of an un-labeled complementary

154

RNA. This un-labeled RNA would anneal to and “trap” any un-labeled strands from the

original double-stranded substrate that are released by helicase activity (Awano et al.,

2010). Products of the reactions would then be analyzed by gel-shift experiments, where

faster migrating bands were released by a putative Dis3 helicase activity. Hence, an ATP-

dependent accumulation of single-stranded RNA in this assay would be indicative of a novel Dis3 helicase activity. These types of in vitro assays will allow us to continue to

build upon our understanding of the fundamental features of Dis3 biochemistry.

IV.C dDis3 in vivo

Although continued in vitro analyses are important for understanding the

biochemistry of dDis3 RNase activities, in vivo analyses will shed light on more complex issues. By using Drosophila melanogaster, we have already moved to a more genetically tractable system that may allow us to more readily address two major questions regarding

Dis3 biology: (1) What are the conserved or unique targets of Dis3 RNase activities in

vivo, and (2) How are Dis3 RNase activities regulated?

IV.C.1 dDis3 substrate specificity

Although our biochemical results demonstrate that dDis3 endo- and 3’Æ5’ exoRNase activities are not substrate specific in vitro, this may or may not be the case in vivo. S. cerevisiae Dis3 has been implicated in the turnover or processing of several specific RNAs. These RNA targets are summarized in Figure 46. To date, only rRNAs have been definitively shown to be conserved Dis3 substrates (Dziembowski et al., 2007;

Staals et al., 2010; Kiss and Andrulis, 2010). However, considering the multitude of

155

PROMPT turnover tRNA met turnover heterochromatic i gene silencing mis-spliced mRNA turnover

mitosis rRNA processing

nucleus exosome binding ran binding importin-α3 binding mRNA transport Dis3 cytoplasm mRNA turnover

mitochondria ??

Figure 46. Summary of Dis3 in vivo functions

All Dis3 functions presented were discovered from studies of S. cerevisiae, S. pombe, D. melanogaster, and H. sapiens homologs. These studies were discussed in earlier text.

156

diverse RNA targets that Dis3 has been linked to in S. cerevisiae, it is likely that Dis3

participates in the turnover and processing of various RNA classes in other organisms as

well. For example, dDis3 could function in Drosophila melanogaster development, as its

expression levels vary during several developmental stages (Cairrao et al., 2005).

Additionally, a microarray study of RNAs extracted from S2 cells, completed by our lab,

showed that ~4% of RNAs affected by dDis3 knock down (via RNAi) are development-

associated (Kiss and Andrulis, 2010). Confirmation of these development-related RNA

targets could reveal novel RNA metabolic pathways in which dDis3 functions. One way

to confirm the involvement of dDis3 in development is to examine the characteristics of

Drosophila embryos in which dDis3 expression is controlled by an inducible promoter system. Our lab possesses transgenic fly lines that could be mated to produce embryos in

which dDis3 is knocked down by inducible RNAi. We could compare the phenotypes of

these embryos to wild-type embryos to determine if dDis3 is required for development in

general. We may also be able to pinpoint the developmental stage(s) or mechanism(s) in

which dDis3 participates based on RNAi knock down phenotypes. In addition, we could

couple this experiment with RNA fluorescence in situ hybridization (FISH) to determine

if specific RNAs are affected by dDis3 knock down during certain stages of development,

and hence may be targeted by dDis3 to regulate known developmental processes. Briefly,

wild-type embryos or dDis3 knock down embryos could be fixed with formaldehyde.

Then, complementary RNA sequences linked to a fluorophore could be hybridized to

RNAs that we suspect may be targeted by dDis3. The levels of these RNAs, and their

localizations could then be examined by fluorescence microscopy. This type of procedure has been used successfully to examine development-associated transcripts in Drosophila

157

embryos at multiple stages of development (e.g. Arvey et al., 2010). Since Dis3 is an

essential protein in yeast (Kinoshita et al., 1991), and Dis3 is required for cell proliferation in Drosophila melanogaster (Kiss and Andrulis, 2009), it is quite possible that Dis3 has essential functions early in development.

In addition to targeting developmental RNAs, Dis3 may target other RNA classes as well. Because our in vivo experiments suggested a possible mitochondrial localization for dDis3, we anticipated that dDis3 may function in mitochondria. Indeed, our group has preliminary data suggesting dDis3 may participate in the turnover of mitoRNAs. We performed RT-PCR analysis of mitochondrial-encoded transcripts, which were extracted from control and dDis3 knock down S2 cells. These experiments showed increased levels of several transcripts involved in the mitochondrial respiratory chain, although these

changes were modest (Turk, Mamolen et al., in preparation). Additional studies are

needed to ascertain dDis3’s role in mitochondrial RNA metabolism. Perhaps further RT-

PCR or northern blotting analyses of RNAs from dDis3 wild-type and knock down S2

cells will reveal additional RNAs that are targeted by dDis3 for processing and/or

turnover. We could also utilize microarray analyses of mitochondrial-encoded transcripts

from dDis3 wild-type and knock down cell lines to elucidate the global effect that dDis3

may have on the mitochondrial RNA metabolism. We envision that discovery of bona

fide dDis3 mitoRNA targets, and other RNA substrates, could uncover additional RNA

turnover and/or processing pathways in eukaryotes.

158

IV.C.2 Regulation of dDis3 RNase activities

IV.C.2a dDis3 localization

With our localization and immunoprecipitation studies, we have already made

strides towards understanding dDis3 in vivo functions, and have established a foundation

for studying how dDis3 activities may be regulated. As mentioned earlier, our results suggest that dDis3 contains multiple localization sequences. Thus, it is possible that dDis3 functions in multiple subcellular compartments, including mitochondria. To date,

no Dis3 homologs, except Drosophila Dis3, have been shown to localize to

mitochondria. However, bioinformatic analysis by our group suggests that this may also

be a conserved property of Dis3. An alignment of Dis3 N-terminal sequences shows that

the MTS in dDis3 is shared by several homologs (Table 2; Turk, Mamolen et al., in

preparation; data not shown). A prediction program (MitoProt) also suggested that many

Dis3 homologs should localize to mitochondria (Table 2; Turk, Mamolen et al., in

preparation). Additionally, each putative MTS contains a conserved glycine residue that

is predicted to be a site of proteolytic cleavage. This cleavage event is known to elicit

mitochondrial localization in other proteins (Li et al., 2010). Thus, this signal could

represent the method by which Dis3 proteins localize to mitochondria, to degrade or

process specific RNAs. This signal could also regulate how and when Dis3 proteins

target mitoRNAs.

