Functional study of MicroRNA 125b in vertebrate development 1

miR-125b is upregulated during differentiation of ReNcell VM cells and miR-125b ectopic expression promotes neurite outgrowth in these cells 48 3.5.. This thesis aims to reveal new func

FUNCTIONAL STUDY OF MICRORNA-125B IN

VERTEBRATE DEVELOPMENT

LE THI NGUYET MINH

(Bachelor of Science (honours) National University of Singapore)

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

IN COMPUTATION AND SYSTEMS BIOLOGY (CSB)

SINGAPORE-MIT ALLIANCE NATIONAL UNIVERSITY OF SINGAPORE

2009

ABSTRACT

microRNAs are a class of small non-coding RNAs that regulate gene expression at the post-transcriptional level This thesis aims to reveal the new functions of miR-125b, a brain-enriched microRNA in vertebrate development First, we demonstrated the important role of miR-125b in differentiation of two human neural cell lines We also found that a subset of miR-125b-repressed targets antagonize neuronal genes in several neurogenic pathways Second, we demonstrated that miR-125b is indispensable for zebrafish embryogenesis, particularly for the survival of neural

cells We identified p53, a key tumor suppressor, as a bona fide target of miR-125b in

both zebrafish and humans miR-125b directly represses p53 and multiple genes in the p53 network while p53 in turn suppressed the expression of miR-125b Together, our study provides a global view of miR-125b function in modulating gene expression to maintain the homeostasis of cell survival, death and differentiation during development

My special thanks are also due to Dr Beiyan Zhou, Dr Cathleen Teh, Dr Henry Yang and Dr Moonkyoung Um who have provided me with patient training and mentoring at different stages of my study I am also very thankful to Huangming Xie, Shyh-Chang Ng, Poh Hui Chia and Dr Pamela Rizk for their help and contribution to some major experiments in this project (more details are included at the end of chapter 3 and 4) I wish to express many thanks to Prof David Bartel, Prof Frank McKeon, Prof Vladimir Korzh and Dr Gerald Udolph for providing me with a lot of advices and useful facilities I am also indebted to Prof Hew Choy Leong, who first offered me the opportunity to join Singapore-MIT Alliance, and over the years, has given me a great deal of good advice, encouragement and care

I would like to extend my sincere thanks to all my colleagues in Biopolis and at the Whitehead Institute, especially Boon Seng Soh, Wai Leong Tam, Philip Gaughwin, Chin Yan Lim, Senthil Raja Jayapal, Yen Sin Ang, Yvonne Tay, Yin Loon Lee, Kar-Lai Poon, Hang Nguyen, Svitlana Korzh, Amanda Goh, Quo Lin, Andrew Thomson, Prakash Rao, Shilpa Hattangadi and Cheng Cheng Zhang for fruitful discussions Thanks to Rani Ettikan, Michael Chin, Li Pin, Adrian Lim, Lingbo Zhang, and Jun-

Liang Tay for their technical supports I also thank the staffs at Biopolis High Content Screening facility, zebrafish facility, confocal microscopy facility and histology facility for providing good services and advices Thanks to Dr Jun Chen and Prof Jinrong Peng for the p53M214K mutant zebrafish and camptothecin; Dr Kian-Chung Lee and Prof Sir David Lane for the anti-p53 antibody and the H1299 cells

I would like to acknowledge with much gratitude the excellent coursework, the generous fellowship and research funding from Singapore-MIT Alliance I also acknowledge the funding from A*STAR that provides the wonderful research facilities and intellectual atmosphere at Genome Institute of Singapore (GIS) I am grateful to the Graduate student committee members and the administrative staffs at Singapore-MIT Alliance and at GIS

Lastly, my heartfelt thanks to my family, especially to my parents whose encouragement has always been much treasured; to my husband, whose tremendous supports are the most essential to all my success; to my baby who has given me the courage to go through the most difficult time during my candidature

1.4 The expression and known functions of miR-125b 14

2.1 Cell culture and differentiation condition 22

2.5 Immunostaining and high-content screening of cells 27

2.7 Image acquisition and microscope settings 29

2.8 Whole mount in situ hybridization 29 2.9 Microinjection in zebrafish embryos 30

2.12 Terminal dUTP nick end labeling (TUNEL) assay 33 2.13 Gene expression microarray and data analysis 33

2.16 Target prediction and Pathway analysis 35

3.3 miR-125b is necessary and sufficient for neurite outgrowth and

neuronal marker gene expression

44

3.4 miR-125b is upregulated during differentiation of ReNcell VM cells

and miR-125b ectopic expression promotes neurite outgrowth in these cells

48

3.5 Profiling the downstream effectors of miR-125b 51 3.6 Identification of direct targets of miR-125b 54 3.7 Pathway analysis and validation of direct miR-125b targets 57

Chapter 4 microRNA-125b is a novel regulator of p53 73 4.1 Loss of miR-125b leads to severe defects in zebrafish embryos 73 4.2 miR-125b binds to the 3’ UTR of human and zebrafish p53 mRNAs 76 4.3 Spatio-temporal expression of miR-125b during zebrafish

Embryogenesis

80

4.4 miR-125b represses endogenous p53 and p53-induced apoptosis in

human neuroblastoma cells

by restoring the normal level of p53

94

4.8 Stress-induced p53 and apoptosis are repressed by ectopic miR-125b 96 4.9 Conservation of miR-125b targets in the p53 network 98

5.1 The function of miR-125b in differentiation of neuronal cells 107 5.2 The function of miR-125b in regulating p53 and p53 dependent-

apoptosis

107 5.3 A global view of miR-125b regulatory network of human cells 109 5.4 The implication of miR-125b in tumorigenesis 113

SUMMARY

microRNAs are a class of small non-coding RNAs that regulate gene expression at the post-transcriptional level Research on microRNAs has highlighted their importance

in many biological processes, especially in development miR-125b is a homolog of

lin-4 which is important for developmental timing in C elegans The expression of

miR-125b is upregulated during embryogenesis and enriched in the nervous system of vertebrate species However, the functions and targets of miR-125b remain poorly understood This thesis aims to reveal new functions of miR-125b in development with focus on two experimental systems: differentiation of human neural cells and zebrafish embryogenesis We first obtained the expression profile of microRNAs during neuronal differentiation in the human neuroblastoma cell line SH-SY5Y Six

microRNAs were significantly upregulated during differentiation induced by

all-trans-retinoic acid and brain-derived neurotrophic factor We demonstrated that ectopic expression of either miR-124a or miR-125b increases the percentage of differentiated SH-SY5Y cells with neurite outgrowth Subsequently, we focused our functional analysis on miR-125b and demonstrated the important role of this miRNA in both spontaneous and induced differentiation of SH-SH5Y cells miR-125b is also upregulated during differentiation of human neural progenitor ReNcell VM cells, and miR-125b ectopic expression significantly promotes neurite outgrowth of these cells

To identify the targets of miR-125b regulation, we profiled the global changes in gene expression following miR-125b ectopic expression in SH-SY5Y cells More than 50%

of the downregulated mRNAs contain the seed match sequence of miR-125b Transcripts with stronger seed matches are repressed with higher fold changes 188 of downregulated transcripts are predicted by TargetScan 5.1 to be direct targets of miR-125b Pathway analysis suggests that a subset of miR-125b-repressed targets

antagonize neuronal genes in several neurogenic pathways, thereby mediating the positive effect of miR-125b on neuronal differentiation We have further confirmed the binding of miR-125b to the microRNA response elements of nine selected mRNA targets and validated the binding specificity for three targets Together, these data reveal for the first time the important role of miR-125b in human neuronal differentiation

