Towards inhibiting Xiwi and analysis of - PDF document

1 Towards inhibiting Xiwi and analys
1 Towards inhibiting Xiwi and analysis of transcripts associated with Xiwi from X.tropicalis Ignatius Ang Undergraduate Advisor: Dr. Nelson C. Lau Brandeis University Department of Biology BIOL99: Research Manuscript Submitted 05/01/2015 2 Towards inhibiting Xiwi and analysis of transcripts associated with Xiwi from X.tropicalis Table of Contents Chapter 1 ……………… Developing RNA aptamers for inhibition of piRNA pathways in Xenopus 05/15/14 – 08/09/14 Pages 4-13 Chapter 2 ………...........Validation of original Xiwi, and Xili CLIP data using RIP, RT-QPCR 08/15/14 – 12/18/14 Pages 14-27 Chapter 3 ………………. Optimizing the XIWI RIP assay with improved IP and RT-QPCR conditions 01/21/15 – 04/30/15 Pages 28-44 3 Acknowledgements I would like to express my gratitude for the help and support to the following people that have made this study possible. I would like to thank Dr. Nelson C. Lau my undergraduate advisor for providing me guidance and support. He helped me think critically about the subject matter, and continually pushed me to excel. Yuliya Sytnikova has been an excellent mentor, who taught me all the research techniques and provided great advice. She also conducted the Xenopus CLIP experiment that these studies are based off. Dr. Chirn has been instrumental in analyzing the CLIP data. Thank you, Joseph Clark for providing advice on experimental procedures, especially QCPR. Nicholas Clark and Thomas Brands completed the SELEX experiments, and proved the aptamers. 4 Chapter 1: Developing RNA aptamers for inhibition of piRNA pathways in Xenopus I. Abstract Piwi proteins are a subfamily of the Argonaute proteins (AGO) that interact with small RNAs to form an RNA induced silencing complexes (RISC). AGO and Piwi protein have two major protein motifs- the PAZ and PIWI.

The PAZ domain binds to the 3ˈ end of s
The PAZ domain binds to the 3ˈ end of small RNAs and the PIWI domain has an RNase H fold that can cleave a target RNA that is complimentary to their bound RNA [8]. This study aims to develop an RNA inhibitor that binds to the PAZ domain of Xiwi - a Piwi homolog found in Xenopus. RNA aptamers were selected with a recombinant Xiwi PAZ domain using systematic evolution of ligands by exponential enrichment (SELEX). A library of RNA containing 1015 random 50 nt RNA sequences were selected for Xiwi PAZ domain binding. I selected 9 candidate RNA aptamers and generated RNAs by in vitro transcription. I characterized RNA aptamer binding ability with an electrophoresis mobility shift essay (EMSA) and in-vitro in Xenopus tropicalis egg extract to check for piRNA degradation due to displacement. EMSA showed that aptamers 2, 5, 8 and 9 bound well to the Xiwi PAZ domain. Aptamers 5 and 9 failed to displace piRNA in complex with Xiwi in X.Tropicalis egg extract. An RNA inhibitor of Xiwi serves as an important molecular tool that can be used to further characterize the functions of Xiwi. II. Introduction The highly conserved argonaute protein (AGO) family binds to small RNA’s to from RNA induced silencing complexes (RISC). AGO’s bind to micro RNAs (miRNAs), short interfering RNAs siRNAs, or PIWI-interacting RNAs (piRNAs) that guide them to their target in RISC. Upon RISC recognition of targets, Argonaute protein ional inhibition, RNA desTableilization and chromatin remodeling through DNA methylation and/inhibits their expression by either cleaving them with its Slicer endonuclease activity, or by inducing translator histone modification [4]. PIWI (P-element induced wimpy testes) proteins are a subfamily of AGO that bind to piRNAs 20-32nt long and are important in the silencing of transposable elements [4]. Transposable elements are selfish nucleic acid elements that can be harmful to the genome when accumulated. PIW have been shown to be key in the germ line development and maintenance. piRNAs are generated u

sing two distinct pathways: the primary
sing two distinct pathways: the primary pathway and the secondary amplification pathway (ping pong amplification loop). The primary pathway generates piRNA from piRNA clusters that are composed of multiple transposon fragments. In the secondary amplification pathway mature sense primary piRNA guide PIWI proteins to complementary antisense sequences and cleave it with slicer to generate a new 5´ end. The 5´ end binds to a new PIWI protein and has its 3´ end trimmed to form a mature antisense secondary piRNA. Antisense secondary piRNA can now guid PIWI proteins to target the sense sequences and cleave it. [6] AGOs are composed of two lobes the N-PAZ and MID-PIWI domains. The PAZ domain anchors to the 3´ end of the RNA, MID domains anchors the 5´ end of the RNA, and the PIWI domain has an RNaseH like endonuclease function [3]. Systematic Evolution of Ligands by Exponential Enrichment (SELEX) is a method used to identify single-stranded nucleic acids that have a high affinity for protein biding 5 [1]. SELEX uses a pool of single stranded nucleic acid consisting of a random region with flanking primers for amplification and transcription. The library is selected for its binding ability to a protein of interest [1]. One round of SELEX includes selecting RNA that bound to the protein, amplification via RT-PCR and transcription. Multiple rounds of selection and amplification enrich the RNA that binds to the protein. Electrophoreses mobility shift assay (EMSA) is used to characterize protein-nucleic acid interactions. EMSA differentiates unbound nucleic acid from nucleic acid in complex with protein by mass, shape and charge [5]. The binding ability of RNA aptamers can be characterized with EMSA. Xenopus have 3 homologs of PIWI: XIWI 1a, XIWI 1b, and XILI. The abundance of PIWI proteins in Xenopus eggs and oocytes, make them ideal for studying PIWI pathways [7]. Egg extracts are noted for their ability to recapitulate many cell-cycle processes in vitro [2]. Xiwi protein is very abundant and sTa

ble in Xenopus oocytes and eggs it is
ble in Xenopus oocytes and eggs it is difficult to deplete it via RNAi or morpholino strategies. Knockout Xenopus are extremely difficult to create. A simple way to knockdown Xiwi would be the use of an RNA inhibitor. In this study SELEX was used to develop an RNA inhibitor for Xiwi. An RNA library of 50 random nt and flanking primers of 20nt was selected for binding to Xiwi PAZ domain. The RNA pool was subjected to 8 rounds of selection and amplification to enrich for RNAs that bind to the Xiwi PAZ domain. 9 candidate RNA aptamers were selected and their binding abilities were characterized in EMSA and X.Tropicalis egg extract. EMSA showed that aptamers 2, 5, 8, and 9 bound well to the Xiwi PAZ domain. We propose that piRNA degradation would occur if it was displaced by the RNA aptamer. Aptamers 2, 5, and 9 were tested in X.Tropicalis egg extract; however no piRNA degradation was detected, showing that the aptamers are unable to displace piRNA from its complex with Xiwi. No suiTable RNA inhibitor was found. An RNA inhibitor would allow further study and characterization of the Piwi pathway and function. III. Materials and Methods T4 vectors of aptamers 1,2,3,4, and 8 (ampR) and TOPO vectors of aptamers 5, 6, 7 and 9 (kanR) cloned by Nicholas Clark were transformed into DH5α competent cells. Cells were plated and cultured (5ml) in respective resistance LB. 5ml cultures were minipreped with BIOBASIC DNA Columns and measured with nanodrop. PCR was done using plasmids from miniprep and primers below in standard conditions. Fwd primer GTAATACGACTCACTATAGGGAGAATTCAACTGCCATCTA Rev primer ACCGAGTCCAGAAGCTTGTAGT PCR products were mixed to a final concentration of 0.3M NaCl and ethanol precipitated. NEB T7 High Yield RNA Synthesis Kit was used to transcribe PCR pellet [BO]. Transcription products were DNase I treated, diluted to 0.3M NH4(OAc), extracted with phenol (pH 4)-chloroform, and measured using NanoDrop Spectrophotometer. Electrophoreses Mobility Shift Assay (EMSA) Binding reaction: 2µl 10X EM

SA buffer, 1 µl 150mM DTT, 0.2 µl RN
SA buffer, 1 µl 150mM DTT, 0.2 µl RNA inhibitor, 5µl of 800ng/µl RNA aptamer, 5µl 6X loading dye, (0, 2, or 5 µl) of 1µg/µl Xiwi PAZ domain, and H2O was added till 22µl. Binding reaction incubated at RT for 30mins. Invitrogen 50bp DNA ladder, BIORAD protein marker, and 22µl of binding reaction was loaded onto gel immediately, ran in 0.5X TBE buffer, 150V, 4⁰C for 4h 30mins. Gel was stained with 1.6X Life Technologies SYBR Gold Stain in 0.5X TBE for 15mins, washed with 0.5X TBE, and visualized with blue fluorescence, 800V. Native gel is composed of 6 6% acrylamide (29:1), 10% glycerol, 96% 0.5X TBE, 0.4% APS, and 0.04% TEMED. 10X EMSA buffer composed of 150mM HEPES pH 7.4, 1000mM, KOAc , 20mM Mg(OAc)2, 0.5% w/v BSA, and H2O. 6X loading dye composed of 50% glycerol, 0.1% Bromophenol Blue, 0.1% Xylene Cyanol, and 50% 1X TBE. Egg Extract 200-300 units of human chorionic gonadotropin hormone (hCG) was injected into the dorsal lymph sac of female X.Tropicalis. Frogs were left in 2L frog system water in dark cool room for 12 hours. Frogs were returned to system. Frog system water was decanted and the eggs were dejelleyed with 2% cysteine in 1/9 Modified Ringer Solution (MRS) for 10 mins. Eggs washed with 0.1X Modified Barth Solution twice and 1/9 MRS twice. Aliquots of 100 healthy eggs were selected under light microscope and flash frozen in dry ice. Eggs were lysed with 1egg:10µl lysis buffer and grounded with dounce homogenizer. Egg extract was centrifuged @4⁰C, 21000crf for 5 mins and clear supernatant was kept; repeated until yolk was removed. Centrifuged @4⁰C, 21000crf for 20 mins, combined clear supernatant, made 100µl aliquots, flash frozen in dry ice, and stored in -80⁰C. Lysis buffer composed of 100 mM NaCl, 20mM NaF, 50mM Tris, pH 7.5, 5mM EDTA, 0.2% (v/v) NP40 (IGEPAL), and 1:50 Protease Inhibitor. Binding Reaction 100µl egg extract was mixed with3- 6µg of respective RNA aptamer and incubated @25⁰C. A time course of the binding reaction was stopped with 500µl Life Technolo

gies TRIzol® Reagent, extracted with
gies TRIzol® Reagent, extracted with chloroform, and precipitated with 1:1 isopropanol. Northern Blot Total RNA from binding reaction mixed with 1:1 2X Urea loading buffer, incubated @95⁰C for 5 mins, placed on ice, loaded onto 15% thick urea gel, ran @20W for 30mins, and 30W for 1hr. Urea gel transfer onto Hybond NX was done with 0.5X TBE using electrophoretic semi dry transfer machine @4⁰C for 4hrs. Nylon membrane was pre-hybridized with salmon sperm DNA, hybridized with radiolabeled γATP X.Tropicalis piRNA probes below, and normalized with U6 probe. Nylon membrane was exposed to storage phosphor screen and visualized with phosphor imager. xtmt45-1anti TGA GCC AGT TGG AGA ATT TTA CCT TGA CA xtmt8-1anti GGG CCA TCA CAA GGG CAG TCT TCC ACT TCA xtc52-2anti TAG AAA GGT GGA TAG GGT AAT TTG TCG CA xtmt1-1anti CAG CAG GAT TAC AGG CAC GCG CCA CTG TGC CCG GC DM-U6 AAA AAT GTG GAA CGT TTC ACG 7 IV. Results Fig. 1a: EMSA of aptamer 1, 2, 3, and 4 8 Figure 1a,b,c–Emsa of aptamer 1- 9- 6% polyacrylamide EMSA gel, 2µg RNA aptamer, gradient Xiwi PAZ domain, @4⁰C, 150v, 0.5X TBE for 4h 30mins. Unbound RNA runs further down the gel, RNA bound with PAZ domains are shifted upwards. 1a) Aptamers 1, 2, 3 and 4 with gradient protein 0µg, 2µg, 10µg. 1b) Aptamers 2,3,4,5 and 6 with gradient protein 0µg, 2µg, 5µg. 1c) Aptamers 7, 8, 9 and negative control with gradient protein 0µg, 2µg, 5µg. Fig. 1b: EMSA of aptamer 2, 3, 4, 5, and 6 Fig. 1c: EMSA of aptamer 7, 8, 9, and negative control 9 Figure 3a- Trial 2 Northern blot showing piRNA- Northern blot showing piRNA levels of samples with total 6µg aptamer 5, 9, 5+9 (3µg each), and control per 100µl of egg extract. Incubated at RT with time course of 0, 1 and 2 hours per aptamer. Total RNA was measured and loaded for northern blot. Equal levels of piRNA were detected in each sample, showing no signs of degradation. Fig. 3a: Trial 2 North