In addition to an MTS, our immunofluorescence data suggest the presence of an

additional nuclear localization element. N-terminally truncated dDis3 mutants were

cytoplasmic, even though they contained NLSs. This suggests that deletion of N-terminal

sequences resulted in (1) deletion of some type of nuclear restriction sequence (NRS)

159

Table 2. Putative MTS alignment

Organism COBALT N‐terminal sequence alignment Predicted mitochondrial localization

S. cerevisiae 1 M—————————S——V——PAIAPRRKRLADG———LSVTOKVFVR—SRNGGATK 34 96%

M. musculus 1 ——————————————————————————MLRSKTFLKKTRAGGVVK 18 96%

H. sapiens 1 ——————————————————————————MLKSKTFLKKTRAGGVMK 18 91%

D. melanogaster 1 ——————————————————————————MQTLREFTRKTKRGNILK 18 96%

160

and/or (2) de-repression of a nuclear export sequence (NES). Bioinformatic analysis has

indicated the presence of two putative NESs in dDis3 and other Dis3 homologs. We are

currently performing additional immunofluorescence analyses of dDis3 N-terminally

truncated mutants in S2 cells treated with the nuclear export inhibitor, leptomycin B

(LMB) to verify the presence of any NESs. Preliminary results have shown that dDis3

localization is sensitive to LMB. However, additional studies with specific point mutants

are necessary to determine an exact NES. Considering dDis3 has multiple putative

localization signals, it is possible that this constitutes a mode of regulation for dDis3

activity in vivo. The activities of other proteins, such as the eukaryotic endo- and 3’Æ5’

exonuclease APE1, are thought to be directly controlled by their subcellular localizations

(Li et al., 2010). It would be interesting to determine if Dis3 is regulated in a similar

manner. Once we define each localization signal, we could also determine if mutations to

these signals result in defects or changes in either of Dis3’s RNase activities.

IV.C.2b dDis3 protein-protein interactions

In our analysis of dDis3, we also examined the requirements for interactions with

exosome proteins. We have uncovered a minimal interacting region, but the necessity for

interactions of any Dis3 homolog with exosome proteins is still unknown. As mentioned

previously, it was once thought that all exosome proteins have 3’Æ5’ exoRNase activity

(Mitchell et al., 1997). Recently, it was suggested that these proteins simply act as a

platform for substrate binding, and that Dis3 is solely responsible for the RNA

degradation and processing functions attributed to the complex (Dziembowski et al.,

2007). It is quite possible that exosome proteins both facilitate RNA binding and regulate

Dis3 activity in vivo. S. cerevisiae Dis3 activity does fluctuate in the presence of exosome

161

proteins in vitro (Liu et al., 2006; Bonneau et al., 2009). However, regulatory roles for

these proteins have not been definitively shown. One way to examine this possibility is

through tethering assays, which could also be coupled with RNAi in Drosophila S2 cells.

RNAs could be tethered to dDis3 in vivo, and then degradation of the RNA could be monitored by northern analysis, as exosome subunits are systematically knocked down by

RNAi. A similar approach has been used to examine the endoRNase activity of the NMD protein SMG6 in vivo (Glavan et al., 2006). Additionally, mutational analysis of dDis3 may be employed in this assay to determine which RNase activity is affected by exosome subunit depletion. Ultimately, this type of assay could facilitate a greater understanding of Dis3 regulation, as there is currently no in vivo study regarding control of Dis3 RNase activities to date.

IV.D Dis3 in the bigger picture: endo-exoRNases in RNA metabolism

There are many additional questions regarding Dis3 function that may be addressed. Although Dis3 has been linked to multiple, diverse RNA metabolic events and mitosis (Figure 46), the way that this enzyme functions in these pathways is not completely understood. Thus, future studies will likely focus on the relationship of Dis3 endo- and 3’Æ5’ exoRNase activities to each of these cellular processes. Both Dis3 activities are already thought to work in rRNA processing (Lebreton et al., 2008;

Schneider et al., 2009). It will be interesting to see if other RNA classes are processed in a similar manner. One can imagine that endo-exoRNases could specifically target and cleave an RNA substrate with endoRNase activity, and then immediately degrade the fragments with exoRNase activity. Thus, these proteins may represent a newly identified

162

group of enzymes that work very specifically and efficiently in multiple pathways.

Additional studies of Dis3 may reveal more conserved features of this exciting class of enzymes.