Further more, we demonstrate that miR-125b is indispensable for zebrafish embryogenesis, particularly for the survival of neural cells during development We

identified p53, a key tumor suppressor, as a bona fide target of miR-125b in both

zebrafish and humans miR-125b-mediated downregulation of p53 is strictly dependent on the binding of miR-125b to a microRNA-response element in the 3’ UTR of p53 mRNA Overexpression of miR-125b represses the endogenous level of p53 protein and suppresses apoptosis in human neuroblastoma cells and human lung fibroblast cells By contrast, knockdown of miR-125b elevates the level of p53 protein and induces apoptosis in human lung fibroblasts and in the zebrafish brain This phenotype can be rescued significantly by either an ablation of endogenous p53 function or ectopic expression of miR-125b in zebrafish Interestingly, miR-125b is downregulated when zebrafish embryos are treated with gamma-irradiation or camptothecin, corresponding to the rapid increase in p53 protein in response to DNA damage Ectopic expression of miR-125b suppresses the increase of p53 and stress-induced apoptosis We also identified seven additional targets of miR-125b in the p53 network and map the connections of miR-125b to many other components of this network Together, our study provides a global view of miR-125b function, as an integrated component of the cellular regulatory network, in modulating gene expression to maintain the homeostasis of cell survival, death and differentiation

LIST OF TABLES

Table 1 Sequences of Northern blot probes and morpholinos 25

Table 3 Differentially expressed genes in SH-SY5Y cells 40 Table 4 The targets of miR-125b predicted TargetScan 5.1 61

Table 6 Percentage of embryos with neural cell death 75 Table 7 Putative miR-125b targets in the p53 pathway 100

LIST OF FIGURES

Figure 1 Neuronal differentiation of SH-SY5Y cells 39 Figure 2 Heat map presentation of miRNA expression level 41 Figure 3 Analysis of miRNA expression in differentiating SH-SY5Y cells 42 Figure 4 Ectopic expression of candidate miRNAs and neurite outgrowth

assay in SH-SY5Y cells

46 Figure 5 The function of miR-125b in SH-SY5Y cell differentiation 47 Figure 6 Expression and function of miR-125b in ReNcell VM cells during

differentiation

50 Figure 7 Profiling the downstream effectors of miR-125b 53

Figure 10 Loss of miR-125b in zebrafish embryos 74 Figure 11 miR-125b binds to the 3’ UTR of zebrafish and human p53

mRNAs

78 Figure 12 miR-125b miRNA response elements (MREs) and luciferase

reporter assays

79 Figure 13 Spatio-temporal expression of miR-125b during zebrafish

embryogenesis

82 Figure 14 Validation of miR-125b overexpression and knockdown in SH-

SY5Y cells and in human lung fibroblast cells

84 Figure 15 miR-125b represses the endogenous p53 expression and

suppresses p53-induced apoptosis in human neuroblastoma

SH-SY5Y cells

85

Figure 16 miR-125b represses the endogenous p53 expression and

suppresses apoptosis in human lung fibroblast cells

87 Figure 17 Cellular responses to different doses of miR-125b and to etoposide

treatments

88 Figure 18 Loss of miR-125b elevates p53 and triggers p53-dependent

apoptosis in zebrafish embryos

91 Figure 19 Developmental onset of apoptosis in miR-125b morphants 92 Figure 20 Rescue of miR-125b morphants by the loss of p53 93 Figure 21 Synthetic miR-125b rescues apoptosis in miR-125b morphants 95 Figure 22 - Overexpression of miR-125b rescues stress-induced apoptosis 97 Figure 23 miR-125b function in the mouse Swiss-3T3 cells 101 Figure 24 miR-125b connections to the p53 network 102

CHAPTER 1 – INTRODUCTION

1.1 Introduction to microRNAs

microRNAs (miRNAs) represent an emerging class of ~22 nucleotide non-coding RNAs that play important roles in post-transcriptional regulation of gene expression (Bartel, 2004) Since the discovery of miRNA in 1993, the number of miRNAs annotated by the miRNA Registry miRBase (http://microrna.sanger.ac.uk) has increased dramatically By March 2009, miRBase has 9539 entries of experimentally identified miRNAs in 43 animals, plants and viruses (miRBase Release 13, 2009) Among these are 706 human miRNAs and 547 mouse miRNAs (miRBase Release 13, 2009) There are many more miRNAs predicted by computational analysis e.g the human genome is predicted to encode one thousand to tens of thousands of miRNAs (Bentwich et al., 2005; Miranda et al., 2006) These miRNAs are to be validated Identified miRNAs are classified into families in which the sequences of mature miRNAs are nearly identical with only one to two nucleotides differences (Bartel, 2004) miRNAs in the same family are thought to have the same set of putative targets (Bartel, 2004) Each miRNA family is predicted to have hundreds to thousands of target mRNAs (Bartel, 2009) Therefore, almost all protein-coding genes with identified 3’ UTRs have some putative binding sites for miRNAs (Bartel, 2009) Moreover, miRNAs are present in all types of tissues and they perform various physiological functions (Ason et al., 2006) While the functions of most miRNAs remain to be discovered, there are increasing examples of important miRNAs involved in different physiological and pathological processes The impact of miRNAs is attracting more and more attention from biological researchers Here we

present a brief introduction to the miRNA pathway in animals, particularly in vertebrate species



1.1.1 Biogenesis of microRNAs

miRNAs are transcribed initially as long RNAs (pri-miRNAs) with stem loop domains by RNA polymerase II or RNA polymerase III (Lee et al., 2004; Borchert et al., 2006) Most pri-miRNAs have only one stem loop to be processed into one mature miRNA whereas; some pri-miRNAs contain a cluster of stem loops which can be processed into multiple mature miRNAs (Altuvia et al., 2005) The stem loop domains are released from pri-miRNAs by the nuclear microprocessor complex which comprises two proteins, the endonuclease Drosha and DGCR8 (DiGeorge syndrome critical region 8) (Denli et al., 2004) For some pri-miRNAs that are transcribed from the introns of protein-coding genes, the stem loops can be excised by the messenger RNAs (mRNAs) splicing machinery (Okamura et al., 2007) The resulting hairpin RNA (called pre-miRNA) is exported to the cytoplasm by Exportin-5 and processed further into a ~22-base-pair double-stranded RNA by Dicer endonuclease (Jiang et al., 2005) Subsequently, one strand (the passenger strand) of the duplex is eliminated and the other strand (the guide strand) is incorporated into the RNA induced silencing complex (RISC) (Wheeler et al., 2006) miRNAs target the RISC to specific mRNAs that contain complementary binding sites (Schratt et al., 2006) In animal cells, this binding leads to a translational suppression or destabilization of the target mRNAs or both (Schratt et al., 2006; Chan et al., 2005b; Bartel, 2004; Lim et al., 2005; Wu et al., 2006)