ern blot showing piRNA levels Fig. 2
ern blot showing piRNA levels Fig. 2: Trial 1 Northern blot showing piRNA levels Figure 2-Trial1 Nortjer blot showing piRNA levels - Northern blot showing piRNA levels of samples with total 3µg aptamer 5, 9, 5+9 (1.5µg each), and control per 115µl of egg extract. Samples incubated at RT over time course of 0, 30 and 60 mins per aptamer. Total RNA concentration was measured and normalized to 21µg. No significant trend in piRNA levels was found. 10 Figure 3b-Trial 2 Northern blot showing U6- Northern blot showing U6 snRNA levels of samples with total 6µg aptamer 5, 9, 5+9 (3µg each), and control per 100µl of egg extract. Incubated at RT over time course of 0, 1 and 2 hours per aptamer. Total RNA was measured and loaded for northern blot. U6 snRNA signal did give strong signal. Normalization of piRNA using U6 snRNA levels did not show any significant trend or signs of piRNA degradation. Fig. 4: Trial 3 Northern blot showing piRNA levels Figure 4- Trial 3 Northern blot showing piRNA- Northern blot showing piRNA levels of samples with total 6µg RNA aptamer per 100µl of egg extract. Incubated at RT over time course of 0, 1 and 2 hours per aptamer. Total RNA was measured and loaded for northern blot. Fig. 3b: Trial 2 Northern blot showing U6 snRNA levels 11 A library of RNA containing 1015 random 50 nt RNA sequence were selected for Xiwi PAZ domain binding through SELEX. 9 candidate aptamers were selected after 8 rounds of SELEX using Xiwi PAZ domain completed by Nicholas Clark and Thomas Brands. Aptamers cloned into vectors were, PCRed then transcribed to RNA. The binding ability of the 9 candidate RNA aptamers to Xiwi PAZ domain were characterized using EMSA. Significant binding in EMSA was determined by a fading band of unbound aptamer and an increase in darkness of shifted aptamer as protein volume increased. Aptamer 1 showed slight shift, however the signal was weak. (Fig. 1a) Aptamer 2 showed good shift in two EMSA’s. (Fig. 1a, b) Aptamers 3 and 4 dis

played minor binding in Fig. 1a but n
played minor binding in Fig. 1a but no binding in Fig. 2a. Aptamer 5 showed significant binding. (Fig. 2b) Aptamer 6 and 7 displayed minor bidnding. (Fig. 1b, c) Aptamers 8 and 9 showed good binding. (Fig. 1c) The negative control aptamer displayed no binding affinity with the Xiwi PAZ domain. (Fig. 1c) Aptamers 2, 5, and 9 that were noted for good binding were tested in-vivo in egg extract as RNA inhibitors. Aptamers 5 and 9 could not displace piRNA in complex with the Xiwi in X.Tropicalis egg extract. piRNA signals in northern blot in trial 1 varied with no trend. (Fig. 2) piRNA northern blot in trial 2 showed no signs of piRNA degradation in any of the aptamer samples. (Fig. 3a) piRNA levels of samples containing aptamers were equal to those with the negative control aptamer or no aptamer. U6 snRNA normalization was difficult due to unusual and weak signals. (Fig. 3b) Even attempted normalization shows no trend in piRNA levels in response to RNA aptamers. Trial 3 northern blot showed no significant variation in piRNA levels except in RNA aptamer 5 at 2 hours. (Fig. 4) V. Discussion The binding ability of the 9 candidate RNA aptamers to Xiwi PAZ domain were characterized using EMSA. Aptamers 2, 5, 8, and 9 showed good binding. Aptamers 1,3,4,6 and 7 showed weak binding. Negative control aptamer showed no binding affinity. Even though the strength of aptamer binding with the Xiwi PAZ domain can be estimated with EMSA, EMSA has not been fully optimized and may not be the best indicator. Optimization of EMSA would include getting all the RNA to migrate through the gel instead of getting stuck in the wells. The Xiwi PAZ domain is also unable to migrate through the gel properly in current conditions. However, the gel cannot be denaturing as binding is required. A wider gradient of Xiwi PAZ domain can be used for the ESMA to get more data on binding ability. Clearer binding results may be obtained by doing an EMSA with radiolabelled RNA aptamers. Lower amounts of radiolabelled RNA apta

mers of Xiwi PAZ domain would be requir
mers of Xiwi PAZ domain would be required for an EMSA. A competition assay could also be used to assess the binding ability of RNA aptamers. Even though aptamers 5 and 9 showed good binding ability in EMSA, they were unable to displace piRNA in complex with the Xiwi in X.Tropicalis egg extract. It is vital for the RNA aptamers to be able to displace piRNA in Xiwi to be able for it to work as an effective inhibitor. If the RNA aptamer displaces piRNA, piRNA would be degraded 12 when it is unbound for Xiwi. In trial 1 the piRNA signal on the northern blot was inconsistent; this may be due to the normalization of RNA concentrations that do not reflect piRNA amounts in the total RNA from the egg extract binding reaction. (Fig. 2) In trial 2 the piRNA signal on the northern blot was the same for all samples, showing no signs of piRNA degradation. (Fig. 3a) Normalization of piRNA levels with U6 snRNA was unsuccessful due to unusual and weak signal. (Fig. 3b) U6 RNA should be abundant in X.Tropicalis egg extract and show strong signal at approximately 74nt. The weak U6 signal may be due to the fact that the probe was for Drosophila Melanogaster and the one nucleotide difference in the probe caused unspecific binding. Trial 3 showed no significant difference in piRNA levels across all samples except aptamer 5 at 2hours. Aptamer 5 at 2 hours is at the edge of the nylon membrane and an incomplete transfer may be the cause for the reduced signal. (Fig. 4) The egg extract results only represent 3 of the many candidate aptamers that could be generated from SELEX. Screening through more candidates could yield a potential RNA inhibitor of Xiwi. Since the RNA aptamer is only targeting the PAZ domain of Xiwi which holds the 3´ end of the piRNA. The inability of the aptamer to displace piRNA could be due to the limitation of the PAZ domain’s role in binding piRNA in Xiwi. Targeting a different region of the Xiwi protein like the PIWI or MID domain for SELEX might yield a more effective RNA inhibitor. Using the entire Xiwi p

rotein or Xiwi in complex with piRNA cou
rotein or Xiwi in complex with piRNA could also be used in SELEX. Once a suiTable RNA aptamer has been developed, in-vivo testing in X.Tropicalis oocytes could be done. Injection of RNA inhibitors into egg oocytes could potentially reveal more characteristics of Xiwi pathway and function. 13 Bibliography 1 Christina Lorenz FvPRS. Genomic Systematic evolution of lignads by exponential enrichment (Genomic SELEX) for the identification of protein-binding RNAs independent of their expression levels. Nature Protocol. 2006 2204-2212. 2 Desai A MAMTWC. The use of Xenopus egg extracts to study mitotic spindle assembly and functions in vitro. Methods Cell Biology. 1999 385-412. 3 Jinek M DJA. A three-dimensional view of the molecular machinery of RNA interference. Nature. 2009 405-412. 4 Kaoru Sato MCS. Piwi-interacting RNAs: Biological functions and biogenesis. Biochemical Society. 2013 39-52. 5 Lorsch J. Methods in ENZYMOLOGY Volume 541. Waltham: Academic Press Elsevier; 2014. 6 Meister G. Argonaute proteins: functional insights and emerging roles. Nature Reviews Genetics. 2013 447-459. 7 Nelson C Lau TOMBREKMDB. Systematic and single cell analysis of Xenopus Piwi-interacting RNAs and Xiwi. The EMBO Journal. 2009 2945-2958. BO New England Biolabs. T7 High Yield RNA Synthesis Kit Instruction Manual. [Internet]. Available from: https://www.neb.com/products/e2040-hiscribe-t7-high-yield-rna-synthesis-kit#Tableselect0. 8 Tolia N.H JT. Slicer and the argonautes. Nature Chemical Biology. 2007 36-43. 14 Chapter 2: Validation of original Xiwi, and Xili CLIP data using RIP, RT-QPCR I. Abstract Piwi proteins are a subfamily of AGO that bind to piRNAs 20-32nt long and are important in the silencing of transposable elements. Piwi also binds to a diverse number of genes. CLIP of Xiwi (Piwi homolog found in Xenopus), and Xili was carried out by Yuliya Sytikova in order to determine genes that interacted more

with Xiwi, and Xili. CLIP data was ana
with Xiwi, and Xili. CLIP data was analyzed based on total CLIP counts in the open reading frame, by Dr. Chirn. The goal of this experiment is to validate the CLIP data through RIP, RT-QPCR. RIPs using X.Tropicalis stage I-IV oocyte extract was being optimized, and validated with Xiwi, and Xili western blots. RT-QPCR surveying a few genes revealed no significant difference in enrichment (% input) between positive and negative controls. Correcting and validating the CLIP data could provide huge insights regarding Xiwi, and Xili gene interactions, and silencing. II. Introduction Piwi (P-element induced wimpy testes) proteins are a subfamily of AGO that bind to piRNAs 20-32nt long and are important in the silencing of transposable elements (1). Transposable elements are selfish nucleic acid elements that can be harmful to the genome when accumulated. PIWI is enriched in the germ line and has been shown to play a role in development and maintenance. . In vertebrates the first letter of the animal “helps” denote the homologs’ origin. Hence the mouse encodes Miwi, Miwi2, and Mili (Miwi-like), while humans encode Hiwi, Hiwi2, and Hili, Ziwi and Zili in Zebrafish, and Xiwi and Xili in Xenopus. The biogenesis of piRNAs is not dependent on Dicer, a key enzyme that matures endo-siRNAs and miRNAS (2). “piRNAs appear to derive from single-stranded transcript precursors that either are non-coding with no annotated features, are transcripts that correspond to the 3'UTR of certain protein coding genes, or are transcripts that bear an unusually high concentration of transposable element (TE)” (3). piRNAs have a distinguished by their tendency to have a U at the 5' end, and a highly conserved 2'O-methylation modification at the 3' end (4) (5). piRNAs are generated using two distinct pathways: the primary pathway and the secondary amplification pathway (ping pong amplification loop). The primary pathway generates piRNA from piRNA clusters that are composed of multiple transposon fragments. In the secondary

amplification pathway mature sense prima
amplification pathway mature sense primary piRNA guide Piwi proteins to complementary antisense sequences and cleave it with slicer to generate a new 5´ end. The 5´ end binds to a second Piwi protein and has its 3´ end trimmed to form a mature antisense secondary piRNA. Antisense secondary piRNA can now guide Piwi proteins to candidate the sense sequences and cleave it (6). The secondary “ping-pong” amplification pathway is theoretical and has not been proven biochemically. However, there is some evidence that indicates its existence. Piwi and piRNAs have been identified in nematodes, Drosophila, zebrafish, mice, rats, anemones, sponges, and Xenopus (3) (7) Drosophila was where the first Piwi protein was discovered and continues to be a great model organism for the study of Piwi and piRNA. The two major piRNA clusters in 15 Drasophila are the Flamenco locus on the X chromosome and the 42AB piRNA cluster locus on chromosome 2R (3). Flamenco generates transcripts from one genomic strand and is mainly expressed in the somatic follicle cells that surround the egg chamber, whereas 42AB generates transcripts from both genomic strands and is mainly expressed in the nurse cells of the female germline and these transcripts are then deposited into the oocyte (8) (9). Eggless, a histone methyltransferase is thought to impact transcription of both 42AB and Flamenco locus through tri-methylating histone H3 on lysine 9 (H3K9Me3) (10). Rhino (Rhi), a protein enriched in the gonads binds to H3K9Me3 and is required for transcription of the 42AB piRNA cluster (11). Cutoff (cuff) a Rail1 nucelase homolog, associates with Rhi at the 42AB cluster and is required for the expression of both 42AB expression and a single-strand piRNA cluster called the 20A cluster (12). Many factors have been identified that affect the biogenesis of either the Flamenco and/or 42AB locus. In Drasophila the secondary piRNA processing “ping-pong” cycle is carried out bu Aubergine (AUB) and AGO3. Squash (squ) and Ma