163

Appendix A

Table A1. Oligonucleotides used to create MBP- and FLAG-tagged constructs

Name Sequence (5' to 3') 1F/BglII GCGAGATCTAAAATGCAAACTTTACGCGAATTTACG 29F/BglII CGCAGATCTAAAATGATCGGCTGCGGCTCCGAGCTGTGC 62F/BglII CGCAGATCTAAAATGCACTATCTCGTTTTGGACACAATG 189F/BglII CGCAGATCTAAAATGGCGGAAGCAGAAGGTATTCTGG 253F/BglII CGCAGATCTAAAATGGGCACTTTTCAGGCATCCAGGG 395F/BglII CGCAGATCTAAAATGCGACAGGCAGCAATGCTGCAGAACC 731F/BglII CGCAGATCTAAAATGTCACTGGACAAGTGTGTCAAGG 188R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCAGCACGATTGGCAGCATCATCTG 252R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCCTGTAGTAGCTTATTTTGGC 319R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCATCGGCATAGACATTCTTCTC 394R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCTGACGTTTCGATACGGATGCGCGG 479R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCATCCCTCAGATCCACACGC 631R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCCTTTTTCAGTATTTTGGCC 732R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCTGAATGCGACAGCTCCAGTCCCG 757R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCTGACTGCATCATACAGCGTGTGGT 798R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCGCGATGTACCATAATGTCGG 855R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCATCCTCCTCCTTGCCGCGG 928R/FLAG/SalI GCGGTCGACCTACTTATCGTCATCGTCCTTGTAGTCAAGACGAACCGTGA 982R/Sal CGCGTCGACTTACTTCTTTTCTTATCCTTC R/C31A GGAAACACTCCCGGCACAGCTCGGATCCGGCGCCGATGTCGTCGCGCAGATAG R/C36A CCTCGTTTTGGAAACACTCCCGGGCCAGCTCGGAGCCGCAGCCGATG R/D67K CGATCTGGTCCAGAACCACATTTGTTTTTAAAACGAGATAGTGCGGGAA R/D183L GCGGTCGACCTAACGATTGGCAGCATCATCTG R/D291N CAAAAGCTCCACGGCCACCAGGTTGCCGTCTACGGCCCGATTAAGAGACTCG R/P419L GAGCGCACAAAATGTCCATGAAGGTACCGCGAGTTGCGCGGCCATGTG

F = forward primer; R = reverse primer.

Table A2. Oligonucleotides used to create GFP-tagged constructs

Name Sequence (5' to 3') 1F/BglII GCGAGATCTAAAATGCAAACTTTACGCGAATTTACG 1F/BglII/myc CGCAGATCTAAAATGGCAGAACAAAAACTTATTTCTGAAGAGGATCTGCAAACTTTACGCGAATTTACGC 1F/BglII/MTS4A CGCAGATCTAAAATGCAAACTTTACGCGAATTTACGGCTGCTACTGCAGCCGGCAACATTCTGAAGATTG 35R/BamHI GCGGGATCCCAGCTCGGAGCCGCAGCCGATGTC

F = forward primer; R = reverse primer.

164

Bibliography

Alexandrov, A., Chernyakov, I., Gu, W., Hiley, S.L., Hughes, T.R., Grayhack, E.J., and Phizicky, E.M. (2006) Rapid tRNA decay can result from lack of nonessential modifications. Molec Cell 21, 87-96.

Allmang, C., and Tollervey, D. (1998) The role of the 3’ external transcribed spacer in yeast pre-rRNA processing. J Molec Biol 278, 67-78.

Allmang, C., Kufel, J., Chanfreau, G., Mitchell, P., Petfalski, E., and Tollervey, D. (1999a) Functions of the exosome in rRNA, snoRNA and snRNA synthesis. EMBO J 18, 5399-5410.

Allmang, C., Petfalski, E., Podtelejnikov, A., Mann, M., Tollervey, D., and Mitchell, P. (1999b) The yeast exosome and human PM-Scl are related complexes of 3’Æ5’ exonucleases. Genes and Dev 13, 2148-2158.

Allmang, C., Mitchell, P., Petfalski, E., Tollervey, D. (2000) Degradation of ribosomal RNA precursors by the exosome. Nuc Acids Res 28, 1684-1691.

Amblar, M., and Arraiano, C.M. (2005) A single mutation in Escherichia coli ribonuclease II inactivates the enzyme without affecting RNA binding. FEBS J 272, 363- 374.

Amblar, M., Barbas, A., Fialho, A.M., and Arraiano, C.M. (2006) Characterization of the functional domains of Escherichia coli RNase II. J Molec Biol 360, 921-933.

Amrani, N., Ganesan, R., Kervestin, S., Mangus, D.A., Ghosh, S., and Jacobson, A. (2004) A faux 3’-UTR promotes aberrant termination and triggers nonsense-mediated mRNA decay. Nature 432, 112-128.

165

Andersen, K.R., Jensen, T.H., and Brodersen, D.E. (2008) Take the “A” tail – quality control of ribosomal and transfer RNA. Bioch et Biophys Acta 1779, 532-537.

Anderson, J.S.J, and Parker, R. (1998) The 3’ to 5’ degradation of yeast mRNAs is a general mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 3’ to 5’ exonucleases of the exosome complex. EMBO J 17, 1497-1506.

Andrulis, E. D., Werner, J., Nazarian, A., Erdjument-Bromage, H., Tempst, P., and Lis, J. T. (2002) The RNA processing exosome is linked to elongating RNA polymerase II in Drosophila. Nature 420, 837-841.

Arvey, A., Hermann, A., Hsia, C.C., Ie, E., Freund, Y., McGinnis, W. (2010) Minimizing off-target signals in RNA fluorescent in situ hybridization. Nuc Acids Res online.

Assenholt, J., Mouaikel, J., Andersen, K.R., Brodersen, D.E., Libri, D., and Jensen, T.H. (2008) Exonucleolysis is required for nuclear mRNA quality control in yeast THO mutants. RNA 14, 1-9.

Auweter, S.D., Obserstrass, F.C., Allain, F.H.T. (2006) Sequence-specific binding of single-stranded RNA: is there a code for recognition? Nuc Acids Res 34, 4943-4959.

Awano, N., Rajagopal, V.., Arbing, M., Patel, S., Hunt, J., Inouye, M., and Phadtare, S. (2010) Escherichia coli RNase R has dual activities, helicase and RNase. J Bact 192, 1344-1352.

Belgrader, P., Cheng, J., Zhou, X., Stephenson, L.S., and Maquat, L.E. (1994) Mammalian nonsense codons can be cis effectors of nuclear mRNA half-life. Molec Cell Biol 14, 8219-8228.

Bernstein, E., Caudy, A.A., Hammond, S.M., and Hannon, G.J. (2001) Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409, 363-366.

166

Boni, I.V., Isaeva, D.M., Musychenko, M.L., Tzareva, N.V. (1991) Ribosome-messenger recognition: mRNA target sites for ribosomal protein S1. Nuc Acids Res 19, 155-162.