The biosynthesis pathway of miRNAs is regulated at multiple steps Same as coding genes, transcription of pri-miRNAs are also modulated by transcription factors in

response to the cellular signalling (He and Hannon, 2004) The maturation of miRNAs by the Drosha and Dicer complexes are subjected to further regulatory mechanisms that are common or specific to individual miRNAs (Winter et al., 2009) There are many examples of feedback loops where miRNA targets in turn regulate miRNA expression and processing (Winter et al., 2009) In this way, miRNAs are well integrated with regulatory proteins in all the signalling network of the cell

1.1.2 Targets of microRNAs

To understand the mechanism of miRNA function, it is the most important to identify the targets of miRNAs In animals, the interaction between a miRNA and its target mRNA does not require a perfect complementarity (Lewis et al., 2003) The sequence from nucleotide two to eight at the 5’ end of a miRNA is considered as the “seed” for its binding to the target mRNAs (Lewis et al., 2003) The binding often requires a perfect match of seven base pairs in the seed region (Lewis et al., 2005) The binding affinity is also determined by additional base pairing, by the accessibility and the position of the binding sites relative to the coding region (Lewis et al., 2005; Grimson

et al., 2007) These criteria are selected and scored differently by the conventional prediction algorithms (Bartel, 2009) To rank the putative targets, most prediction methods also examine the conservation of the target sites across species based on sequence alignment (Bartel, 2009) Sites that are conserved through out evolution are considered to be more important and more likely to be true (Lewis et al., 2005) In addition, the number of binding sites for a miRNA in each mRNA is another factor that determines the probability of targeting (Bartel, 2009) The target mRNA that has multiple binding sites for a particular miRNA is more likely to be regulated by this miRNA (Bartel, 2009) Typically, miRNAs binding sites are found only in 3’ UTRs

of mRNAs but recently, miRNAs are also found to bind to the 5’ UTRs and the coding regions (Lytle et al., 2007; Tay et al., 2008a)

Each miRNA has hundreds to thousands of putative targets, depending on the stringency of selection Even the prediction by TargetScan, the most reliable and stringent method, estimates an average of 300 conserved targets for each miRNA (Friedman et al., 2009) Indeed, the potential of miRNA to regulate multiple targets has been demonstrated by genomic and proteomic profiling experiments (Baek et al., 2008; Selbach et al., 2008) A microarray analysis by Lim et al showed that ectopic expression of miR-124a or miR-1 in HeLa cells downregulated about 100 transcripts, most of which contain seed matches for the respective miRNAs (Lim et al., 2005) This study suggests that miRNAs are able to target a large number of transcripts through the mRNA-degradation mode Since then, mRNA microarray analysis has become a popular approach for miRNA target identification This approach may however miss the targets that are regulated only by translational inhibition As the ultimate indicator for miRNA targeting is the reduction in protein levels of the target mRNA, a better experimental approach to identify miRNA targets is to profile miRNAs-induced changes of the cellular proteome The Bartel group and the Rajewsky group have recently developed two similar amino-acid labelling and mass spectrometry methods for this purpose (Baek et al., 2008; Selbach et al., 2008) Both groups observed a wide-spread downregulation of protein expression by their miRNAs of interest (Baek et al., 2008; Selbach et al., 2008) Majority of the downregulated proteins contain putative binding sites for the miRNAs and most of them are also downregulated at the transcript level; only a few proteins are downregulated without any change in their mRNA levels (Baek et al., 2008; Selbach

et al., 2008) More importantly, most of the downregulated proteins exhibited very small fold changes (less than 50%), consistent with previous reports on the commonly observed modest effect of miRNAs on their target protein levels (Baek et al., 2008; Selbach et al., 2008) This supports the hypothesis that miRNAs may act on only a few targets in a switch-like mode, where miRNAs “turn off” the expression of the targets completely, but on many other targets in a failsafe or fine-tuning mode, where miRNAs slightly dampen the target expression to an optimized level (Bartel, 2009)

To follow up on a prediction of miRNA targets and/or a global profiling of induced changes in mRNA/protein levels, researchers often choose to focus on one or

miRNA-a few tmiRNA-argets for further vmiRNA-alidmiRNA-ation Results of the micromiRNA-arrmiRNA-ay or proteomic miRNA-anmiRNA-alysis of the target expression are often combined with the prediction of targets by conventional methods This helps to identify the most promising direct target candidates However, a miRNA-target interaction is considered to be meaningful only when it is shown to occur in a physiologically relevant context The interaction between a miRNA and a particular target is often determined by the availability, subcellular localization and accessibility of the target mRNA (Bartel, 2009) Importantly, these factors can vary in a temporal and spatial-specific manner (Bartel, 2009) In addition, the regulation of each target gene can be a net result of several mechanisms mediated by multiple miRNAs and proteins Therefore, it is important to examine miRNA-mediated gene regulation in a network context

1.2 The role of microRNAs in development

1.2.1 MicroRNA functions in embryogenesis

The miRNA pathway is essential for all stages of metazoan development The expression of most miRNAs is regulated during embryogenesis (Wienholds et al., 2005) Strikingly, inactivation of miRNA biogenesis by the loss of Dicer, a key enzyme in the processing of miRNAs, leads to severe defects in zebrafish embryos and in mouse prenatal lethality (Giraldez et al., 2005; Bernstein et al., 2003)

In the maternal-zygotic dicer zebrafish mutants, the embryos are viable but defective

in gastrulation, somitogenesis, heart and brain morphogenesis (Giraldez et al., 2005) Interestingly, injection of synthetic miRNA duplexes mimicking miR-430 family members rescued most of the brain defects (Giraldez et al., 2005) A subsequent study reveals that miR-430 miRNA family is also essential for a wide-spread clearance of maternal mRNAs by direct targeting for deadenylation during the maternal-zygotic transition in zebrafish embryos (Giraldez et al., 2006) At a later stage of zebrafish development, miR-430 plays a role in mesoderm formation by balancing the expression of Nodal agonist and antagonist (Choi et al., 2007) Hence, miR-430 family is a multifunctional regulator of zebrafish embryogenesis

In early embryogenesis, miRNA-mediated gene regulation is particularly important for developmental timing and patterning A classical example of miRNA function in development is the role of Hox miRNAs in regulating Hox genes that play critical roles in anatomic patterning Hox miRNAs refer to miRNAs encoded in the Hox gene clusters, including miR-10 and miR-196 families (Yekta et al., 2008) These miRNAs modulate the expression of Hox protein-coding genes for precise anterior-posterior

programmes in Bilateria species (Yekta et al., 2008) In chick embryos, miR-196

suppresses Hoxb8 that is also regulated at a transcriptional level to ensure a tight control of the Hoxb8 expression domain for the normal development of hindlimb in

the chick embryos (Hornstein et al., 2005)

Many miRNAs are expressed in a tissue-specific manner and the function of an individual miRNA may be determined by the target mRNAs it regulates in each tissue Tissue-specific ablation of miRNA processing by conditional knockdown of Dicer reveals the key role of miRNAs in organogenesis Dicer knockout in the mouse pancreas leads to a reduction of pancreatic endocrine cells and depletion in insulin secretion (Lynn et al., 2007) miR-375 is the most abundant miRNA in the pancreatic islets of human, mouse and zebrafish (Joglekar et al., 2009; Wienholds et al., 2005) Morpholino-mediated knockdown of miR-375 in zebrafish leads to a severe delay and defect in the development of both the endocrine and exocrine pancreas (Kloosterman

et al., 2007), whereas knockout of miR-375 in mouse resulted in a specific reduction

of -islet cells and impaired glucose metabolism (Poy et al., 2009)