elstrom (mael) are two known effectors
elstrom (mael) are two known effectors of Piwi-mediated gene silencing because mutations or knockdowns greatly impacted fertility yet most piRNAs were sustained in these animals or gonadal cells (12) (13) (14)It is still unclear how 5' ends of primary piRNAs is defined and why the 3'UTR of certain genic transcripts are selected for primary piRNA biogenesis. Perhaps this gap is attributed to an incompleteness of the list of genes that do impinge upon the Piwi pathway. The Drosophila OSS cell line is derived from follicle cells that express only primary piRNAs and the single PIWI protein. (15) (16) (14). A similar ovarian somatic cells (OSCs) cell line is also used to study Piwi pathways. A genome wide survey of OSS and OSC cells with various deep sequencing methods revealed that transposable elements (TE) are the major transcripts regulated by Piwi; the TE landscape between OSS and OSC cell line are different; and de novo TEs appeared to stimulate the expression of novel candidate long noncoding RNAs (lncRNAs) in a cell lineage-specific manner some of which associated with PIWI and overlapped PIWI-regulated genes (17). There are two species of xenopus that have been used as animal models: X. Laevis and X.Tropicalis. Currently there is a favor to work with X.Tropicalis because it is diploid compare to the allo-tetraploid X.Laevis. Xenopus have 4 homologs of PIWI: Xiwi 1a, Xiwi 1b, Xiwi 2 and Xili. Egg extracts are noted for their ability to recapitulate many cell-cycle processes in vitro [2]. Xenopus tropicalis has well characterized piRNA clusters conFig.ured in an analogous fashion to mammalian clusters. piRNAs are restricted to the germline and found in high copy numbers or high repetitive elements in the genome, however individual frogs have unique cluster expressions (7) (19). Large-scale piRNA sequencing in vertebrate cells reveals that the 3’UTR-directed piRNA pathway is conserved in vertebrates such as mice and Xenopus (20). The abundance of Xiwi 1b and Xili proteins

in Xenopus oocytes and early embryos
in Xenopus oocytes and early embryos make them ideal for studying Piwi pathways (21). Xiwi1 is a nucleo-cytoplasmic protein while Xili is purely cytoplasmic; neither of the proteins interacts with a cap-binding complex (21). Xiwi 1b has been found to be associated with microtubules and the meiotic spindle, and is localized to the germ plasm—a cytoplasmic determinant of germ cell formation (19). Xiwi associates with translational regulators in an RNA-dependent manner, but Xenopus tudor interacts with Xiwi independently of RNA (19). Xiwi was co-purified with 16 CPEB a major translational regulator, however the interaction is unknown (21). Xiwi 1a and Xili are involved in the secondary processing of piRNAs in the “ping-pong” amplification cycle. Xiwi is also highly involved in the primary processing of piRNAs. piRNAs are extremely diverse compared to the miRNA population, and it is a challenge to determine the full breadth of candidates that are regulated by piRNAs (3). High-throughput deep sequencing technologies have enabled better characterization of piRNAs. Various methods like CLIP-Seq, ChIP-Seq, Nascent-Seq, and RNA-Seq have been used in combination with high-throughput deep sequencing to construct libraries (sup1). Cross-linking-immuno-precipitation (CLIP) enriches transcript associated with Piwi proteins. Chromatin immune-precipitation (ChIP) enriches genomic DNA regulated by Piwi complexes. Nascent RNA analyses reinforce a transcriptional gene silencing role of Piwi. RNA-seq analyzes total RNA, including mRNA and piRNA (3). These new approaches are highly dependent on bioinformatics that require sequenced and assembled animal genomes (3). Accurate data management and analysis of high-throughput deep sequencing data require many bioinformatics tools. These processes can create highly variable data sets, depending on the method used to analyze the data. It is important to confirm data sets using other in-vitro or in-vivo experiments on the Piwi pathway. Xiwi CL

IP was carried out by Yuliya Sytnikova.
IP was carried out by Yuliya Sytnikova. The CLIP data aimed to reveal potential genes that interact with Xiwi more strongly than others. Initial analysis of CLIP by Dr. Chirn was based on total CLIP counts in the open reading frame of the gene; however it failed to include the CLIP counts in the 3’ UTR. Validation of the CLIP data with RIP and RT-QPCR was carried out. Genes with high CLIP counts were selected as positive; genes with total CLIP counts below 100 were selected as negative controls. Transposon candidates were selected based on CLIP counts. Xiwi, and Xili RIP of stage I-IV oocyte from X.Tropicalis was used for RT-QPCR. Xili RIP did not work. RT-QPCR of Xiwi RIP surveying 14 positive, 9 negative and 5 transposons revealed no significant difference in enrichment (% input) among genes. Supplementary 1: Sequencing Technology Supplementary 1: Sequencing Technology- Systems biology approaches applied to study Piwi pathways Four main biochemical techniques now enable system-wide analysis of the Piwi-piRNA pathway. Cross-Linking Immuno-Preciptiation (CLIP) enriches transcripts associated with Piwi proteins. Chromatin Immuno-Precipitation (ChIP) enriches genomic DNA regulated by Piwi complexes. Global Run On (GRO) and Nascent RNA analyses reinforced a transcriptional gene silencing role by Piwi, while messenger RNA and piRNA analyses were facilitated by high-throughput deep sequencing of cDNA libraries. (2) 17 III. Materials and Methods RNA immune-precipitation of X.Tropicalis Stage I-IV oocyte was done, and RNA was extracted using TRIzol l-Chloroform Extraction. Reverse transcription using Promega M-MLV Reverse Transcriptase was carried out using Input, Xiwi RIP, IgG RIP RNA. RT was diluted 1/50x for QPCR. QCPR was carried out using Promega GoTaq® qPCR Master Mix (7.5 µl GoTaq, 1.5 µl H2O, 1 µl F+R Primer 10 µM, 5 µl RT product). QPCR data was analyzed using Opticon Manager 3. Xiwi pull down in RIP was confirmed with Western Blot. Sup

plementary 2: Experimental Scheme Sup
plementary 2: Experimental Scheme Supplementary 2: Experimental Scheme- Stage I-IV oocytes harvested from X.Tropicalis were made into extract and flash frozen. Xiwi, Xili, and IgG RIPs were carried out by pre-binding antibodies to Protein A/G magnetic beads. RNA was extracted with TRIzol-chloroform. RT-QPCR of RIPs were normalized to input and analyzed for enrichment (% input). 18 X.Tropicalis Stage I-IV oocyte extraction 3L of 2g/L EMSS in frog water (pH=7.5) used for anesthetizing frog 15min. Frog was monitored for reflexes. Once unconsciousness confirmed, frog was decapitated with garden scissors. Double Y-cut from armpits to sternum, from sternum to groin, and from groin to thighs was made on skin. Same cut was performed on muscle. Ovaries were removed from abdominal cavity using thin tweezers and scissors, and place in a beaker of 1x MBS. Washed 3x with 1xMBS. Oocytes transferred to multiple petri dishes with 1xMBS. Oocytes cut into smaller chunks with tweezers (~4mm3). 1xMBS decanted. Pre-chilled 3mg/ml Type IV collagenase in 1xMBS added, rotated slowly at RT. Fresh collagenase added every 30-40 min for until oocytes are singular. Washed 3x with 1xMBS. Separate clumps with 850µm mesh, stage VI with 600µm mesh, stage V 425µm mesh. Mesh was held in place in over petri dish with 1xMBS. Oocytes were transferred on to mesh and 1xMBS was pipetted on top to push oocytes through. Oocytes the did no go through were collected in a separate petri dish with 1xMBS. Filtered oocytes were transferred to the next mesh and separated as above. All oocytes were kept in 15ml falcon tube, supernatant removed and flash frozen on dry ice. 1ml of 0.1M KOAc QColumn buffer (0.2% NP40, 0.5mM DTT, 1x Protease inhibitor) per 100µl oocyte was added. Dounce homogenized 50x. 1ml aliquots were made. Sonicated 2x (2 on, 2 off for 80sec). Centrifuged at 21,000 crf for 10min, 4⁰C to separate yolk, clear supernatant kept in fresh tube (avoid yolk). Repeat centrifugation step to remove majority of yolk. Stored at -80⁰C. RNA I

mmuno-precipitate of Stage I-IV oocy
mmuno-precipitate of Stage I-IV oocyte 25µl Life Technologies Pierce™ Protein A/G Magnetic Beads were washed 5x with 1ml 1xPBS (mixed by inversion, placed on magnet 1min, removed FT). 10µg IgG antibody (anti Xiwi, anti IgG) rotated with beads in 150µl 1x PBS for 2hr RT. Aliquot 100µl of stage 1-IV oocyte extract for input RNA and add 500µl trizol (freeze). Aliquot 40µl of oocyte extract for input western blot and add SDS LD. Beads washed 4x with 1xPBS. 1ml 0.1M KOAc QColumn buffer (0.2% NP40, 0.5mM DTT) added then transferred to fresh eppi. Placed on magnet 1 min, FT removed. 100µl oocyte extract + 900µl 0.1M KOAc Qcolumn buffer (0.2% NP40, 0.5mM DTT, 1X protease inhibitor, 10µl/ml ribolock) mixed with beads. Rotated at 4°c for 3 hrs. Kept FT, washed 3x 5min rotating and 2x quick wash with 0.2M KOAc QColumn Buffer (0.2%NP40, 0.5mM DTT). 700µl of RIP aliquot for RNA extraction, 300µl for western blot after final wash before placing on magnet. Complete wash and remove FT. 500µl Life Technologies TRIzol® Reagent added to RNA aliquot, 40µl SDS LD added to Western blot aliquot. TRIzol l-Chloroform Extraction 100 µl 24:1 Chloroform: Isoamyl alcohol added to RIP protein A/G magnetic beads in 500 µl TRIzol® Reagent. Vortexed for 1min. Centrifuged at 21,000 crf for 5min, 4⁰C. Transferred clear supernatant to fresh eppi. 1:1 volume 24:1 Chloroform: Isoamyl alcohol added to supernatant. Vortexed 1min, centrifuged at 21,000 crf for 5min, 4⁰C. Kept clear supernatant in fresh eppi. 1:1 volume of isopropanol, 1 µl glycogen (20mg/ml) was added. Precipitated at -20⁰C for 30min. Centrifuged at 21,000 crf for 30 min, 4⁰C. Remove supernatant carefully without losing pellet. Washed 2x with 100 µl 70% ethanol (Centrifuged at 21,000 crf for 5 min, 4⁰C, carefully remove supernatant). Air dried pellet 19 in fume hood for 5 min. 20 µl of nuclease free water added. Concentration of RNA measured using NanoDrop Spectrophotometer. Reverse Transcription using Promega M-MLV Reverse Transcriptase 1

st Mix: 1 µl Random Hexamer, 2.5
st Mix: 1 µl Random Hexamer, 2.5 µl dNTP 10x, 4.5 µl H2O, 2 µl RNA (Input, Xiwi RIP, IgG RIP). Placed in PCR machine at 65⁰C for 10min, cool to 4⁰C. Add 2nd mix: 4µl MMLV RT Buffer 5x, 2µl 100mM DTT, 0.3µl RiboLock RNase Inhibitor, 0.2µl BSA 100x, 1µl MMLV Reverse Transcriptase, 2µl H2O. 25⁰C 10min, 42⁰C 50min, 70⁰C 15min, 4⁰C 15min. Negative control used Input RNA and had no MMLV Reverse Transcriptase added; volume compensated with H2O. QPCR 7.5 µl Promega GoTaq Master Mix, 1.5 µl H2O, 1 µl F+R Primer 10 µM, 5 µl RT product. 95⁰C 2min, (95⁰C 30sec, 58⁰C 30sec, 72⁰C 30sec, read) 50x, Melting Curve. If product has significant primer dimers add 80⁰C 15 sec before read. SYBR green was detected using FAM channel of QPCR machine. Western Blot 10% SDS PAGE loaded with 10µl RIP magnetic beads in SDS LD, 7.5µl BIORAD Precision Plus Protein™ Unstained Standards, ran at 80v 30min, 130v until end. Gel washed 2x 5min in 1xTBE. Transferred to Whatman Nitrocellulose Membrane using semidry apparatus in 1xTBE, 0.16A 30min, 4⁰C. Ponceau staining done to confirm transfer. Washed 5min in 1xTBST. Membrane blocked in 5% Milk 1xTBST 90min. 1⁰ antibody 1:5000 Xiwi Antibody in 5% Milk 1xTBST O/N, shaking, 4⁰C. Washed 5 x 5min in 1xTBST. 2⁰ AB, 1:10,000 αHRP IgG Rabbit 1h, shaking, RT. Washed 5min 1xTBST. Incubated in Amersham ECL Prime Western Blotting Detection Reagent for 2min. Exposed in film cassette (30sec, 1min) and developed. 0.1M QColumn Buffer KOAc 20mM Hepes pH 7.9 (with KOH) 10% glycerol 100 mM KOAc 0.2 mM EDTA 1.5 mM MgCl2 Modified Barth's Saline (MBS) Reagent Quantity (for 1 L) Final concentration (1X) NaCl 5.143 g 88 mM KCl 0.075 g 1 mM MgSO4 0.120 g 1 mM HEPES 1.192 g 5 mM NaHCO3 0.210 g 2.5 mM CaCl2, dihydrate 0.103 g 0.7 mM H2O to 1 L 20 Table 1a: Positive candidates’ primer design Name Sequence Size Comments XT_aurkb_F CTGTGCCAGTCAAGGCTACA