Bonneau, F., Basquin, J., Ebert, J., Lorentzen, E., and Conti, E. (2009) The yeast exosome functions as a macromolecular cage to channel RNA substrates for degradation. Cell 139, 547-559.

Bousquet-Antonelli, C., Presutti, C., and Tollervey, D. (2000) Identification of a regulated pathway for nuclear pre-mRNA turnover. Cell 102, 765-775.

Buhler, M., Wilkinson, M.F., Muhlemann, O. (2002) Intranuclear degradation of nonsense codon-containing mRNA. EMBO Rep 3, 646-651.

Cairrao, F., Arraiano, C., and Newbury, S. (2005) Drosophila gene tazman, an orthologue of the yeast exosome component Rrp44p/Dis3, is differentially expressed during development. Dev Dyn 232, 733-737.

Callahan, K.P., and Butler, J.S. (2008) Evidence for core exosome independent function of the nuclear exoribonuclease Rrp6. Nuc Acids Res 36, 1-11.

Callahan, K.P., and Butler, J.S. (2010) TRAMP complex enhances RNA degradation by the nuclear exosome component Rrp6. J Biol Chem 285, 3540-3547.

Chan, W.K., Huang, L., Gudikote, J.P., Chang, Y.F., Imam, J.S., MacLean, J.A., and Wilkinson, M.F. (2007) An alternative branch of the nonsense-mediated decay pathway. EMBO J 26, 1820-1830.

Chen, C. Y., Gherzi, R., Ong, S. E., Chan, E. L., Raijmakers, R., Pruijn, G. J., Stoecklin, G., Moroni, C., Mann, M., and Karin, M. (2001) AU binding protein recruit the exosome to degrade ARE-containing mRNAs. Cell 107, 451-464.

167

Chen, H.W., Rainey, R.N., Balatoni, C.E., Dawson, D.W., Troke, J.J., Wasiak, S., Hong, J.S., McBride, H.M., Koehler, C.M., Teitell, M.A., and French, S.W. (2006) Mammalian polynucleotide phosphorylase is an intermembrane space RNase that maintains mitochondrial homeostasis. Molec Cell Biol 26, 8475-8487.

Cheng, Z., and Deutscher, M.P. (2002) Purification and characterization of the Escherichia coli exoribonuclease R, comparison with RNase II. J Biol Chem 277, 21624- 21629.

Claverie-Martin, F., Wang, M., and Cohen, S.N. (1997) ARD-1 cDNA from human cells encodes a site-specific single-strand endoribonuclease that functionally resembles Escherichia coli RNase E. J Biol Chem 272, 13823-13828.

Clissold, P.M., and Pontig, C.P. (2000) PIN domains in nonsense-mediated mRNA decay and RNAi. Curr Biol 10, 888-890.

Cochrane, J.C., and Strobel, S.A. (2008) Catalytic strategies of self-cleaving ribozymes. Acc Chem Res 41, 1027-1035.

Conrad, N.K., Mili, S., Marshall, E.L., Shu, M.D., and Steitz, J.A. (2006) Identification of a rapid mammalian deadenylation-dependent decay pathway and its inhibition by a viral RNA element. Molec Cell 24, 943-953.

Couttet, P., and Grange, T. (2004) Premature termination codons enhance mRNA decapping in human cells. Nuc Acids Res 32, 488-494.

Custodio, N., Carmo-Fonseca, M., Geraghty, F., Pereira, H.S., Grosveld, F., and Antoniou, M. Inefficient processing impairs release of RNA from the site of transcription. (1999) EMBO J 18, 2855-2866.

168

Das, B., Das, S., and Sherman, F. (2006) Mutant LYS2 mRNAs retained and degraded in the nucleus of Saccharomyces cerevisiae. Molec Cell Biol 23, 5502-5515.

Dez, C., Houseley, J., and Tollervey, D. (2006) Surveillance of nuclear-restricted pre- ribosomes within a subnucleolar region of Saccharomyces cerevisiae. EMBO J 25, 1534- 1546.

Doma, M.K., and Parker R. (2006) Endonucleolytic cleavage of eukaryotic mRNAs with stalls in translation elongation. Nature 440, 561-564.

Doma, M.K., and Parker, R. (2007) RNA quality control in eukaryotes. Cell 131, 660- 668.

Dupureur, C.M. (2008) Roles of metal ions in nucleases. Curr Opin Chem Biol 12, 25-255.

Dziembowski, A., Lorentzen, E., Conti, E., and Seraphin, B. (2006) A single subunit, Dis3, is essentially responsible for yeast exosome core activity. Nat Struct and Molec Biol 14, 15-22.

Eberle, A.B., Lykke-Andersen, S., Muhlemann, O., and Jensen, T.H. (2009) SMG6 promotes endonucleolytic cleavage of nonsense mRNA in human cells. Nat Struct Molec Biol 16, 49-55.

Estevez, A. M., Lehner, B., Sanderson, C. M., Ruppert, T., and Clayton, C. (2003) The roles of intersubunit interactions in exosome stability. J Biol Chem 278, 34943-34951.

Fang, F., Hoskins, J., and Butler, J.S. (2004) 5-fluorouracil enhances exosome-dependent accumulation of polyadenylated rRNAs. Molec Cell Biol 24, 10766-10776.

169

Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E., and Mello, C.C. (1998) Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 744-45.

Frazao C, McVey CE, Amblar M, Barbas A, Vonrhein C, Arraiano CM, Carrondo MA. (2006) Unraveling the dynamics of RNA degradation by ribonuclease II and its RNA- bound complex. Nature 443, 110-114.

Frischmeyer, P.A., van Hoof, A., O’Donnell, K., Guerrerio, A.L., Parker, R., and Dietz, H.C. (2002) An mRNA surveillance mechanism that eliminates transcripts lacking termination codons. Science 295, 2258-2261.

Gatfield, D., and Izaurralde, E. (2004) Nonsense-mediated messenger RNA decay is initiated by endonucleolytic cleavage in Drosophila. Nature 429, 575-578.