Similarly, miRNA function in cardiogenesis was revealed by Dicer deletion specific

to the heart in mouse As the consequence, the embryos died from cardiac failure by embryonic day 12.5 (Zhao et al., 2007) The muscle-specific miR-1-2 and miR-133 are the most prominent miRNAs in heart development miR-1-2 knockout in mouse results in hyperplasia of the heart whereas suppression of miR-133 by infusion of its antisense antagomir into the heart leads to sustained cardiac hypertrophy (Zhao et al., 2007; Care et al., 2007)

1.2.2 MicroRNA functions in stem cell development

The role of miRNAs during development has also been notably featured in the identity and behaviours of stem cells, from embryonic stem cells (ESCs) to all types

of adult stem cells (Gangaraju and Lin, 2009; Foshay and Gallicano, 2007) Embryonic stem cells derived from Dicer-null mice are defective in their cell cycling and differentiation capability (Bernstein et al., 2003) They are not able to differentiate even after a prolonged exposure to differentiation inducers (Bernstein et al., 2003) The expression profile of miRNAs in wildtype ESCs exhibits specific and significant changes before and after the induction of differentiation (Houbaviy et al., 2003; Tay et al., 2008b) Particularly, a number of miRNAs including miR-21, miR-

134 and miR-145 promotes differentiation of ESCs by repressing the key transcription factors Oct4, Nanog, Sox2 and Klf4 that are essential for the pluripotency and self-renewal capability of ESCs (Tay et al., 2008a; Xu et al., 2009) Interestingly, these transcription factors in turn promote the expression of ESC-specific miRNAs, forming multiple feedback loops in an integrated gene regulation circuitry (Marson et al., 2008)

Similar to ESCs, the self-renewal and differentiation functions of adult stem cells, including skin stem cells, muscle stem cells, haematopoietic stem cells, mesenchymal stem cells and neural stem cells, also involve specific sets of miRNAs miRNAs are especially essential for the transition between sequential steps of differentiation and for cell fate determination during lineage specification For example, the multiple stages of haematopoiesis are marked by distinct sets of miRNAs and the progression from one stage to another is assisted by several key miRNAs (Chen et al., 2004) miR-106a and miR-20a promote the differentiation of granulocyte-macrophage progenitor

cells into granulocytes but not into monocytes by repressing AML1, a specific transcription factor (Fontana et al., 2007) Likewise, miR-150 is important for the derivation of pro-B cells but blocking the transition from pro-B cells to pre-B cells (Zhou et al., 2007)

monocyte-miRNAs are also important for the development of germ line, particularly the specification and maintenance of germline stem cells Loss of maternal Dicer1 in

Drsosophila represses the formation of primordial germ cells which normally give rise to germline stem cells Similar to the mouse Dicer-null ESCs, Drosophila Dicer1-

null germline stem cells are also defective in cell division, hence, not able to derive normal germ cells (Megosh et al., 2006)

1.2.3 microRNA functions in neurogenesis

miRNAs have been demonstrated to be essential for neural development Previous reports have highlighted the abundant and diverse expression of miRNAs in the central nervous system (CNS) (Krichevsky et al., 2003; Sempere et al., 2004; Miska

et al., 2004a; Wheeler et al., 2006; Kim et al., 2004; Kosik, 2006) Mammalian brain tissues express about 70% of experimentally detectable miRNAs, many of which are developmentally regulated (Krichevsky et al., 2003; Sempere et al., 2004; Miska et al., 2004a; Wheeler et al., 2006; Kim et al., 2004; Kosik, 2006) Recently, more than

400 new miRNAs were cloned from human and chimpanzee brains and interestingly, many of them are specific to mammals or to primates only (Berezikov et al., 2006) This study suggests that new miRNAs arose during the brain evolution and these miRNAs should be important for the more sophisticated development and functions

of the brains in higher organisms (Berezikov et al., 2006) Moreover, evidences for

the role of miRNA in neurogenesis have been provided by the loss of Dicer

experiments in different species In maternal-zygotic zebrafish dicer mutants,

deficiency in Dicer-mediated biogenesis of miRNAs leads to severe defects in brain

morphogenesis (Giraldez et al., 2005) Similarly, the loss of Dicer in sca3-mutant Drosophila enhances neurodegeneration (Bilen et al., 2006) Specific knockdown of

Dicer in mouse midbrain dopaminergic neurons resulted in progressive loss of these cells (Kim et al., 2007)

The most abundant and well-studied miRNA in the central nervous system (CNS) is miR-124 This miRNA is upregulated during CNS development and during differentiation of adult neural stem cells and ESC-derived neural progenitor cells (Smirnova et al., 2005; Krichevsky et al., 2003; Krichevsky et al., 2006; Cheng et al., 2009) The expression pattern of miR-124 is regulated by the RE-1 silencing

transcription repressor (REST) which binds to the promoter of mir-124 gene and

suppresses the expression of miR-124 in neural stem cells (Conaco et al., 2006; Cheng et al., 2009) Several reports have demonstrated the important role of miR-124

in neuronal identity, neuronal differentiation and CNS development Ectopic expression of miR-124 in Hela cells represses the expression of ~100 non-neuronal transcripts with seed matches to miR-124 and drives the gene expression profile towards a brain-like transcriptome (Lim et al., 2005) Consistent with this result, knockdown of miR-124 in mature neurons upregulates multiple non-neuronal mRNAs (Conaco et al., 2006) A few targets of miR-124 have been validated carefully and

correlated to its function Makeyev et al showed that miR-124 targets ptbp1, which

encodes a global repressor of alternative splicing therefore enables neuronal specific splicing during neural cell differentiation (Makeyev et al., 2007) In the spinal cord of

chick embryos, miR-124 is found to repress the anti-neuronal gene scp1 to modulate neurogenesis timing (Visvanathan et al., 2007) Most recently, Chen et al reveals that

miR-124 drives neural stem cells to cell cycle exit and through sequential stages of

differentiation in adult mouse subventricular zone by suppressing jag1, dlx2 and sox9

(Cheng et al., 2009)

Recent studies have also elucidated the contribution of other miRNAs in various aspects of neural development For example, mir-9a regulates the organizer function

of the zebrafish midbrain-hindbrain boundary (Leucht et al., 2008) In C elegans,

lsy-6 and mir-273 determine the cell fate of chemoreceptor neurons (Johnston and Hobert, 2003) miR-7 regulates differentiation of photoreceptor neurons in

Drosophila (Li and Carthew, 2005) The miR-200 family regulates the terminal

differentiation of olfactory neurons in both mouse and zebrafish (Choi et al., 2008) In addition, miRNAs play important roles in neuronal function and survival In

Drosophila, miR-8 prevents neuronal apoptosis by suppressing the proapoptotic gene atrophin (Karres et al., 2007) In mature rat neurons, mir-134 localizes to dendrites and regulates spine size (Schratt et al., 2006) In C elegans, mir-1 regulates MEF-2

dependent retrograde signalling at the neuromuscular junctions (Simon et al., 2008)