96 7481 CLIP counts, exonic amplicon
96 7481 CLIP counts, exonic amplicon XT_aurkb_R TGCTTGAAAGGCATTGCTAC XT_ccna1 _3_F GTAGGCACAGCCGCTATCTT 181 266 CLIP counts, exonic amplicon XT_ccna1 _3_R GCAGGAACTGGCTGATTGTAG XT_ccna1_F TCCAGATGCTGTAGCTGTGTC 96 266 CLIP counts, exonic amplicon XT_ccna1_FR CGCATGTAATAAGCCTTTGGT XT_ccnb1_3_F CAGCAGTCGACAGATGAAGG 62 3701 CLIP counts, exonic amplicon XT_ccnb1_3_R GGCTTTCCACAAGCTCAGAC XT_ccnb1_F CATGGAAATGCAAATCCTTAGA 92 3701 CLIP counts, exonic amplicon XT_ccnb1_R CCTCCCCAATTTTAGATGCTC XT_cpeb1_F GTGGGATTTTTACCAGTTCCTTT 130 355 CLIP counts, exonic amplicon XT_cpeb1_R GCAGCTCCCAGAAACAGAAA XT_elf1_F AAAGATGCATTAAAACATAACAGGTG 109 140 CLIP counts, exonic amplicon XT_elf1_R GCAAATTTTCAGAGTTGCAGATAATA XT_fam46b_F CTGTGCCCACTTTTACTTGCT 150 260 CLIP counts, 3'UTR amplicon XT_fam46b_R AGCCAGGATCTCAGTTGCAG XT_foxh1_F CTCAAAGGCTTGGCTTTCC 102 1397 CLIP counts, exonic amplicon XT_foxh1_R GGGGCTCAGTGGGAGAATAG XT_mkrn3_F CTGTTCTAGCATGGCACAGC 97 24842 CLIP counts, exonic amplicon XT_mkrn3_R CTTATATCCCCCACAAAGGATG XT_papss1_F ACCACTGTTGTGGCCATCTT 65 114 CLIP counts, exonic amplicon XT_papss1_R TGCCACTGAACCTCTGTAGGA XT_plk3_3_F CCATATGAAGGCTTTTGCATC 132 107 CLIP counts, exonic amplicon XT_plk3_3_R TGTGCATGCAATATTAAAATGTGTT XT_plk3_F GCAGCTGCAGGGTCAGAT 107 107 CLIP counts, exonic amplicon XT_plk3_R GAGTCTGGCTTCCTTCCTCA XT_spdyc_F TGAAATCTTTAGGCATGAGTCCA 197 152 CLIP counts, exonic amplicon XT_spdyc_R CCATCTAAATCTGCAGCCCC XT_tpd52_F CACGATGTTGCCTCCACTAC 108 133 CLIP counts, exonic amplicon XT_tpd52_R TTTTGTTATGACAGAGCCAACTG XT_velo1_F CCAGAAGAAGTAGAAACATGGGA 103 20701 CLIP counts, exonic amplicon XT_velo1_R ACAGTTTGCCATTCATCTACTTC XT_zfp36l2.2_F TCTGTTCTCGGCTAACTTTTTATG 109 127 CLIP counts, 3'UTR amplicon XT_zfp36l2.2_R TCTTCTCACTGCTGGCACAT 21 Table 1b: “Negative” candidates’ primer design Name

Sequence Size Comments XT_amy2b_F
Sequence Size Comments XT_amy2b_F CCACCCAACCTGTCACTATTAAA 114 70 CLIP counts, exonic amplicon XT_amy2b_R GGCCATCAACCACATTCC XT_atp5s_F CTGTATTATCGTTGCGCTCTGT 137 17 CLIP counts, 3' UTR amplicon XT_atp5s_R CACAATGCTGCTTCGAGCTA XT_haghI_F CTTCCAGGCTTCTTCTGCAC 138 12 CLIP counts, exonic amplicon XT_haghI_R CGTCTCTCGTGAGTCCATTG XT_ndufa13_F CTTCGACAGGGGCTTAACTAAA 89 69 CLIP counts, 3'UTR amplicon XT_ndufa13_R CTTGTAATCCACAGGACCATAGC XT_pmm1_F TGAGGTGTAAATCCAGTTTTTATACAG 148 23 CLIP counts, exonic amplicon XT_pmm1_R AAAGTGCCTACACTTTAGCTTTCC XT_rpl18_F AACAAGGACCGCAAGGTG 88 71 CLIP counts, exonic amplicon XT_rpl18_R TGGCCAAGAAACGATAAAGC XT_rpl23_F CGACAACACAGGTGCAAAGA 94 81 CLIP counts, exonic amplicon XT_rpl23_R CATGTCTCCAACACCAGCA XT_rps11_F GGTTCAAGACCCCCAGAGA 69 25 CLIP counts, exonic amplicon XT_rps11_R CGTTGCCAGTAAAGGGACAT XT_tacc3_F AGCTCTGCAGGCGACACT 108 81 CLIP counts, exonic amplicon XT_tacc3_R TCATCACATATCTTTGTCAGCTCAT Table 1c: Transposon primer design Gene Sequence Size Comments XT_cr1_2_F AGGCCAAGACTAACCCCAAA 241bp 14.4 CLIP, RT-QPCR primer of cr1_2 transposon XT_cr1_2_R ACCATATCCTGTGCCAACCA XT_piggyback_N2_F GTGTCTCATACGTCGTTGGC 262bp 2.1 CLIP, RT-QPCR primer of piggybacN2 transposon XT_piggyback_N2_R CCTGTTCCTGAAGACCCGAT XT_chap4a_F AGGGGTTTGTTCTGGGAGTT 194bp 0.2 CLIP, RT-QPCR primer of chap4a transposon XT_chap4a_R CCTCACTGCGGGGATATCAA XT_helintronN1A_F TACAACACCCAAACTCCCCA 233bp 4.7 CLIP, RT-QPCR primer of heltinronN1A transposon XT_helintronN1A_R ATGTCGCTGTTTCATTCGGG XT_cr1_1a_F GCAGTAAGCTGGGTGACAGTT 96bp 25.1 CLIP, RT-QPCR primer of cr1_1a transposon XT_cr1_1a_R CAATCTGGCAAATCTGTTGG Table 1a- Positive candidates’ primer design- Positive candidates’ design based on total CLIP counts above 100; Amplicons spanned either exons or 3’UTR with length of 62-150bp. Table 1b- “Negative”

candidates’ primer design- Negativ
candidates’ primer design- Negative candidates’ design based on total CLIP counts below 100; Amplicons spanned either exons or 3’UTR with length of 69 – 148bp. Table 1c- Transposon primer design Transposon candidates’ design based on total CLIP counts; Amplicons were 96bp – 262bp in length. 22 IV. Results RNA Immuno-precipitate of Xiwi, Xili, and IgG control was carried out on Stage I-IV oocytes from X. Tropicalis. Pull down of Xiwi was confirmed on a Western Blot (Fig.1A). Xili RIP was unsuccessful, and little or no Xili was obtained (Fig.1B). cpeb did not co-precipitate with either Xiwi or Xili in RIP experiments (Fig. 1C). Fig.1: Western Blot of Xiwi, Xili, and CPEB of 10/27/14 RIP Positive mRNA and negative controls were selected based on CLIP data, and mRNA sequencing of Xiwi completed by Yuliya Sytnikova and analyzed by Dr. Chirn. The original CLIP data included CLIP counts only in the open reading frame of the genes but not in the 3’ UTR. Positive candidates include: aurkb, ccna1 , ccnb1, cpeb1, elf1, foxh1, papss1, plk3, tpd52, fam46b, mkrn3, spdyc, velo1, and zfp36l2.2 (Table 1a). Transposons: cr1_1a, cr1_2, chap4a, helintronN1A, and piggybacN2 (Table 1b). “Negative” controls: amy2b, atp5s, haghl, ndufA13, pmm1, rpl18, rpl23, rps11, tacc3, and tsen54 (Table 1c). Figure1- Western Blot of Xiwi, Xili, and CPEB of 10/27/14 RIP- A. Xiwi present in input. Fraction of Xiwi (100kDa) was pulled down in RIP, remaining Xiwi in FT. IgG did not pull down Xiwi, Xiwi remains in IgG FT. Small amounts of Xili co-precipitated with Xiwi. B. Xili present in input. Xili was not pulled down in RIP, Xili remains in FT. IgG did not pull down Xili, Xili remains in FT. C. CPEB present in input. CPEB did not co-precipitate with Xiwi, Xili or IgG. CPEB remains in Xiwi, and IgG FT. A C B 23 Fig.1a Fig.1b Figure 1a-10/29/14 RIP- 10/29/14 RIP 0.1M KOAc wash; input, Xiwi RIP, IgG RIP RNA DNase I treated; no large difference in en

richment among positive, negative, and t
richment among positive, negative, and transposon candidates. Figure 1b- 11/26/14 RIP- 11/26/14 RIP 0.2M KOAc wash; input RNA DNase I treated; no large difference in enrichment among positive, negative, and transposon candidates. Refer to supplemental file 102414 to 120814 Consolidated QPCR data for details. 24 The 10/29/14 RIP was washed with 0.1M KOAc Q-Column Buffer after protein binding reaction; while the 11/26/14 RIP was washed with 0.2M KOAc Q-Column biffer after protein binding reaction. In 11/26/14 RIP the 0.2M wash conditions showed reduce enrichment of positive, negative, and transposon candidates. Extracted RNA concentration was measured with nanodrop spectrophotmeter. Input showed much greater levels of RNA compared to Xiwi, and Xili RIP. No RNA was detected in IgG, or Xili RIP. Two different dilutions of RT were made: Normal RT using 10% Input vs 9% RIP RNA. Diluted RT using 2% Input vs 45% RIP RNA. QPCR was done and enrichment (% input) was calculated, and corrected for dilutions. Candidate genes showed Xiwi RIP enrichment (% Input) from 22% - 90% in 10/29/14 RIP, and 6% - 31% in diluted RT (Fig. 1a, b). IgG RIP showed no enrichment (% Input) in both normal RT and diluted RT. Average enrichment with standard error showed no significant difference among positive, negative, and transposon candidates in both 10/29/14, and 11/26/14 RIP (Fig.1a, b). More details on primers can be found in QPCR primers Fall and Spring excel. QPCR enrichment data can be found in Fall 2014 Consolidated QPCR Enrichment excel. More detail on other RIPs, and Western Blots can be found in lab note book, and Fall CLIP RIP presentation PowerPoint. CLIP data can be found in Combine Comparative CLIP excel. V. Discussion Validation of original CLIP data based on total CLIP counts was done by RIP then RT-QPCR. Positive and negative controls were selected based on total CLIP counts. Negative controls had CLIP counts below 100. Amplicons were designed to span the exons. ndufa13, atp5s, zfp36l2.2, and fam46b