Gavin, A. C., Aloy, P., Grandi, P., Krause, R., Boesche, M., Marzioch, M., Rau, C., Jensen, L. J., Bastuck, S., Dumpelfeld, B., Edelmann, A., Heurtier, M. A., Hoffman, V., Hoefert, C., Klein, K., Hudak, M., Michon, A. M., Schelder, M., Schirle, M., Remor, M., Rudi, T., Hooper, S., Bauer, A., Bouwmeester, T., Casari, G., Drewes, G., Neubauer, G., Rick, J. M., Kuster, B., Bork, P., Russell, R. B., and Superti-Furga, G. (2006) Proteome survey reveals modularity of the yeast cell machinery. Nature 440, 631-636.

Gavin, A. C., Bosche, M., Krause, R., Grandi, P., Marzioch, M., Bauer, A., Schultz, J., Rick, J. M., Michon, A. M., Cruciat, C. M., Remor, M., Hofert, C., Schelder, M., Brajenovic, M., Ruffner, H., Merino, A., Klein, K., Hudak, M., Dickson, D., Rudi, T., Gnau, V., Bauch, A., Bastuck, S., Huhse, B., Leutwein, C., Heurtier, M. A., Copley, R. R., Edelmann, A., Querfurth, E., Rybin, V., Drewes, G., Raida, M., Bouwmeester, T., Bork, P., Seraphin, B., Kuster, B., Neubauer, G., and Superti-Furga, G. (2002) Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141-147.

170

Glavan, F., Behm-Ansmant, I., Izaurralde, E., and Conti, E. (2006) Structures of the PIN domains of SMG6 and SMG5 reveal a nuclease within the mRNA surveillance complex. EMBO J 25, 5117-5125.

Graham, A.C., Kiss, D.L., and Andrulis, E.D. (2006) Differential distribution of exosome subunits at the nuclear lamina and in cytoplasmic foci. Molec Biol Cell 17, 1399-1409.

Graham, A. C., Davis, S. M., and Andrulis, E. D. (2009a) Interdependent nucleocytoplasmic trafficking and interactions of Dis3 with Rrp6, the core exosome, and Importin-α3. Traffic 10, 499-513.

Graham, A. C., Kiss, D. L., and Andrulis, E. D. (2009b) Core exosome-independent roles for Rrp6 in cell cycle progression. Molec Biol Cell 20, 2242-2253.

Granneman, S., and Baserga, S.J. (2004) Ribosome biogenesis: of knobs and RNA processing. Ex Cell Res 296, 43-50.

Graumann, P.L., and Marahiel, M.A. (1998) A superfamily of proteins that contain the cold-shock domain. Trends Biochem Sci 23, 286-290.

Gupta, R.S., Kasai, T., and Schlessinger, D. (1977) Purification and some novel properties of Escherichia coli RNase II. J Biol Chem 252, 8945-4949.

Hall, G.W., and Thein, S. (1994) Nonsense codon mutations in the terminal exon of the β-globin gene are not associated with a reduction in β-mRNA accumulation: a mechanism for the phenotype of dominant β-Thalassemina. Blood 83, 2031-2037.

Hieronymus, H., Yu, M. C., and Silver, P. A. (2004) Genome wide mRNA surveillance is coupled to mRNA export. Genes and Dev 18, 2652-2662.

171

Hollis, G.F., Gazdar, A.F., Bertness, V., Kirsch, I.R. (1988) Complex translocation disrupts c-myc regulation in a human plasma cell myeloma. Molec Cell Biol 8, 124-129.

Hilleren, P., McCarthy, T., Rosbash, M., Parker, R., and Jensen, T.H. (2001) Quality control of mRNA 3’-end processing is linked to the nuclear exosome. Nature 413, 538- 542.

Hilleren, P.J., and Parker, R. (2003) Cytoplasmic degradation of splice-defective pre- mRNAs and intermediates. Molec Cell 12, 1453-1465.

Hodgkin, J., Papp, A., Pulak, R., Ambros, V., Anderson, P. (1989) A new kind of informational suppression in the nematode Caenorhabditis elegans. Genetics 123, 301- 313.

Hopper, A.K., and Phizicky, E.M. (2003) tRNA transfers to the limelight. Genes and Dev 17, 162-180.

Houalla, R., Devaux, F., Fatica, A., Kufel, J., Barrass, D., Torchet, C., and Tollervey, D. (2006) Microarray detection of novel nuclear RNA substrates for the exosome. Yeast 23, 439-454.

Houseley, J., and Tollervey, D. (2009) The many pathways of RNA degradation. Cell 136, 763-776.

Huntizinger, E., Kashima, I., Fauser, M., Sauliere, J., and Izaurralde, E. (2008) SMG6 is the catalytic endonuclease that cleaves mRNA containing nonsense codons in metazoan. RNA 14, 2609-2617.

Ibrahim, H., Wilusz, J., and Wilusz, C.J. (2007) RNA recognition by 3’-to-5’ exonucleases” the substrate perspective. Bioch et Biophys Acta 1779, 256-265.

172

Inada, T., and Aiba, H. (2005) Translation of aberrant mRNAs lacking a termination codon or with a shortened 3’-UTR is repressed after initiation in yeast. EMBO J 24, 1584-1595.

Isken, O., and Maquat, L.E. (2007) Quality control of eukaryotic mRNA: safeguarding cells from abnormal mRNA function. Genes and Dev 21, 1833-3856.

Kadaba, S., Krueger, A., Trice, T., Krecic, A. M., Hinnebusch, A. G., and Anderson, J. (2004) Nuclear surveillance and degradation of hypomodified initiator tRNA Met in Saccharomyces cerevisiae. Genes and Dev 18, 1227-1240.

Kadaba, S., Wang, X., Anderson, J.T. (2006) Nuclear RNA surveillance in Saccharomyces cerevisiae: Trf4p-dependent polyadenylation of nascent hypomethylated tRNA and an aberrant form of 5S rRNA. RNA 12, 508-521.