1.3 The role of microRNAs in diseases

The essential roles of miRNAs in various physiological processes suggest that their absence or misexpression would affect development and result in pathological conditions Indeed, miRNAs have been implicated in many human diseases, including tumorigenesis, diabetes, obesity, neurological disorders, cardiac hypertrophy and

myopathies, etc (Garzon et al., 2006; Poy et al., 2009; Xie et al., 2009; Jin et al., 2004; Care et al., 2007; Zhao et al., 2007)



Differences in miRNA expression profiles define the signatures of various cancers, and miRNA dysregulation can lead to all the hallmarks of cancer (Zhang et al., 2007; Garzon et al., 2006) In 2005, Lu et al published the first report suggesting a classification of cancers based on miRNA expression patterns They performed microarray profiling of miRNAs in a large number of human tumour samples and found that the expression pattern of miRNAs clearly reflects the cancer types, the tumour origins and differentiation stages (Lu et al., 2005) This report was supported

by many subsequent studies on miRNA profiles of cancers, signifying miRNAs as a powerful set of new biomarkers for cancer diagnosis Further more, many miRNA genes are located in regions associated with frequent rearrangement, deletion or amplification in cancers miRNAs are known to be both regulators and targets of oncogenes and tumor-suppressor genes (Garzon et al., 2006) The first examples of tumour-suppressing miRNAs are miR-15 and miR-16, both of which are frequently deleted in B-cell chronic lymphocytic leukemia (Cimmino et al., 2005) These two miRNAs directly repress the anti-apoptotic factor Bcl-2, and promote apoptosis (Cimmino et al., 2005) The family of let-7 miRNAs are also postulated to be tumour suppressors (Takamizawa et al., 2004) let-7 miRNAs are downregulated in several types of cancers and it has been used as a prognostic marker (Takamizawa et al., 2004; Johnson et al., 2005; Yu et al., 2007) They suppress the proliferation of cancer cells, particularly of tumour-initiating cells, by blocking two potent oncogenes, Ras and HMGA2 (Johnson et al., 2005; Yu et al., 2007) Similarly, the miR-34 family is reported recently as a key downstream effector of the p53 tumour suppressor, mediating p53 function in inducing apoptosis and cell cycle arrest (He et al., 2007;

Bommer et al., 2007; Chang et al., 2007; Tarasov et al., 2007) In contrast of the tumour suppressing miRNAs, oncogenic miRNAs are observed with anti-apoptotic and pro-proliferation properties The miR-17-92 cluster is an example of miRNAs with these properties (Matsubara et al., 2007; Lu et al., 2007) This miRNA cluster is frequently amplified in several human cancers and their endogenous expression is

driven by the powerful oncogene c-myc (Venturini et al., 2007; O'Donnell et al.,

2005) miR-20, a member of the cluster, targets and suppresses the transcription factor E2F family (O'Donnell et al., 2005) Interestingly, E2F proteins in turn promote the expression of miR-17-92 cluster (Sylvestre et al., 2007)

Unlike for cancers, it is much more difficult to identify miRNA profiles that are representative for neurological disorders because it is not possible to collect disease and normal tissue samples from the same patients with these disorders Moreover, sporadic cases of neurological disorders like Alzheimer’s diseases (AD) and amyotrophic lateral sclerosis (ALS) arise by different causes and probably progress through different mechanisms Recent studies on miRNAs in AD patients reveal different sets of AD-associated miRNAs with little overlap In stead, miRNAs in these diseases are characterized with more success in animal genetic models For example,

dicer-1 mutation in a Drosophila model of spinocerebellar ataxia shows enhanced polyQ-mediated neurodegeneration (Bilen et al., 2006) Also in this model, bantam

miRNA was identified as a key protective factor that is able to suppress both polyQ and tau toxicity (Bilen et al., 2006) Another miRNA, miR-133, is downregulated in the mouse model of Parkinson’s disease where it acts in a negative feedback loop with

a transcription factor, Pix3, for the maintenance of dopamine neurons in the midbrain (Kim et al., 2007) miRNAs also associate with the key proteins involved in

neurological diseases Argonaute, the core component of RISC, is found to interact directly with FMR1, a protein that is often mutated in fragile X syndrome (Jin et al., 2004) miR-29a/b is downregulated in AD patient samples and is found to target BACE1 which mediates the cleavage of amyloid precursor protein in the major pathogenic pathway of AD (Hebert et al., 2008) Further understanding of the function and the regulation of miRNAs in both physiological and pathological conditions would provide new approaches for the diagnosis and therapies against these diseases

1.4 The expression and known functions of miR-125

The discovery of miRNAs started with lin-4 miRNA which suppresses lin-14 and modulates timing of developmental stages in C elegans (Olsen and Ambros, 1999) miR-125 is a homolog of lin-4 (82% identical) and is highly conserved from flies to

humans (100% identical) In species that have additional miRNAs with nearly identical sequence to miR-125, the conserved miR-125 is called “miR-125b” miR-

125 or miR-125b have been found in 42 species (miRBase Release 13, 2009) The similar miRNAs, miR-125a (with two nucleotides different from miR-125b) is annotated in 16 species and miR-125c (with one nucleotide different from miR-125b)

is found only in zebrafish 125b has the same seed sequence as 125a, 125c and miR-351 hence, they are classified into the same family and predicted by TargetScan to have the same targets (Friedman et al., 2009) Apart from the region

miR-corresponding to the mature miR-125b sequence, other regions of mir-125b precursors are not highly conserved Some species have multiple copies of pre-mir- 125b with variations in their sequences and encoded by different loci e.g humans and mouse express two copies of pre-mir-125b; zebrafish expresses three copies of pre-

mir-125b (miRBase Release 13, 2009) In several species including human and mouse, both strands of pre-mir-125b are processed into a mature miRNAs: the one

with conserved miR-125 sequence is called miR-125b-5p; the antisense one is called

miR-125b-3p or miR-125b* (miRBase Release 13, 2009) In Drosophila, mir-125 is transcribed together with let-7 and mir-100 then processed by Dicer into individual mature miRNAs (miRBase Release 13, 2009) In other species, some but not all mir- 125a/b/c genes are also found in close proximity to let-7 genes; however, it is not known whether mir-125b and let-7 are transcribed together in these cases (miRBase

Release 13, 2009)

The expression of miR-125 is often upregulated during development C elegans lin-4 and Drosophila miR-125 are found only in post-embryonic stages (Olsen and Ambros, 1999; Caygill and Johnston, 2008) lin-4 is expressed in C.elegans from the first larval stage and upregulated quickly to “turn off” the expression of its target lin-

14 (Olsen and Ambros, 1999) Drosophila miR-125 expression is initiated from the

L3 pupal transition and reaches the highest level in pupae during metamorphosis (Caygill and Johnston, 2008) In contrast, the expression of miR-125 in vertebrate species is observed from early embryogenesis Zebrafish miR-125a and miR-125b are expressed from 24 hours post fertilization (hpf), upregulated until 96 hpf and remain high in the adults (Wienholds et al., 2005) miR-125a/b expression is enriched in the brain, spinal cord and cranial ganglia of zebrafish embryos (Wienholds et al., 2005) During mouse embryogenesis, miR-125b expression increases gradually from embryonic day twelve till birth (Miska et al., 2004b) The highest expression of miR-