had amplicons that spanned the exon
had amplicons that spanned the exon and 3’ UTR of the gene. RIP of Xiwi, and Xili were carried out initially. Xiwi RIP was successful in pulling down Xiwi, however not 100% of Xiwi was pulled down from the Stage I-IV oocyte extract. RT-QPCR was carried out using Xiwi RIP. RIP can still be optimized by increasing the amount of Protein A/G Magnetic beads, and anti Xiwi IgG antibody. Xili RIP was not successful and low levels or no Xili was pulled down. Xili antibody worked well in the Western Blot. The Xili antibody used might be able to bind to a motif of Xili that is exposed when denatured like in the Western Blot; however it is unable to bind to Xili in its folded naïve state. The inability of the Xili antibody might have caused the inefficient Xili RIP. A cepb western blot was carried out to screen for cpeb co-precipitation with Xiwi or Xili. No co-precipitation of cpeb with Xiwi or Xili was detected at 56KDa. However, the cpeb western blot showed multiple bands, and it is difficult to distinguish which was the correct one. The IgG heavy chains from the antibodies are 50KDa and may have prevented proper visualization of cpeb. A higher percentage SDS PAGE could be done to get clearer resolution of the bands. 25 The washes carried out after the protein binding reaction was done with either 0.1M or 0.2M KOAc Q-Column buffer. DNase I treatment of input RNA was done to reduce DNA signal in QPCR. DNaseI treatment of Xiwi, and IgG RIP RNA potentially caused loss of RNA. With 0.1M wash for RIP and DNase treatment of input, Xiwi RIP, and IgG RIP RNA; Enrichment of mRNA is low compared to negative control, but transposons showed high enrichment. With 0.2M wash for RIP and DNase treatment of input RNA only. Enrichment for mRNA is better compared to negative control, but transposon show less enrichment. The lack of standardization in RIP experiments due to optimization made it difficult for cross comparison of absolute enrichment levels of genes. 10/29/14 RIP, RT used 5% input vs 10.5% RIP. 11/26/14 RIP, RT used 1% input

vs 7% RIP. All dilutions were corrected
vs 7% RIP. All dilutions were corrected for during calculation of enrichment (% input). However, the differences in dilution made it difficult to compare enrichment across RIPs. Standardizing the dilution factor of input vs RIP would be important in analyzing biological replicates. Few technical replicates of QPCR were carried out to reinforce enrichment values. QPCR duplicates were not always the same, and could cause abnormal enrichment. There was no significant difference in enrichment (% input) between positive, negative, and transposon candidate in both all RIP, RT-QPCR. After surveying nine potential negative controls, no representative candidate could be found. The RT-QPCR enrichment does not correspond to the original CLIP data. Further analysis of the CLIP data might yield a more accurate CLIP rank of genes. Current CLIP data is based on total CLIP counts in the open reading frame of the gene, and fails to include those in the 3’ UTR. A noise filter to include the 3’ UTR is being applied on the CLIP data by Dr. Chirn. Different positive and negative candidates may be revealed after the noise filter. Surveying those genes with RIP, RT- QPCR would be important for the validation of the Xiwi CLIP. 26 Bibliography 1. Piwi-interacting RNAs: Biological functions and biogenesis. Kaoru Sato, Mikiko C.Siomi. 2013, Biochemical Society, pp. 39-52. 2. Biology of PIWI-interacting RNAs: new insights into biogenesis and function inside and outside of germlines. Ishizu H, Siomi H, Siomi MC. 2012, Genes & development, Vol. 26, pp. 2361-2373. 3. Piwi Proteins and piRNAs step onto the Systems Biology Stage. Clark, Joesph P and Lau, Nelson C. 2014, Advances in Experimetnal Medicine and Biology, Vol. 825, pp. 159-197. 4. The Drosophila RNA methyltransferase, DmHen1, modifies germline piRNAs and single-stranded siRNAs in RISC. Horwich, MD, et al. 2007, Current biology , Vol. 17, pp. 1265-1272. 5. the Drosophila homolog of HEN1,mediates 2'-O-methylation of Piwi- inte

racting RNAs at their 3' ends. Saito K,
racting RNAs at their 3' ends. Saito K, Sakaguchi Y, Suzuki T, Siomi H, Siomi MC. Pimet,. 2007, Genes & development, Vol. 21, pp. 1603-1608. 6. Argonaute proteins: functional insights and emerging roles. Meister, Gunter. 2013, Nature Reviews Genetics, pp. 447-459. 7. Abundant and dynamically expressed miRNAs, piRNAs, and other small RNAs in the vetebrate Xenopus tropicalis. Armisen, Javier, et al. 2009, Genome Research, Vol. 19, pp. 1766-1775. 8. Uniting germline and stem cells: the function of Piwi proteins and the piRNA pathway in diverse organisms. Juliano, C, Wang, J and Lin, H. 2011, Annual review of genetics, Vol. 45, pp. 447-469. 9. PIWI-interacting small RNAs: the vanguard of genome defence. Siomi, MC, et al. 2011, Nat Rev Mol Cell Biol, Vols. 246-258, p. 12. 10. piRNA production requires heterochromatin formation in Drosophila. Rangan, P, et al. 2011, Current biology, Vol. 21, pp. 1373-1379. 11. The Drosophila HP1 homolog Rhino is required for transposon silencing and piRNAproduction by dual-strand clusters. Klattenhoff C, Xi H, Li C, Lee S, Xu J, Khurana JS, Zhang F, Schultz N, Koppetsch BS, Nowosielska, A. 2009, Cell, Vol. 138, pp. 1137-1149. 12. The Cutoff protein regulates piRNA cluster expression and piRNA production in the Drosophila germline. Pane, A, et al. 2011, The EMBO journal, Vol. 30, pp. 4601–4615. 13. Specialized piRNA pathways act in germline and somatic tissues of the Drosophila ovary. Malone, CD, et al. 2009, Cell, Vol. 137, pp. 522-535. 14. Probing the initiation and effector phases of the somatic piRNA pathway in Drosophila. Haase, AD, et al. 2010, Genes and Development, Vol. 24, pp. 2499-2504. 15. EsTablelishment of sTable cell linesof Drosophila germ-line stem cells. Niki, Y, Yamaguchi, T and Mahowald, AP. 2006, Proc Natl Acad Sci, Vol. 103, pp. 16325-16330. 16. Abundant primary piRNAs, endo-siRNAs, and microRNAs in a Drosophila ovary cell line. Lau, NC, et al. 2009, Genome Res, Vol. 19, pp. 1776-1785. 27 17. T

ransposable element dynamics and PIWI re
ransposable element dynamics and PIWI regulation impacts lncRNA and gene expression diversity in Drosophila ovarian cell cultures. Sytnikova, Yuliya A, et al. 2014, Genome Research, Vol. 24, pp. 1977-1990. 18. The use of Xenopus egg extracts to study mitotic spindle assembly and functions in vitro. Desa, i A, et al. 1999, Methods Cell Biology, pp. 385-412. 19. Systematic and single cell analysis of Xenopus Piwi-interacting RNAs and Xiwi. Lau, Nelson C, et al. 2009, The EMBO Journal, Vol. 28, pp. 2945-2958. 20. A Broadly Conserved Pathway Generates 3' UTR-Directed Primary piRNAs. Robine, Nicolas, et al. 2009, Current Biology, Vol. 19, pp. 2066-2076. 21. Two Piwi proteins, Xiwi and Xili, are expressed in the Xenopus female germline. Wilczynska, Anna, et al. 2009, RNA, Vol. 15, pp. 3337-345. 28 Chapter 3: Optimizing the XIWI RIP assay with improved IP and RT-QPCR conditions I. Abstract In chapter 2 original CLIP data based on CLIP counts in the open reading frame of the gene did not reveal any significant difference between positive and negative gene candidates. A noise filter that included CLIP counts in the 3’ UTR but remove open reading frame counts was applied by Dr. Chirn. 10 positive, 11 negative, and 5 transposon candidates were surveyed using RIP, RT-QPCR. 2 dilutions of RNA were used for RT: Normal RT using 10% input vs 9% of RIP RNA, and Diluted RT using 2% Input vs 45% RIP RNA. QCPR of Normal RT showed a directly proportional higher enrichment compared to Diluted RT. Both QPCR of Normal RT, and Diluted RT showed no significant difference in enrichment (%input) between positive, negative, and transposon candidate. Similar to the results in chapter 2, RT-QCPR data did not correlate with noise-filtered CLIP data. Further analysis of the CLIP data to include both 3’UTR and open reading frame CLIP tags could reveal another set of positive and negative genes. Validation of the newly analyzed CLIP data with RIP, RT-QPCR would provide insight

into Xiwi gene interactions. II. In
into Xiwi gene interactions. II. Introduction Original CLIP data based on open reading frame CLIP counts could not be validated in chapter 2. There was no significant difference in enrichment between genes candidates with high CLIP counts and genes with low CLIP counts. An addition of a noise filter on CLIP data was done by Dr. Chirn. The noise filter included CLIP counts from the 3’ untranslated region (UTR) of the gene, but removed CLIP counts from the open reading frame. The noise filtered data provided a new list of positive and negative candidate genes. New positive candidates were selected based on high CLIP counts, and positive mRNA seq. Negative controls had no CLIP counts, and varied in mRNA seq. Positive candidates had amplicons spanning the 3’ UTR according to CLIP tags. Negative candidates had amplicons that spanned the exon, based on mRNA sequence. 10 positive, 11 negative, and 5 transposon candidates were surveyed using RIP, RT-QPCR. Extraction of stage I-IV ooctytes from X. Tropicalis differed from the chapter 2. During size sorting, oocytes were pipetted onto the mesh, exposed to air and filtered by pipetting buffer onto mesh in chapter 2. This time oocytes were kept submerged in buffer during sorting. Exposure to air for an extended period of time would stress the oocytes. X.Tropicalis stage I-IV oocyte extract may be higher quality in these experiments. RIP was modified and standardized in these experiments compared to those in the chapter 2. 20µg instead of 10µg of antibodies for pre-binding to the Protein A/G magnetic beads was used. This would make RIP more efficient and increase the amount of Xiwi pulled down. After the magnetic antibody-Xiwi binding reaction is completed it is washed with 0.2M instead of 0.1M KOAc Q-Column buffer. 4x 5min, and 3 x quick washes were done instead of 3x 5min, and 2xquick wash. The increased salt concentration and number of washes was aimed to reduce non-specific bound RNA to the antibody magnetic beads. Previous 70% of total RIP was us

ed for RNA extraction; however that was
ed for RNA extraction; however that wasted too much RNA. These experiments use 90% of the RIP for RNA 29 extraction. 03/12/15 RIP utilized 70% of RIP for RNA extraction; however this dilution is corrected for in all RIPs. 2 dilutions of RNA were used for RT: Normal RT using 10% input vs 9% of RIP RNA, and Diluted RT using 2% Input vs 45% RIP RNA. Decreased input and increased RIP RNA was done in an attempt to get a more accurate sample population for RT. RT was diluted 1/50x previously and sometimes yielded poor CT values in QPCR. RT was diluted 1/20x before use in QPCR. QPCR revealed that enrichment of Normal RT was proportionally higher than Diluted RT. Both QPCR of Normal RT, and Diluted RT showed no significant difference in enrichment (%input) between positive, negative, and transposon candidate. Similar to the results in chapter 2, RT-QCPR data did not correlate with noise-filtered CLIP data. Validation of CLIP data is would allow further analyses of candidates to understand Xiwi gene interactions. III. Materials and Methods X.Tropicalis Stage I-IV oocyte extraction 3L of 2g/L EMSS in frog water (pH=7.5) used for anesthetizing frog 15min. Frog was monitored for reflexes. Once unconsciousness confirmed, frog was decapitated with garden scissors. Double Y-cut from armpits to sternum, from sternum to groin, and from groin to thighs was made on skin. Same cut was performed on muscle. Ovaries were removed from abdominal cavity using thin tweezers and scissors, and place in a beaker of 1x MBS. Washed 3x with 1xMBS. Oocytes transferred to multiple petri dishes with 1xMBS. Oocytes cut into smaller chunks with tweezers (~4mm3). 1xMBS decanted. Pre-chilled 3mg/ml Type IV collagenase in 1xMBS added, rotated slowly at RT. Fresh collagenase added every 30-40 min for until oocytes are singular. Washed 3x with 1xMBS. Separate clumps with 850µm mesh, stage VI with 600µm mesh, stage V 425µm mesh. Mesh was held in place in a large beaker filled with 1xMBS. Oocytes were transferred on to mesh and separated i