Kadowaki, T., Chen, S., Hitomi, M., Jacobs, E., Kumagai, C., Liang, S., Schneiter, R., Singleton, D., Wisniewska, J., and Tartakoff, A. M. (1994) Isolation and characterization of Saccharomyces cerevisiae mRNA transport-defective (mtr) mutants. J Cell Biol, 126, 649-659.

Kinoshita, N., Goebl, M. and Yanagida, M. (1991) The fission yeast dis3+ gene encodes a 110-kDa protein implicated in mitotic control. Molec Cell Biol 11, 5839-5847.

Kiss, D.L., and Andrulis, E.D. (2010) Genome-wide analysis reveals distinct substrate specificities of Rrp6, Dis3, and core exosome subunits. RNA 16, 781-791.

Kong, J., and Liebhaber, S.A. (2007) A cell-type restricted mRNA surveillance pathway triggered by ribosome extension into the 3’-untranslated region. Nat Struct Molec Biol 14, 670-676.

173

Kuai L., Fang, F., Butler, J.S., and Sherman, F. (2004) Polyadenylation of rRNA in Saccharomyces cerevisiae. Proc Natl Acad Sci 101, 8581-8586.

Korennykh, A.V., Plantinga, M.J., Correll, C.C., Piccirilli, J.A. (2007) Linkage between substrate recognition and catalysis during cleavage if sarcin/ricin loop RNA by restrictocin. Biochemistry 46, 12744-12756.

LaCava, J., Houseley, J., Saveanu, C., Petfalski, E., Thompson, E., Jacquier, A., and Tollervey, D. (2005) RNA degradation by the exosome is promoted by a nuclear polyadenylation complex. Cell 121, 713-724.

LaRiviere, F.J., Cole, S.E., Ferullo, D.J., and Moore, M.J. (2006) A late-acting quality control process for mature eukaryotic rRNAs. Molec Cell 24, 619-626.

Lebreton, A., Tomecki, R., Dziembowski, A., and Seraphin, B. (2008) Endonucleolytic RNA cleavage by a eukaryotic exosome. Nature 456, 993-996.

Lejeune, F., Ranganathan, A.C., and Maquat, L.E. (2004) eIF4G is required for the pioneer round of translation in mammalian cells. Nat Struct Molec Biol 11, 992-1000.

Li., W.M., Barnes, T., and Lee, C.H. (2010) Endoribonucleases- enymes gaining spotlight in mRNA metabolism. FEBS J 277, 627-641.

Libri, D., Dower, K., Boulay, J., Thomsen, R., Rosbash, M., Jensen, T.H. (2002) Interactions between mRNA export commitment, 3’-end quality control, and nuclear degradation. Molec Cell Biol 22, 8254-8266.

Lim J, Kuroki T, Ozaki K, Kohsaki H, Yamori T, Tsuruo T, Nakamori S, Imaoka S, Endo M, Nakamura Y. (1997) Isolation of murine and human homologues of the fission-yeast dis3+ gene encoding a mitotic-control protein and its overexpression in cancer cells with progressive phenotype. Cancer Res 57, 921-925.

174

Lim, S.K. and Maquat, L.E. (1992a) Human beta-globin mRNAs that harbor a nonsense codon are degraded in murine erthyroid tissues to intermediates lacking regions of exon 1 or exons 1 and 2 that have a cap-like structure at the 5’ termini. EMBO J 11, 3271-3278.

Lim, S.K., Sigmund, C.D., Gross, K.W., and Maquat, L.E. (1992b) Nonsense codons in human beta-globin mRNA result in the production of mRNA degradation products. Molec Cell Biol 12, 1149-1161.

Lis, J.T., Mason, P., Peng, J., Price, D.H., Werner, J. (2000) P-Tefb kinase recruitment and function at heat shock loci. Genes and Dev 14; 792-803.

Liu, J., Carmell, M.A., Rivas, F.V., Marsden, C.G., Thomson, J.M., Song, J.J., Hammond, S.M., Joshua-Tor, L., and Hannon, G.J. (2004) Argonaute2 is the catalytic engine of mammalian RNAi. Science 305, 1437-1441.

Liu, Q., Greimann, J.C., Lima, C. (2006) Reconstitution, activities, and structure of the eukaryotic RNA exosome. Cell 127, 1223-1237.

Lorentzen, E., Basquin, J., Tomecki, R., Dziembowski, A., and Conti, E. (2008). Structure of the active subunit of the yeast exosome core, Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family. Molec Cell 29, 717-728.

Mamolen, M., and Andrulis, E. D. (2009) Characterization of the Drosophila melanogaster Dis3 ribonuclease. Biochem Biophys Res Commun 390, 529-534.

Mamolen, M., Smith, A., and Andrulis, E.D. (2010) Drosophila melanogaster Dis3 N- terminal domains are required for ribonuclease activities, nuclear localization and exosome interactions. Nuc Acids Res, in press.

175

Matskevich, A.A., and Moelling, K. (2007) Dicer is involved in protection against influenza A virus infection. J Gen Vir 88, 2627-2635.

Messias, A.C., and Sattler, M. (2004) Structural basis of single-stranded RNA recognition. Acc Chem Res 37, 279-287.

Mitchell, P., Petfalski, E., and Tollervey, D. (1996) The 3’ end of yeast 5.8S rRNA is generated by an exonuclease processing mechanism. Genes and Dev 10, 502-513.

Mitchell, P., Petfalski, E., Shevchenko, A., Mann, M., and Tollervey, D. (1997) The exosome: a conserved eukaryotic RNA processing complex containing multiple 3’Æ5’ exoribonucleases. Cell 91, 457-466.

Mitchell, P., Petfalski, E., Houalla, R., Podtelejnikov, A., Mann, M., and Tollervey, D. (2003) Rrp47p is an exosome-associated protein required for the 3’ processing of stable RNAs. Molec Cell Biol 23, 6982-6992.

Mitchell, P., and Tollervey, D. (2003) An NMD pathway in yeast involving accelerated deadenylation and exosome-mediated 3’Æ5’ degradation. Molec Cell 11, 1405-1413.