125b was observed by in situ hybridization in the midbrain-hindbrain boundary of

mouse embryos (Ason et al., 2006) However, miR-125b expression in chick embryos

observed by the same method is a lot more ubiquitous with enrichment in the brain,

motor horns, gonads and pharyngeal arches (Ason et al., 2006) Further more, microarray and sequencing data from different reports show that mammalian miR-125a/b are expressed in many different tissues and cell types with the highest expression in the brain, skin, testis, ovary, spleen and thyroid (Landgraf et al., 2007) Smirnova et al reveals that miR-125b is much higher in neurons than in astrocytes (Smirnova et al., 2005) Other studies show that miR-125a/b is upregulated during neural differentiation of mouse embryonic stem cells (Krichevsky et al., 2006), and upon RA treatment of embryonic carcinoma cells (Sempere et al., 2004) and of neuroblastoma SK-N-BE cells (Laneve et al., 2007), suggesting a role of miR-125b in differentiation In contrast, miR-125b is downregulated in splenocytes of mice injected with lipopolysacharide or in the lung of mice exposed to cigarette smoke, suggesting a role of miR-125b in the innate immune response and stress responses (Tili et al., 2007; Izzotti et al., 2009) Of note, the expression pattern of miR-125a and miR-125b are not identical miR-125b is often expressed at a higher level than miR-125a in neural cells/tissues and the upregulation of miR-125b during differentiation is also more profound (Wienholds et al., 2005; Sempere et al., 2004)

The expression pattern of miR-125 family suggests its important role during

development As described earlier, miR-125b homolog lin-4 plays a critical role in developmental timing of C elegans (Olsen and Ambros, 1999) In Drosophila, deletion of mir-125 and let-7 together does not alter developmental timing but

affecting metamorphosis in the pupal stage (Caygill and Johnston, 2008) This mutant displays defective cell cycling in the wings and defective cell differentiation in the neuromuscular junction of the abdominal muscles; however, it is not clear whether

mir-125 or let-7 or both are responsible for these phenotypes (Caygill and Johnston,

2008) Loss-of-function study has not been performed on miR-125 in vertebrate species However, the function of this miRNA has been investigated in several mammalian cell culture systems Mizuno et al demonstrated that miR-125b suppresses proliferation and thereby inhibiting the differentiation of the mouse mesenchymal stem cells ST2 (Mizuno et al., 2008) When overexpressed together with miR-125a and miR-9, miR-125b also reduces the proliferation of the human neuroblastoma SK-N-BE cells by suppressing truncated tropomyosin kinase C (Laneve et al., 2007) The most well known target of miR-125b in mammalian cells is

lin-28, that is also targeted by lin-4 in C elegans (Wu and Belasco, 2005) Lin-28

mRNA contains two binding sites for miR-125b in its 3’ UTR (Wu and Belasco, 2005) The protein level of lin-28 is suppressed by miR-125b through both

translational inhibition and deadenylation of lin-28 mRNA in the mouse embryonic

carcinoma P19 cells (Wu et al., 2006) Indeed, lin-28 is a reprogramming factor that, together with Oct4, Sox2 and Klf4, is able to turn skin cells into induced pluripotent stem cells (Takahashi and Yamanaka, 2006) lin-28 is also known to block the

processing of let-7 precursor in mouse (Viswanathan et al., 2008) During the

differentiation process from embryonic stem cells (ESCs) to neural progenitor cells, miR-125b is upregulated and gradually reduces the level of lin-28 that leads to an

increase in let-7 expression (Rybak et al., 2008) This suggests that miR-125b may be

important to suppress pluripotency and promote differentiation of ESCs

miR-125b is also associated with cancers It is downregulated in ovarian carcinoma and thyroid carcinoma (Nam et al., 2008; Iorio et al., 2007; Volinia et al., 2006) but upregulated in pancreatic cancer, oligodendroglial tumors, prostate cancer,

myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) (Bloomston

et al., 2007; Nelson et al., 2006; Shi et al., 2007; Bousquet et al., 2008; Sonoki et al., 2005a) miR-125b was shown to suppress cell cycling in hepatocellular carcinomas (Li et al., 2008) but to promote proliferation of prostate cancer cells (Shi et al., 2007) Therefore, miR-125b may have a dual role in tumorigenesis, acting as both an oncogene and a tumour suppressor

1.5 Motivation of the thesis

As we reviewed, the role of miRNAs in development and diseases has been described extensively by the recent literatures However, there are a lot more to learn about miRNA functions, especially how they are integrated into the gene regulatory network

of each cellular process We began this project with the desire to understand miRNA expression in human neuronal differentiation Although prior reports have demonstrated the roles of a few miRNAs in mammalian neurogenesis, they hardly examine the functions of these miRNAs in human cells or tissues To investigate

miRNA expression in human neuronal differentiation, we started with a simple in vitro model, the human neuroblastoma SH-SY5Y cells When sequentially treated with all-trans-retinoic acid (RA) and brain-derived neurotrophic factor (BDNF), SH-

SY5Y cells give rise to fully differentiated neuron-like cells (Encinas et al., 2000) These differentiated SH-SY5Y cells are withdrawn from the cell cycle, express various neuronal markers, and exhibit carbachol-evoked noradrenaline release (Encinas et al., 2000) Moreover, as no glial cell is derived by this process, it is a robust and homogenous model system for investigating neuronal differentiation (Encinas et al., 2000) Using microarrays and Northern blots, we identified a group of miRNAs that are significantly upregulated in differentiated SH-SY5Y cells

Subsequently, we tested the function of the candidate miRNAs by a neurite outgrowth assay and found that one of these miRNAs, miR-125b, significantly enhances differentiation and neuronal morphogenesis of SH-SY5Y cells We decided to focus our further study on miR-125b because many prior reports have suggested a role of miR-125b in vertebrate development (as reviewed in the previous session) but none of them have studied these functions in depth From our preliminary data, we hypothesized that (i) miR-125b plays an important role in neuronal differentiation of human cells; (ii) this function may be conserved in other vertebrate species; (iii) the function of miR-125b may be mediated by multiple targets or pathways Therefore,

we are interested in understanding the functions and the mechanism of action of 125b

miR-1.6 Objectives of the thesis

1.6.1 To demonstrate the function of miR-125b in neuronal differentiation of human

neural cells and to identify miR-125b downstream targets in these cells

1.6.2. To investigate the function and the targets of miR-125b in vivo, using

zebrafish as a model

1.7 Project workflow

According to the objectives for the project, we performed functional analysis of

miR-125b both in vitro and in vivo First, we demonstrated the important role of miR-miR-125b

this miRNA in both spontaneous and induced differentiation of SH-SH5Y cells 125b is also upregulated during differentiation of human neural progenitor ReNcell

miR-VM cells, and miR-125b ectopic expression significantly promotes neurite outgrowth

changes in gene expression following miR-125b ectopic expression in SH-SY5Y cells miR-125b represses 188 genes that are predicted by TargetScan 5.1 to be the direct targets of miR-125b Pathway analysis suggests that a subset of miR-125b-repressed targets antagonize neuronal genes in several neurogenic pathways, thereby mediating the positive effect of miR-125b on neuronal differentiation We have further validated the binding of miR-125b to the microRNA response elements of nine selected mRNA targets and confirmed the binding specificity for three targets by mutagenesis Together, these data demonstrate for the first time the important role of miR-125b in human neuronal differentiation