n large beaker by moving mesh up and do
n large beaker by moving mesh up and down. Filtered oocytes were collected in beaker were transferred to smaller mesh and separated as above. All oocytes were kept in 15ml falcon tube, supernatant removed and flash frozen on dry ice. 1ml of 0.1M KOAc QColumn buffer (0.2% NP40, 0.5mM DTT, 1x Protease inhibitor) per 100µl oocyte was added. Dounce homogenized 50x. 1ml aliquots were made. Sonicated 2x (2 on, 2 off for 80sec). Centrifuged at 21,000 crf for 10min, 4⁰C to separate yolk, clear supernatant kept in fresh tube (avoid yolk). Repeat centrifugation step to remove majority of yolk. Stored at -80⁰C. RNA Immuno-precipitate of Stage I-IV oocyte 25µl Life Technologies Pierce™ Protein A/G Magnetic Beads were washed 5x with 1ml 1xPBS (mixed by inversion, placed on magnet 1min, removed FT). 20µg IgG antibody (anti Xiwi, anti IgG) rotated with beads in 150µl 1x PBS for 2hr RT. Aliquot 100µl of stage 1-IV oocyte extract for input RNA and add 500µl trizol (freeze). Aliquot 40µl of oocyte extract for input western blot and add SDS LD. Beads washed 4x with 1xPBS. 1ml 0.1M KOAc QColumn buffer (0.2% NP40, 0.5mM DTT) added then transferred to fresh eppi. Placed on magnet 1 min, FT removed. 100µl oocyte extract + 900µl 0.1M KOAc Qcolumn buffer (0.2% NP40, 0.5mM DTT, 1X protease inhibitor, 10µl/ml ribolock) mixed with beads. Rotated at 4°c for 3 hrs. Kept FT, washed 4x 5min rotating and 3x quick wash with 0.2M KOAc QColumn Buffer (0.2%NP40, 0.5mM DTT). 30 900µl of RIP aliquot for RNA extraction, 100µl for western blot after final wash before placing on magnet. Complete wash and remove FT. 500µl Life Technologies TRIzol® Reagent added to RNA aliquot, 20µl SDS LD added to Western blot aliquot. RIPS before 0312 used 10µg IgG antibody (anti Xiwi, anti IgG). Washed with 0.1M KOAc QColumn Buffer (0.2%NP40, 0.5mM DTT), 3x 4min, 2x quick. 700µl of RIP aliquot for RNA extraction, 300µl for western blot.31 Table 1a: Positive mRNA CLIP candidates Name Sequence Size Comments XT_ago4_2_F ACCTTGGGAC

CTGTAAAGCA 223 61.5 pseudoentichment
CTGTAAAGCA 223 61.5 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate ago4 from noise filtered list rank 2, (-) Strand. Based on CLIP reads. XT_ago4_2_R GGAGATTTGTTGCAGCACGA XT_cnot6l_2_F GGTTAGCCTCCCAAGCTTTG 153 24.1 pseudoentichment, RT-QPCR primers of mRNA CLIP candidate cnot6l from noise filtered list rank 1, (-) Strand. In unannotated 3'UTR. Based on CLIP reads. XT_cnot6l_2_R CCAAGCATCCCAGTTTGCAT XT_ddx3x_2_F TCATTCCTGCCCTACACTCC 153 2.2 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate ddx3x from noise filtered list rank 9, (-) Strand. 3' UTR. Based on CLIP reads. XT_ddx3x_2_R GCCCTTGTACATTTTGCCCA XT_gsk3a_F GTCGTGGCTGCTCTATAGGT 156 4.4 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate gsk3a from noise filtered list rank 48, (+) strand. Unannotated 3'UTR. Based on CLIP reads. XT_gsk3a_R TCCCAGCACACAAAGGGTTA XT_Larp4_2_F AGCGTGGAGAATGGTCAAGA 223 57.8 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate Larp4 from noise filtered list rank 7, (-) Strand. Unannotated 3'UTR. Based on CLIP reads. XT_Larp4_2_R CTGCTAGTCTCTCTGCGTGA XT_mos_F CCCAGGTGTCAATCTTGCAG 170 39.6 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate mos from noise filtered list rank 24, (-) Strand. Based on CLIP reads. XT_mos_R ACATCTTAAGTGGCCCCTCC XT_pum2_2_F CTGGAGGAATTCATGGAGGA 169 19. pseudoentichment, RT-QPCR primer of mRNA CLIP candidate pum2 from noise filtered list rank 41, (-) Strand. Unannotated 3'UTR. Based on CLIP reads. XT_pum2_2_R AATGCACCTTTTGTGCTTCC XT_sp1_2F GATGTTGCAGGTAGTGGTCG 221 75.9 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate sp1 from noise filtered list rank 16, (+) strand. Unannotated 3'UTR. Based on CLIP reads. XT_sp1_2R GCAGTTTTGACAGTGATCGGA XT_yap1_3_F CCAACTTCAGGCAGCTTGTT 225 2.1 pseudoentichment, RT-QPCR primer of mRNA CLIP candidate yap1 from noise filtered list rank 21, (-) Strand. Unannotated 3'UTR. Based on CLIP reads. XT_yap1_3_R TAAGCAGA

CACGCACAGGTA XT_TNRC6A_2F TCTCCGAC
CACGCACAGGTA XT_TNRC6A_2F TCTCCGACGCATTAGTCCTT 212 22. pseudoentichment, RT-QPCR primer of mRNA CLIP candidate TNRC6A from noise filtered list rank 10, (-) Strand. Unannotated 3'UTR. Based on CLIP reads. XT_TNRC6A_2R CTTGCTGTTGGAATGGGTTT Table 1a- Positive mRNA CLIP candidates- Positive candidates’ design based on pseudoenrichment above 0; Amplicons were designed based on CLIP tags, and spanned the 3’UTR with length ranging from 153 - 225bp. Table 1b- “negative” control CLIP candidates- Negative candidates’ design based on total pseudoenrichement of 0, and positive mRNA seq RPM; Amplicons spanned the exons or 3’ UTR with length ranging from 88 - 216bp. Table 1c-Transposon candidates- Transposon candidates’ design based on total CLIP counts; Amplicons were 90bp – 155bp in length. 32 Table 1b: “negative” control CLIP candidates Name Sequence Size Comments XT_ccnb1_F CATGGAAATGCAAATCCTTAGA 92 (-) control, ccnb1, 1945 RPM no enrichment, 3' end exon XT_ccnb1_R CCTCCCCAATTTTAGATGCTC XT_tsen54_F ACGGGACAATTTTGGTTTTG 113 (-) control, tsen54, 520 RPM no enrichment, 3' end exon XT_tsen54_R AATTGTAGGACACTTTAGCCCTATTT XT_tuba1a_F AACCTGAACCGTCTGATTGG 216 (-) control, tub1a, 1046 RPM no enrichment, 3' end exon XT_tuba1a_R CTCAAAGCAAGCATTGGTGA XT_ICT1a_F AGCAAAGCCGTTACCAGATG 188 (-) control, ict1a, 223 RPM no enrichment, 3' end exon XT_ICT1a_R GCCGGCTCTGTTTTATGGTA XT_dtnbp1_F CCGAGTTTGATTGCCACAGT 185 (-) control, dtnbp, 153 RPM no enrichment, 3' end exon XT_dtnbp1_R GGTAAGAACAGAGTCCGCCT XT_gemin6_F AAGTGGTGAAGGGAGCAGAT 198 (-) control, gemin6, 106 RPM no enrichment, 3' end exon XT_gemin6_R TTAATACACCAGCCACGCAC XT_atp5s_2_F AGGACACTTGCCTGGAGAGA 207 (-) control, atp5s, 82 RPM no enrichment, 3' end exon XT_atp5s_2_R AGGGATTTGCTGTTTGAAGC XT_haghl_F CTTCCAGGCTTCTTCTGCAC 138 (-) control, haghl, 57 RPM no enrichment, 3' UTR XT_haghl_R

CGTCTCTCGTGAGTCCATTG XT_haghl_2_F
CGTCTCTCGTGAGTCCATTG XT_haghl_2_F CCAAGCAATGAGAAGGTGAAG 224 (-) control, haghl, 57 RPM no enrichment, 3' end exon XT_haghl_2_R AACCGATCCTTTGGCTTTTT XT_amy2b_F CCACCCAACCTGTCACTATTAAA 114 (-) control, amy2b, 1228 RPM no enrichment, exon XT_amy2b_R GGCCATCAACCACATTCC XT_rpl18_F AACAAGGACCGCAAGGTG 88 (-) control, rpl18, 1288 RPM no enrichment, exon XT_rpl18_R TGGCCAAGAAACGATAAAGC Table 1c: Transposon candidates Name Sequence Size Comments Chap4a_3_F GTTGATATCCCCGCAGTGAG 90 RT-QPCR primer of chap4a transposon Chap4a_3_R CACACCCACTTCAAACCACA cr1_1a_2_F AACTGGGAGAATGAGTGGCA 155 RT-QPCR primer of cr1_1a transposon cr1_1a_2_R CCTCTAGTGCTCTAACGCGA Piggybac N2_2_F TCTGGGGTGAGCGTACTTTT 144 RT-QPCR primer of piggybacN2 transposon Piggybac N2_2_R ACCTAAGTATGTGGCATGTAGGG 33 IV. Results Positive mRNA and negative controls were selected based on CLIP data, and mRNA sequencing of Xiwi completed by Yuliya Sytnikova and analyzed by Dr. Chirn. A noise filter was applied to the original CLIP data. The original CLIP data included CLIP counts only in the open reading frame of the genes but not in the 3’ UTR. The noise filter included CLIP counts in the 3’ UTR but removed those that were in the open reading frame of the genes. The noise filter dramatically changed the CLIP rank of gene candidates, thus new positive and negative gene candidates were selected. Positive controls include: ago4, cnot6l, ddx3x, gsk3a, larp4, mos, sp1, TNRC6A, yap1, and pum2 (Table 1a). Transposons: cr1_1a, chap4a, and piggybacN2 (Table 1c). “Negative” controls: amy2b, atp5s, ccnb1, dtnbp1, fam136a, gemin6, haghl, ict1a, rpl18, and tsen54 (Table 1b). Fig. 1: Western blots confirming efficacy of Xiwi IP in RIP experiments A B C Figure 1- Western blots confirming efficacy of Xiwi IP in RIP experiments- Input, Xiwi RIP flow through, IgG RIP, and Xiwi RIP ran on 10% SDS PAGE gel, 80v 30min, 130v to end of gel. Transferred onto nitr