Montgomery, M.K., Xu, S., and Fire, A. (1998) RNA as a target of double-stranded RNA-mediated genetic interference in Caenorhabditis elegans. Proc Natl Acad Sci 95, 15502-15507.

Moore, M.J. (2005) From birth to death: the complex lives of eukaryotic mRNAs. Science 309, 1514-1518.

Morrow, G., Inaguma, Y., Kato, K., and Tanguay, R.M. (2000) The small heat shock protein Hsp22 of Drosophila melanogaster is a mitochondrial protein displaying oligomeric organization. J Biol Chem 275, 31204-31210.

176

Muhlrad, D., and Parker, R. (1994) Premature translational termination triggers mRNA decapping. Nature 370, 578-581.

Murakami, H., Goto, D.B., Toda, T., Chen, E.S., Grewal, S.I., Martienssen, R.A., and Yanagida, M. (2007) Ribonuclease activity of Dis3 is required for mitotic progression and provides a possible link between heterochromatic and kinetochore function. PLoS One 2, 1-12.

Murzin, A.G. (1993) OB (oligonucleotide/oligosaccharide binding)-fold: common structural and functional solution for non-homologous sequences. EMBO J 12, 861-867.

Noguchi, E., Hayashi, N., Azuma, Y., Seki, T., Nakamura, M., Nakashima, N., Yanagida, M., He, X., Mueller, U., Sazer, S., and Nishimoto, T. (1996) Dis3, implicated in mitotic control, binds directly to Ran and enhances the GEF activity of RCC1. Embo J 15, 5595- 5605.

Ohkura H, Adachi Y, Kinoshita N, Niwa O, Toda T, Yanagida M. (1988) Cold-sensitive and caffeine-supersensitive mutants of the Schizosaccharomyces pombe dis genes implicated in sister chromatid separation during mitosis. EMBO J 7, 1465-1473.

Orban, T.I., and Izaurralde, E. (2005) Decay of mRNAs targeted by RISC requires XRN1, the Ski complex, and the exosome. RNA 11, 459-469.

Page, M.J., and DiCera, E. (2006) Role of Na+ and K+ in enzyme function. Phys Rev 86, 1049-1092.

Papadimitriou, A., and Gross, H.J. (1996) Pre-tRNA 3’-processing in Saccharomyces cerevisiae. Purification and characterization of exo- and endoribonucleases. Eur J Biochem 242, 747-759.

177

Parker, R., and Song. H. (2004) The enzymes and control of eukaryotic mRNA turnover. Nat Struc and Molec Biol 11, 121-127.

Piwowarski, J., Dziembowski, A., Dmochowska, A., Minczuk, M., Tomecki, R., Gewartowski, K., and Stepien, P.P. (2004) RNA degradation in yeast and human mitochondria. Toxic Mech Methods 14, 53-57.

Preker, P., Nielsen, J., Kammler, S., Lykke-Andersen, S., Christensen, M.S., Mapendano, C.K., Schierup, M.H., and Jensen, T.H. (2008) RNA exosome depletion reveals transcription upstream of active human promoters. Science 322, 1851-1854.

Prokisch, H., Scharfe, C., Camp, D.G., Xiao, W., David, L., Andreoli, C., Monroe, M.E., Moore, R.J., Gritsenko, M.A., Kozany, C., Hixson, K.K., Mottaz, H.M., Zischka, H., Ueffing, M., Herman, Z.S., Davis, R.W., Meitinger, T., Getner, P.J., Smith, R.B., and Steinmetz, L.M. (2004) Integrative analysis of the mitochondrial proteome in yeast. PLoS Biol 6, 795-804.

Raines, R.T. (1998) Ribonuclease A. Chem Rev 98, 1045-1066.

Reinders, J., Zahedi, R.P., Pfanner, N., Meisinger, C., and Sickmann, A. (2006) Toward the complete yeast mitochondrial proteome: multidimensional separation techniques for mitochondrial proteomics. J Prot Res 5, 1543-1554.

Rougemaille, M., Gudipati, R.K., Olesen, J.R., Thomsen, R., Seraphin, B., Libri, D., and Jensen, T.H. (2007) Dissecting mechanisms of nuclear mRNA surveillance in THO/sub2 complex mutants. EMBO J 26, 2317-2326.

Rozenblum E, Vahteristo P, Sandberg T, Bergthorsson JT, Syrjakoski K, Weaver D, Haraldsson K, Johannsdottir HK, Vehmanen P, Nigam S, Golberger N, Robbins C, Pak E, Dutra A, Gillander E, Stephan DA, Bailey-Wilson J, Juo SH, Kainu T, Arason A, Barkardottir RB, Nevanlinna H, Borg A, Kallioniemi OP. (2002) A genomic map of a 6-

178

Mb region at 13q21-q22 implicated in cancer development: identification and characterization of candidate genes. Hum Genet 110, 111-121.

Schaeffer, D., Tsanova, B., Barbas, A., Reis, F. P., Dastidar, E. G., Sanchez-Rotunno, M., Arraiano, C. M., and van Hoof, A. (2009) The exosome contains domains with specific endoribonuclease, exoribonuclease and cytoplasmic mRNA decay activities. Nat Struct Mol Biol 16, 56-62.

Schmid, M., and Jensen, T. H. (2008) The exosome: A multipurpose RNA-decay machine. Trends Biochem Sci 33, 501-510.

Schneider, C., Anderson, J.T., and Tollervey, D. (2007) The exosome subunit Rrp44 plays a direct role in RNA substrate recognition. Molec Cell 27, 324-331.

Schneider, C., Leung, E., Brown, J., and Tollervey, D. (2009) The N-terminal PIN domain of the exosome subunit Rrp44 harbors endonuclease activity and tethers Rrp44 to the yeast core exosome. Nuc Acids Res 37, 1127-1140.

Schwer B., Mao, X., and Shuman, S. Accelerated mRNA decay in conditional mutants of yeast mRNA capping enzyme. Nuc Acids Res 26, 2050-2057.