Second, we examined the function of miR-125b in zebrafish development Morpholino-mediated knockdown of miR-125b leads to severe defects in zebrafish embryos, where neural cell death is the most apparent phenotype This phenotype resembles Mdm2 morphants in which the p53 pathway is activated; hence, we postulated that the function of miR-125b in zebrafish embryos is mediated by the p53 pathway Interestingly, p53 is predicted to be a target of miR-125b in both zebrafish and humans We validated this prediction by a luciferase reporter assay and mutagenesis Our data reveal that miR-125b binds directly to the 3’ UTR of human and zebrafish p53 mRNAs, and represses p53 protein levels in a manner dependent on its binding site in the p53 3’UTRs miR-125b-mediated regulation of p53 is critical for modulating apoptosis in human cells and in zebrafish embryos during development and during the stress response We also identified seven additional targets of miR-125b in the p53 network and a feedback loop from p53 that suppresses miR-125b expression in human cells

Microarray analysis identified 388 genes downregulated

by miR-125b in SH-SY5Y cells

Prediction of p53 as a target of miR-125b

in humans and zebrafish

Project flowchart

Characterization of miR-125b functions in neuronal

differentiation of SH-SY5Y cells by neurite outgrowth

assay and immunostaining

Characterization of miR-125b expression and functions in neuronal differentiation of the neural progenitor RVM cells by qRT-PCR and neurite outgrowth assay

Microarray and Northern blot analysis identified six

miRNAs upregulated during neuronal differentiation in

the neuroblastoma SH-SY5Y cells

Functional screening by neurite outgrowth assay showed

that miR-124a and miR-125b promotes neurite

outgrowth of SH-SY5Y cells

Identification of miR-125b putative targets among the

388 genes by TargetScan

Selection of nine miR-125b putative targets that may

antagonize known neurogenic pathways

Validation of the nine targets by qRT-PCR, mutagenesis

and luciferase reporter assays

Validation of miR-125b binding to p53 mRNAs by luciferase reporter assays, mutagenesis and western blots

Demonstrating the regulation of p53 and apoptosis by miR-125b in SH-SY5Y cells and human lung fibroblasts

Knockdown of miR-125b in zebrafish embryos by morpholinos

Demonstrating the regulation of p53 and apoptosis by miR-125b in zebrafish

embryos

Demonstrating miR-125b function in regulating p53 and apoptosis during stress

response Mapping the human miR-125b network

CHAPTER 2 – MATERIALS AND METHODS

2.1 Cell culture and differentiation condition

Human HEK-293T cells, human neuroblastoma SH-SY5Y cells, p53-null human lung carcinoma H1299 cells, mouse Swiss-3T3 cells and human lung fibroblast cells were maintained in Dulbecco's Modified Eagle medium (DMEM) or RPMI media, supplemented with 10% heat-inactivated fetal bovine serum (GIBCO), 110 mg/L sodium pyruvate (GIBCO), 2 mM L-glutamine (GIBCO) and 1% penicillin-streptomycin (Invitrogen)

For differentiation, SH-SY5Y cells were seeded on collagen-coated plates (BD Biosciences) at an initial density of 104 cells/ cm2 All-trans-retinoic acid (Sigma) was

added at a final concentration of 10 μM on the next day after plating After five days, the cells were washed three times with DMEM and incubated with 50 ng/ml brain-derived neurotrophic factor (Sigma) in growth medium without serum for seven days

ReNcell VM (RVM) cells were cultured in laminin-coated plates in DMEM/F12 (1:1) medium (Invitrogen), supplemented with 10% B27 medium (Invitrogen), 10 μg/ml Gentamycin (GIBCO), 10 units/ml Heparin (Sigma), 20 ng/ml epidermal growth factor and 10 ng/ml basic fibroblast growth factor (Invitrogen) For differentiation, the growth medium was replaced with the Neurobasal medium (Invitrogen), supplemented with 10% B27 medium (Invitrogen), 10 μg/ml Gentamycin (GIBCO) and 10 units/ml Heparin (Sigma)

2.2 miRNA expression profiling

Total RNA samples were extracted from untreated SH-SY5Y cells, cells treated with

RA for 5 days and from cells treated subsequently with BDNF in serum-free medium for 7 days Small RNA was purified, labeled and subjected to an oligonucleotide-based microarray as previously described (Baskerville and Bartel, 2005) Briefly, two 32

P-labeled RNA markers of 18 nt and 24 nt were co-loaded with total RNA samples and used as indicators to identify the small RNA population on the gel separating 100

μg of total RNA 18-24 nt RNAs were gel-purified and sequentially ligated to 3’- and

a 5’-end adaptors Ligated products were gel-purified, reverse transcribed, amplified and labeled with Cy3 The labeled sense strand was then gel-purified and applied to the array A set of synthetic reference oligonucleotides (with a uniform amount of oligonucleotides corresponding to every probe) was processed similarly but labeled with Cy5 These Cy5-labeled reference oligonucleotides were applied concurrently with the Cy3-labled samples to a DNA oligonucleotide-based array, serving as internal hybridization controls This array (provided by the Bartel laboratory at Whitehead institute) contains ~600 DNA probes, including probes for

PCR-175 human miRNAs (Baskerville and Bartel, 2005) The obtained signals were normalized to the total intensity of all non-cognate probes (corresponding to the nematode miRNAs that are not conserved in human) Subsequently, signals from the biological samples were normalized to the corresponding references as Cy3/Cy5 ratios The final reading was the average normalized intensity of four replicates (two biological replicates each with two technical replicates) Microarray data was deposited into Gene Expression Omnibus (accession number: GSE14787)

2.3 Northern blot analysis

10 – 40 μg of each total RNA sample and a 33P –labeled Decade® RNA marker (Ambion) were separated on a 15% denaturing gel, transferred to a Genescreen Plus membrane (PerkinElmer), UV-cross-linked and baked at 80ºC for 30 minutes DNA probes with the sequences complementary to the miRNAs were synthesized (Invitrogen) and labeled with 32P--ATP (Amersham) U6 RNA and 5S RNA probes were used to determine loading equity The probe sequences are provided in Table 1 The membrane was prehybridized in PerfectHyb buffer (Sigma) with 1 mg of freshly added sheared salmon sperm DNA (Sigma) for two hours at 48 ºC Subsequently, the labeled probes were added; hybridization was carried out overnight at 48 ºC The membrane was then washed and developed according to the Bartel laboratory Northern blot protocol (http://web.wi.mit.edu/bartel/pub/)