ocellulose using semi-dry apparatus i
ocellulose using semi-dry apparatus in 1xTBE, 0.16A 30min. 1:5000 Rabbit IgG Xiwi AB in 5% Milk 1x TBST. 1:10,000 IgG αHRP Rabbit AB. ECL soln 2 min. Exposed 1 min. A) 03/12/15 RIP. Input contains large amount of Xiwi (100kDA). Majority of Xiwi was pulled down during RIP, small amounts of Xiwi remain in Xiwi FT. IgG RIP did not pull down Xiwi. IgG heavy chain (60kDA) can be seen in both IgG and Xiwi RIP. B) 03/23/15 RIP. Input has 2 bands of Xiwi; IgG RIP has no Xiwi, and faint heavy chain band. Xiwi RIP was pulled down 100%, with no Xiwi in FT. C) 03/30/15 RIP Input showed 4 bands of Xiwi; IgG RIP has no Xiwi, and has heavy chain band. Xiwi RIP has Xiwi band, and heavy chain band. Xiwi FT was not loaded. 34 RNA Immuno-precipitate of Xiwi, and IgG control was carried out on Stage I-IV oocytes from X. Tropicalis. Pull down of Xiwi was confirmed on a Western Blots (Fig. 1). Xiwi was pulled down in all three RIPs. The 03/23/15 RIP was the most efficient and 100% Xiwi pull down was achieved, as no Xiwi remained in the FT. The heavy chain band of 03/23/15 IgG RIP might indicate low levels of antibody. In the 03/12/15 RIP 70% of RIP Protein A/G magnetic beads was used for RNA extraction. In both 03/23/15 and 03/30/15 RIP 90% of RIP Protein A/G magnetic beads were used for RNA extraction. Extracted RNA concentration was measured with nanodrop spectrophotmeter. Input showed much greater levels of RNA compared to Xiwi RIP. No RNA was detected in IgG RIP. Pseudoenrichment was calculated by dividing Total CLIP counts by the mRNA seq RPM (Table 2). Pseudoenrichment varied depending on the type of analysis used on the CLIP data. Pseudoenrichment of the total CLIP counts was the highest. Pseudoenrichment of the noise filtered CLIP data depended on the number of CLIP counts at the 3’ UTR of the gene. Pseudoenrichment of the gene centric analysis depended on CLIP counts in the exons of the gene. The positive candidates selected in this chapter have higher, and non 0 pseudoenrichment values compared to the c

andidates selected in chapter 2. Enr
andidates selected in chapter 2. Enrichment (% input) levels of positive candidates were positive, however no representative negative controls were found. Pseudoenrichment of negative candidates were 0 or below 1. Xiwi RIP enrichment (%input) levels of negative candidates were high, and were not representative of genes that did not interact with Xiwi.Transposons CLIP data was only analyzed once, as there are no 3’UTR’s in transposons. Pseudoenrichment of transposon ranged from 0.2 – 25.1; even though enrichment levels correlate to pseudoenrichment, the difference between them is very low. 35 Gene XIWI CLIP PseudoEnrichment Calculations Enrichment All Counts Noise_Filter Gene Centric 10/29/14 RIP 11/26/14 RIP 12/04/14 RIP 12/08/14 RIP Normal RT Normal RT St Dev Diluted RT Diluted RT StDev Positive Candidates ago4 208.5 61.6 35.9 52% 42% cnot6l 79.4 24.1 10.9 2% 44% 10% 23% 10% ddx3x 11.8 2.2 5.6 4% 45% 7% 29% 19% gsk3a 22.1 4.4 7.5 42% 4% 15% 9% larp4 236.7 57.8 43.8 1% 57% 24% 25% 12% mos 235.2 39.6 83.1 41% 9% 16% 10% pum2 84.4 19.1 19.8 34% 29% sp1 267.9 75.9 24.1 30% 14% 22% 8% TNRC6A 11.8 2.1 3.3 12% 6% 7% 6% yap1 78.0 22.7 9.7 8% 38% 12% 10% 8% ccna1 2.4 0.0 3.8 31% 18% 8% 11% cpeb1 0.9 0.0 1.4 33% 5% elf1 1.9 0.0 3.8 24% 8% fam46b 13.4 1.7 13.7 6% foxh1 15.8 0.2 17.2 17% 13% 12% papss1 9.7 0.0 17.9 27% 23% plk3 5.5 0.0 19.2 29% 8% spdyc 1.9 0.2 3.5 26% 8% 10% velo1 0.6 0.0 16.8 22% 7%

4% zfp36l2.2 6.9
4% zfp36l2.2 6.9 0.0 21.1 18% 12% 5% Negative Candidates amy2b 0.0 0.0 0.0 34% 4% 5% 73% 28% atp5s 0.1 0.0 0.1 27% 17% 63% 26% aurkb 0.4 0.0 0.8 32% 18% ccnb1 0.1 0.0 0.3 21% 12% 8% 49% 22% 12% 5% dtnbp1 0.1 0.0 0.1 38% 14% 28% 13% fam136a_2 0.1 0.0 0.1 68% 31% 19% 10% gemin6 0.1 0.0 0.1 41% 18% 16% 7% glod5 0.0 0.0 0.0 12% haghl 0.1 0.0 0.1 31% 3% 24% 44% 7% 16% 7% ict1 0.1 0.0 0.0 47% 4% 6% 6% mkrn3 0.2 0.0 0.6 19% 9% 8% ndufa13 0.1 0.0 0.0 23% 6% pmm1 0.3 0.0 0.0 60% rpl18 0.0 0.0 0.0 27% 40% 12% 16% 6% rpl23 0.0 0.0 0.0 25% rps11 0.1 0.0 0.1 39% 8% 3% 3% tacc3 0.4 0.0 0.6 28% 12% tpd52 0.1 0.6 0.0 30% 14% tsen54 0.1 0.0 0.0 61% 22% 31% 28% tuba1a 0.0 0.0 0.0 Transposons chap4a 0.2 49% 8% 3% 31% 16% 17% 17% cr1_1a 25.1 66% 16% 4% 13% 35% 14% 23% 7% cr1_2 14.4 90% 10% 2% Helitron-N1A 4.7 75% 18% 4% 13% PiggybacN2 2.1 79% 11% 2% 4% 21% 14% 21% 19% Table 2- Pseudoenrichment, and Enrichment (%input) of Fall and Spring RIPs- List of positive, negative and transposon targets used for RIP, RT-QPCR analysis. Pseudenrichment = (total CLIP counts)/ (mRNA seq RPM +1). No filter total CLIP includes CLIP counts from open reading frame, and 3’UTR of genes. Noise Filter CLIP accounted for 3â

€™UTR CLIP tags but reduced open readin
€™UTR CLIP tags but reduced open reading frame counts. Gene centric analysis focused on CLIP counts in open reading frame, but failed to account for 3’UTR. Pseudoenrichment varied largely across different analyses. Enrichment (% input) of 10/29/14, 11/26/14, Normal RT, and Diluted RT RIPs for each gene is presented. Normal RT and Diluted RT utilized 3 RIPs: 03/12/15, 03/23/15, and 03/30/15. 10/29/2014 RT, 5% Input vs 10.5% RIP. 11/26/14 RT used 1% input vs 7% RIP. Normal RT, 10% Input vs 9% RIP. Diluted RT 2% Input vs 45% RIP. Enrichment levels in 10/29/14 RIP were higher than 11/26/14 RIP. Enrichment levels of Normal RT were higher than Diluted RT. Pseudoenrichment calculation using CLIP did not correlate with enrichment (%input) of RIP, RT-QPCR in all trials; Positive, negative, and transposon targets had enrichment levels that were not significantly different. Genes in red were positive candidates used in chapter 2. Table 2: Pseudoenrichemnt of all count, noise filtered, and gene centric CLIP data. Enrichment (% input) of 10/29/14, 11/26/14, 12/04/14, 12/08/14, Normal RT, and Diluted RT 36 Table 3a: QPCR enrichment (%Input) of normal RT Candidate % Input Xiwi Rip Average St Dev % Input IgG Rip Average St Dev 0312 RIP 0323 RIP 0330 RIP St Error 0312 RIP 0323 RIP 0330 RIP mRNA Ago4_2 68% 84% 4% 52% 42% 24% 0.10% 0.48% 0.01% 0% 0% cnot6l_2 39% 56% 39% 44% 10% 6% 0.10% 0.35% 0.18% 0% 0% ddx3x_2 42% 53% 40% 45% 7% 4% 0.15% 0.39% 0.35% 0% 0% gsk3a 45% 44% 38% 42% 4% 2% 0.02% 0.22% 0.09% 0% 0% larp4_2 58% 80% 32% 57% 24% 14% 0.16% 0.23% 0.02% 0% 0% mos 47% 45% 30% 41% 9% 5% 0.05% 0.20% 0.00% 0% 0% sp1_2 28% 45% 18% 30% 14% 8% 0.01% 0.07% 0.02% 0% 0% TNRC6A_2 9% 18% 7% 12% 6% 3% 0.01% 0.02% 0.18% 0% 0% pum2_2 28% 65% 9% 34% 29% 17% 0.02% 0.13% 0% 0% yap1_3 26% 50% 37

% 38% 12% 7% 0.09% 0.19% 0.0
% 38% 12% 7% 0.09% 0.19% 0.01% 0% 0% Transposons cr1_1a_2 32% 50% 22% 35% 14% 8% #DIV/0! 0.06% 0.00% #DIV/0! #DIV/0! chap4a_3 19% 49% 23% 31% 16% 10% 0.06% 0.18% 0.01% 0% 0% PiggybacN2_2 16% 37% 12% 21% 14% 8% 0.27% 0.02% #DIV/0! #DIV/0! #DIV/0! Negative Controls amy2b 45% 101% 74% 73% 28% 16% #DIV/0! 0.04% 0.11% #DIV/0! #DIV/0! atp5s_2 44.20% 81.62% 63% 26% 15% 0.00% #DIV/0! #DIV/0! #DIV/0! ccnb1 25.52% 54.04% 68.87% 49% 22% 13% 0.09% 0.25% 0.23% 0% 0% dtnbp1 25.96% 36.40% 52.92% 38% 14% 8% 0.01% 0.22% 0.10% 0% 0% fam136a_2 32.41% 84.21% 87.48% 68% 31% 18% #DIV/0! 0.92% 0.24% #DIV/0! #DIV/0! gemin6 23.89% 39.15% 59.54% 41% 18% 10% 0.05% 0.19% 0.25% 0% 0% haghl 32% 69% 50% 50% 19% 11% 0.04% 0.02% 0.02% 0% 0% haghl_2 36.46% 48.53% 48.03% 44% 7% 4% 6.63% 1.88% #DIV/0! #DIV/0! #DIV/0! ict1a 51.75% 43.59% 47.04% 47% 4% 2% 0.44% 0.57% 0.35% 0% 0% rpl18 30% 54% 38% 40% 12% 7% 0.03% 0.06% 0.25% 0% 0% tsen54 35% 77% 69% 61% 22% 13% 0.11% 0.07% 1.03% 0% 1% RT of three RIPs (03/12/15, 03/23/15, and 03/30/15) was done. Two different dilutions of RT were made: Normal RT using 10% Input vs 9% RIP RNA and Diluted RT using 2% Input vs 45% RIP RNA. QPCR was done and enrichment (% input) was calculated, and corrected for dilutions. The average, standard deviation, and standard error were calculated based on enrichment of the three RIPs. Candidate genes showed Xiwi RIP enrichment (% Input) from 12% - 73% in normal RT, and 6% - 31% in diluted RT (Table 3a, b). IgG RIP showed little to no enrichment in both Normal and Diluted RT, supporting the fact that the RNA is coming from Xiwi. 37 Table 3b: QPCR enrichment (%Input) of diluted RT Candidate % Input Xiwi Rip % Input IgG Rip 0312 RIP 032