Shiomi, T., Fukushima, K., Suzuki, N., Nakashima, N., Noguchi, E., and Nishimoto, T. (1998) Human Dis3p, which binds to either GTP- or GDP-Ran, complements Saccharomyces cerevisiae Dis3. J Biochem (Tokyo) 123, 883-890.

Sickmann, A., Reinders, J., Wagner, Y., Joppich, C., Zahedi, R., Meyer, H.E., Schonfisch, B., Perschil, I., Chacinska, A., Guiard, B., Rehling, P., Pfanner, N, and Meisinger, C. (2003) The proteome of Saccharomyces cerevisiae mitochondria. Proc Natl Acad Sci 100, 13207-13212.

179

Singer, M.F., and Tolbert, G. (1964) Specificity of potassium-activated phosphodiesterase of Escherichia coli. Science 145, 593-595.

Slomovic, S., Laufer, D., Geiger, D., and Schuster, G. (2006) Polyadenylation of ribosomal RNA in human cells. Nuc Acids Res 34, 2966-2975.

Spahr, P.F. (1964) Purification and properties of ribonuclease II from Escherichia coli. J Biol Chem 239, 3716-3726.

Spahr, P.F., and Schlessinger, D. (1963) Breakdown of messenger ribonucleic acid by a potassium-activated phosphodiesterase from Escherichia coli. J Biol Chem 238, 2251- 2253.

Staals, R.H., Bronkhorst, A.W., Schilders, G., Slomovic, S., Schuster, G., Heck, A.J., Rajimakers, R., and Pruijn, G.J. (2010) Dis3-like 1: a novel exoribonuclease associated with the human exosome. EMBO J. online.

Steitz, T.A., and Steitz, J.A. (1993) A general two-metal-ion mechanism for catalytic RNA. Proc Natl Acad Sci 90, 6498-6502.

Suzuki, N., Noguchi, E., Nakashima, N., Oki, M., Ohba, T., Tartakoff, A., Ohishi, M., and Nishimoto, T. (2001) The Saccharomyces cerevisiae small GTPase, Gsp1p/Ran, is involved in 3' processing of 7S-to-5.8S rRNA and in degradation of the excised 5'-A0 fragment of 35S pre-rRNA, both of which are carried out by the exosome. Genetics 158, 613-625.

Symmons, M.F., Williams, M.G., Luisi, B.F., Jones, G.H., and Carpousis, A.J. (2002) Running rings around RNA: a superfamily of phosphate-dependent RNases. TRENDS 27, 11-18.

180

Theobald, D.L., Mitton-Fry, R.M., and Wuttke, D.S. (2003) Nucleic acid recognition by OB-fold proteins. Ann Rev Biophys and Biomolec Struct 32, 115-133.

Tomecki, R., Kristiansen, M.S., Lykke-Andersen, S., Chlebowski, A., Larsen, K.M., Szczesny, R.J., Drazkowska, K., Pastula, A., Andersen, J.S., Stepien, P.P., Dziembowski, A., and Jensen, T.H. (2010) The human core exosome interacts with differentially localized processive RNases: hDis3 and hDis3L. EMBO J. online.

Vanacova, S., Wolf, J., Martin, G., Blank, D., Deftwiler, S., Friedlein, A., Langen, H., Keith, G., Keller, W. (2005) A new yeast poly(A) polymerase complex involved in RNA quality control. PLoS Biol 3, e189.

Van Hoof, A., Frischmeyer, P.A., Dietz, H.C., and Parker, R. (2002) Exosome-mediated recognition and degradation of mRNAs lacking a termination codon. Science 295, 2262- 2264.

Vincent, H.A., and Deutscher, M.P. (2009) Insights into how RNase R degrades structured RNA: Analysis of the nuclease domain. J Molec Biol 387, 570-583.

Vogtle, F.-N., Wortekamp, S., Zahedi, R.P., Becker, D., Leidhold, C., Gevaert, K., Kellerman, J., Voos, W., Sickmann, A., Pfanner, N., and Meisinger, C. (2009) Global analysis of the mitochondrial N-proteome identifies a processing peptidase critical for protein stability. Cell 139, 428-439.

Wall, D., and Kaiser, D. (1999). Type IV pili and cell motility. Molecular Microbiology 32, 1-10.

Wang, H.W., Wang, J., Ding, F., Callahan, K., Bratkowski, M.A., Butler, J.S., Nogales, E., and Ke, A. (2007) Architecture of the yeast Rrp44 exosome complex suggests routes of RNA recruitment for 3’ end processing. Proc Natl Acad Sci 104, 16844-16849.

181

Williamson, J.R. (2000) Induced fit in RNA-protein recognition. Nat Struct Molec Biol 7, 834-837.

Win, T.Z., Draper, S., Read, R.L., Pearce, J., Norbury, C.J., and Wang, S.W. (2006) Requirement of fission yeast Cid14 in polyadenylation of rRNAs. Molec Cell Biol 26, 1710-1721.

Yang, W., Lee, J.Y., and Nowotny, M. (2006) Making and breaking nucleic acids: two- Mg2+-ion catalysis and substrate specificity. Molec Cell 22, 5-13.

Yang, W. (2008) An equivalent metal ion in one- and two-metal ion catalysis. Nat and Struct Molec Biol 15, 1228-1231.

Yang, X.C., Sullivan, K.D., Marzluff, W.F., Dominski, Z. (2009) Studies of the 5’ exonuclease and endonuclease activities of CPSF-73 in histone pre-mRNA processing. Molec Cell Biol 29, 31-42.

Zuo, Y., and Deutscher, M.P. (2001) Exoribonuclease superfamilies: structural analysis and phylogenetic distribution. Nuc Acids Res 29, 1017-1026.

Zuo Y, Vincent HA, Zhang J, Wang Y, Deutscher MP, Malhotra A. (2006) Structural basis for processivity and single-strand specificity of RNase II. Molec Cell 2, 149-156.

182