Table 1 Sequences of Northern blot (NB) probes and morpholinos

Sequences are written from 5’ to 3’ Lower case letters indicate mismatches in the misMO sequence in comparison to m125bMO sequence

microRNA/Control RNA Probe sequences

miR-124 NB probe GGCATTCACCGCGTGCCTTA

miR-125b NB probe TCACAAGTTAGGGTCTCAGGGA

miR-199a NB probe GAACAGGTAGTCTGAACACTGGG

miR-199a* NB probe AACCAATGTGCAGACTACTGTA

miR-214 NB probe CTGCCTGTCTGTGCCTGCTGT

miR-189 NB probe ACTGATATCAGCTCAGTAGGCAC

miR-27a NB probe GCGGAACTTAGCCACTGTGAA

miR-143 NB probe GAGCTACAGTGCTTCATCTCA

miR-106a NB probe GCUACCUGCACUGUAAGCACUUUU

2.4 Transfection and drug treatments

For immunostaining, SH-SY5Y cells (passage number less than 25) were seeded as 80,000 cells/well in collagen-coated 12-well plate (BD Bioscience) On the next day, using 4 μl lipofectamineTM2000 reagent (Invitrogen) per well according to the manufacturer’s instruction, the cells were transfected with one of these RNA oligos at

80 nM final concentration: BlockITTM fluorescent oligo (Invitrogen), negative control duplexes, miRNA duplexes or miRNA antisense oligonucleotides (Ambion) After five hours, the transfection medium was replaced by fresh growth medium either with

or without 10 μM RA 125,000 ReNcell VM (RVM) cells were transfected in suspension with 80 nM RNA oligos in the same manner as for SH-SY5Y cells After five hours of transfection, the medium was changed to fresh RVM growth medium or differentiation medium and the cells were plated in laminin-coated plates

For Western blots, H1299 cells, SH-SY5Y cells and human lung fibroblast cells were transfected in suspension with 4 x 105 cells per well in 6-well plates using lipofectamin-2000 (Invitrogen) Plasmids (human/zebrafish wild-type or mutant p53 constructs) were transfected into H1299 cells at a final concentration of 0.5 g/ml miRNA duplexes and antisense oligonucleotides were transfected at a final concentration of 80 nM and 100 nM respectively (unless otherwise stated) p53 siRNA (Dharmacon) was transfected at 60 nM final concentration

For luciferase reporter assays, HEK-293T cells were seeded overnight in a black well plate at a density of 50,000 cells per well and transfected with 10 ng of each psiCHECK-2 construct and 10 nM miRNA duplexes or 100 nM miRNA antisenses using lipofectamin-2000 (Invitrogen) After 48 hours, the cell extract was obtained;

96-firefly and Renilla luciferase activities were measured with the Dual-Luciferase

reporter system (Promega) according to the manufacturer’s instructions

1-(5-isoquinoline sulphonyl)-2-methyl piperazine (H-7) and etoposide (Sigma) were dissolved in water and dimethyl sulfoxide (DMSO) respectively SH-SY5Y cells (untransfected or 24 hours after transfection with miRNA duplexes) were treated with

10 μM H-7 or 10 μM etoposide for 24 hours Control cells were treated with water or DMSO respectively

2.5 Immunostaining and high-content screening of cells

Four days after transfection, SH-SY5Y cells or RVM cells were fixed with 4% paraformaldehyde for 15 minutes, followed by three washes with phosphate buffer saline (PBS) After one-hour blocking with 0.2% TritonX100 and 3% goat serum in PBS, the cells were incubated with primary antibodies overnight at 4 ºC The primary antibodies used in this study include: mouse monoclonal III-tubulin antibody (Abcam, 1:1000 dilution), mouse monoclonal Map2ab (Sigma, 1:1000 dilution), mouse monoclonal pan-axonal neurofilament antibody (Covance, 1:1500 dilution), goat monoclonal synaptotagmin V antibody (Santa Cruz 1:1000 dilution), rabbit polyclonal musashi-1 antibody (Abcam 1:1000 dilution), rabbit active-caspase-3 antibody (BD Biosciences, 1:500 dilution) Subsequently, the cells were washed with PBS for three times, and then incubated with Alexa Fluor® 488 goat-anti-mouse, Alexa Fluor® 568 goat-anti-rabbit or Alexa Fluor® 488 donkey-anti-goat secondary antibody (Invitrogen) for an hour Hoechst dye (Invitrogen) was added for five minutes The cells were then washed with PBS for three times For high-resolution imaging, the cells were observed with a Zeiss DUO inverted confocal microscope

(Carl Zeiss Vision GmbH) For quantitative imaging, fluorescent images of the cells were collected automatically by the Cellomics® high content screening system using a 10x objective lens For neurite outgrowth assays, images of III-tubulin and Hoechst staining were analyzed using the Neuronal Profiling BioApplication software (Cellomics) Differentiated SH-SY5Y cells with neurite outgrowth were defined as

III-tubulin positive cells with neurites longer than 30 m For RVM cells, the percentage of differentiated cells with neurite outgrowth was defined as the percentage of III-tubulin positive cells having neurites longer than 20 m (RVM cells are smaller than SH-SY5Y cells so we applied a lower threshold of neurite length) For quantification of neuronal marker staining, the images were analyzed by Target Activation BioApplication (Cellomics) In all the high content screening assays, a cell was considered as positive for a specific staining only if its fluorescent intensity was equal or higher than the mean intensity plus two times the standard deviations of the respective negative control replicates

2.6 Immunostaining of zebrafish embryos

Embryos were dechorionated and fixed in 4% paraformaldehyde at 4oC overnight After three washes with PBDT (PBS containing 2% BSA, 1% DMSO and 0.5% Triton X-100), the embryos were incubated with cold acetone at -20ºC for 20 minutes followed by three additional PBDT washes Subsequently, the embryos were blocked with 1x blocking buffer (Roche) for one hour then incubated with mouse anti-acetylated tubulin monoclonal antibody, 1:200 (Sigma) overnight at 4ºC The embryos were washed extensively in PBDT (30 minutes x 6 times) and incubated with Alexa Fluor 568 goat-anti–mouse IgG antibody, 1:200 (Molecular Probe) for 4

hours After 5 washes in PBDT (30 minutes each), the embryos were re-fixed in 4% paraformaldehyde at 4oC overnight

2.7 Image acquisition and microscope settings

Fluorescent images of the TUNEL assays and the acetylated tubulin staining were obtained with a LSM510 confocal laser-scanning microscope (Carl Zeiss Vision GmbH) A bright-field image was acquired at the same time as the fluorescent image Projection of image stacks was made by the Zeiss image browser Images were then imported into Adobe Photoshop for cropping, resizing, and orientation Contrast and brightness were adjusted equally for all images of the same figure

Images of live embryos were obtained by a SZX12 stereomicroscope (Olympus) and a MagnaFIRE SP camera (Olympus) The embryos were mounted in 3% methyl-cellulose Images were acquired with a 65x objective, at a resolution of 1280x1024, with ~100 ms exposure and 8 bit depth at room temperature The image set of each embryo was combined, resized, cropped, and oriented using Adobe Photoshop

2.8 Whole mount in situ hybridization

Whole mount in situ hybridizations with double-Dig-labeled miR-125b miRCURY™

LNA probe (Exiqon) on zebrafish embryos were performed essentially as described (Wienholds et al., 2005) Modifications to the protocol include an incubation of the

19, 22 and 24hpf embryos for 30 seconds and of the 30hpf embryos for 1 minute with PCR-grade proteinase K (Roche) after fixing The hybridization mix was prepared by adding 20 pmol of miR-125b doubled-labeled LNA probe to every 1 ml of hybridization solution The hybridization temperature used was 20qC below the