3 RIP 0330 RIP Average St Dev
3 RIP 0330 RIP Average St Dev St Error 0312 RIP 0323 RIP 0330 RIP Average St Dev mRNA cnot6l_2_Dil 26% 32% 12% 23% 10% 6% #DIV/0! 0% 0% #DIV/0! #DIV/0! ddx3x_2_Dil 42% 7% 37% 29% 19% 11% 0% 0% 1% 1% 0% gsk3a_Dil 14% 25% 7% 15% 9% 5% 0% 0% 0% 0% 0% larp4_2_Dil 33% 32% 11% 25% 12% 7% 0% 0% 0% 0% 0% mos_Dil 19% 25% 5% 16% 10% 6% 0% 0% 0% 0% 0% sp1_2_Dil 24% 29% 13% 22% 8% 5% 0% 0% 0% 0% 0% TNRC6A_2_Dil 4% 14% 2% 7% 6% 4% 1% 0% #DIV/0! #DIV/0! #DIV/0! yap1_3_Dil 10% 18% 2% 10% 8% 5% 0% 0% 0% 0% 0% Transposons chap4a_3_Dil 35% 14% 2% 17% 17% 10% 2% 1% 0% 1% 1% cr1_1a_2_Dil 16% 30% 24% 23% 7% 4% 1% 1% 1% 1% 0% piggybacN2_2_Dil 2% 39% 23% 21% 19% 11% 1% 3% 0% 1% 2% Negative Controls ccnb1_Dil 13% 17% 7% 12% 5% 3% 0% 0% 0% 0% 0% dtnbp1_Dil 28% 41% 15% 28% 13% 7% 0% 1% 3% 1% 2% fam136a_2_Dil 23% 27% 7% 19% 10% 6% 0% 0% 0% 0% 0% gemin6_Dil 18% 21% 8% 16% 7% 4% 0% 0% 0% 0% 0% haghl_2_Dil 18% 22% 8% 16% 7% 4% 0% 0% #DIV/0! #DIV/0! #DIV/0! ict1a_Dil 5% 12% 0% 6% 6% 4% 0% 0% 0% 0% 0% rpl18_Dil 16% 22% 9% 16% 6% 4% 0% 0% 0% 0% 0% tsen 54_Dil 56% 35% 1% 31% 28% 16% 1% 0% #DIV/0! #DIV/0! #DIV/0! Table 3a- QPCR enrichment (%Input) of normal RT- Average enrichment of 03/12/15, 03/23/15, and 03/30/15, with standard deviation, and standard error using Normal RT (10% input vs 9% of RIP). Xiwi RIP showed positive enrichment, IgG RIP showed low to no enrichment. Table 3b- QPCR enrichment (%Input) of diluted RT- Average enrichment of 03/12/15, 03/23/15, and 03/30/15, with standard deviation, and standard error using Normal RT (2% input vs 45% of RIP). Xiwi RIP showed positive

enrichment, IgG RIP showed low to no
enrichment, IgG RIP showed low to no enrichment. 38 Figure 2: Average Enrichment (% Input) of all candidates with Standard Error Bars Figure 2- Average Enrichment (% Input) of all candidates with Standard Error Bars-RT-QPCR enrichment (% input) values of all candidates. Normal RT using 10% input vs 9% of RIP. 2a) Average enrichment with standard error showed no statistically significant difference among positive, negative, and transposon candidates. 39 Figure 3a: Average Enrichment (% Input) of selected candidates using Normal RT Figure 3b: Average Enrichment (% Input) of selected candidates using Diluted RT Figure 3a, b-Average Enrichment (% Input) of selected candidates using Normal RT and Diluted RT- Normal RT, 10% Input vs 9% RIP. Diluted RT 2% Input vs 45% RIP. 3a) Average enrichment with standard error showed no statistically significant difference among positive, negative, and transposon candidates. 3b) Average enrichment with standard error showed no statistically significant difference among positive, negative, and transposon candidates. Average enrichment levels using Diluted RT are proportionally lower than Normal RT. 40 Table 3: Cross comparison of enrichment (%input) across various RT dilutions Enrichment levels in Normal RT are proportionally higher than Diluted RT in most genes (Fig. 3a, b Table 3). Cross – comparison of enrichment between chapter 2 and chapter 3 RIPs revealed that absolute enrichment levels vary widely across RIPs (Table 2). Average enrichment with standard error of Normal RT showed no statistically significant difference among positive, negative, and transposon candidates (Fig. 2). The same trend is noted in the Diluted RT (Fig. 3b). Pseudoenrichment values do not correlate with Xiwi RIP enrichment levels in all RIPs, and validation of CLIP has not been accomplished (Table 2). More details on primers can be found in QPCR primers Fall and Spring excel and Spring 2015 prime

rs. QPCR enrichment data can be found in
rs. QPCR enrichment data can be found in Compiled Spring QPCR RIP, and Normal vs Diluted RT excel files. More detail on other RIPs, and Western Blots can be found in lab note book, and Spring CLIP RIP presentation PowerPoint. CLIP data can be found in Combine Comparative CLIP excel. V. Discussion Positive controls were selected based on high average CLIP counts from the noise filtered CLIP data. The noise filtered CLIP data is significantly different from the original CLIP data. Genes like ccnb1 was used as a positive control in the original CLIP data, but was a negative control in the noise filtered CLIP data. Negative controls had low or no CLIP counts, and various levels of mRNA sequence reads assigned per million mapped reads (RPM). Xiwi enrichment (% input) was not statically different between positive and negative controls. The difference in enrichment does not reflect the CLIP data. The positive candidates had primers designed based on the CLIP data. The CLIP data showed most counts at the 3’ untranslated region (UTR) of the mRNA. According the X.Tropicalis daTablease, some CLIP counts were detected after the 3’ end of the gene. This may be due to the fact that the 3’UTR has not been fully annotated in X.Tropicalis. The primer amplicon not being within the exons of the gene might affect enrichment levels of positive candidates. However, that portion can still be amplified well in both input and Xiwi RIP. This provides some validity to the CLIP experiments, as gene found in CLIP were also found in RIP. Further analysis of the CLIP data might provide some insight into why the current positive and negative Gene Name Enrichment (%Input) 10/29/2014 RT Normal RT Diluted RT ccnb1 21% 49% 12% rpl18 27% 40% 16% haghl 31% 44% 16% Table 3- Cross comparison of enrichment (%input) across various RT dilutions- 10/29/2014 RT, 5% Input vs 10.5% RIP. Normal RT, 10% Input vs 9% RIP. Diluted RT 2% Input vs 45% RIP. The enrichment (% input) of ccnb1, rpl18, and haghl changes proportio

nally depending on the dilution used.
nally depending on the dilution used. 41 controls show no significant difference in enrichment. Since the negative candidates have no CLIP counts, the primers are designed in the exons of the gene. Even a housekeeping gene like rpl18 showed low level of enrichment. The high background of negative candidates could indicate that Xiwi interacts loosely with many genes. According to Armisen, piRNAs are found in low, mixed, and high copy number blocks in the germline (1). The diversity of RNA candidates of Xiwi makes it difficult to differentiate between genes that are positive candidates’ versus negative candidates. In Drosophila germ cells there is a protein Mael that is involved in the production of piRNAs, and is required for germline-specific transposon silencing (3). There is a homolog of maelstrom in Xenopus (4). If a protein similar to maelstrom regulates gene silencing in the germline; it is possible that Xiwi could interact with many genes at a certain level, and silencing is activated by another regulatory protein. There is a large difference between CLIP, and RIP. The UV cross-linking done in CLIP could potentially introduce false candidates that are bound to Xiwi. UV cross-linking could also potentially enhance binding for genes that interact highly with Xiwi, while reducing those that do not interact as much with Xiwi. RNase treatment in CLIP on the other hand could destroy RNA candidates that were lowly expressed. If this occurred, the CLIP data would have underestimated and overestimated the significance of some gene candidates. Conducting RT-QPCR using CLIP would help clarify the results from RIP. CLIP without RNase treatment could also be done, to prevent destruction of RNA targets. Even though TNRC6A is ranked the 10th highest based on average CLIP counts, it showed the lowest enrichment compared to all other positive and negative candidates. The highest ranked gene cnot6l was not significantly higher than any of the negative controls. Even among positive genes,

rank did not directly correlate to enri
rank did not directly correlate to enrichment. Negative controls have enrichment levels similar to or higher than the positive candidates. Negative controls also vary widely in enrichment level. The positive candidates selected in this chapter have higher, and non 0 pseudoenrichment values compared to the candidates selected in chapter 2 (Table 2). Enrichment (% input) levels of positive candidates were positive, however no representative negative controls were found. Pseudoenrichment of negative candidates were 0 or below 1. Xiwi RIP enrichment (%input) levels of negative candidates were high, and were not representative of genes that did not interact with Xiwi. Cross-comparison of RIP data from chapter 2 shows large variation in Xiwi RIP enrichment levels for the same gene (Table 2). These variations could be due to extract quality, RIP standardization, and amount of RNA used in RT for input and RIP. However, even with standardized procedures in 03/23/15, and 03/30/15 there is discrepancy in enrichment levels (Table 3a, b). This may be due to the use of frozen stage I-IV oocyte extract that may decrease in quality over time. RT products also degrade over time and should be used for QPCR as soon as possible. Extraction of stage I-IV ooctytes from X. Tropicalis differed from the previous chapter. During size sorting, oocytes were pipetted onto the mesh, exposed to air and filtered by pipetting buffer onto mesh in chapter 2. In this chapter oocytes were kept submerged in buffer during sorting. Exposure to air for an extended period of time would stress the oocytes. Stage I-IV oocyte extract may be higher quality in these experiments. 10/29/14 RIP showed higher Xiwi RIP enrichment levels compared to 11/26/14, 42 12/04/14, and 12/08/14 (Table 2); this could be due to the fact that the latter three RIPs used the same batch of frozen extract. This shows that the quality of extract could severely affect the efficiency of RIP, RT-QPCR. RIP was modified and standardized in these experiments compared to th

ose in the previous chapter. I modifie
ose in the previous chapter. I modified and optimized the RIP experiment in this chapter compared to the previous chapter to attempt to achieve clearer Xiwi transcript enrichment. First, I used 20ug of antibody per IP instead of 10ug in binding to the Protein A/G beads. Second, I used 0.2M KOAc Q-Column buffer instead of a 0.1M KOAc Q-Column buffer for washes after the Xiwi protein-antibody binding reaction. Third, the percentage of RIP used for RNA extraction was increased from 70% to 90%. 03/12/15 RIP utilized 70% of RIP for RNA extraction; however this dilution is corrected for in all RIPs. Freshly made 0.1M Q-Column buffer was also made for these experiments. The fresh buffer might have helped the RIP efficiency in these experiments as it is used for both the protein binding reaction, and washes. Various dilutions of RT were used to for QPCR. Normal RT used 10% input vs 9% of RIP RNA. Diluted RT used 2% Input vs 45% RIP RNA. 10/29/2014 RT used 5% Input vs 10.5% RIP RNA. The standard used by Sigma Aldrich RIP use 1% input vs 100% RIP RNA (2). However, when % input is calculated the dilution is factored in to bring input to 100%. Xiwi, and IgG RIP also have dilutions factored in, and are brought to 100%. The diluted RT was done in an attempt to separate the positive and negative controls. Using a high amount of input RNA in the RT could overwhelm the reaction and cause unequal levels of transcription. RNAs that are highly expressed and abundant could be falsely enriched due to availability of RT in the reaction. RNAs expressed at low levels could be under represented in the RT reaction. Using a more dilute input RNA could provide a more representative RT. Diluting the input RNA too much could cause the CT values in the QPCR to be too high. Using dilute RIP RNA could underrepresent highly expressed RNA, and over represent lowly expressed RNA. This would cause enrichment levels of highly expressed and lowly expressed RNAs to look similar in profile. Accounting for the input and RIP dilutions caused different l

evels of enrichment. A cross compari
evels of enrichment. A cross comparison of ccnb1, rpl18, and haghl using different normal, diluted, and 10/29/14 RT showed that enrichment level per gene increases proportionally to the dilution factor (Table 2). The RT-QPCR data from each RIP experiment is consistent relative to each gene candidate. In the future more gene candidates can be surveyed. The CLIP data should also be reanalyzed to include both open reading frame and 3’ UTR CLIP counts. A new set of positive and negative genes could be revealed. As shown in the normal CLIP data versus the noise filtered CLIP data. Primer design for all candidate genes should be based on exon and mRNA sequences. The RIP can still be optimized as not 100% of Xiwi is being pulled down from input (Fig. 3). More Protein A/G Magnetic beads and Xiwi antibody could be used to increase efficiency. DNase treatment of input fraction would help eliminate DNA signals; however the previous experiment showed that DNase treatment did not show a large effect on enrichment levels. The amount of RNA used in RT should be 1% input vs 100% RIP in the future. This could potentially represent the RNA population in the RT better, and allow positive and negative candidates to be distinguished. An important experiment would be to use CLIP, RT-43 QPCR for validation of the CLIP data. This would reduce the confounders between RIP and CLIP. 44 Bibliography 1. Abundant and dynamically expressed miRNAs, piRNAs, and other small RNAs in the vetebrate Xenopus tropicalis. Armisen, Javier, et al., et al. 2009, Genome Research, Vol. 19, pp. 1766-1775. 2. Sigma Aldrich. Imprint® RNA Immunoprecipitation (RIP) Kit. Sigma Aldrich. [Online] 2013. [Cited: April 20, 2015.] http://www.sigmaaldrich.com/catalog/product/sigma/rip?lang=en®ion=US. 3. Functional and structural insights into the piRNA factor Maelstrom. Sato, K and Siomi, M.C. March 2015, FEBS Letters. 4. Xenbase. Mael. [Online] 2015. [Cited: April 30, 2015.] http://www.xenbase.org/gene/showgene.do?method=display&geneId=X